By Dr. Tariq Rahman, Fulbright Visiting Fellow

What was the language of the Indus Valley, present-day Pakistan, in the pre-Islamic period? Did this region have one language or many? Did it have one language family or many? In which script, or scripts, were they written? These questions cannot be answered by the linguist alone. To answer them one needs the help of the archaeologist, the historian and the anthropologist. Let us then begin with the evidence about the Indus Valley civilization brought to light by the archaeologists first.

The Indus Valley Civilization

The Indus Valley, or Harrapan, civilization 1 was discovered in 1920-21 when engraved seals were discovered near present-day Sahiwal in Pakistani Punjab at a place called Harappa.2 Later Rakhal Das Banerjee, John Marshall, E.J.H Mackay and M.S.Vats carried out excavations at Mohenjodaro in Sind and discovered the buried remains of a civilization with a pictographic script.3Many archaeologists , including the celebrated Sir Mortimer Wheeler, added to our knowledge of this civilization .4 We now know that it extended to the Yamuna 5 along the bed of the river Ghaggar in Rajhastan, Gujrat and upto the mouths of the rivers Narbada and Tapati.6

It does appear, however, that the major sites of this civilization are in Pakistan. In fact it is in Pakistan that an earlier phase of it has also been unearthed. This happened between 1955-57 when a Pakistani archaeologist, F.A.Khan, discovered a town of the pre-Indus period (c. 3300-2800 B.C) at Kot Diji in Khairpur, Sind.7 Such sites were also discovered by Rafique Mughal in Bahawalpur, especially in the Cholistan desert, extending the area of this culture to the whole of southern Pakistan.8 The area was further extended by Professor Ahmad Hasan Dani, the famous Pakistani archaeologist and Sanskritologist, when he discovered the sites of this civilization at Gumla, seven miles from Dera Ismail Khan.9 In fact Dani identified six cultural periods and Professor Farzand Ali Durrani , who excavated Rahman Dheri which is fourteen miles north of Dera Ismail Khan city, provided more details about the extension of this civilization in the North West Frontier Province. 10

Archaeologists disagree whether the Kot -Diji type of cultural artifacts constitute a separate civilization or an early phase of the same civilization. Rafique Mughal, citing evidence from the excavations at Bahawalpur and Cholistan, concluded that 'all Kot Diji-related sites together constitute an Early Harappan or early urban, formative phase of the Indus Civilization'.11 However, Parpola argues that the term is misleading because it 'suggests discontinuity, like pre-Aryan vs. Aryan'.12 In fact, many scholars treat the latter culture as a changed form of the earlier one. This is significant because, if the dates of the Indus Valley Culture are approximately 2300-2000 B.C, and the dates of the kot Diji one are c. 3300-2800, then the length of the period of urban civilization in South Asia have been pushed back a thousand years. The area of the early culture is given by Mughal as follows:

... the central-northern areas of Baluchistan, the greater portion of Sind and the Punjab, Kalibangan on the Indian side, and the south-western part of the Frontier Province are the regions which are likely to have been comprised within the limits of the Kot Dijian culture. 13

Thus , one may suggest that the area now called Pakistan had some sort of cultural similarity as early as three thousand years before the birth of Christ. Whether it also had linguistic similarity is a question which needs to be answered.

The Language of the Indus Valley

Unfortunately the few symbols on the ceramics of the Kot Dijian culture have not been deciphered. F.A.Durrani, following B.B. Lal and B.K.Thapar, suggests that these symbols may be the beginning of writing in the Indus Valley.14 There are, however, nearly 4,000 specimens of a script from the Indus Valley Civilization carved on stone, fragments of pottery and other objects.15 They have not been deciphered satisfactorily but a history of the attempts at such decipherment is available in Asko Parpola's most recent book on the subject.16 The script, or at least the pictographs, appear to have been uniform but that is not proof that the language too was one. In fact, as in all parts of the world, the language must have been divided in dialects or area-bound varieties. It is possible, however, that these were varieties of a language belonging to one language family. The question then is what that language family was?

