Film: Junoon (1978)
Singer: Rafi
Music: Vanraj Bhatia
Lyrics: Jigar Moradabadi
*ing: Shashi Kapoor, Nafisa Ali, Shabana, Naseeruddin,
Jennifer, Pearl Padamsee, Deepti Naval, Benjamin Gilani
From the opening qawwali, the film is touched with a dark hand of
overwhelming passion and obsession, a masterful touch by master
director Shyam Benegal.
The same mood pervades this song, too. The brooding, charged air is
conveyed by the abrupt phrases sounding stridently from an organ and
from the short phrases into which the lyrics are broken. The ghazal
selected is also one with very short lines which lends itself to the
melodic phrasing selected. Rafi sings at a higher octave, which also
contributes to the tension in the song. All said and done, this is a
love song, but a love which is on the edge of a sword, a tale written
in blood and tested by fire.
dil hai kisii kaa raaz\-e\-haqiikat
raaz\-e\-haqiikat kyaa kahiye
hairat\-e\-jalavaa mohar\-b\-lab hai
jalavaa\-e\-hairat kyaa kahiye
us pe kisii kii tiir\-e\-sitam sii
mashq\-e-\siyaasat kyaa kahiye
Handpicking a set of songs already written to match the mood and
context of a storyline is no mean task, and the Shyam Benagal-Shashi
Kapoor-Vanraj Bhatia team does a wonderful task of it here.
The "lyricists" for the film read make for an impressive list - Kabir,
Khusro, Jigar and Shakespeare. The last in a sweet song "Come live with
me and be my love". The one relative unknown who appears in the
lyricists credits is Yogesh Praveen who wrote saavan kii aayii bahaar
re, sung by Asha Bhosale in a voice more reminiscent of a Vanraj Bhatia
favourite - Preeti Sagar.
Vijay
Sent via Deja.com http://www.deja.com/
Before you buy.
highly digressive from the original post - but i just wanted to ask
something. just for the sake of it. shyam benegal and vanraj bhatia have
been working together since so long.......junoon, manthan, discovery of
india, and trikaal, sardari begum etc....how would vanraj bhatia have
felt when benegal used ar rehman for zubeida. forget commercial value
and the rehman brand - i guess benegal and bhatia are good friends,
having worked for so long. taal to ar rehman - lp? actually govind
nihalani wanted rehman to do drohkaals background score, and only when
he refused did he turn to vanraj bhatia. btw, are vanraj bhatia and
tushar bhatia related....err tushar bhatia of andaz apna apna (and tv
serials) fame (op nayar style tonga music).
A small correction. "Come live with me and be my love"
has been attributed to Shakespeare erroneously. The
poem is in fact by Christopher Marlowe. The first
stanza goes something like :
Come live with me and be my love,
And we will all the pleasures prove
That valleys, groves, hills, and fields,
Woods, or sleepy mountain yields.
Also, the third line of the Jigar ghazal should
preferably be transcribed as :
Sunne wala koi nahin hai (rather than "sunane")
Afzal
I took the Geet Kosh as an authority on this. I don't claim great
familiarity with either Shakespeare or Marlowe's works but I did think
that this must be from one of Shakespeare's other poems and not from
one of his plays.
>
> Also, the third line of the Jigar ghazal should
> preferably be transcribed as :
>
> Sunne wala koi nahin hai (rather than "sunane")
>
I follow ITRANS notation for practically all my transcriptions.
Strictly speaking, it IS sunane vaalaa, because every character in
Hindi has an implicit "a" at its end. Using "sunne" implies a halant
after the first n which is not correct. ITRANS uses this same
principle, only it makes the use of "a" mandatory. Thus, sunne would be
interpreted as n with halant in the first case. use of _ is an
acceptable alternatice sun_ne vaalaa koii nahiin hai. ITRANS treats a
space at the end of a word to be an implicit "a" and a _ anywhere in
the middle of the word as the same. If you see the ISB, you will find
enough examples of all the above. In any case, sunne is wrong, although
I understand perfectly what you are trying to say.
Vijay
There was certainly no error on your part.
I suppose the movie's audio cassette or the
credit titles may also convey the same
(erroneous) information that the Geet Kosh did.
I was merely trying to set the record straight.
Besides, the reference to this song reminded me
of this fine poem which I had first read nearly
half a century back, hence the response.
>
> >
> > Also, the third line of the Jigar ghazal should
> > preferably be transcribed as :
> >
> > Sunne wala koi nahin hai (rather than "sunane")
> >
> I follow ITRANS notation for practically all my transcriptions.
