style="font-size:15px;"> sit
fee
put
boo
vocalic r, American pur-
dy or English pretty
lengthened r
vocalic l, able
made, esp. in Welsh pro-
nunciation
bite
rope, esp. Welsh pronun-
ciation; Italian solo
sound
anusvara nasalizes the pre-
ceding vowel
visarga, a voiceless aspira-
tion (resembling the En-
glish h), or like Scottish
loch, or an aspiration with
a faint echoing of the last
element of the preceding
vowel so that taih is pro-
nounced taihi
luck
blockhead
go
bighead
anger
chill
matchhead
jog
aspirated j, hedgehog
canyon
retroflex t, try (with the
tip of tongue turned up
to touch the hard palate)
same as the preceding but
aspirated
retroflex d (with the tip
vii
CSL CONVENTIONS
of tongue turned up to
touch the hard palate)
same as the preceding but
aspirated
retroflex n (with the tip
of tongue turned up to
touch the hard palate)
French tout
tent hook
dinner
guildhall
now
pill
upheaval
before
abhorrent
mind
yes
trilled, resembling the Ita-
lian pronunciation of r
linger
word
shore
retroflex sh (with the tip
of the tongue turned up
to touch the hard palate)
hiss
hood
CSL Punctuation of English
The acute accent on Sanskrit words when they occur outside of the Sanskrit text itself, marks stress, e.g., Ramayana. It is not part of traditional Sanskrit orthography, transliteration, or transcription, but we supply it here to guide readers in the pronunciation of these unfamiliar words. Since no Sanskrit word is accented on the last syllable it is not necessary to accent disyllables, e.g., Rama.
The second CSL innovation designed to assist the reader in the pronunciation of lengthy unfamiliar words is to insert an unobtrusive middle dot between semantic word breaks in compound names (provided the word break does not fall on a vowel resulting from the fusion of two vowels), e.g., Maha·bharata, but Ramayana (not Rama·ayana). Our dot echoes the punctuating middle dot (·) found in the oldest surviving samples of written Indic, the Ashokan inscriptions of the third century bce.
The deep layering of Sanskrit narrative has also dictated that we use quotation marks only to announce the beginning and end of every direct speech, and not at the beginning of every paragraph.
viii
CSL CONVENTIONS
CSL Punctuation of Sanskrit
The Sanskrit text is also punctuated, in accordance with the punctuation of the English translation. In mid-verse, the punctuation will not alter the sandhi or the scansion. Proper names are capitalized. Most Sanskrit meters have four “feet” (pada); where possible we print the common sloka meter on two lines. In the Sanskrit text, we use French Guillemets (e.g., «kva samcicirsuh?») instead of English quotation marks (e.g., “Where are you off to?”) to avoid confusion with the apostrophes used for vowel elision in sandhi.
sandhi
Sanskrit presents the learner with a challenge: sandhi (euphonic combination). Sandhi means that when two words are joined in connected speech or writing (which in Sanskrit reflects speech), the last letter (or even letters) of the first word often changes; compare the way we pronounce “the” in “the beginning” and “the end.”
In Sanskrit the first letter of the second word may also change; and if both the last letter of the first word and the first letter of the second are vowels, they may fuse. This has a parallel in English: a nasal consonant is inserted between two vowels that would otherwise coalesce: “a pear” and “an apple.” Sanskrit vowel fusion may produce ambiguity.
The charts on the following pages give the full sandhi system.
Fortunately it is not necessary to know these changes in order to start reading Sanskrit. All that is important to know is the form of the second word without sandhi (pre-sandhi), so that it can be recognized or looked up in a dictionary. Therefore we are printing Sanskrit with a system of punctuation that will