Beginning from Sir John Marshall, who was the first to suggest that the language of the Indus Civilization was Dravidian 17, most scholars have taken the 'Dravidian hypothesis' seriously. Piero Meriggi, a scholar who contributed towards the decipherment of the Hittite hieroglyphs, opined that Brahvi, the Dravidian language spoken even now in part of Balochistan, must be the original Harappan language 18. However, Brahvi has changed so much and become so Balochified, as Elfenbein points out 19, that it cannot give clear evidence of any sort in this case. Another scholar, the Spanish Jesuit Henry Heras, 'turned more than 1,800 Indus texts into "Proto-Dravidian" sentences' 20 but his decipherment and linguistic theories were not accepted. Later Soviet scholars headed by Yurij V. Knorozov, carried on a very rigorous computer analysis of sign distribution in the Indus texts coming to the conclusion that it belonged to the Dravidian language family 21. However, Kamil Zvelebil, also a Russian scholar came to the conclusion that 'the Dravidian affinity of the Proto-Indian language remains only a very attractive and quite plausible hypothesis.22 Indeed, the plausibility of the hypothesis is such that many people, such as Iravatham Mahadevan, a scholar of old Tamil epigraphy, have used it to offer readings of the Indus script 23. F.C.Southworth and D.Mc Alpin used the Dravidian roots to reconstruct the language of the Indus Valley.24 Walter A. Fairservis, another specialist in this area, stated with considerable certainty that 'the Harappan language was basically an early Dravidian language'.25 Even Parpola, after much careful and detailed sifting of the evidence, opines 'that the Harappan language is most likely to have belonged to the Dravidian family'.26

The Dravidian Influence on Pakistani Languages

If the Harrapan language family was Dravidian, then the first languages of the area of present-day Pakistan was not Indo-Aryan but Dravidian. Such a claim has been made in an extreme and unsubstantiated form by Ainul Haq Faridkoti, a Pakistani philologist, in his several publications.27 Other scholars have used the theory of linguistic 'transfer' or 'interference' to explain the presence of Dravidian elements in the languages of present-day Pakistan which are generally said to be the daughters of Sanskrit, an Indo-Aryan language. 'Transfer' or 'interference' refers to the influence of the rules of one's first language on another language one learns later. Thus, if Pakistanis learn English, they speak it more or less according to the rules of their first language. As they get more and more exposed to the rules of English, they will speak like native speakers. However, some characteristics of the mother tongue of the speaker will remain which is what we call a Punjabi, Pashto, Sindhi or Urdu accent.28

Sometimes an old language dies out and all its speakers learn a new language. But the way they use this language is influenced by the rules of their old language. The new language, then, has a 'substratum' of the old language. If we apply this theory to old Indo-Aryan we can hypothesize that the Harappan language, which was probably Dravidian, influenced old Aryan. Thus Pakistani languages have a Dravidian substratum. The evidence for the presence of this substratum, according to Bertil Tikkanen, is the presence of retroflex consonants which do not exist in Iranian or European members of the Indo-European family of languages.29 Another clue may be the existence of consonantal clusters in the beginning and end of words in Iranian, European , Dardic languages and even Sanskrit. Thus Sanskrit has /p r e m/ which means love. But Hindi-Urdu speakers call it /p i r e m/. They insert the vowel /i/ between the two word-initial consonants /p/ and /r/ because their own rules of pronunciation (called phonological rules) do not allow word-initial consonantal clusters. Similarly, speakers of Urdu, Punjabi and Sindhi separate consonants in word such as 'school', 'stool' and 'small' etc.30 It may be that this splitting of consonantal clusters comes into some of the languages of South Asia from languages older than Sanskrit. This, however, is a suggestion by the present author which needs much research by linguists for substantiation.