> Strictly speaking, it IS sunane vaalaa, because every character in
> Hindi has an implicit "a" at its end. Using "sunne" implies a halant
> after the first n which is not correct. ITRANS uses this same
> principle, only it makes the use of "a" mandatory. Thus, sunne would
be
> interpreted as n with halant in the first case. use of _ is an
> acceptable alternatice sun_ne vaalaa koii nahiin hai. ITRANS treats a
> space at the end of a word to be an implicit "a" and a _ anywhere in
> the middle of the word as the same. If you see the ISB, you will find
> enough examples of all the above. In any case, sunne is wrong,
> although
> I understand perfectly what you are trying to say.
I understand that you always use the ITRANS
notation for lyrics. Unfortunately, I do not
quite follow it. Besides, I am not too sure
that ITRANS is suitable for transcribing lyrics
or poems/ghazals which are in Urdu. This
language loses quite a bit of its charm if the
pronunciation does not conform to phonetic
correctness. Also, any reference to "halant" and
other grammatical/phonetic terms pertaining to
Sanskrit or Hindi is not exactly germane to
Urdu pronunciation principles. "Sunne" may be
incorrect as per the ITRANS or Hindi pronunciation
pattern, but I cannot accept that it is wrong
as per the Urdu pronunciation rules. In fact,
any Urdu knowing person would immediately under-
stand the word. Also, I had used the word
"preferably" in my response, because a casual
reader can confuse "sunane" with "sunaane".
In another NG, there was a post today involving
the word "samajhe". Any ordinary reader would
perhaps (mis)pronounce the "meem" with a "zabar",
which would be absolutely wrong. In my view, it
is better or "preferable" to transcribe this
word as "samjhe", (with "meem" carrying a "jazm")
which doesn't leave any room for confusion.
I am sure you will appreciate that it is not my
intention to rake up the Urdu/Hindi question.
I only wanted to express my views about this.
Afzal
>
> Vijay
> I understand that you always use the ITRANS
> notation for lyrics. Unfortunately, I do not
> quite follow it.
That is fine....
> Besides, I am not too sure
> that ITRANS is suitable for transcribing lyrics
> or poems/ghazals which are in Urdu.
....but this statement needs some justification. I know you
have provided some pointers (such as a halaa.nt being
irrelevant to Urdu) but IMO they are not quite enough to
make the above statement. Unfortunately, I do not know
Urdu. For all I know, I ought to be nodding my head in
agreement. So, for the benefit of us ignorami, could you
elaborate a little?
> This
> language loses quite a bit of its charm if the
> pronunciation does not conform to phonetic
> correctness.
I am sure that statement is true of all languages. What
does that have to do with ITRANS?
> "Sunne" may be
> incorrect as per the ITRANS or Hindi pronunciation
> pattern, but I cannot accept that it is wrong
> as per the Urdu pronunciation rules.
First of all, you have to accept that transliterating
Urdu or Hindi or Sanskrit into English has severe
limitations. I say that ITRANS defines a set of
rules that reduce this limitation for Hindi/Sanskrit,
sometimes at the expense of readability. You say
it is not adequate for Urdu. Fine. But without
defining a set of rules - a transliteration scheme -
for Urdu, how can you pronounce (no pun
intended) judgment on the in/correctness of a
particular transliteration attempt?
I understand and appreciate the readibility issue,
but that concern is a secondary one as far as
transliteration schemes are concerned. There
can be alternatives such as "sun_ne" which are
correct *and* readable. But we must never forget
that correctness and *unambiguity* are the
primary goals of the transliteration scheme.
C
Chetan,
I was merely emphasising the need for phonetic
correctness insofar as Urdu lyrics and ghazals
etc. are concerned. I understand that ITRANS
is perhaps the only transliteration scheme
devised and available at present which can
convey the broad sense of Sanskrit and Hindi
script to Roman script readers. I would go
even further than that. If the lyrics are
in "plain Urdu", or "Hindustani" if you please,
these too can perhaps be conveyed properly
to the readers. But it is difficult to
transcribe pure or chaste Urdu as per the
ITRANS format. I gave a few examples, like
"sunne" and "samjhe" (rather than "samajhe").
There can ne a hundred others. Some two years
back, I would interject a few comments about
phonetic correctness when some Urdu words were
mis-transcribed in RMIM postings. But then I
gave it up. It is for experts in phonetics to
devise another format that would cater to the
requirements of Urdu/Persian/Arabic terms.
I am merely a "practitioner" of the language
and not qualified to undertake this job.
