# The 1999 Ling-TeX Archives

### 18th January 1999: 1999 info-message on Ling-TeX

From: Christina Thiele (cthiele@ccs.carleton.ca)
Subject: 1999 info-message on ling-tex

To *subscribe* (or un-subscribe, for that matter) to this list, send

ling-tex-request@ifi.uio.no

and once you've received our canned welcoming message, then you can
post to the real thing:

ling-tex@ifi.uio.no

Notice also that there's a web site, set up by Dag Langmyhr, the
fellow who runs/hosts the ling-tex list itself:

www.ifi.uio.no/~dag/ling-tex.html

One really nice feature is the bundling of past ling-tex e-mail by
year -- makes it really really easy to look at all the topics, and the
messages related to it (done via the Subject: line, I suspect).

As well, there's an inventory of style files and such that many
members of this list have put together over the years and offered up
to the TeX-using linguistics community.

You can ftp to your nearest CTAN site (go to www.tug.org for a listing
of sites, and links to them) and in the /info subdirectory, you'll
find the following file:

ling-mac.tex

Not totally up to date but the information does cover many of the
linguistics-related macros and style files that are out there. Yes
... style files'. I'm not sure which, if any, have undergone
conversion to LaTeX2e format -- anyone know of such upgrades? [Hint --
that'd be a really good reason for me to get off my duff and actually
post an updated ling-mac.tex file ;-))) ]

The ling-mac.tex file is the same version which appeared in TUGboat
16,1 (1995), pp.42-44.

The web page has a link directly to the file (via ftp.dante.de), so
maybe it's time to put it into your bookmarks.

Christina (whose machine is now 2e-compliant,
even if my macros aren't ;-) )

### 18th January 1999: Interlinearization in LaTeX

From: Ashild Naess (ashild.nass@ilf.uio.no)
Subject: Interlinearization in LaTeX

I'm working on a collection of texts in the little-described Polynesian
language Pileni. The texts will be transcribed from my field tape
recordings and given with interlinear morpheme-by-morpheme translations as
well as translation into plain English. Since interlinearization in Word is
a messy and unrewarding job, to say the least, I have started looking at
LaTeX as a possible alternative.

I am absolutely new to TeX/LaTeX and will have to start learning it from
scratch if I am to use it. My question is: Has anyone out there ever used
TeX/LaTex for a similar project, and can they recommend it? Any tips as to
how it's best done? I am aware of the covington.sty package which supports
interlinearization in examples, but has anyone used it, or something else,
with longer texts? The aim here is to produce a book with an introduction
about the language and its speakers, about 20 texts of varying lengths
(some fairly short, some up to about 45 minutes on tape) and probably some
comments about the different kinds of stories and recurring themes found in
the material.

So, will LaTeX facilitate this work enough to make it worthwhile taking the
time to learn it? And if so, how can it best be done?

Ashild Naess, University of Oslo

### 18th January 1999: Re: Interlinearization in LaTeX

From: Christina Thiele (cthiele@CCS.CARLETON.CA)
Subject: Re: Interlinearization in LaTeX

Ashild Naess writes:
>
> I'm working on a collection of texts in the little-described Polynesian
> language Pileni. The texts will be transcribed from my field tape
> recordings and given with interlinear morpheme-by-morpheme translations as
> well as translation into plain English. Since interlinearization in Word is
> a messy and unrewarding job, to say the least, I have started looking at
> LaTeX as a possible alternative.
>
> I am absolutely new to TeX/LaTeX and will have to start learning it from
> scratch if I am to use it. My question is: Has anyone out there ever used
> TeX/LaTex for a similar project, and can they recommend it? Any tips as to
> how it's best done? I am aware of the covington.sty package which supports
> interlinearization in examples, but has anyone used it, or something else,
> with longer texts? The aim here is to produce a book with an introduction
> about the language and its speakers, about 20 texts of varying lengths
> (some fairly short, some up to about 45 minutes on tape) and probably some
> comments about the different kinds of stories and recurring themes found in
> the material.
>
> So, will LaTeX facilitate this work enough to make it worthwhile taking the
> time to learn it? And if so, how can it best be done?
>
> Ashild Naess, University of Oslo

Talk about asking how to learn a new language! Or ... maybe that's not
as difficult a task for you, given the subject matter ... ;-)

Anyways, to start answering a few questions ...

Yes, TeX and LaTeX have been used for glossed examples -- that is, the
three-line variety:

foreign language
unit-by-unit glosses
English translation

we've been doing such stuff with TeX since the 1984 volume; also,
volumes of the annual Papers of the Algonquian Conference. And those
are just journals I've worked on. Others on this list will probably be
able to cite more examples.

Now, CJL and ALG, until 1990-something ... , were done in plain
TeX. We had custom macros for glosses, with all kinds of
variations. Since 1990-something, I've been using LaTeX for CJL (2.09,
not 2e, but in the main, I don't recall seeing anyone reporting any
difficulties in making the 2.09 macros known to ling-tex readers/users
work with 2e.

Uh -- Ashild ... that last paragraph won't make any sense to you just
yet ... ;-) LaTeX underwent a major upgrade/overhaul in 1994 but some
of us (... ok, me!) haven't moved all operations over to the new
setup. But as I said, I haven't seen any complaints posted to ling-tex
so I'll just breeze on by this upgrade business and say it shouldn't
affect anything you plan on doing. And if it does, this list is the
best place to be for getting advice and help in dealing with any
hurdles. So we'll now close that little door and move on to the real
issues, ok?

You may want to fetch an old file from CTAN which is our inventory of
style files and such:

ftp ftp.dante.de (Germany's CTAN site is closest for you, I guess)

and then go to tex-archive/info/ and get the file ling-mac.tex. I just
checked -- it's there.

You'll find a number of options available, not just Michael's
covington.sty. However, I _believe_ all these options are not able to
do automatic line-breaks for overlong glossed examples.

It's always been a problem -- the manual breaking of examples when
they're just a bit too wide for the page, because the macros usually
are based on the {tabular} environment -- setting  up a table with
tab-markers to separate the aligned items.

Does anyone have a solution to that one?

Our plain TeX approach for CJL was a bit different: each set of
glossed elements (the foreign word/phrase and its glosses) were done
up as a unit, one after the other. When we needed to introduce a line
break, then it was just done by inserting a \newline equivalent
between the appropriate sets.

I don't know if those macros would work with LaTeX. And I don't know
that that's the better solution either.

===

So, I guess the summary is:

1. yes, it has been done in LaTeX and yes, your Pileni work can also
be handled by LaTeX

2. yes, covington.sty will do it, among other options

3. the members of this list are a great resource, so ask away!
Details and advice are all here.

Ch.

### 18th January 1999: Re: Interlinearization in LaTeX

From: "Michael A. Covington" (mc@ai.uga.edu)
Subject: Re: Interlinearization in LaTeX

I'd be inclined to put some kind of preprocessor (maybe a Perl program)
ahead of TeX, rather than making TeX do all the work (which is doubtless
possible but may not be easy).  There's no reason you can't have the text
consist of "words" where each word is actually a 3-line-high vbox.

Michael A. Covington  /  AI Center  /  The University of Georgia
http://www.ai.uga.edu/~mc  http://www.mindspring.com/~covington   <><

### 18th January 1999: Re: Interlinearization in LaTeX

From: Christina Thiele (cthiele@CCS.CARLETON.CA)
Subject: Re: Interlinearization in LaTeX

Michael A. Covington writes:
>
> I'd be inclined to put some kind of preprocessor (maybe a Perl program)
> ahead of TeX, rather than making TeX do all the work (which is doubtless
> possible but may not be easy).  There's no reason you can't have the text
> consist of "words" where each word is actually a 3-line-high vbox.
>
> Michael A. Covington  /  AI Center  /  The University of Georgia
> http://www.ai.uga.edu/~mc  http://www.mindspring.com/~covington   <><

Hmmmm. An approach I hadn't thought of -- and since I don't know any
programs other than TeX, not all that surprising, perhaps ;-))

I haven't had to do great stretches of glossed materials, so the
inconvenience of having to manually insert line breaks hasn't been a
big issue. But if this Polynesian text does involve doing lots of
glossed lines, then a more efficient approach would be nice.

Has anyone done this?

Ch.

### 18th January 1999: Re: Interlinearization in LaTeX

From: Christina Thiele (cthiele@CCS.CARLETON.CA)
Subject: Re: Interlinearization in LaTeX

Soren Dayton writes:
>
> Christina Thiele  writes:
>
>
> > Does anyone have a solution to that one?
> >
> > Our plain TeX approach for CJL was a bit different: each set of
> > glossed elements (the foreign word/phrase and its glosses) were done
> > up as a unit, one after the other. When we needed to introduce a line
> > break, then it was just done by inserting a \newline equivalent
> > between the appropriate sets.
> >
> > I don't know if those macros would work with LaTeX. And I don't know
> > that that's the better solution either.
>
> I would be interested in looking at them because I expect to have an
> analogous problem in the future.  I might try to make them work well for
> LaTeX2e.
>
> Thanks,
> Soren Dayton

Soren, I can hardly believe it!  I seem to have done all the necessary
work of extracting the old plain TeX codes and even making up a sample
file for someone a few years back. The glories of being a pack-rat!