The Coming of the Indo-Aryan Languages

Contrary to the popular myth in Pakistan, the Aryans did not roll down the northern mountains like a tidal wave carrying all the Dravidians before them. According to some scholars they came in at least two major waves in Pakistan as well as small trickles. The first wave came 'around 2000 B.C, and the second some six centuries later'. After the second wave, when they became dominant, their language too spread over northern India. It is this language, or rather a number of dialects, which we call Old Indo Aryan for convenience. The language of the first wave, which remained confined to the Pamir mountains of Pakistan, is identified as Dardic while the second one may be called Indic.31 The chart given in Figure -1, based on George Grierson's classification, may be useful in illustrating the hypothesized relationship.


Note: This chart is abstracted from Grierson, Vol. 1 (P. 120).

The languages not used in Pakistan have generally been left out.

This chart does not take the Dravidian element into account (see Figure 2 for it) but it cannot be ignored in the face of considerable evidence. The evidence of such an element is clearer in vocabulary than in syntax or pronunciation. For instance there are Dravidian loan words in the Rigvedic language:

...including phalam '(ripe) fruit'... as well as mukham 'mouth' and khala - 'threshing-floor'. The Harappans used a plough, and the Rigvedic word for 'plough', langala, is probably derived from Proto-Dravidian * nangal / *nangal.... 32.

The number of Dravidisms -- in all aspects of the language including phonology -- increased in the post-Rigvedic era. According to Burrow:

The large majority appear first in the classical language, but in its early stage, being first recorded in Panini, Patanjali, Mahabharata, Srautasutra, etc. The majority appear also in Pali, which is important for dating since these canonical texts take us back to a period from 500-300 B.C. 33

According to Parpola, by 1800 B.C 'Mohenjo-daro was abandoned' and a 'cultural fragmentation in the Greater Indus Valley' took place.34 When this happened the Harappan language remained as a substratum in the language of the Aryan civilization of the Indus Valley of about 1200 to 1000 B.C when the Rigveda was largely composed in the plains of the Punjab.35

The Sanskritic Legacy

Sanskrit became the elitist language of the Indus Valley from about 1000 B.C and remained in use in some domain or the other, generally religion and the state, till the Muslim conquest when Persian took its place. Thus, although the Prakrits which finally changed into the vernacular languages of the people of Pakistan were simultaneously in use as I will argue later, let us look into the development of Sanskrit first. The Rigveda itself gives importance to language which is personified as a goddess. In Esa Itkonen's translation it glorifies itself as follows:

I gave birth to the father on the head of this world. My womb is in the waters, within the ocean. From there I spread out over all creatures and touch the sky with the crown of my head.

I am the one who blows like the wind, embracing all creatures. Beyond the sky, beyond this earth, so much have I become in my greatness. 36

Language was sacred and change was seen as corruption. But all living languages change and the spoken languages of the people, the Prakrits, changed all the time. This threat was countered by making grammatical rules which would petrify language. The most well known of this set of rules was made by the great grammarian Panini who was born at 'Salatura' which is about twelve miles from Jahangira near the Attock bridge in the North West Frontier Province of Pakistan. In those days this village was part of Gandhara which, according to Panini, comprised 'the valley of the Kabul river, with its frontier outpost at Takshasila'.37 Panini's grammar contains about 4000 rules which were memorized and orally transmitted 'for a couple of hundred years' and was not written down at all. 38 So sacred was the language of the religious texts, Sanskrit, that the grammar itself acquired a central and almost sacrosanct place in the education system of the Indus Valley Aryans. 39

Since Panini lived in what is now Pakistan it was the speech of the elite of this region that was considered 'correct' and it was this that he wrote about. 40 There are, indeed, passages in the Sanskritic texts which bear this out. The following quotations from them are in Hock's translation:

(1) In the northern region, speech is spoken particularly distinct(ly). People go to the north to learn speech. Or if someone comes from there, they like to hear/ learn from him ... For this is known as the region of speech (Kausitaki-Brahmana 7.6).