But till that happens, we can use ITRANS and
think about ways to bring about changes or
improvements that can take care of word-
situations like "sunne" and "samjhe". If
"sun_ne" is acceptable, it is certainly an
improvement on "sunane". If we write a simple
sentence like "Woh kuchh banna chahta hai",
it would not do, IMHO, to write it as "Woh
kuchh banana chahata hai". In Hindi or Sans-
krit, we can talk about "halant". In Urdu/
Persian/Arabic, there is "tashdeed" which
defines this particular rule of pronuncia-
tion---a letter or sound which is pronounced
twice. There are other rules of pronunciation,
peculiar to these languages which are ignored
some times by people who are not familiar with
them. If you are a cricket follower, you
must have heard commentators talking about
Abdu"l" Razzaq, whereas it should be pronounced
as "Abdur" Razzaq. (This has to do with the
distinction between "Huroof-e-shamsi" and
"Huroof-e-qamari"). I suppose one could write
a small book on these rules, but that is
neither here nor there. Our only concern at
the moment is to ensure that an Urdu or
Persian word is transcribed in such a way that
it conforms to the actual/correct pronunciation.
And, for this purpose, I would even "amend" the
ITRANS principles. Ironically, words like
"sunne" and "samjhe" are not typically difficult
Persianised Urdu words ! They are simple words
used by almost 100 % of people speaking Urdu or
Hindi. I pointed out the difference only because
the "song" in question is an Urdu ghazal by one
of the most eminent Urdu poets of the 20th
century. And if you pronounce this word as
"sunane" (with the slight accentuation of the "a"
sound), it would perhaps alter the metre and, in
any case, it would not "sound" like Urdu.
Now for some other points in your response :
I made a specific response to Urdu (as a
language losing its charm) because of two
reasons : (a) it is the language under
discussion in this thread. There was no need
to make a generalization. (b) Urdu has a
unique charm of its own, which is something
not shared by many other languages. Persian
too has the same sort of peculiar charm. I
would not say the same about Arabic. Amongst
European languages, one can perhaps include
French in this category, but not German, in
my view.
Secondly, I was merely expressing my views,
not "pronouncing a judgment", even if the
pun was intended ! Even you seem to be in
agreement, as the following excerpt will
indicate :
I say that ITRANS defines a set of rules
that reduce this limitation for Hindi/
Sanskrit, sometimes at the expense of
readability. You say it is not adequate
for Urdu. Fine.
I have already spoken about my inability (and
lack of qualification) for devising an alter-
native sheme of transliteration.
In conclusion, I can do no better than quote
the last sentence in your response, viz.
"We must never forget that correctness and
'unambiguity' are the primary goals of the
transliteration scheme." (I would perhaps
replace the last "the" with "any").
"Sunne" and "sunane" are two different "sounds"
or pronunciations for the same word. And there
you have the "ambiguity". The second "sound"
or pronunciation negates "correctness".
With all good wishes,
Afzal
P.S. It might interest you to know that Iranis
have a very distinctive accent. Their
pronunciation of Perisan or Farsi is
rather different from the way Indians/
Pakistanis speak this language.
For those who have the time and inclination, here is a little tale...
When Shiva's wife was killed, he went into an uncontrollable rage and
danced the taaNdav nR^itya. He also played his Damaruu as he danced.
From the Damaruu, emerged two sets of syllables. The first set started
with "a" and ended with "ch", the second set started with "h" and ended
with "l". It so happens that all the vowels (of the Sanskrit script)
are contained in the first set and all the consonants in the second. The
two sets are called "ach" and "hal", respectively.
In Sanskrit (more so than in Hindi), every syllable ends with a vowel,
and the default vowel is "a". Thus, Vijay is strictly to be pronounced
as "vi-ja-ya". Syllables which end with a consonant are to be explicitly
marked as such, with a diagonal line slanting towards the right at the
bottom of the character. In the ITRANS scheme, this is denoted by a .h
This special diacritical mark is called a halant - a syllable that ends
with a "hal" syllable, that is, a syllable ending in a consonant. It
does not have any thing to do with the instance of a syllable/ letter/
character repeating itself in a word.
When a consonant appears in the middle of a word without an explicit or
implicit vowel transformer, one uses the "half" character (for example
with k in vaakya, t in tatva, sh in ashva and so on...) There are a few
characters for which there is no "half" character like T and D. When
such a character appears without a conjoint vowel in the middle of the
word, a halant is used ("khaD.hg", where you don't want to say "khaDag")
The assumption that Afzal makes is a special case of this (as in
"guD.hDaa" or "laT.hTuu"). Again, ITRANS seems to be handling this
intelligently without the use of an explicit .h, so laTTuu and laT.hTuu
are both right.
