Faultlines of Hindi and Urdu

Published : Aug 10, 2012 00:00 IST

One example of the huge wall conventional literary history has erected between related linguistic traditions is that of Hindi and Urdu.

The mind has faculties which are universal, but our habits are singular.

Rabindranath Tagore

One of the major issues that Satya P. Mohanty analyses in his interview ( Frontline, April 6) is the issue of cultural chauvinism and how it has come to plague literary studies, a discipline whose ultimate goal often seems to be to produce insular national or regional literary histories, each celebrating the greatness of its own literary tradition and the linguistically defined culture that produced it. In order to battle this tendency, he calls for more comparative textual analysis across linguistic traditions and argues that a more adequate literary history will be possible once we have transcended not only the artificial opposition between high and low cultures but also the huge wall conventional literary history erects between different though related linguistic traditions. 1

One such example of the huge wall conventional literary history has erected between related linguistic traditions or rather within one linguistic tradition is that of Hindi and Urdu. What was once a shared common language of people of India stretching from Peshawar to the borders of Bengal split into two languages, Urdu and Hindi, towards the end of the 19th century. As a result, there arose two artificially separated literary cultures, each harking back to a different literary past due to the chauvinistic attitudes both of Hindus and Muslims. This cultural chauvinism was to subsequently embroil them in a practice of divisive politics, and each language became a marker of religious identity. With the passage of time the differences between the two sides became so irreconcilable that it led to the creation of a separate homeland for Muslims.

In order to understand this contested history, we need to trace the genealogies of these two terms, Urdu, the Persianised Khari Boli in Persian script, and Hindi, the Sanskritised Khari Boli in Devanagari script. The term Urdu is of fairly recent provenance. We find it being used for the first time in 1780 in the first Diwan of Mushafi, but not in the sense of a language different from Hindi. In fact, the terms Urdu and Hindi continued to be used interchangeably until the first half of the 19th century and, in some instances, even later up to the early decades of the 20th century. Initially, Urdu, or Urdu-e-Muallah, used to refer to the city of Shahjehanabad (rather than a military encampment or bazaar attached to the Mughal court as some historians suggest) and not to a language. The term used for the language practiced in the city was Zuban-e-Urdu-e-Muallah (speech of the exalted city). With the passage of time Zuban-e-Urdu-e-Muallah was first shortened to Urdu-e-Muallah and finally to Urdu. Until the third quarter of the 18th century Urdu-e-Muallah usually referred to Persian. 2

In the earlier centuries, the names of the language of the north were Hindi or Hindavi. But Hindi or Hindavi did not just refer to this newly emergent lingua franca or Khari Boli, but instead to the whole gamut of north Indian vernaculars and dialects, including Braj and Avadhi, in order to distinguish them from Persian on the one hand and Sanskrit and Prakrit on the other. This rudimentary Khari Boli was to develop as a language suitable for literary use not in the north, its birthplace 3, but in the Deccan where it had spread with the Sufis in the late 14th century in the wake of the Muslim conquest. 4 It flourished under the patronage of the rulers of the Bahmani Kingdom. This language, to be called Dakhini, soon developed a distinct style of its own with words drawn from local vernaculars and dialects and relatively free of the influence of Persian and Arabic. Quli Qutb Shah (r.1580-1611), the fifth sultan of Golkonda (the Bahmani kingdom broke up into five different states after 1518 A.D.), was a famous Dakhini poet.

It returned to Delhi in the early 18th century with Vali Dakhini, who created a fresh poetic idiom in this language with a mix of Persian, Sanskrit, Gujari and Dakhini on a Khari Boli template and started the tradition of Ghazal in what was then called Rekhata and not Urdu (the latter gained currency only in the early 19th century). This new idiom caught the imagination of the poets of the imperial city, who adopted it as a medium of poetic expression.

A parallel development took place by late 18th and early 19th centuries, which would sow the seeds of differentiation between Urdu and Hindi. In the courtly centres, first in Delhi and then in Lucknow, there developed an exclusive elite culture with a set of carefully cultivated manners and etiquette to be known as courtly tehzeeb (culture) or Mirzaness, which acted as an elitist class marker to set the nobility apart from the common people. 5 They promoted a heavily Persianised idiom, with emphasis on correctness of style and usage, which was carefully purged of all words of Dakhini, Hindavi and Sanskrit origin to distinguish it from ordinary Hindavi. Since it was largely used by the Muslim nobility attached to the courts, it came to be associated with Muslims. 6 But this was not exclusively Muslim or religious in character, since it remained inaccessible to common Muslims, and even Hindus attached to the court practised it. If, at all, there was a distinction at this point of time between two idioms/registers, it was between urban and rural, or between informal language used in the courtly circles and outside the courtly circles. It was not differentiated along religious lines, and the script was not an issue.