(2) Through Pathya Svasti they recognized the northern quarter/ region. Therefore there speech speaks better, among the Kuru-Panchalas. For she is really speech (Satapatha-Brahmana 41

Panini was not memorized in isolation. Katayayana (c.250 B.C) and Patanjali (c. 150 B.C), who wrote commentaries on his work, were also part of the canon which aspiring scholars at great centres of Brahmanical learning like Taxila had to learn. 42

The Emergence of the Prakrits

In all probability the Indo-Aryans did not speak one uniformly standardized language but mutually intelligible non-standardized dialects. The process of standardization must have been started by the Brahmins earlier but Panini perfected it in about 400 B.C so that this polished (samskrita) language did not change and was considered superior to the ever-changing dialects which were spoken by the people. As the elite looked down upon the uneducated people, it also held their languages in contempt. Thus the Prakrits were a sign of rusticity and illiteracy as the languages of the ordinary people are even nowadays. But the termprakrriti means 'root' or 'basis' according to Katre who suggests that they existed when Sanskrit was standardized. 43 It is in the light of this insight that we can study the development of the Prakrits into the vernaculars spoken in Pakistan today.

According to George Grierson the Primary Prakrits were living languages in Vedic days. Later they were also fixed by grammarians who wrote their grammars and the living languages of the people were called Secondary Prakrits or 'Sauraseni'. When even these were fossilized by grammarians the Tertiary Prakrits or 'Apabhramasas' were born. By 1000 A.D even the tertiary Prakrits became dated and from this time onward, as we shall see, the modern Pakistani vernaculars emerged. 44 But before we come to the actual emergence of the Pakistani languages let us look at the language of Gandhara.

The Language of Gandhara

According to A.H.Dani 'the new cultural trends of the centuries were identified in the swat, Dir, and Peshawar valleys, and because of its original location in that area, it was termed "Gandhara Grave Culture"'. 45 This region was inhabited by the Dasas who worshipped the snake and must have spoken the Indus Valley's Dravidian languages(s) before the Aryans established their supremacy here. 46 By the first millennium B.C, however, 'the Aryanization of most of the population of the northern areas of the subcontinent was complete'. 47 The elite used Sanskrit as we have seen but the common people used what scholars have called 'North-Western Prakrit' or the 'language of Gandhara'. 48 This language, opines Gankovsky, was probably made up of elements from the languages of the 'local pre-Indo-European population and Indo-Aryan tribes, as well as the Dardic and East-Iranian ethnic elements'. 49

Among the pre-Vedic languages the Dardic languages of the first wave of Aryans who settled down in the Pamir mountains were mentioned earlier. These languages influenced the Indo-Aryan language of Gandhara as the language of the Gandhari Dhammapada bears out. This Buddhist text was written in the Kharoshthi script, which was derived from Armaic and will be dealt with in more detail later, and was discovered in the Chinese Turkestan. The dates of this text is c. 269 A.D. and the language:

agrees closely with the (Post-Asokan) Kharoshthi inscriptions from N.W.India and (slightly less closely) with the Prakrit version of the Dhammapada. Moreover, it shows sufficient characteristics in common with the modern Dardic languages to be assigned definitely to that group, and among these languages it would seem to be most closely allied to Torwali.'50

Torwali is still spoken in the Kohistan region of Pakistan. But Dardic is not the only influence on the Gandharan language. Another influence was Persian.