When a half-consonant occurs at the end of a word, a halant is always
used, because of the implicit "a" at the end. Thus, it is "dhik.h" and
"dhat.h" (I cannot think of non-trivial examples of this case at the
moment)
So much for that. Interesting discussion, probably, but RMIM may not be
the place for this.
> Afzal
Vijay
> I was merely emphasising the need for phonetic
> correctness insofar as Urdu lyrics and ghazals
> etc. are concerned. I understand that ITRANS
> is perhaps the only transliteration scheme
> devised and available at present which can
> convey the broad sense of Sanskrit and Hindi
> script to Roman script readers. I would go
> even further than that. If the lyrics are
> in "plain Urdu", or "Hindustani" if you please,
> these too can perhaps be conveyed properly
> to the readers. But it is difficult to
> transcribe pure or chaste Urdu as per the
> ITRANS format. I gave a few examples, like
> "sunne" and "samjhe" (rather than "samajhe").
There is absolutely no difficulty in transcribing these two words into
iTrans. From the point of view of iTrans, "sun_ne" and "sam_jhe" are
correct. That is to say, any iTrans rendering engine will render them into
correctly spelled Hindi words in Nagari script.
True, from the point of view of iTrans, "sunane" and "samajhe" are also
correct, and equivalent to "sun_ne" and "sam_jhe" respectively. So what?
Also, the two words are used in Urdu too. Again, so what?
It seems to me that what you're complaining about, Afzal, is not the
limitations of iTrans. What you're complaining about is the fact that some
iTranscribers use "sunane" when from *your* point of view, the equally
correct "sun_ne" is preferable.
That is to say, the point of view of somebody unfamiliar with iTrans, who
might mistake "sunane" for "sunaane," or who might fully pronounce the
full "na" instead of correctly truncating it. This is not iTrans's
problem. The problem is a lack of familiarity with iTrans, and perhaps a
lack of understanding of what iTrans is actually supposed to do.
iTrans simply provides ASCII equivalents to elements of the Hindi script.
In the Hindi script, the word "sun_ne" is written "su + na + ne," and so
"sunane" is a perfectly correct transcription. In fact, it is The correct
transcription, because it perfectly mirrors what is actually written in
Hindi. Being able to write "sun_ne" instead of "sunane" is just a nice
bonus.
> we can use ITRANS and
> think about ways to bring about changes or
> improvements that can take care of word-
> situations like "sunne" and "samjhe". If
> "sun_ne" is acceptable, it is certainly an
> improvement on "sunane".
And it is already in iTrans. In fact, it is there precisely because
"sunane" might confuse some folks and "sun_ne" indicates the pronunciation
better. So what more improvement do you think is needed?
> If we write a simple
> sentence like "Woh kuchh banna chahta hai",
> it would not do, IMHO, to write it as "Woh
> kuchh banana chahata hai".
"banana" is wrong from any standpoint, except if the guy you're talking
about was in fact desirous of obtaining some fruit. In iTrans the correct
transcription would be "bananaa." Or "ban_naa," to alleviate the
discomfort of those who cringe at "bananaa."
It's worth pointing out that this discomfort is transitory. As you get
habituated to iTrans, it becomes more natural. Any sort of scheme like
iTrans is not going to feel perfectly natural as soon as one first starts
learning it. But there is no reason to project this discomfort onto iTrans
itself, and claim that iTrans is inherently flawed.
Nobody says iTrans is perfect. Nobody even says it is natural. It is,
however, remarkably flexible, accurate, and reliable. If you believe that
Urdu is not well served by iTrans, Afzal-bhaai, you may be right. But
examples like "sun_ne," "ban_naa," etc. won't make your case. Come up with
better evidence, if you can.
Until you get to know iTrans better, however, you might want to reserve
judgment. And you might find that your real argument is not with iTrans,
but with the Nagari script as a way of rendering Urdu.
-s
>-s
That is a red herring. Both "sunanaa" and "samajhanaa" are native
words; to my knowledge, neither comes from Persian or Arabic. So,
the problem is with the Nagari script as a way of rendering Hindi.
I doubt that Arabic script does any better job of rendering Urdu.
I do know that Roman script is totally inadequate as a way of
rendering English. Any script would be inadequate to the task
of faithfully rendering all the nuances of speech.
This whole issue is pointless distraction, stemming from ignorance
regarding the role of transcription software and perhpas of script
itself.
Ashok