Anyways, below is the file. Make of it what you will. I'm leaving the
opening commentary in, since it provides a bit of context ...

I'm also cc'ing this to the ling-tex list, in case there's anyone else
out there who might be interested. These definitions aren't mine, by
the way! They were originally devised by Prof. Jean-Pierre Paillet and
the variants are from Michael Dunleavy (both were at Carleton
University in the mid-1980s). And notice that the sets' are only of
two elements: the foreign language element and then its gloss; the
translation is added afterwards, separately. You can argue for or
against this independence; sometimes it's just better to not try to
wrap up too much stuff into one macro ... and sometimes it is.

If Soren's willing to do the conversion to make these macros workable
with 2e, how about passing on suggestions directly to him (I'll assume
this is a guy's name -- makes me think of the Star Trek movie,
Generations), since he's offering to take a crack at it ... ;-)

Ch.

======   old-cjl-glosses.tex

%% 21 MAY 93: you were asking for macros to do linguistics glosses.
%%            Here's what we use in CJL (Canadian Journal of
%%            Linguistics). Originally, I had prepared this as a
%%            plain.tex file for Tim Curling in Australia. He found it
%%
%%            Your recent post to the German list (copy sent to me by
%%            on some LaTeX commands (the def's are of course still in
%%            plain.tex), just to see if things still work.
%%
%%            I've also revised the sample a bit, to make the format
%%            samples a bit more obvious to read. But these macros
%%            should do the trick for you. Enjoy!
%%
%%
%%            Christina Thiele

%%  From cthiele Mon Aug 10 23:18:27 1992
%%  Subject: Here are the gloss macros...
%%  To: curtling@fac.anu.edu.au
%%  Date: Mon, 10 Aug 92 23:18:27 EDT
%%  X-Mailer: ELM [version 2.3 PL11]
%%  Status: OR

%% 10 AUG 92: Tim, TeX this file, and see what you get. I sure hope
%%            it's the same as what I have on my screen! If it doesn't
%%            work, please let me know right away... (Ch.)

%%            I've just done a quick test: by inserting the usual
%%            LaTeX commands: \documentstyle{article}
%%                            \begin{document}
%%                                ....
%%                            \end{document}
%%            I get more or less the same results. Differences seem
%%            to be in the parindent being smaller in LaTeX, and
%%            spaces around cglossall (see example) are reduced to
%%            nothing -- you'd probably be better off using lglossall.
%%
%%            Have fun!
%%
%%%%%%%%%%%%%%%%%%%%%%%%%%
%%
%% Various types of glosses'. All def's have two arguments:
%%   -- the first contains the morpheme, word or phrase,
%%   -- the second contains the translation, to appear immediately
%%      below the first.

%% The gloss definitions also require these additional def's:

\documentstyle{article}

\textwidth=16cm  %% NOTE: for journals, text width is usually around 11cm
\textheight=24cm %%                          height                  18cm
\hoffset=-2cm
\voffset=-2cm

\def\nipar{\par\noindent\ignorespaces}%
\def\ipar{\par}%

%% \lglossit -- the two elements are aligned at the left, with the
%%              first element on the baseline, the second element
%%              below it.
%%           -- only for one set of elements. Multiple
%%              sets should use \lglossall.

\def\lglossit#1#2{{\setbox0=\hbox{#1\strut}%
\setbox1=\hbox{#2\strut}%
\hbox{\vtop{\box0\box1}}}}%

\def\lgloss#1#2|{\hbox{\vtop{\hbox{\ignorespaces#1\strut}%
\hbox{\ignorespaces#2\strut}}}\enspace}%

\def\lglosses#1@{\leavevmode%
\ifx>#1\let\next=\nipar\vskip2pt%     % was \smallskip
\else\ifx\nipar#1\nipar\let\next=\lglosses%
\else\lgloss#1|\let\next=\lglosses%
\fi\fi\next}%

\def\lglossall#1{{\raggedright\lglosses#1@>@}}%

%% \overglossit -- the two elements are aligned at the left, with the
%%                 *second* element on the baseline, and the first element
%%                 appears *above* it.

\def\overglossit#1#2{\setbox1=\hbox{\ignorespaces#1}%
\setbox2=\hbox{\ignorespaces#2}%
\ifdim\wd1>\wd2%
\setbox3=\hbox to \wd1{\hss #2\hss}%
\setbox4=\hbox{\ignorespaces#1}%
\hbox{\strut\vbox{\box4\box3}}%
\else%
\setbox5=\hbox to \wd2{\hss #1\hss}%
\setbox6=\hbox{\ignorespaces#2}%
\hbox{\strut\vbox{\box5\box6}}%
\fi}%

%% \flushglossit -- where there is no space added between
%%                  successive sets of elements.

\def\flushglossit#1#2{\setbox1=\hbox{\ignorespaces#1}%
\setbox2=\hbox{\ignorespaces#2}%
\ifdim\wd1>\wd2%
\setbox3=\hbox to \wd1{\hss #2\hss\strut}%
\setbox4=\hbox{\ignorespaces#1\strut}%
\hbox{\vtop{\box4\box3}}%
\else%
\setbox5=\hbox to \wd2{\hss #1\hss\strut}%
\setbox6=\hbox{\ignorespaces#2\strut}%
\hbox{\vtop{\box5\box6}}%
\fi}%

%% \cglossit -- the two elements are centred, one above the
%%              other, first element on the baseline, the
%%              second appears below it.
%%           -- only for one set of elements. Multiple
%%              sets should use \cglossall.

\def\cglossit#1#2{\setbox1=\hbox{\ignorespaces#1}%
\setbox2=\hbox{\ignorespaces#2}%
\ifdim\wd1>\wd2%
\setbox3=\hbox to \wd1{\hss #2\hss\strut}%
\setbox4=\hbox{\ignorespaces#1\strut}%
\hbox{\vtop{\box4\box3}}\enspace%
\else%
\setbox5=\hbox to \wd2{\hss #1\hss\strut}%
\setbox6=\hbox{\ignorespaces#2\strut}%
\hbox{\vtop{\box5\box6}}\enspace%
\fi}%

\def\cgloss#1#2|{\setbox1=\hbox{\ignorespaces#1}%
\setbox2=\hbox{\ignorespaces#2}%
\ifdim\wd1>\wd2%
\setbox3=\hbox to \wd1{\hss #2\hss\strut}%
\setbox4=\hbox{\ignorespaces#1\strut}%
\hbox{\vtop{\box4\box3}}%
\else%
\setbox5=\hbox to \wd2{\hss #1\hss\strut}%
\setbox6=\hbox{\ignorespaces#2\strut}%
\hbox{\vtop{\box5\box6}}
\fi}%

\def\cglosses#1@{\leavevmode
\ifx>#1\let\next=\nipar\smallskip%
\else\ifx\nipar#1\nipar\let\next=\cglosses%
\else\cgloss#1|\let\next=\cglosses%
\fi\fi\next}%

\def\cglossall#1{{\raggedright\cglosses#1@>@}}%

%% END OF MACROS

\begin{document}

\section{Glosses for linguistic \newline
morpheme-by-morpheme presentations}

%% Sample use of these macros...

\overfullrule=0pt

Some text here \lglossit{\bf mot}{\it word} for an example
of {\tt lglossit}. Notice that the words are aligned on the left, and
that there is a bit of an extra gap between mot'' and for''. Just
to make the samples easier to identify, I'm putting in overt coding
for boldface and italics. Neither is usually used when doing
linguistic examples --- the formatting already signals the various
elements in the display.

\vskip2pc

Some text here \cglossit{\bf Ein Wort}{\it word} for an example
of {\tt cglossit}. Notice that there seems to be that extra space
again after Wort''. When using these gloss macros in a regular line
of text, the {\tt flushglossit} macro might be more pleasing
aesthetically.

\vskip2pc

Some text here \flushglossit{\bf palabra}{\it word} for an
example of {\tt flushglossit}: notice that there's no additional space
following palabra'', which is not the case with the use of {\tt
cglossit}, just above.

\vskip2pc

Some text here \overglossit{\bf Some language}{\it translation} for
an example of {\tt overglossit}.

\vskip2pc

\lglossall{\bf First}{\it Second}@{\bf element}{\it
item} Here in an example using {\tt lglossall}. Notice
that when using the recursive def, the macro automatically inserts a
new line once the closing {\tt Nlglossall} has been read. If you
really must have a couple of glosses on a line {\bf without} this
forced line break, then you should probably just use a series of
single-purpose {\tt *glossit}s.