Colonial language policy

Through the colonial language policy, it was the British who officially differentiated between Urdu and Hindi. As is well known, they began with the idea that Hindus and Muslims were two separate races, each with its own history, culture and language. They attempted to categorise the languages of Hindustan on the basis of religion and culture. It was John Borthwick Gilchrist who first identified language with script and religion. He identified three different styles of Hindustani: first, a highly Persianised and urbane variety of Hindustani in Persian script practised in the courtly centres with a large concentration of Muslims, which he associated with Muslims; second, a rustic and rural Hindi/Hindavi largely free of the influence of Persian and Arabic words, spoken largely in the countryside with a predominantly Hindu population, which he identified with Hindus; and a third, a middle style between the two which was neither heavily Persianised nor rustic, but was close to the polite speech with an admixture of Persian and Arabic words assimilated into it. He called this middle style Hindustanee and advocated its promotion as the standard language that would cater to both the Muslim and Hindu populations. However, during his stint at the Fort William College as Professor of Hindustanee, he actively promoted two different styles as two different languages Hindustanee in Persian script, which came to be associated with Urdu and Hindavi/Hindui in Nagari script, from which all foreign (Arabic/Persian) words were purged. This differentiation and dichotomy were to prove providential and to influence and shape subsequent colonial language policies.

With the adoption of the policy of replacing Persian with vernacular languages as the language of the lower court in November 1837, the British decided to replace Persian with Hindustanee in Persian script in the North-Western Province, Bihar and Central India. Their intention was to promote the middle standard Hindustanee, as recommended by Gilchrist, comprehensible to common people, both Hindus and Muslims. But what this policy in effect resulted in was the use of a heavily Persianised Urdu, hardly distinguishable from Persian, in the name of Hindustanee, since Perso-Arabic terminology of Persian remained intact with only Hindustanee verbs substituting for Persian verbs.

The patronage of this heavily Persianised Urdu as the language of administration had a three-fold consequence. One, the majority of people who were familiar with Hindi in Nagari script were denied the opportunity of government employment. Two, the whole purpose of introducing the vernacular as the language of the lower court and administration so as to make it comprehensible to people was defeated, since a heavily Persianised Urdu continued to be used. Three, this led to a further dichotomisation between Urdu and Hindi. The official patronage of Urdu created a sense of discontent among the emerging Hindi vernacular elite and middle classes, which led to a demand for introducing Nagari script as the court character (language). This demand was to develop into a full-scale agitation in the 1860s and onwards in which supporters of both Urdu and Hindi were to be aligned on opposite sides making claims and counterclaims in support of their respective languages. The real force behind this agitation was the conflict in the social, cultural, economic and political interests of a newly emergent Hindi vernacular class on the one hand, which demanded a share in the lower administration, and of the upper-class Muslims with a handful of Hindus, mostly Kayasthas, on the other, who were loath to let go off this privilege.

New equation

Ostensibly, the controversy arose over the choice of script, but the issues involved were much deeper. As a subsequent turn of events reveals, the issue of script got intimately linked with those of language, identity and nation. In the process, the two scripts came to be associated not only with Urdu and Hindi respectively but also with the two religious communities, Muslims and Hindus. A new equation emerged: Persian/NastalikhUrduMuslim and NagariHindiHindu. This newly invented equation, which conflated script and language (linguistic identity) with religious identity, was seized and subsequently expanded by the later-day cultural/religious nationalists, Hindu as well as Muslim, into constructing separate and exclusive national identities, which would seal the fate of the nation, dividing it into a Muslim Pakistan and a largely Hindu India. The Hindu nationalists increasingly viewed themselves as the original inhabitants of the nation and thus its true inheritors. As an extension of this logic, tracing a direct line of descent from Devbhasa Sanskrit they claimed pure Hindi as their true language, which was contaminated with Arabic and Persian words due to the presence of Muslims on the Indian soil who were now seen as invaders and usurpers of the true national heritage. Hence, the need for recovering and restoring pure Hindi to its true national status by purging and excising it of all foreign influences.

