avoid errors and imposition. People who are used to pronouncing
it as ‘Ashoka’ are free to continue doing so.
Even the important dental–retroflex distinction (which can only
be indicated by diacriticals or awkward-looking capitalising) has
both regional and social variations. For example, in modern
Marathi, the ‘pani-loni’ words won a certain fame in the context of
the non-Brahman movement: Brahmans pronounce a retroflex ‘n’;
non-Brahmans a dental ‘n’ (though today some overcompensate
and use the retroflex too extensively, where the ‘correct’ version is
dental) Beyond this there are complex north–south gradations and
Dravidian elements in the use of retroflex sounds. In this case any
imposition of a ‘correct’ pronunciation—which the use of diacriti-
cals would force us to choose—would be wrong; the choice itself
should be avoided as going counter to the linguistic insight that
there is no correct versions of a language; we only have to do jus-
tice to the regularities which govern how people speak, regularities
which themselves vary. Leaving the dental–retroflex distinction
ambivalent for the whole series of ‘t’, ‘d’, ‘n’ may thus paradoxi-
cally be more accurate for a book that aspires to be read through-
out India.
I have used the Pali forms for most words. For many place names
they are generally the more accurate even today. Two important
examples are ‘Paithan’ and ‘Taxila’. Even well-known archaeolo-
gists such as Bridget and Raymond Allchin use the form ‘Pratisthana’
(perhaps a compromise with the Sanskrit ‘Pratishthana’) for the
ancient city that was Patitthana in a Sutta Nipata story that records
the route travelled by the followers of the Brahman Bavari to meet
the Buddha. Over a period of nearly 2500 years, the Pali form
remains more accurate than the Sanskrit; it is unlikely that the city
was ever known as ‘Pratisthana’ except to the small Brahman
minority of its population. Similarly, the Pali ‘Takkasila’ is much
closer to the modern ‘Taxila’ and it is doubtful if very many of its
inhabitants ever called it ‘Takshashila’. For many other terms, Pali
is the closest to the spoken languages of the period and areas which
are covered in the book.
Thus I have used ‘samana’ rather than ‘shramana’ and ‘Gotama’
rather than ‘Gautama’. While discussing Mahayana, however, I
have followed Sanskrit spelling, though again it should be remem-
bered that many existing Mahayana texts are usually Sanskrit (or
Chinese/Tibetan translations of Sanskrit) renderings of texts thatThe publication of this book is thus a culmination of over three
years of research and writing, and a much longer period of discussion,
debate and research on the issues involved. In some ways it is being
published too soon and I am aware of the incomplete, even tenta-
tive nature of the project. Just as Buddhist stupas, carvings and
caves of all kinds are continually being unearthed in various parts
of India, old sources on the themes of this book are constantly
coming to my attention and new studies and interpretations are
constantly being published. Wherever you dig in the soil of India,
Buddhist remains are likely to be found; wherever you look in the
fields of scholarship and intellectual life, something new and arrest-
ing is being said. However, given the growing demand for literature
on Buddhism and the Buddhist aspects of Indian history, I have felt
the need to stop, to publish and throw the work open for debate.
This volume is designed both as an academic and a popular book.
The aim is to communicate with a wide variety of readers of English
in India, and not just with those in the universities. Therefore, dia-
criticals are not used in the text. There are two major reasons
for this.
First, for the Indian intellectual readership, which is much
wider and in some ways wiser than the academic readership, nor-
mally seen as the target of scholarly works on Indian culture and
history, diacriticals are an obstacle. At the same time, they are in
a way unnecessary; in most cases people generally know (or have
a fairly good idea of) the pronunciation and meaning of terms
used.
There is, however, a more important intellectual and methodo-
logical reason for avoiding diacriticals. The use of diacriticals very
often leads to an over-precise and sometimes incorrect identifica-
tion of the pronunciation of a word. This is related to the tendency
to Sanskritise terminology, but it goes beyond this. Take the exam-
ple of ‘Asoka’. The name of India’s great Buddhist emperor is com-
monly pronounced ‘Ashoka’ and written by contemporary
historians and scholars with a diacritical mark over the ‘s’. Yet the
languages that were used throughout most of India at the time
would have pronounced it as ‘Asoka’; ‘Ashoka’ would at best have
been a regional (eastern) variant. The use of the Sanskritised ver-
sion borrowed from the Puranic king lists, which tried to avoid all
recognition of this greatest of Indian rulers, is both an anachronism
and an insult. The avoidance of diacriticals in this way helps to
x Buddhism in India Preface xi