\vskip2pc

\cglossall{\bf First}{\it Second}@{\bf element}{\it
item} Here in an example using {\tt cglossall}. Same as
above: once the recursive cycle of {\tt cglossall ... Ncglossall} has
been read by \TeX, there is an automatic line break. But this isn't so
catastrophic: if you're doing a series of glosses, it's usually within
an example format, so there's no following text.

\vskip1pc

In all cases, glosses are usually followed by a third line of text,
which is usually a proper translation of the sentence being analysed
with the glosses:

\lglossall{\it Il}{They}@{pensent}{think}@{pas}{not}@{\it
personne}{anyone}

None of them thinks.'

\noindent or:

\cglossall{\it Il}{They}@{pensent}{think}@{pas}{not}@{\it
personne}{anyone}

None of them thinks.'

\vfill\eject

Then again, you might also find it simpler to just use {\tt tabular}
with left, right, or centred alignments!  There is a lot less coding
involved, and you do have control over spacing between columns of text
(by inserting {\tt tabcolsep=} x just above the \verb|\begin{tabular}|
line --- the default value in La\TeX\ is 5pt, so reducing this to 3pt
is often all you need.

Additionally, you can reduce interline spacing in a La\TeX\ table by
adding [-2pt] after the double-backslash. By the way, this is the
reason that you cannot begin a row in the {\tt tabular} environment
with a [ --- you have to enclose such items within an {\tt mbox}.
Just in case you need to do some tables where the first column
contains phonetic transcriptions within [ ... ] braces.

Can't think of any other tips right now.

Next item on the list is to send you a copy of my paper. It was done
with macros and a version of \TeX\ from 1986--1987! I'll have to first
check that it still runs!

Bye for now...

P.S. I'm no doctor''! I've my M.A., and then I got out to earn
money... theory isn't what I wanted to do with linguistics.
And that's sort of how I ended up doing \TeX\ in linguistics.

\end{document}  %%  \bye

### 18th January 1999: Re: Interlinearization in LaTeX

From: Kevin Russell (krussll@CC.UMANITOBA.CA)
Subject: Re: Interlinearization in LaTeX

krussll writes:
Subject: Re: Interlinearization in LaTeX
To: cthiele@CCS.CARLETON.CA (Christina Thiele)

Christina Thiele writes:
> Michael A. Covington writes:
> > I'd be inclined to put some kind of preprocessor (maybe a Perl program)
> > ahead of TeX, rather than making TeX do all the work (which is doubtless
> > possible but may not be easy).  There's no reason you can't have the text
> > consist of "words" where each word is actually a 3-line-high vbox.
...
>
> Has anyone done this?

The quick-and-dirty way I usually do interlinear glosses is with tables.
I'm remembering/making-this-up off the top of my head, so don't quote me
if doesn't work:

\newcommand{\ilp}[2]{\begin{tabular}[t]{@{}l@{}}#1\\{}#2\end{tabular}}

where \ilp is my dumb mnemonic for "interlinear pair". (The meaning and
definition of my corresponding 3-line "\ilt" command is left as an
And then (grabbing a random book off my shelf):

\ilp{Qanuq-}{how-}%
\ilp{it-}{be-}%
\ilp{tu-}{{\sc rel:intrans}-}%
\ilp{mik}{{\sc inst:sg}
\ilp{qimmi-}{dog-}%
\ilp{qar-}{have-}%
\ilp{p-}{{\sc inter}-}%
\ilp{i-}{{\sc intrans}-}%
\ilp{t}{{\sc 2sg}

\sentgloss{What kind of dog do you have?'}

Lines with non-word-final morphemes end with a % to suppress the space.
Word wrapping will happen automatically (though the sentence-by-sentence
running translation has to be separate).

I've done tons of paper examples like this (up to "paragraph" length).
I've done a few longer texts like it too, but only with a word-for-word
gloss.  I get a Perl program to break text lines apart into words and
write out:

\ilp{Qanuqittumik}{}
\ilp{qimmiqarpit}{}
...

and I just go through filling in the second set of braces.

I tremble at the thought of doing an entire text morpheme-by-morpheme,
even with mechanical help.

-- Kevin

-------------
Kevin Russell
Linguistics, University of Manitoba

### 18th January 1999: Re: Interlinearization in LaTeX

From: Kirk Lowery (KirkLowery@xc.org)
Subject: Re: Interlinearization in LaTeX

Christina Thiele wrote:
>
> Ashild Naess writes:
> >
> > I am absolutely new to TeX/LaTeX and will have to start learning it from
> > scratch if I am to use it. My question is: Has anyone out there ever used
> > TeX/LaTex for a similar project, and can they recommend it?

I've played around with doing it with ancient near eastern texts, but got hung
up with the right-to-left issue (ArabTeX is wonderful, but chokes on reversing
entire lines of macros)....Michael Covington is right--a preprocessor is the way
to go. And perl is perfect for the job, and far easier to parse and manipulate
strings than TeX...I've used perl to merge files from several sources and
emitted TeX codes for the macros I wanted. Of course, you have to know *both*
TeX and perl (or the programming language of your choice).

> > I am aware of the covington.sty package which supports
> > interlinearization in examples, but has anyone used it, or something else,
> > with longer texts? The aim here is to produce a book with an introduction
> > about the language and its speakers, about 20 texts of varying lengths
> > (some fairly short, some up to about 45 minutes on tape) and probably some
> > comments about the different kinds of stories and recurring themes found in
> > the material.

If a camera-ready copy of a formatted book is where you're headed, then LaTeX is
the way to go. But there are alternative software solutions for the
interlinearization of texts, especially if you have a lot of it. Check out
where they have two major (and free :-) software packages:
Shoebox and IT. They produce the interlinearized text in ASCII form, but going
from that to TeX shouldn't be hard. But these are serious software packages
which have learning curves to use them effectively. I've used both, and they do
the job well. But it's only worth it if interlinearization is a major part of
what you do. If all you want is a one-time shot, then I'd recommend you
transcribe your material in straight ASCII files, and write a program to massage
the data into the TeX code you want.

Hope this helps,

Kirk
_____________________________________________________________________________
Kirk E. Lowery, Ph.D.                              voice: (215) 572-3845
Associate Director of Academic Computing             fax: (215) 887-5404
Adjunct Professor of Old Testament                 email:
Westminster Theological Seminary

### 18th January 1999: Re: Interlinearization in LaTeX

Subject: Re: Interlinearization in LaTeX

Am I missing something?  The gloss macros (a version of which is included in
covingtn.sty) are not based on the {tabular} environment, and they break
overlong lines just fine.  (The resulting lines are not right-adjusted, but
this can easily be changed).

If you use gloss.sty or cgloss4e.sty instead of Covington's version, the
terminator/separator is \\, not newline as in covingtn.sty, and hence it's
easy to run glosses one paragraph at a time.  I suppose TeX might run out of
something or other with extra-long paragraphs, but anyway everything worked
fine when I tried it with cgloss4e.sty and paragraphs of up to 400 words
(which were also broken across pages with no problems :-).

The gloss-type macros are for text arranged like this:
\gll Text in original language \\
word for word gloss of the text \\
\glt Non-literal translation.'

If someone prefers to type in the glosses after each word, in the form
word/gloss word/gloss word/gloss,
well, \gll is not the right tool, but this is a much easier task, no?

alexis@ling.upenn.edu

### 19th January 1999: Re: Interlinearization in LaTeX

From: Christina Thiele (cthiele@CCS.CARLETON.CA)
Subject: Re: Interlinearization in LaTeX

>
> Am I missing something?  The gloss macros (a version of which is included in

Nope. You aren't. I am, though ... ;-(

I just checked: I'm not using Michael C's stuff, but rather
cm-lingmacros, Chris Manning and Avery Andrews' modified version of
Emma Pease lingmacros. Apologies to all concerned ...

> covingtn.sty) are not based on the {tabular} environment, and they break
> overlong lines just fine.  (The resulting lines are not right-adjusted, but
> this can easily be changed).

Hmmm ... I didn't know that.

> If you use gloss.sty or cgloss4e.sty instead of Covington's version, the
> terminator/separator is \\, not newline as in covingtn.sty, and hence it's
> easy to run glosses one paragraph at a time.  I suppose TeX might run out of

And this sounds even better.

> something or other with extra-long paragraphs, but anyway everything worked
> fine when I tried it with cgloss4e.sty and paragraphs of up to 400 words
> (which were also broken across pages with no problems :-).
>
> The gloss-type macros are for text arranged like this:
> 	\gll Text in original language \\
> 	word for word gloss of the text \\
> 	\glt Non-literal translation.'
>
> If someone prefers to type in the glosses after each word, in the form
> 	word/gloss word/gloss word/gloss,
> well, \gll is not the right tool, but this is a much easier task, no?
>
>
> alexis@ling.upenn.edu
>

Alexis ... since you seem to be familiar with all these gloss options
-- and I'm clearly not -- would you be at all interested in writing up
a sort of survey article for TUGboat? I'd be more than happy to lend a
hand, if that would help. I'm sure we could also find a way to then
post the article somewhere on the web to make it more widely
available.