The Persian Influence

This was hardly surprising because the Gandhara region was ruled by the Persians some time in the sixth century B.C. This is evidenced in the

inscriptions of Darius in which 'clear mention has been made of Hi (n) du, that is, the Punjab territory, as a part of the realm'. 51Further evidence comes from the discovery of an Armaic-Greek inscription of Asoka, the great Buddhist ruler of around 250 B.C., a few miles west of present-day Kandahar in April 1957. Carratelli, writing on this discovery remarks:

... the region had been an old Iranian province and it is logical to assume that the tradition of the Achaeminian state language was maintained. Satrapal offices must have survived during Macedonian domination (when Greek was added) and continued their use of Armaic when the Mauryas took over. The importance of Armaic for administration purposes in the former Iranian provinces is borne out by the Taxila and the Pul-i-Durunteh inscriptions.52

Armaic 'came to the fore' at 'the time of the Assyrian empire and became the principal means of communication in the Persian empire'.53 It was a kind of lingua franca in Gandhara and Bactria (part of present-day Afghanistan). King Ashoka (spelled as Asoka by other writers) used it presumably because the people of Kandahar at that time understood Armaic. However, as Gankovsky points out, the common peoples' language did not become Persianized. 54 Even so the script in which their Prakrit was written came from the Persian empire: it was Kharoshthi.

This script, like other Middle Eastern scripts, was written from right to left and A.H.Dani gives the values of its symbols in Arabic letters in his Kharoshthi Primer. 55 The script was used not only by the Iranian kings who ruled

this part of the world but even by the Mauryas who succeeded them. In fact the script of the famous edicts of Ashoka at Shahbazgarhi and Mansehra (Pakistan) is Kharoshthi whereas the edicts in the rest of India are in Brahmi. 56 This script was 'mainly prevalent in the places which are known as the North Western Province, the Punjab and the Ganges-Jumna Doab roughly from the third century A.D.'57 Latter the script remained in use in Chinese Turkestan for at least a century. But, whatever the script, the edicts of Ashoka are in the Prakrit of Gandhara and not in an Iranian language. Before we go on to the language of Sind, let us look at the Greek influence on the area we now call Pakistan at the linguistic level.

The Greek Influence on the Languages of Ancient Pakistan

The Urdu word for Greek is Yunani (Yunani Tib = Greek medicine). This is 'derived from the Persian Yauna, meaning Ionian'. 58 As the Ionian Greeks -- the Greeks settled in Ionia which is present-day Turkey -- were the first to be encountered by the Persians, they called them, and by extension all Greeks, Ionian or Yunani. From this root comes the Sanskrit word Yavana which one encounters in ancient Sanskrit sources including Panini's grammar. Thus, according to Agrawala, 'the yavanani lipi was known only in Gandahara and the north-west at Panini's time' [lipi = edict].59 This is not surprising because there were Greek settlements in the Hindu Kush even when Alexander entered that area in 327 B.C. Although Alexander did not stay long in India, he left his representatives and the Greeks established their rule in Bactria.

By the third century B.C. the Mauryan kings (ruled c. 317-180 B.C.), of whom Ashoka was so illustrious an example, were losing their grip over the northern part of the subcontinent. The Greek kings of Bactria now seized the Western provinces of the Mauryas and by 180 B.C. the Greek language came to be used in some domains such as coins. King Menander (d. 130 B.C) inherited 'western Punjab and Gandhara up to the Indus, with its capital at Taxila'.60 Under him 'Pushkalawati -- present-day Charsadda near Peshawar -- began its period of prominence as a Greek centre'.61 Since the coins of these Greek kings bear Kharoshthi -- and sometimes Brahmi -- inscriptions, it is evident that this script was never suppressed. Similarly, the local languages continued to be used.62 However, Greek too found a place of prominence and came to be used at least in the elitist domains. According to Woodcock:

For at least a century and a half, in fact, Greek remained not only the commercial but also the patrician lingua franca of the Kabul valley and of Gandhara at least as far as Taxila. Merchants and kings learnt it as a matter of course, as is shown by the experiences of Appolonius of Tyana when he journeyed to Taxila in 44 A.D. 63

