Ietf-languages Digest, Vol 81, Issue 40

Frank Bennett biercenator at
Sat Sep 26 15:04:50 CEST 2009

Mark Crispin's information on the system used in the Jorden textbooks,
on informal practices, and on his own preferred method of romanizing
Japanese text well illustrates the variety of practice that I tried to
convey in the descriptions accompanying the three filings.  I am
impressed by the thoroughness of the review process, and grateful for
the time taken over them.

There is just one point to which I would take exception.

> Date: Fri, 25 Sep 2009 22:18:00 -0700 (PDT)
> From: Mark Crispin <mrc+ietf at Panda.COM>
> Subject: Re: Status of Japanese requests
> To: John Cowan <cowan at>
> Cc: ietf-languages at, Doug Ewell <doug at>
> Message-ID: <alpine.OSX.2.00.0909252133140.93829 at>
> Content-Type: text/plain; charset="utf-8"


> Anyway, I'm not advocating any specific course of action other than to
> leave some expansion and not assume that a three-way Hepburn/Nihon/kunrei
> addresses Japanese romanization as it is actually used.

With respect, I don't have a need to tag Japanese romanization in the
wild, "as it is actually used".  I'm not even sure it would be
feasible (let alone useful) to embrace the full-blown variety of
real-world possibilities with a tagging scheme like RFC 4646.  I tried
to make that clear in the descriptions that I appended to the
three-part filing, but in case it's not clear, I'm needing these tags,
and the ja-Latn-hepburn-heploc tag in particular, for a more
restricted purpose.

This filing originated with a desire to add multi-lingual
functionality to Zotero, an open-source reference manager product that
has attracted a significant and increasingly international following.
The idea is that Japanese resources (in this case) will be stored in
the original script, with translations and various forms of
transliteration attached to a primary title or author name field as
child objects.  A multi-lingual formatting engine will then be able
select the appropriate forms from the database when composing cites in
a particular style.

In practice, the forms of transliteration registered in the database
will be dictated by the target publication or publisher to which a
given manuscript, containing a Zotero-generated bibliography, is to be
submitted.  The romanization method is effectively a standard shared
between the author and the publisher; and the reference manager needs
to be able to comply with that standard.  In about 99.9% of imaginable
cases, the romanization method of choice will be "Hepburn".  If the
publisher has their act together, they will refer the author to some
reasonably comprehensive, reasonably well circulated document that
defines the standard.  In the case of ja-Latn-hepburn-heploc, I have
referenced the ALA-LOC romanization tables, which are freely available
and provide a clear set of rules.

In the case of kunrei (which, yes, simply means an administrative
order), I referred to ISO 3602, as a similarly unambiguous statement
of the rules for this particular type of romanization.  That said, the
call for it in the real world of international publication (with which
I am most immediately concerned) is nearly nil.  I made the filing
because (a) kunrei romanization is taught in Japanese schools, (b)
there is a small but vocal population of persons in Japan who advocate
kunrei as somehow more appropriate or more Japanese than Hepburn, and
(c) a question was raised over the initial proposal to file Hepburn
only, without distinguishing it from something else -- kunrei is the
something else.

Nihon-shiki (or nihon-siki, or if you prefer then nihonshiki, or even
nihonsiki) is even less likely to be used in overseas publication than
kunrei.  If it is not specified in ISO 3602 (and I apologize for not
popping for the fifty bucks needed to discover that fact), then as far
as I know at present, the only other sources that define it are
written in Japanese.

It will be no skin off my nose if the nihon tag is dropped at this
point.  I do not have a sufficient incentive to argue the case for it,
or to track down the resources necessary to support the filing.

I'm not sure what the standing of kunrei will be in light of the
latest round of discussion; but if it is dropped, this too will not be
a show-stopper for the application that I am attempting to support.

Hepburn, however, _is_ needed, and for the reasons stated above, I
hope that it will be deemed suitable to put it forward.

Begging your indulgence and goodwill,

Frank Bennett

More information about the Ietf-languages mailing list