So, Ashild, looks like there are a number of choices to make, judging
from the number of replies your query's elicited. And what's clear is
that TeX is definitely a reasonable route for you to consider in
page.

Meanwhile, I'd better go look into some of Alexis' suggestions about
those glossing options ... ;-))

Ch.

### 19th January 1999: Re: Interlinearization in LaTeX

From: Steve McConnel (steve@mcconnel.ac.sil.org)
Subject: Re: Interlinearization in LaTeX

> From: Kirk Lowery (KirkLowery@xc.org)
> Subject: Re: Interlinearization in LaTeX
> ...
> If a camera-ready copy of a formatted book is where you're headed, then
> LaTeX is the way to go. But there are alternative software solutions for
> the interlinearization of texts, especially if you have a lot of it. Check
> out  where they have two major (and free :-) software
> packages: Shoebox and IT. They produce the interlinearized text in ASCII
> form, but going from that to TeX shouldn't be hard. But these are serious
> software packages which have learning curves to use them effectively. I've
> used both, and they do the job well. But it's only worth it if
> interlinearization is a major part of what you do. If all you want is a
> one-time shot, then I'd recommend you transcribe your material in straight
> ASCII files, and write a program to massage the data into the TeX code you
> want.

We (SIL) also have a TeX macro package for typesetting interlinear texts,
bundled with programs to convert IT/Shoebox files to TeX input.  The primary
macro package is in plain TeX, but there's a stripped down LaTeX style
package as well.  The LaTeX macros were designed for 2.09, but I think they
work for 2e as well.  The TeX package is called ITF (for Interlinear Text
Formatting).  It uses the nested vbox/hbox approach, and handles an
arbitrary number of word or morpheme gloss lines.

When we first produced this package in 1990, i meant to write it up for
TUGboat but never got around to it (partly because i developed CTS shortly
after that).  If you'd like to look at this, it's at URL

ftp://ftp.sil.org/software/dos/itf11.zip

Don't worry about the "dos" -- all that means is that the conversion
programs are MS-DOS executables, which make sense because IT and (the
original version of) Shoebox are MS-DOS programs.  The TeX macros are, of
course, system independent.  The sources to the programs are also available
upon request.  :-)
--
Summer Institute of Linguistics        or Stephen_McConnel@sil.org
7500 W. Camp Wisdom Road           phone: (972)708-7361
Dallas, TX 75236  U.S.A.           fax:   (972)708-7363

### 19th January 1999: Re: Interlinearization in LaTeX

From: "Michael A. Covington" (mc@ai.uga.edu)
Subject: Re: Interlinearization in LaTeX

Thanks for the update.  covingtn.sty is long overdue for an update and I may
well tell people to use gloss.sty rather than the kluge that I put together
(from an early gloss.sty with the author's permission).  I actually had
never tried to run across a line break and couldn't remember what it would
do.

Michael A. Covington  /  AI Center  /  The University of Georgia
http://www.ai.uga.edu/~mc  http://www.mindspring.com/~covington

### 19th January 1999: Re: Interlinearization in LaTeX

From: Serge Rosmorduc (rosmord@iut.univ-paris8.fr)
Subject: Re: Interlinearization in LaTeX

Using a different table for each glossed part works. The following
macro works under latex :
(use : \GLO/WORD,transcription,translation/)

\def\GLO/#1,#2,#3/{{%
\setbox6=\hbox{{#1}}%
\setbox2=\hbox{{#2}}%
\setbox4=\hbox{\emph{#3}}%
\ifdim\wd6 > \wd2\relax\dimen8=\wd6\else\dimen8=\wd2\fi%
\dimen4=\wd4%
\ifdim\dimen8 < \wd4\relax\dimen8=\dimen4\else\fi%
\leavevmode\hbox to\dimen8{%
\vbox{%
\hbox to \dimen8{\hfil\box6\hfil}%
\hbox to \dimen8{\hfil\box2\hfil}%
\vskip 3mm%
\hbox to \dimen8{\hfil\box4\hfil}}}}}

regards,
Serge Rosmorduc

### 19th January 1999: Re: Interlinearization in LaTeX (Perl)

From: Avery Andrews (andaling@pretty.anu.edu.au)
Subject: Re: Interlinearization in LaTeX (PERL)

So, has anyone written a PERL module to support preprocessing of Tex/Latex?
What I have in mind, perhaps not very coherently, is a scheme wereby

(a) by default the text is just passed thru
(b) you can write code for actually processing the arguments of
specific macros, including the possibility of using wierd
argument-delimeters.

Once upon a time I wrote an interlinearization preprocessor for Scribe,
& built into it a fair amount of knowledge about the syntax of scribe
macros, but Tex/Latex is less constrained in this respect.  I'm also
interested in the general issue of Latex->HTML/XML conversion, having
some success with Eitan Gurari's tex4ht package, although HTML's
deficiencies as a typesetting language make me suspect that some
latex preprocessing, HTML postprocessing or both would help smooth things
out.

Avery.Andrews@anu.edu.au

### 19th January 1999: Re: Interlinearization in LaTeX (Perl)

From: Kirk Lowery (KirkLowery@xc.org)
Subject: Re: Interlinearization in LaTeX (PERL)

Avery Andrews wrote:
>
> So, has anyone written a PERL module to support preprocessing of Tex/Latex?
> What I have in mind, perhaps not very coherently, is a scheme wereby
>
>   (a) by default the text is just passed thru
>   (b) you can write code for actually processing the arguments of
>         specific macros, including the possibility of using wierd
>         argument-delimeters.
>

A quick search on "TeX" in CPAN yields the following perl modules:

TeX::DVI: write out TeX's DVI (DeVice Independent) file
TeX::DVI::Parse: parse TeX's DVI output file
TeX::Hyphen: hyphenate words using TeX's patterns
Text::BibTeX::Entry: read and parse BibTeX files
Text::BibTeX::File: interface to whole BibTeX files
Text::BibTeX::Structure: check/enforce database structure in BibTeX entries
Text::TeX: Perl module for parsing of TeX

I've never used or looked at any of these. Until Avery suggested it, I never
thought to look. (I look there for everything else before I program. Duh! ;-)

I noted along the way that there are a *lot* of modules for text manipulation of
many different types. If you can program in perl, these are definitely worth
checking out.

Thanks, Avery!

Kirk
_____________________________________________________________________________
Kirk E. Lowery, Ph.D.                              voice: (215) 572-3845
Associate Director of Academic Computing             fax: (215) 887-5404
Adjunct Professor of Old Testament                 email:
Westminster Theological Seminary

### 19th January 1999: Metrical schemes in LaTeX

From: Stefan Baums (sb51@soas.ac.uk)
Subject: Metrical schemes in LaTeX

Hello!

Does anyone on this list know a package to typeset metrical schemes
with LaTeX? The output should looks somewhat like

u-u-|-uu|-u-u-

for a scheme "light syllable - heavy syllable ... caesura ..."

In a more complicated scenario, two light syllables might be equivalent
to one long one:

__ __ | __ __
uu uu| uu uu

(which is supposed to be represent _one_ line of verse).

What I need is a) the symbols for heavy and light syllables, and b) maybe
a more convenient way to pile them up than a tabular environment.

Have a nice day,
Stefan

### 20th January 1999: Re: Interlinearization in LaTeX

From: "Christopher D. Manning" (cmanning@sultry.arts.usyd.edu.au)
Subject: Re: Interlinearization in LaTeX

Some people might feel that this thread has gone on long enough, but

If you want to do whole paragraphs of interlinear text, making boxes for
each glossed item is definitely the way to go.

As far as I'm aware, the style files for conveniently entering stuff for
glossing into (La)TeX files that do this all ultimately derive from
Marcel R. van der Goot's plain TeX Midnight macros package, gloss.tex.
(Very few other people have ever gotten their head around the looping
primitives of TeX.).  It surprised me, but they are still available at:
ftp://csvax.cs.caltech.edu/pub/tex/Midnight/
That's gloss.tex, and gloss.doc (documentation even!).  It has never
been updated as far as I'm aware.

Deriving from that source there's at least:
covingtn.sty, covington.sty
cgloss4e.sty
gloss.sty	 by Alexander Holt
gloss.sty	 by me

All of these are basically similar.  I think mine has a few additional
options (allowing 3 aligned lines instead of 2, and doing "morpheme
glossing" where additonal space is not put between morphemes).  It's
available at: http://www.sultry.arts.usyd.edu.au/cmanning/tex/

Chris Manning

### 29th January 1999: Wanted: a *dictionary.cls!

From: "BP.Jonsson"
Subject: Wanted: a *dictionary.cls!

Dear ling-TeXers,

I have been asked to help with the technical/typographical
aspects of editing an etymological dictionary and preparing it
for the press.  Since the people involved in the project are
spread over three different operating systems and as many
countries I suggested that LaTeX should be used for the task.  To
my surprise I found no tools apt for this kind of project, or
even an ordinary dictionary, in the CTAN archive.
I suppose what is really needed is a LaTeX class for
dictionaries!  Hpefully there are several people on this list
that would have an interest in rectifying this situation, and/or
the technical skills and the motivation to join a a project/group
effort create such a tool.  I unfortunately lack the necessary
LaTeX know-how as yet.