By the middle of the first century B.C., Greek rule in Gandhara had come to an end except for an enclave around Peshawar. The Sakas, who were from Central Asia and spoke an Iranian tongue, came to rule Gandhara by 32 B.C.64 Later they left their original language and became strong supporters of Sanskrit.65 They did not, however, stop the use of Kharoshthi or the Greek language altogether. In fact 'the local Saka ruler of Ujjain' sent a letter to Augustus Ceasar in 24 B.C. in Greek.66 The Saka kings also inscribed Greek legends, as well as Kharoshthi ones, on their coins. It is also reported that 'the women of Surastra continued to use the Greek form of greetings' for quite some time.67 The Sakas did, however, become Indianized and language reflects this. According to Chattopadhyaya:

The inscriptions of the successors of Rudradama are also mostly written in Sanskrit. On the contrary, the inscriptions of the contemporary Satavakanus are written in Prakrit, which seems to have been the language of the common folk. The later coins of Damaghasada, son of Rudradaman, are in pure Sanskrit, and the use of Sanskrit legends on the coins was continued by his son Satyadaman also.68

Rudradaman ruled about 130 A.D. and it was during his reign that the 'Sindhu-Sauvira region', which, as we shall see in more details later, has been identified with modern Sind and the lower Pakistani Punjab, was conquered from the Kushanas who had been ruling it earlier. In short, around 200 A.D. Sanskritization was being encouraged at the highest level and classical Sanskrit drama was developing.69 By 295 A.D. the Sakas were subordinates of the Iranian Sassanid kings and by 400 A.D. they had been replaced by the Gupta kings who also patronized Sanskrit.

According to A.H.Dani, Gandhara and the Punjab were ruled by three families by the third and the early part of the fourth century A.D. The Kidara Kushanas too penetrated up to the Hindu Kush and may have acknowledged the suzerainty of the Sassanians or Chandragupta Vikramaditya (ruled c. 375-413 A.D). 70 The point which is relevant for us is the increasing Indianization as Brahmi replaces Kharoshthi on coins and non-Buddhist Indian religions replace Buddhism and Jainism which had favoured the Prakrits initially. The White Huns even destroyed Gandhara, a centre of Buddhist civilization, in 450 A.D. 71 and Sung-Yun, the Chinese traveller who visited this area in 520 A.D., found Taxila being ruled by Mihirakula (d. circa 532 A.D.) who worshipped the Hindu deity Shiva and used the Brahmi script.72 In the 8-10th centuries, opines Dani, 'Taxila went into the hands of the Shahis'.73 The Hindu Shahis, who also ruled part of Afghanistan, probably spoke the prevalent Prakrit of the North Western part of the subcontinent. This could be a descendant of the language of Gandhara which Ashoka used about 250 B.C. Let us now come to the language of Sindh which will lead us on to the languages of Pakistan at about 1000 A.D. when Muslim rule was established in this area.

The Language of Sind

Although the Arabs attacked Sindh earlier, it was Mohammad bin Qasim who conquered it and ruled it for about three years (712-715 A.D) before being recalled and killed.74 The north was conquered by the Turks beginning a little before 1000 A.D. when Mahmud Ghaznavi first entered northern India.75 However, for many centuries before the Muslim conquest, the cultural development of the northern and southern parts of Pakistan would appear to be different. However, it is not easy to assert that Sind has had a different development from the Punjab and the Frontier because it is not always easy to determine the meaning of Sind.

The countries of Sindhu and Sauvira are mentioned in the Mahabharata and have been taken to be roughly the present province of Sind and lower (i.e. Siraiki) Punjab. Some scholars, however, consider them 'neighbouring countries of the Punjab' with Sindhu on the west and Sauvira on the east of the Indus.76 A.H.Dani, however, locates Sindhu roughly in the province of Sind and Sauvira, in his opinion, 'definitely lay to the east of the river Indus much higher up.' This leads to interesting linguistic hypotheses which are best given in Dani's own words as follows:

If we accept this suggestion, it is not difficult to understand why the Sindhi language is confined to the lower Indus while Saraiki is now spoken in much the same area where Sauvira is located by Alberuni. With this understanding of the Saraiki-speaking area, we can now say that the very name Saraiki is probably a corruption of the original term Sauviraki.77

At different periods in history, however, the boundaries of Sind have been shifting. The Achaeminian kings 'of 2500 years ago provide us, in their rock inscriptions, with some thirty names of sixteen Aryan provinces. Among them, we have Hindu (Sindhu) and its adjective Hinduya (Sindhi)'.78 The Muslim historians, however, differentiate between Sind and Hind but their Sind extends up to northern Pakistan. For instance Ibn Khurdadba and Al-Masudi count Kandahar, Multan and Kanauj among the countries of Sind.79 Rashid ud Din, whose work is based upon Al-Beiruni, says:

Hind is surrounded on the east by Chin and Machin, on the west by Sind and Kabul, and on the South by the sea. On the north lie Kashmir, the country of the Turks, and mountain of Meru.80

But Kashmir, according to the evidence of Hsien Tsiang, appears to have been larger than it is now. Around A.D.640, even the Punjab -- or some part of it-- was a dependency of that kingdom as king Kanishka ruled there and Taxila (Ta ch'a shi lo) was a 'tributary to kia-shi-mi-lo (Kasmir)'. 81 This evidence suggests that, for some centuries before the Arab conquest of Sind and Multan and about three centuries after it, we should look for one kind of cultural development in Sind and another kind in Kashmir, parts of Punjab and the North West Frontier Province. In Sind the Rai dynasty and then the Brahmin Chach's line held sway till the eighth century. In Afghanistan and part of Punjab the Hindu Shahis ruled till the Turk Subuktagin defeated Jayapala around 992 A.D and wrested away all the territory west of the Indus including Peshawar.82 In Kashmir a certain Muslim Rajput adventurer called Shah Mir (also called Shams ud Din) ascended the throne around 1339 A.D.83 By this period the indigenous languages of Pakistan were emerging.

The Languages of Pakistan on the Eve of the Muslim Conquest

One of the factors which make it difficult to ascertain the names of the languages of the subcontinent is that the Muslims used to refer to all these languages as Hindi, Hindui or Hindawi. For instance, according to Al-Badaoni, the Commander of the fort of Kalinjar 'composed a poem in Hindi in praise of the Sultan [Mahmud], and sent it to him'.84 Later, in the reign of Mahmud's grandson, the poets Masud Saad Salman and Ustad Abul Faraj Runi both had poetic collections of verse (diwans) in Hindi as well as Persian and Arabic.85 Both these poets lived, at least for some period of their lives, in Lahore around 1114 A.D. and if they wrote in the language of Lahore it could hardly have been what we now understand as Hindi. Badaoni also tells us that Shamsuddin Ayaltimish 'with the assistance of Hindu pundits translated 32 stories about him which are a wonder of relation and strange circumstance, from the Hindui into the Persian tongue and called it Nama-i-Khirad Afza.'86 As this book has not been discovered so far it is impossible to say whether the original was from Kalidasa's Sanskrit as George Ranking, the translator of Badaoni's history, suggests or some Prakrit work.87 Another work called the Mujmalu-t Tawarikh was, however, translated by Abu Salih bin Shuaib bin Jami 'into Arabic from the Hindwani language' in 1026 A.D. Here by 'Hindwani' it is probably Sanskrit which is meant.88

The Muslims did, however, know the differences between some of the Indian languages and despite the generic use of the term Hindi, referred to these differences in some writings. Al-Masudi tells us, for instance, that 'the language of Sind is different from that of India',89 while Ibn Haikal tells us that after the Islamic conquest 'the language of Mansura, Multan, and those parts is Arabic and Sindian. In Makran they use Persian and Makranic.'90 In short the Sindhi language had been identified as a distinct language a little after the Muslim conquest of Southern Pakistan.