Similarly, a true national literature was to be created, which would act as the sure and secure foundation of a true national culture through which the nation could attain the highest social and cultural ideals of its existence. If this was the claim preferred by the Hindu nationalists (they were mostly upper-class-upper-caste Hindi vernacular elite), then Muslims, on the other hand, contended that they were a separate race and nation with their own national language and literature. It was in this increasingly shrill climate of competing nationalisms that Hindi and Urdu got split along religious lines with Hindi in the Nagari script becoming identified with Hindus, and Urdu in Persian script identified with Muslims. They would then set about inventing two separate linguistic and literary pasts.

Attempts were made to carve out monolithic literatures in terms of singular and exclusive literary traditions. The result has been production of insular literary canons and histories with wide gaps, discontinuities and often distortions through gross simplification, which prevent a genuine understanding of our literary past. These literary histories refuse to take into account the deeply plural and heterogeneous nature of pre-colonial Indian society, marked as it was by a wide variety of linguistic, literary and cultural traditions and practices at local, regional and supra-regional/cosmopolitan levels. Together, they were constituted into multiple interacting layers of Indian culture/society; each layer, though distinct and autonomous to some extent, was shaped and influenced by other layers in this formation. The boundaries between these layers were fluid and porous, each flowing seamlessly into others. This applied to linguistic and literary practices and traditions of north India, too. Different vernaculars and dialects Khari Boli, Awadhi, Braj, Rajasthani, Bangdu, Maithili, Bhojpuri, and so on not only existed in a shared common space but also freely interacted, mixed and borrowed from one another, leading to significant overlaps. There were no sharp lines of demarcation, nor were there distinct speech communities. The same speaker could be found to be using different vernaculars/dialects or a mix of them on different occasions. During the early centuries of the second millennium when these north Indian Aryan vernaculars and dialects, including Khari Boli, were in the process of emergence from Apabhramsas and Prakrits, they came into contact with Persian and Arabic languages and developed in close interaction with them. It was not that these vernaculars were fully developed and had existence before the arrival of Perso-Arabic language with the invasion of Muslims. 7 But literary historians, especially of Hindi, in their search for pure mythic origins would have us believe that Khari Boli existed before the arrival of Muslims and it was only the subsequent addition of Perso-Arabic vocabulary to it that gave rise to Urdu. These literary enthusiasts, while tracing the beginning of Hindi literature to Amir Khusrau (1253-1325), forget to mention that he was a Muslim Sufi saint who wrote mainly in Persian with only occasional odd verse compositions such as couplets, rhymes and proverbs in Khari Boli. Even a cursory look at these occasional verse compositions which have been attributed to him clearly shows that these examples of Khari Boli were odd mixtures of Persian, Braj and Khari Boli. While literary historians of Hindi acknowledge that they carried an imprint of Braj (since they appropriated Braj into Hindi canon), they deliberately ignored the Persian influence.

Yet, what the compositions of Khusrau and others of this period reveal is that these vernaculars/dialects did not exist in pure form but were always in a fluid state, mixed in various proportions depending on the context of the speaker. This becomes clear even from the very examples these literary historians cite in support of Khari Boli, but they wilfully turn a blind eye to this phenomenon. In fact, this situation obtained not only in the early period of the development of these vernaculars but even as late as the early 19th century. While citing the example of Simhasan Battisi, which was produced at Fort William College, Alok Rai points out: The language of Simhasan Battisi reflects the glorious confusion of the common tongue of north India, drawing freely not only from the classical founts of Sanskrit and Arabic and Persian, but also from the hybrid descendants of a whole range of Prakrits and other linguistic influences. 8

Further, these literary histories of both Urdu and Hindi leave a number of gaps and discontinuities that are hard to explain when looked at in terms of single traditions. For example, not much is known about the state of Urdu in the north between its beginning in the 13th century with Khusrau and its re-efflorescence around 1700 A.D. with Vali coming to Delhi. Was it that after Khusrau Urdu ceased to be practiced in the north for almost close to four centuries? Could it be possible? But the literary historians of Urdu would have us believe that this was precisely the case, and during this period Urdu migrated from its birthplace in the north to Gujarat and Deccan, where it was practiced as Gujari and Dakhini.