Some features that distinguish dictionaries from books in general
(listed in no particular order):

# the _entries_ of a dictionary are treated typographically
rather different from paragraphs in a book, yet they are
structurally equivalent to each other.

# entries are separated by a much smaller vertical skip than
paragraphs, or by none at all.

# dictionaries are _as a rule_ set in double columns.

# the keywords of the first and last entry of a dictionary page
should be displayed in the page-header.

# the entries of a dictionary may contain several kinds of
information which are distinguished typographically:
(1) keyword,
(2) pronunciation,
(3) grammatical information,
(4) translation(s)/definition(s),
(5) information on derivation and cognates,
(7) examples -- often with abbreviated references to
authors and works from which they were taken.
Further there are often subentries, in which the keyword and
pronunciation are usually truncated.

# the entries of a dictionary are sorted alphabetically.

Different special-purpose dictionaries will usually have further
specific features.  For an etymological dictionary these may be:

# each entry as a rule contains several cognates and examples
from different languages, each of which is tagged by an
abbreviation.

# there may be a separate index of words from each language
cited in the entries, apart from the "main" language of the
dictionary.  If so it is common for each entry to have a
number
by which it is referenced in the indices.

Come to think of it, separate indices for cited words from
different languages may come in handy in many kinds of linguistic
publications.  So may separate indices for e.g. topics and
languages mentioned in the text.

Regards,
B.Philip Jonsson
----------------------------------------------------
Solitudinem faciunt pacem appellant!
(Tacitus)

### 29th January 1999: Re: Wanted: a *dictionary.cls!

From: Emma Pease
Subject: Re: Wanted: a *dictionary.cls!

I've assisted on one dictionary (an Hausa-English one back in the
mid-80's done in TeX), so your message caught my eye.  Most of my work
on that was metafont hackery not TeX hackery though.

1. You may not want to work with latex directly but rather with some
database program (and/or perl) that manipulates then converts what you
are working on into latex (letter-a.tex, letter-b.tex...).  Perhaps in
the form

\keyword{word}\pronounce{...}...

You only handwrite the top level file that includes letter-a.tex, ....

A database program would talk care of alphabetizing, etc.  A good one
would also allow you to have much information, only some of which you
might choose to have in the final dictionary.  The author of the Hausa
dictionary, if I remember correctly, was using sed and awk (choices
have expanded a lot since then).

2. You might want to look at the packages

multicol (for mulicolumn)

3. If I were doing this, I would probably use the book document class
(suitably modified) and then add packages for doing double-column,
fancy running heads, dictionary entries (which would include the \mark
needed for the fancy running heads).  Writing the dictionary package
would be necessary but probably not too difficult.

4. There is a discussion back in 1997 in comp.text.tex on
dictionaries.  You might want to hunt the thread up in dejanews as a
few of the messages were by people who had done dictionaries in
latex/tex.

Subject line was "typesetting a Huge dictionnary"

Emma

### 29th January 1999: Re: Wanted: a *dictionary.cls!

From: "Harold F. Schiffman"
Subject: Re: Wanted: a *dictionary.cls!

Here at the U. of Pennsylvania we have been working on an English
Dictionary of the Tamil Verb and have been using LaTeX to format the
output with pretty good success.  Tom Ridgeway at the U. of Washington (we
started the project there) wrote some macros to give us the kind of
two-column output we want.  The data are in  a d-base where we do input
and editing etc. and then it is extracted to LaTeX for hard copy
formatting.  We also can produce an html version which we have on our
website and will share the URL with people who have special interest in
Tamil.  We do have a gif file of one LaTeX page we can also share (I'm
writing this from home so haven't got the exact url's in front of me).

Vasu Renganathan is in charge of d-base management and can dredge up the
.sty file that we use.  It is double-column but not right-justified since
that would require hyphenation routines for Tamil that we never
could work out right.  (People know intuitively where Tamil words should
be split but can't give us explicit rules, so we put in soft hyphens in
very long words and hope for the best.) And in any event, there is a lot
of indentation in a given entry, example sentences, synonyms etc. that
gives a ragged left margin anyway, so a ragged right doesn't look so bad
after all.  The Tamil font we use was also developed  by Ridgeway
(Wntamil) with input from me about shapes and aesthetics, and has been
modified and adapted by Vasu to give us various sizes and thicknesses.

Hal Schiffman

On Fri, 29 Jan 1999, BP.Jonsson wrote:

>   # the _entries_ of a dictionary are treated typographically
>     rather different from paragraphs in a book, yet they are
>     structurally equivalent to each other.
>
>   # entries are separated by a much smaller vertical skip than
>     paragraphs, or by none at all.
>
>   # dictionaries are _as a rule_ set in double columns.
>
>   # the keywords of the first and last entry of a dictionary page
>     should be displayed in the page-header.
>
>   # the entries of a dictionary may contain several kinds of
>     information which are distinguished typographically:
>         (1) keyword,
>         (2) pronunciation,
>         (3) grammatical information,
>         (4) translation(s)/definition(s),
>         (5) information on derivation and cognates,
>         (7) examples -- often with abbreviated references to
>         authors and works from which they were taken.
>     Further there are often subentries, in which the keyword and
>     pronunciation are usually truncated.
>
>   # the entries of a dictionary are sorted alphabetically.
>
> Different special-purpose dictionaries will usually have further
> specific features.  For an etymological dictionary these may be:
>
>   # each entry as a rule contains several cognates and examples
>     from different languages, each of which is tagged by an
>     abbreviation.
>
>   # there may be a separate index of words from each language
>     cited in the entries, apart from the "main" language of the
>     dictionary.  If so it is common for each entry to have a
> number
>     by which it is referenced in the indices.
>
> Come to think of it, separate indices for cited words from
> different languages may come in handy in many kinds of linguistic
> publications.  So may separate indices for e.g. topics and
> languages mentioned in the text.
>
> Regards,
>
>     B.Philip Jonsson
>     ----------------------------------------------------
>         Solitudinem faciunt pacem appellant!
>                                             (Tacitus)

### 29th January 1999: Re: Wanted: a *dictionary.cls!

From: "Harold F. Schiffman"
Subject: Re: Wanted: a *dictionary.cls!

I've located the gif file that we have that is a picture of (one column
only) of our LaTeX output of the first entry in our English Dictionary of
the Tamil Verb.

The URL is:

http://ccat.sas.upenn.edu/plc/dictionary/

Scroll down to the bottom of the page and click on "A sample hard copy
output"  the url for which is

http://ccat.sas.upenn.edu/plc/dictionary/dict2.gif

As you can see, we still have some formatting problems, such as placement
of commas after a change of font from Tamil to English.

There is a textual description of the project at

http://ccat.sas.upenn.edu/plc/dictionary/doc/tamildic.htm

(This is of course a LaTeX document converted to html by Latex2html).

H. Schiffman

=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
------------------------------------------------------------------------------
Harold F. Schiffman
Henry R. Luce Professor of Language Learning		Penn Language Center
Dept. of South Asia Regional Studies		       715-16 Williams  Hall
820 Williams Hall, Box 6305					   Box  6305

University of Pennsylvania

Phone:  (215) 898-5825	   		       	              (215) 898-6039
Fax:  (215) 573-2138				          Fax (215) 573-2139

Email:  haroldfs@ccat.sas.upenn			      plc@ccat.sas.upenn.edu
WWW:  http://ccat.sas.upenn.edu/~haroldfs/    http://ccat.sas.upenn.edu/~plc/

The Penn Language Center is a facility supporting language teaching in the
less-commonly-taught languages, as well as research in language pedagogy
and interdisciplinary language-related issues.
------------------------------------------------------------------------------

### 17th February 1999: Dotted lines with Emma Pease tree macros

From: Christina Thiele
Subject: dotted lines with Emma Pease tree macros

I had a query yesterday for a quick way to get a dotted line when
using Emma Pease tree macros. It's not in the documentation, so
perhaps it's of interest to others as well.

Ch.

==============

> \nodecurve ... \nodeconnect ... you're using Emma's stuff, right?
>
> Here's what I have marked in my documentation. Did you print hers up?
> On p.7, just above Section 7 Errors, there's a piece on how to make a
> dashed line. Well, I experimented with that and came up with a value
> for dotted lines ...
>
> If you don't have the documentation in hardcopy, here's what it says,
> and then what I've added:
>
> In your file, in the section after the tree, where you start defining
> the nodecurves and nodeconnects, you have to make a group (enclosed
> with curly braces), where the line being described by the
> nodecurve/nodeconnect sets will be dashed ... or dotted:
>
> ....
> \nodeconnect{a}{b} ... etc.
> %
> {\makedash{4pt}
>  \nodecurve{x}{y}
> }
>
>
> That sets the line to have the dash length that shown on p.6 of the
> documentation.
>
> I reset it to 1pt, and that yielded essentially a dotted line:
>
> {\makedash{1pt}
>  \nodecurve{m}{q}
> }
>
> See if that works for you and your printer.
>
> Ch.

