Romanization when?

[Dominik Wujastyk]

Richard Hayes said,
 
\begin{quotation}
 
 > I would like to pose a question to Dominik or to
 > others who now routinely print in devanagarii.
 
Actually, I am an unreconstructed tranliterationist!  For most
purposes, anyway, such as articles and papers that I read
at conferences.  For anything text-length, though (about over
a page), I do use Velthuis's Devanagari.  My original musings
were prompted by discussions with colleagues in other language
fields where the practices and sensibilities about script
seem quite different.  It made me think seriously of changing
my own practice.  And I think I probably will move towards
using Devanagari more commonly.
 
 > If one is writing
 > in English and quotes a single "sloka or one sentence or just a phrase,
 > should one romanize or devanagarize (pardon my mlecchisms)? What if one
 > is citing a single word? In other words, is there ever an occasion for
 > using romanized Sanskrit? I am interested in hearing what habits others
 > have developed in this area.
 
Yes.  It's rather a "how long is a piece of string" situation, isn't it.
If we look at publications from India, practice seems to vary
very widely, almost crazily.  I have no answer.  I think I would
consider a sloka, or even a pada, worth casting into Devanagari.
 
\end{quotation}
 
What I think we would all agree on is that tranliteration badly
done -- with wrong or missing diacritics -- is the worst of all
worlds.  Perhaps a general drift towards using Devanagari in our
profession would separate the sheep from the goats a bit.  Or is this
po-faced elitist nonsense?
 
Dominik