Canonisation of Hindi literature

Similarly, historians of Hindi account for the absence of Khari Boli literature during this period by turning to Awadhi and Braj. Even here, the appropriation of Awadhi and Braj literary traditions is selective. Christopher King says: Part of the process of defining Hindi, then, involved affirming the earlier literary heritage of other regional dialects in the past, but rejecting literary creations in the same dialects in the present. 9 Awadhi and Braj, which had evolved as full-fledged languages and had rich and flourishing literary traditions to their credit, were now reduced to being mere local dialects and were even cannibalised by Hindi. Similar is the fate of the Kaithi script at the hands of Devanagari. 10 The process of canonisation of Hindi literature is marked by such selective appropriations and exclusions.

Such gaps and discontinuities in literary histories of both Urdu and Hindi are quite revealing. The reason for such gaps and discontinuities is obvious; it is ideological and it has to do with their refusal to engage with each others traditions. Instead of searching for evidence of Khari Boli in Devanagari script only, if the historians of Hindi literature had (as Francesca Orsini suggests) looked for evidence of Khari Boli writing in the Urdu script and not limited their search to the categories of devotional [Bhakti] and ornate [Riti] poetry (the language of which was likely to be Braj), they would have found it. 11

Similarly, if the historians of Urdu literature had taken into consideration examples of what Imre Bangha calls macaronic (mixed) Rekhata experiments in not only Persian script by Sufis but also in Nagari and even Gurumukhi and Kaithi scripts by Nirgun sant poets like Darya Das, Panjabi poets such as Miharavan Sodhi and Krishna Bhaktas such as Swami Haridas, they would have discovered a continuous tradition of Urdu, however slender. Bangha suggests that in all its various forms, Rekhata literature, though neglected by modern scholarship, is more than one of the most important meeting points between Hindi and Urdu; it is the shared early life of the two gradually separated languages. 12

Another problem that plagues these literary histories is that the script is not always the basis on which these literary canons seem to have been drawn up. The equation script-language-religion is not always at work. How else is one to explain positioning of Mulla Daud and Malik Muhammad Jayasi in the Hindi canon? Both Daud and Jayasi happen to be Muslims, they used Persian script in their literary compositions, they wrote their famous works Candayan and Padmavat respectively in the Persian Masnavi tradition, and yet they are not part of the Urdu canon but are positioned within the Hindi canon. The case of Rahim is more interesting. He was a Muslim noble in the court of Akbar who was conversant with many languages, including Persian, Turki, Sanskrit, Avadhi, Braj and Khari Boli, and was said to have learnt even Portuguese, but he chiefly wrote Braj poetry in Nagari script. While he is squarely placed in the Hindi canon of writers, what has generally been overlooked is his cosmopolitanism and multilingualism both reflected in the bold experiments that he made in his poetry by blending Persian vocabulary and motifs with those of vernaculars and Sanskrit. As Allison Busch in her analysis of Rahims poetry shows, Rahims literary talents in Hindi ranged across many dialects (Avadhi, Braj and Khari Boli), and within these, various lexical registers ranging from sanskritised to tadbhava to pure Persian are all attested. 13

Kabir is an even more interesting example since in his case there had been no script to contend with; his verses were oral and not written, and they reveal the glorious confusion of vernaculars/dialects so much so that one cannot place Kabir in any one tradition. In his verses we find use of a wide variety of registers of different vernaculars/dialects, inextricably mixed together. Even though he is a Muslim by birth, he is not accepted into the canon of Urdu writers.

The above examples attest to the fact that neither the script nor the language was a marker of ones religious identity, and alternatively, ones religious identity did not constrain ones choice of either language or script. There were borrowings, mixings and interactions taking place freely. A Rahim would use Turki at home, Persian in the court, Khari Boli/Urdu in the informal circles of the court, local vernaculars/dialects in his dealings with the common people, and Braj with a mixed register in his literary writings. This phenomenon points to the existence of a syncretic and composite culture that evolved and took shape in the north during the early modern period. But that is not to suggest that there were no contentions or conflicts, and ones religious identity did not matter. Of course, there were notions of religious identity, but they were more localised and sectarian in nature and these religious identities were often run through other kinds of local identities. There was no sense of a pan-regional or pan-Indian religious identity. This was only a later construction, and it was imposed on a past that was much more complex and variegated. In writing of literary histories in terms of separate single language and literary traditions, this sense of a complexly interwoven, variegated and nuanced cultural past is either lost or misrepresented through gross simplification.