### 25th March 1999: Special charsets with virtual fonts

From: BP Jonsson
Subject: special charsets with virtual fonts

Dear Ling-TeXers,

it's me with the etymological dictionary again.  I got the idea that the
work of entering the material might be made much easier by defining virtual
fonts with ligtables such that sequences of ASCII characters in the source
are converted into characters with diacritics, special letters and the like.

Unfortunately I lack the degree of TeX literacy required for defining
special charset virtual fonts with fontinst (or manually), so I wonder if
there is anyone who has experience an is willing to help?

Regards,

/BP

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
B.Philip Jonsson

Solitudinem faciunt pacem appellant!
(Tacitus)

### 25th March 1999: Re: Special charsets with virtual fonts

Subject: Re: special charsets with virtual fonts

> it's me with the etymological dictionary again.  I got the idea that the
> work of entering the material might be made much easier by defining virtual
> fonts with ligtables such that sequences of ASCII characters in the source
> are converted into characters with diacritics, special letters and the like.

This approach was used with the Greek environment for latex (or at least
with some version of it--I won't swear it's in the babel version).  Perhaps
you can use as a worked example?

Alexis

### 16th April 1999: Footnotes at page bottom

From: Martin Haase
Subject: footnotes at page bottom

Hello!
I'm using LaTeX2e (together with the Koma-Script package). A common
LaTeX-problem is that footnotes are not really at the bottom of the page,
if the page cannot be filled completely. Unfortunately, my publisher is
very keen on having inevitable empty space between the text and the
footnote so that the footnote closes the page. I'm really at a loss... and
under time pressure!
Martin
---
mailto:haase@uos.de
http://www.cl-ki.uni-osnabrueck.de/~haase/
Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

### 16th April 1999: Re: Footnotes at page bottom

From: Steve Grathwohl
Subject: Re: footnotes at page bottom

>Hello!
>I'm using LaTeX2e (together with the Koma-Script package). A common
>LaTeX-problem is that footnotes are not really at the bottom of the page,
>if the page cannot be filled completely. Unfortunately, my publisher is
>very keen on having inevitable empty space between the text and the
>footnote so that the footnote closes the page. I'm really at a loss... and
>under time pressure!
>Martin
>---
>mailto:haase@uos.de
>http://www.cl-ki.uni-osnabrueck.de/~haase/
>Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

Try using Robin Fairbairn's footmisc package with the "bottom"
option. This will force footnotes to the bottom of the page, even
when \raggedbottom is in effect. I think that I've used footmisc with
koma-script, but I can't be sure.

Steve

--
Steve Grathwohl
Duke Mathematical Journal * International Mathematics Research Notices
grath@math.duke.edu * +001 919-687-3634 * fax: +001 919-688-5595

### 17th April 1999: Re: Footnotes at page bottom

Subject: Re: footnotes at page bottom

> >LaTeX-problem is that footnotes are not really at the bottom of the page,
> >if the page cannot be filled completely. Unfortunately, my publisher is
> >very keen on having inevitable empty space between the text and the
> >footnote so that the footnote closes the page. I'm really at a loss... and
>
> Try using Robin Fairbairn's footmisc package with the "bottom"
> option. This will force footnotes to the bottom of the page, even
> when \raggedbottom is in effect. I think that I've used footmisc with
> koma-script, but I can't be sure.

Or, start by using \flushbottom (or the [twoside] option, which turns it on).
Perhaps that will do the job?

Alexis

### 19th April 1999: Floating tables, footnotes

From: Martin Haase
Subject: floating tables, footnotes

Thank you for your help with my footnotes problem. Since I got several
replies both publicly and privately, here's a summary:
If you want your footnotes at the bottom of the page even with
\raggedbottom, Robin Fairbairn's footmisc package (on CTAN) provides an
option for that.
And now for something completely different: :-)
I have floating tables which appear somewhere within my text (which is what
I want), but in one case a line with only three words follows the table and
then comes a new paragraph. That looks awful! Is there a means to keep
floatings tables within the text, but having at least one and a half line
after it?
Martin
- ---
mailto:haase@uos.de
http://www.cl-ki.uni-osnabrueck.de/~haase/
Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

### 19th April 1999: A phonetic problem (using TIPA)

From: Martin Haase
Subject: a phonetic problem (using TIPA)

My publisher wants me to typeset my book in Times (and not in Computer
Modern), which is no problem thanks to the times-package. My phonetic
transcriptions (with TIPA) are automatically adapted using the xipa-fonts
(which looks acceptable with times). But there is no 'ng'-sign (a
right-tailed n for a velar n). What can I do?
Thanks!
Martin
- ---
mailto:haase@uos.de
http://www.cl-ki.uni-osnabrueck.de/~haase/
Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

### 19th April 1999: Re: A phonetic problem (using TIPA)

From: FUKUI Rei
Subject: Re: a phonetic problem (using TIPA)

> My publisher wants me to typeset my book in Times (and not in Computer
> Modern), which is no problem thanks to the times-package. My phonetic
> transcriptions (with TIPA) are automatically adapted using the xipa-fonts
> (which looks acceptable with times). But there is no 'ng'-sign (a
> right-tailed n for a velar n). What can I do?

It seems strange to me that no 'ng' sunbol is found in the
xipa fonts. tipa and xipa have exactly the same encoding.

Fukui Rei

### 19th April 1999: Re: A phonetic problem (using TIPA)

From: Christina Thiele
Subject: Re: a phonetic problem (using TIPA)

> > My publisher wants me to typeset my book in Times (and not in Computer
> > Modern), which is no problem thanks to the times-package. My phonetic
> > transcriptions (with TIPA) are automatically adapted using the xipa-fonts
> > (which looks acceptable with times). But there is no 'ng'-sign (a
> > right-tailed n for a velar n). What can I do?
>
> It seems strange to me that no 'ng' sunbol is found in the
> xipa fonts. tipa and xipa have exactly the same encoding.
>
> Fukui Rei

It's probably there ... but perhaps the code to access it is where the
problem lies?

When I have problems getting at a char (often happens when I want to
get a math/symbol char in a different size), I resort to a very
rude/crude manoeuver: I locate the char on the chart and use
\fontname\char'XXX to get it to print (where \fontname has been def'd
at the specific size/style I want).

So, looking at Fukui's article in TUGboat 17,2, p.114, I find that eng'
is \char'116.

Uh ... let's say I've def'd the TIPA \font

\font\engtipa=tipa10

Then I'd make up a definition

\def\ENG{{\engtipa\char'116}}

So, my text file looks like this -- I haven't got tipa in 2e, only in
2.09. But the principle should not be any different. I tested the
following it I did get my eng's -- try it:

\documentstyle{article}

\begin{document}

\font\engtipa=tipa10

\def\ENG{{\engtipa\char'116}}

\ENG\ here and there \ENG

\end{document}

%% END OF FILE

Ch.

### 19th April 1999: Lacking ng (TIPA)

From: Martin Haase
Subject: lacking ng (TIPA)

Hello again!
Since I was not clear enough about my phonetic symbol problem (no ng-sign
with times), here is a test file and the dvips output. It is probably a
dvips problem.
Best wishes,
Martin

\documentclass[12pt]{article}
\usepackage{times,tipa}

\begin{document}

Hello world: \textipa{MNO}.

\end{document}
dvips output:
C:\Eigene Dateien\Forschung\HAB\Stauffenburg>dvips test
leye.com)
' TeX output 1999.04.19:2301' -> test.ps
- make_pk_font (ptmr8r, 720, 600, ljfour)
name: ptmr8r, dpi=720, bdpi=600, mag=magstep(1.0), mode=ljfour
- C:\texmf\miktex\bin\makemf.exe --verbose ptmr8r
Cannot find ptmr8r source file.
- C:\texmf\miktex\bin\ttf2pk.exe -q -n ptmr8r 720
Consulting C:\texmf\miktex\config\miktex.map...
Creating ptmr8r.pk...
searching FontOutline for Times-Roman...
No FontOutline found for Times-Roman!
be l
eft blank.
- make_pk_font (xipa10, 720, 600, ljfour)
600 ins
C:\TEXMF\MIKTEX\BIN\DVIPS.EXE: Such scaling will generate extremely poor
output.

. [1]

---
mailto:haase@uos.de
http://www.cl-ki.uni-osnabrueck.de/~haase/
Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

### 19th April 1999: Re: Lacking ng (TIPA)

From: FUKUI Rei
Subject: Re: lacking ng (TIPA)

> Hello again!
> Since I was not clear enough about my phonetic symbol problem (no ng-sign
> with times), here is a test file and the dvips output. It is probably a
> dvips problem.
> Best wishes,
> Martin
>
>     \documentclass[12pt]{article}
>     \usepackage{times,tipa}
>
>     \begin{document}
>
>     Hello world: \textipa{MNO}.
>
>     \end{document}

I have no problem in compiling, previewing (by xdvi) and
printing (by dvips) the above source file on a Unix machine.
The ng' symbol is there, of course. Your case seems to be
a dvips or Postscript problem.