According to Grierson the mother of Sindhi was Vrachda. It was the spoken language, or Apabhramsa, 'of the country round the lower Indus.'91 It was also the mother of what Grierson calls Lahnda and what are now known as Siraiki and Hind Ko.92According to the same author 'India had left the Prakrit stage, and had reached the stage of the tertiary Prakrits, i.e. of the modern Indo-Aryan vernaculars, by the year 1000 A.D.' 93. It is possible then, as A.L.Turner opines, that Sindhi must have separated from the mass of related languages sometime between 250 B.C. and the first century A.D.94 John Bordie, using linguistic evidence of loss of certain words per thousand years, suggests that Sindhi and Punjabi separated between A.D.750 to 1400 and that the implosives of the Sindhi language 'came into existence prior to A.D. 1400 and subsequent to the separation of Sindhi from the mass of related languages.'95 Since Siraiki too has implosive sounds, it too may have become a separate language around this period. But Siraiki shares its vocabulary, or at least a major part of the core vocabulary, with Punjabi so that the present writer is unsure whether Siraiki is a sister of Punjabi which picked up some features of the Sindhi sound-system (phonology) or a sister of Sindhi which picked up Punjabi words as Grierson suggests.

In Grierson's opinion Punjabi is the descendant of the Takka Apabhramasa of the North Central Punjab and the Upanagara Apabhramasa of the Southern Punjab.96 However, the language is not mentioned by this name till after the Muslim conquest. In fact the very term Punjab is from Persian (Punj = five and ab = water or river). Since five rivers flow in this region the Persian chroniclers called it 'Punj-ab' and the name replaced the earlier names of the region. However, Amir Khusrau, writing in 1317, calls the language of Lahore not Punjabi but 'Lahori'. Khusrau also mentions Sindhi and Kashmiri but not Pashto or Balochi. It is Abul Fazal who mentions 'Afghan' (Pashto) in the Ain-i-Akbari in the sixteenth century.97

In short, Sindhi, Punjabi, Siraiki and some form of Hind Ko as well as the Dardic languages were spoken in some form or the other in the area now comprising Pakistan. Pashto and Balochi are not part of this article since they were only on the fringes of the boundaries of present-day Pakistan in the tenth century when the Muslim conquests took place. However, they will be mentioned in passing here. According to Gankovsky the Pakhtuns moved in the plains of Peshawar, Kohat and Bannu in the thirteenth and fourteenth centuries, to Swat, Kurram and Panjkora as well as to Zhob, Loralai and Quetta in the fifteenth and the sixteenth centuries. 98 The Baluchi language, which too is descended from the Iranian branch of the Indo-European languages, is a late comer in its present location. However, in the tenth century when the Arabs ruled parts of Baluchistan, they occupied Kalat which was probably Brahvi-speaking even then.99 Baluchi too may have been present on the peripheries but it spread all over Pakistani Balochistan and elsewhere with the raids of the Ghaznavids and the Ghorids.100


Pakistan is heir to some of the most ancient civilizations of the world. Its languages, which were part of the culture of the people of this region, too have ancient roots. These languages have not generally been used in the domains of power because the rulers of this region were generally foreigners. But the foreigners -- whether Achaeminian Iranians, Greeks or Muslim Arabs, Turks and Pathans as well as the British -- have also enriched the indigenous languages so that their vocabulary is multilingual and varied. As the people of this area converted to Islam the Arabic and Persian words became part of their Islamic identity and remain so. In a sense it is their very presence as well as the Arabic-based scripts of all Pakistani languages which give them a kind of cultural unity. Linguistically, then, Pakistan faces two directions: India - because the roots of its languages are Dravidian as well as Indo-Aryan; and the Middle East - because its scripts and vocabulary owe much to Arabic and Persian. To deny any of these directions out of ideological zeal is historically incorrect to say the least. As for the linguist, the chart given as Figure 2 gives a more accurate picture of the development of Pakistani languages than the chart based on Grierson given in Figure 1 earlier.