What is required is empirical study of literary texts of this period, much in the same fashion as Satya Mohanty suggests in his interview, to understand this complex cultural phenomenon of patterns of linguistic choice and hybridity: how in different localities/sub-regions of north India one vernacular mixed with others, what kinds of mixing were taking place and in what manner, and how different idioms, styles and registers, and forms were emerging.

The case has not been very different even in more recent times after the formal separation between Hindi and Urdu. There is a lot in common that creative writers of both languages share despite the difference in scripts. Except for the use of Persian script, there is no reason why Rajinder Singh Bedi, Krishan Chandar or Balwant should be considered only as a part of Urdu canon of writers and not of Hindi. There is no reason why Ismat Chugtai, Sahadat Hasan Manto, Sajjad Zaheer, Nazir Akbarabadi, Faiz Ahmad Faiz, Akbar Allahabadi, Quruwatull in Haider and Kaifi Azmi should be read as part of Urdu canon and not of Hindi, as they use a language which is not heavily Persianised but has words which have long been assimilated into the common spoken idiom. Similarly, there is no reason why Rahi Masoom Raza and Abdul Bismillah should be considered as part of Hindi canon and not of Urdu. Going by the logic of cultural chauvinists, their writings seem to be closer to Islam/Urdu in their cultural orientation. The list is only indicative and not exhaustive. The case of Premchand is quite peculiar as he is considered as belonging to both canons, Urdu and Hindi, but on the basis of his Urdu and Hindi writings separately. So, half of Premchand belongs to the Urdu canon and the other half to the Hindi. The carving of Premchands body of works into two separate canons is symbolic of the division of the body politic of the nation. Hindi and Urdu writers belonging to the Progressive Writers Movement (most of the writers mentioned above belong to this movement), despite the similarities in their concerns and choice of themes and subject matter and their treatment, are studied separately. Modernism in Hindi literature is studied in reference to European modernism but never in reference to modernism in Urdu or other vernacular literatures. Ghazal, once a quintessential Persian/Urdu form, is now practiced in Hindi as well, but they are never studied together. The list of things that we miss out on because of our myopic and insular approach is practically endless.

Even now both the languages at the spoken level continue to be mutually intelligible and even similar, so much so that at times it is difficult to tell one from the other. At the level of popular culture, especially music and films, they still share a lot in common. Qawali continues to be extremely popular across the border, so are Bollywood films, which are appreciated and enjoyed as much by audiences in Pakistan as in India. It is said that late President Zia-ul-Haq, during whose regime there was a revival of Islamic agenda in Pakistan, used to enjoy Bollywood films secretly, in the privacy of his presidential quarters.

However, all is not yet lost and there seems to be a new ray of hope. Contemporary creative writers in both Urdu and Hindi are turning more and more towards the spoken idiom, the middle standard. This way, they avoid explicit Perso-Arabic and Sanskrit vocabulary and use words and idioms in popular use, blurring the distinction between Hindi and Urdu. I will cite here a few examples from contemporary Urdu poets 14:

Roti bankar aa jata hai, Chhand mere chappar ke upar Lal pari mere bacchon ko, Ambar thapki de jati hai

Ambar Bahraichi

(In the shape of a roti, the moon comes over my thatched roof, The red fairy visits and comforts, Ambar, my children.)

Dukanein sahar mein sari nayi hain, Hamein sab kuch purana chahiye tha

Shuza Khabar

(All the shops in the city are new, I wanted everything old.)

And

Suna hai gaon ke peepal ke paas ek patthar, Bahut dino se mera intezaar karta hai.

Khursi Akbar

(I have heard that a stone under the peepal tree in the village, Has been waiting for me for a long time.)

The significance of this moment should not be lost on us. We need to seize it to make a new beginning towards re-imagining our literary pasts in order to refashion our literary present and future. I feel that it is in this context that the call given by Satya Mohanty becomes both relevant and urgent. I will end this piece with a sher from Nazir Akbarabadi:

Kaash Sheikh au Barahman, mil kar karein kuch rok tham, Varna Bharat par koi, bhaari azaab aane ko hai.