Rei

### 19th April 1999: Re: Lacking ng (TIPA)

From: Christina Thiele
Subject: Re: lacking ng (TIPA)

Martin Haase writes:
>
> Hello again!
> Since I was not clear enough about my phonetic symbol problem (no ng-sign
> with times), here is a test file and the dvips output. It is probably a
> dvips problem.
> Best wishes,
> Martin
>
>     \documentclass[12pt]{article}
>     \usepackage{times,tipa}

I have the TeXLive 3 distribution -- but tipa.sty isn't there. Did you
have to fetch it specially?

>     \begin{document}
>
>     Hello world: \textipa{MNO}.

I don't know the package but ... \text + tipa would be \texttipa, no?
It's not a typo?

>     \end{document}

> dvips output:
> C:\Eigene Dateien\Forschung\HAB\Stauffenburg>dvips test
> This is dvips 5.83 (MiKTeX 1.20) Copyright 1998 Radical Eye Software
> leye.com)
> ' TeX output 1999.04.19:2301' -> test.ps
> - make_pk_font (ptmr8r, 720, 600, ljfour)
> name: ptmr8r, dpi=720, bdpi=600, mag=magstep(1.0), mode=ljfour
> - C:\texmf\miktex\bin\makemf.exe --verbose ptmr8r
> Cannot find ptmr8r source file.
> - C:\texmf\miktex\bin\ttf2pk.exe -q -n ptmr8r 720
> Consulting C:\texmf\miktex\config\miktex.map...
> Creating ptmr8r.pk...
> searching FontOutline for Times-Roman...
> No FontOutline found for Times-Roman!
> be l
> eft blank.
> - make_pk_font (xipa10, 720, 600, ljfour)
^
tipa or xipa?

> 600 ins
> C:\TEXMF\MIKTEX\BIN\DVIPS.EXE: Such scaling will generate extremely poor
> output.
>
> . [1]
>
> ---
> mailto:haase@uos.de
> http://www.cl-ki.uni-osnabrueck.de/~haase/
> Tel. (mobil): +49-171-7085525, PGP ID 0xFC48BE0F

Ch.

### 20th April 1999: Re: Lacking ng (TIPA)

From: Dirk Kussin
Subject: Re: lacking ng (TIPA)

I would say, its a problem of the programm maketexpk (or similar;
probably make_pk_font on your equipment) which is started by dvips; it
doesn't find the source file, probably the mf-file (or the tfm-file),
and thus no pk-file will be generated. Perhaps just a problem of the
path-variable. Or the suitable mf-file is not on the disc. Good luck.

Dirk

dvips output:
C:\Eigene Dateien\Forschung\HAB\Stauffenburg>dvips test
leye.com)
' TeX output 1999.04.19:2301' -> test.ps
- make_pk_font (ptmr8r, 720, 600, ljfour)
name: ptmr8r, dpi=720, bdpi=600, mag=magstep(1.0), mode=ljfour
- C:\texmf\miktex\bin\makemf.exe --verbose ptmr8r
Cannot find ptmr8r source file.
- C:\texmf\miktex\bin\ttf2pk.exe -q -n ptmr8r 720
Consulting C:\texmf\miktex\config\miktex.map...
Creating ptmr8r.pk...
searching FontOutline for Times-Roman...
No FontOutline found for Times-Roman!
be l
eft blank.
- make_pk_font (xipa10, 720, 600, ljfour)
600 ins
C:\TEXMF\MIKTEX\BIN\DVIPS.EXE: Such scaling will generate extremely poor
output.

. [1]

- --
Fachbereich 17 Mathematik   Raum D2.323
Universität-GH Paderborn    Tel. (+49) (5251) 60-2636

### 21st April 1999: LI style

From: "Benjamin T. Bruening"
Subject: LI style

Dear Ling-Texers:

I have a bib style file called linquiry.bst (and .sty) that is supposed to
format the bibliography according to Linguistic Inquiry's style guidelines.
Unfortunately the guidelines it conforms to are from back in vol. 15 (which
was in 1984 or so; I believe the file was created in 1993).  Does anyone
know if an updated version of this exists, and if so where I can get it?
The style has changed dramatically.

Thanks,

Benjamin Bruening

-------------------------------------------
Benjamin Bruening
Dept. of Linguistics
MIT E39-245
77 Massachusetts Ave.
Cambridge, MA 02139

### 22nd April 1999: New scripts

From: Johannes Reese
Subject: New scripts

Hallo,

I have an especially good text-editor, with which I can write a lot of
languages that currently are not to be typeset with TeX, like Thai,
Burmese, Nabataean, Khmer, etc. They usually look really good, too.

All it can do is writing text. I would like to input this text into
LaTeX-documents. The easiest way could be as graphics. But I want to use
the scripts as real scripts and part of my document. The fonts are in
ttf-format, so it is no problem to make TeX-fonts out of them by
ttftopk.

But I am no computer specialist, my skills are restricted to using
LaTeX. Is there an easy way for me to produce the necessary *.sty, *.fd
and *.mf-files that we need for LaTeX2e?

Johannes

### 22nd April 1999: Re: New scripts

From: Anshuman Pandey
Subject: Re: New scripts

> From: Johannes Reese
> Subject: New scripts
>
> I have an especially good text-editor, with which I can write a lot of
> languages that currently are not to be typeset with TeX, like Thai,
> Burmese, Nabataean, Khmer, etc. They usually look really good, too.

I believe Yannis Haralambous developed a Khmer package sometime ago. I
don't think the package was released on CTAN, but I do recall looking at
the METAFONT sources and documentation at his home site. The address of
the site escapes me at the moment, but I will search for it. I think the
package lacks a preprocessor.

There are also two Thai packages on CTAN. Look under language/thai/ or
fonts/thai/.

> All it can do is writing text. I would like to input this text into
> LaTeX-documents. The easiest way could be as graphics. But I want to use
> the scripts as real scripts and part of my document. The fonts are in
> ttf-format, so it is no problem to make TeX-fonts out of them by
> ttftopk.

One method is to use ttf2mf. I have done this for a Gurmukhi (Panjabi)
font. The results depend on the quality of the original TrueType font, of
course. The font I worked with had the virtue of being well-designed, and
so the print quality is amazingly high, even at exceptionally large
magnifications.

> But I am no computer specialist, my skills are restricted to using
> LaTeX. Is there an easy way for me to produce the necessary *.sty, *.fd
> and *.mf-files that we need for LaTeX2e?

As above, ttf2mf will produce a METAFONT source of the TrueType font.
However, don't expect to be able to modify the resulting source; it is
rather convoluted. Font definition (fd) files can then be created based on
the output file.

Regards,
Anshuman Pandey

### 22nd April 1999: Metrical schemes in LaTeX

From: Stefan Baums
Subject: Metrical schemes in LaTeX

Hello!

Does anyone on this list know a package to typeset metrical schemes
with LaTeX? The output should looks somewhat like

u-u-|-uu|-u-u-

for a scheme "light syllable - heavy syllable ... caesura ..."

In a more complicated scenario, two light syllables might be equivalent
to one long one:

__ __ | __ __
uu uu | uu uu

(which is supposed to be represent _one_ line of verse).

What I need is a) the symbols for heavy and light syllables, and b) maybe
a more convenient way to pile them up than a tabular environment.

Have a nice day,
Stefan

### 22nd April 1999: Re: Metrical schemes in LaTeX

From: Anshuman Pandey
Subject: Re: Metrical schemes in LaTeX

This was posted to the list sometime ago. I don't recall if there were any
answers. I hope no one minds my reposting the message... I am interested

Regards,
Anshuman Pandey

> From: Stefan Baums
>
> Hello!
>
> Does anyone on this list know a package to typeset metrical schemes
> with LaTeX? The output should looks somewhat like
>
>   u-u-|-uu|-u-u-
>
> for a scheme "light syllable - heavy syllable ... caesura ..."
>
> In a more complicated scenario, two light syllables might be equivalent
> to one long one:
>
>  __ __ | __ __
>  uu uu | uu uu
>
> (which is supposed to be represent _one_ line of verse).
>
> What I need is a) the symbols for heavy and light syllables, and b) maybe
> a more convenient way to pile them up than a tabular environment.
>
> Have a nice day,
> Stefan

### 22nd April 1999: Re: Metrical schemes in LaTeX

From: Christina Thiele
Subject: Re: Metrical schemes in LaTeX

Anshuman Pandey writes:
>
> This was posted to the list sometime ago. I don't recall if there were any
> answers. I hope no one minds my reposting the message... I am interested
> in any answers as well.
>
> Regards,
> Anshuman Pandey

Hi, Anshuman!