(If only Sheikh and Brahman could come together and do something,

Or else a grave calamity is to befall Bharat.)

Sanjay Kumar is an associate professor at Banaras Hindu University. He is deputy coordinator of Inter-Cultural Studies Research Centre, a multi-disciplinary research initiative of Banras Hindu University engaged in a comparative study of South Asian vernacular and folk literary and cultural traditions as sites of articulation of alternative modernities.

End notes

1. Bhatnagar, Rashmi Dube, and Kaur Rajender. Literature to Combat Cultural Chauvinism: From Indian Literature to World Literature: Frontline, April 6, 2012: page 90.

2. For more details about the genealogy of Urdu see Faruqi, Shamsur Rahman. Early Urdu Literary Culture and History (2001). History, Faith, Politics Origin Myths of Urdu and Hindi. pages 21-42.

3. Most historians claim that it got eclipsed in the north between the 14th and 17th centuries, a claim which I examine later in this paper.

4. It had also spread to Gujarat, again with the Sufis, in the late 13th and early 14th centuries after the annexation of Gujarat by Alauddin Khalji (r.1296-1316) and was practised as Gujari.

5. Premchands short story Shatranj ke Khiladi gives a picture of this decadent culture.

6. While describing it as the root cause of the subsequent separation between Hindi and Urdu, Amrit Rai and Tariq Rahman blame the Mughal nobility, which deliberately promoted this culture to compensate for its declining political fortunes. Rai, Amrit. A House Divided: The Origin and Development of Hindi Urdu (1984); Rahman, Tariq. From Hindi to Urdu: A Social and Political History (2011).

7. These Muslims came from different parts and cultures and in different waves. One cannot treat them as one homogeneous or monolithic group. Faruqi (2001).

8. Rai, Alok. Making a Difference: Hindi, 1880-1930. Annual of Urdu Studies 10 (1995): page 138.

9. King, Christopher. Forging a New Linguistic Identity: The Hindi Movement in Banaras,1868-1914 in Freitag, Sandria B. (Ed.). Culture and Power in Banaras: Community, Performance and Environment, 1800-1980. page 192.

10. Kaithi was more popular than Devanagari, but since it was used both by common Hindus as well as Muslims, this was not acceptable to Hindu nationalists.

11. Orsini, Francesca. (Ed.). Before the Divide: Hindi and Urdu Literary Culture (2010)

12. Bangha, Imre. Rekhta: Poetry in Mixed Language, in Orsini, page 83.

13. Busch, Allison. Riti and Register: Lexical Variation in Courtly Braj Bhasa Texts, in Orsini, page 89

14. I am grateful to Professor Aftab Ahmad, my colleague in the Urdu Department of Banaras Hindu University, for drawing my attention to these examples.

Literature & cultural chauvinism

"Literature to combat cultural chauvinism" said the headline of the interview with Satya P. Mohanty, Professor of English, Cornell University, that Frontline published in its April 6, 2012, issue. Scholars and writers in English, among others, were quick to take notice, and soon the interview took on, in Mohanty's words, "a life of its own outside India".

Importantly, it caught the attention of the Kenyan writer and scholar Mukoma Wa Ngugi, who decided to organise a "Global South Cultural Dialogue" and invite scholars from around the world to talk about their own cultural contexts and present their views on the dialogue. He got responses from writers in India, China, Trinidad, Kenya, Morocco, and other parts of the global South.

Frontline took the opportunity to carry on the debate that Mohanty initiated by publishing, with Mukoma's permission, the response from the renowned Kenyan writer Ngugi wa Thiong'o, who has taught in universities in India and is an important figure in the global South. "Asia in my life", as his essay was titled, appeared in the June 1 issue of Frontline. It talks, among other things, about his Indian and more generally Asian connection.

The essay featured here is part of the first forum organised by the Global South Cultural Dialogue project scheduled to be published in late August in The Journal of Contemporary Thought. It takes forward the theme of literature and cultural chauvinism. The author, Sanjay Kumar of the Department of English, Banaras Hindu University, Varanasi, breaks down for readers the wall conventional literary history erected between Hindi and Urdu, languages that have their origin in one linguistic tradition.

You have exhausted your free article limit.
Get a free trial and read Frontline FREE for 15 days
Signup and read this article for FREE

More stories from this issue

Get unlimited access to premium articles, issues, and all-time archives