I had to do this back in 1993 (just checked the file) -- but I see
that I did no better: it was also done using a table layout.

My def's at the time were:

\def\mac{\lower1.3ex\hbox{\bigrmn\={}}}    %% for macron
\def\br{\lower1.3ex\hbox{\bigrmn\u{}}}     %% for breve

where \bigrmn was just cmr at 14.40 (i.e., 10pt at magstep2 or 12pt at
magstep1).

Each line of poetry had these scansion lines above it, so I separated
sets of lines by [-4pt] (the poetry was set in 9pt).

And each set was its own table.

So ... pretty primitive. I did a quick search on poetry' and
scansion' on Peter Flynn's CTAN site but no luck -- you'd have to
have an exact filename to get anything, I guess. I also checked the
old Jones Index but again, zip.

An interesting situation, though. I can mention it in my next Treasure
Chest column in TUGboat ... maybe something will turn up.

Ch.

=========================
>
> > From: Stefan Baums
> >
> > Hello!
> >
> > Does anyone on this list know a package to typeset metrical schemes
> > with LaTeX? The output should looks somewhat like
> >
> >   u-u-|-uu|-u-u-
> >
> > for a scheme "light syllable - heavy syllable ... caesura ..."
> >
> > In a more complicated scenario, two light syllables might be equivalent
> > to one long one:
> >
> >  __ __ | __ __
> >  uu uu | uu uu
> >
> > (which is supposed to be represent _one_ line of verse).
> >
> > What I need is a) the symbols for heavy and light syllables, and b) maybe
> > a more convenient way to pile them up than a tabular environment.
> >
> > Have a nice day,
> > Stefan

### 26th April 1999: Re: Metrical schemes in LaTeX

From: Peter Schmitt
Subject: Re: Metrical schemes in LaTeX

On Thu, 22 Apr 1999, Christina Thiele wrote:

> My def's at the time were:
>
>    \def\mac{\lower1.3ex\hbox{\bigrmn\={}}}    %% for macron
>    \def\br{\lower1.3ex\hbox{\bigrmn\u{}}}     %% for breve
>
> > > From: Stefan Baums
[]
> > > In a more complicated scenario, two light syllables might be equivalent
> > > to one long one:
> > >
> > >  __ __ | __ __
> > >  uu uu | uu uu
[]
> > > What I need is a) the symbols for heavy and light syllables, and b) maybe
> > > a more convenient way to pile them up than a tabular environment.
> > >

I think these are indeed two problems:
(1) how should the ideal symbols look like
-- are constructions like that given above satisfactory,
or should symbols be defined using MetaFont?
(2) what input syntax is most convenient?
-- Only after deciding that one can think
about designing macros (an environment) to realize it.

- ---
Peter Schmitt                 Peter.Schmitt@ap.univie.ac.at
Institute of Mathematics                   Strudlhofgasse 4
Universit"at Wien                      A-1200 Wien, Austria

### 26th May 1999: Accent over accent over letter in Plain TeX

From: BP Jonsson
Subject: accent over accent over letter in PlainTeX

Dear LingTeXers,

does anyone know a trick to get an accent over another accent over a letter
in PlainTeX' text mode?  I've tried different groupings etc. only to find
that the accents are set on the side of each other...

/BP

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
B.Philip Jonsson

Solitudinem faciunt pacem appellant!
(Tacitus)

### 26th May 1999: Re: Accent over accent over letter in Plain TeX

From: Christina Thiele
Subject: Re: accent over accent over letter in PlainTeX

BP Jonsson writes:
> Dear LingTeXers,
>
> does anyone know a trick to get an accent over another accent over a letter
> in PlainTeX' text mode?  I've tried different groupings etc. only to find
> that the accents are set on the side of each other...
>
> /BP
>
>  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>   B.Philip Jonsson
>
>         Solitudinem faciunt pacem appellant!
>                                             (Tacitus)

Yup. Use the following -- it works in both plain and 2.09 and I'll
assume 2e, since no-one's complained ;-)

It's a very old macro (written ca 1984 or 85) called \diatop.

Code:
----

\def\diatop[#1|#2]{\leavevmode{\setbox1=\hbox{{#1{}}}%
\setbox2=\hbox{{#2{}}}%
\dimen0=\ifdim\wd1>\wd2\wd1\else\wd2\fi%
\setbox1=\hbox to1\dimen0{\hss#1\hss}%
\rlap{\raise1\dimen1\box1}%
\hbox to1\dimen0{\hss#2\hss}}}%

Usage:
-----

\diatop[#1|#2]   where: #1 is one of the diacritics
#2 is the other diacritic PLUS the letter

e.g., \diatop[\'|\"u]

will yield an umlauted u' with an additional accute accent above it.

There are times when the order of diacritics is important -- you'll
just have to test things.

But that should do the trick.

====

As I said, this is an old old dog ... anyone got anything newer and
better?

Ch.

### 26th May 1999: Re: Accent over accent over letter in Plain TeX

From: Anshuman Pandey
Subject: Re: accent over accent over letter in PlainTeX

On Wed, 26 May 1999, BP Jonsson wrote:

> does anyone know a trick to get an accent over another accent over a letter
> in PlainTeX' text mode?  I've tried different groupings etc. only to find
> that the accents are set on the side of each other...

I believe diatop.tex does what you want. Example, \diatop[\=|\d{r}] will
give you a r-macron-underdot.

Regards,
Anshuman

### 26th May 1999: Re: Accent over accent over letter in Plain TeX

From: Jed Parsons
Subject: Re: accent over accent over letter in PlainTeX

I use \twoacc supplied by covington.sty

e.g., \twoacc[\'|\u\i] for an acute over a breve over an i

Cheers,

Jed

Jed Parsons (Just another P[erl|lautus] hack ...)    jed@socrates.berkeley.edu
http://www.ocf.berkeley.edu/~jparsons/

[         My superbrain will find a way to cope.'' -- Supergrover          ]
- ------------------------------------------------------------------------------

% end of macros

### 29th October 1999: Multi-line glosses

From: "J.Kremers"
Subject: multi-line glosses

Hi everyone,

This is the first time I am posting a message on the list, so a short
introduction may be in order. I am a PhD student at the University of
Nijmegen, The Netherlands. I have a master in Arabic (specialization in
linguistics) and I am currently working on a project The noun phrase in
Arabic'. Just recently I have started using LaTeX to write my papers. I use
the extensions qtree and linguex.
When I converted my WordPerfect documents to LaTeX, I ran into a problem.
In some examples, I use multi-line glosses to indicate the features of the
lexical items. However, I could not get linguex to produce such glosses. To
give you an idea, this is what I want:

to	islands	preceding		mentioning-NOM-their
fem		masc			masc
pl		sg			sg
indef		indef			def
gen		gen			nom
to aforementioned islands'

(I hope the tabs are not distorted too much...) I've tried groupings in the
gloss with newlines in them, (i.e. {islands \\ fem \\ pl \\ indef \\ gen }
etc.) but I was not surprised to see that did not work.

I am now using a tabular environment to get the effect, but that puts the
number of the example too low; not in front of the Arabic phrase, but
before the line with pl   sg   sg'. Has anyone encountered and hopefully
solved this problem?

Joost Kremers

-------------------------------------------------------
Joost Kremers
University of Nijmegen - The Netherlands
Department of Languages and Cultures of the Middle-East
Erasmusplein 1
6525 HT Nijmegen - The Netherlands
phone: +31 24 3612996
fax: +31 24 3611972

### 29th October 1999: Re: Multi-line glosses

From: Christina Thiele
Subject: Re: multi-line glosses

J.Kremers writes:
>
> Hi everyone,
>
>   This is the first time I am posting a message on the list, so a short
> introduction may be in order. I am a PhD student at the University of
> Nijmegen, The Netherlands. I have a master in Arabic (specialization in
> linguistics) and I am currently working on a project The noun phrase in
> Arabic'. Just recently I have started using LaTeX to write my papers. I use
> the extensions qtree and linguex.

Welcome to the list! While I myself don't use either of the packages
you mention, I'm quite sure there are a good number of list readers
who do, so help will surely be on the way.

>   When I converted my WordPerfect documents to LaTeX, I ran into a problem.
> In some examples, I use multi-line glosses to indicate the features of the
> lexical items. However, I could not get linguex to produce such glosses. To
> give you an idea, this is what I want:
>
> (1)  'ilâ  jazâ'ir-a mutaqaddim-i 	dhikr-u-hâ
> 	to	islands	preceding		mentioning-NOM-their
> 		fem		masc			masc
> 		pl		sg			sg
> 		indef		indef			def
> 		gen		gen			nom
> 	to aforementioned islands'

Ahhhh ... multi-depth glosses ...

Have you considered stringing your attributes (fem, pl, indef, gen)
into one line (which is how I see them regularly done in linguistics
articles for the Cdn Jrnl of Ling, which I typeset).

The above would then revert to a 3-line entry of the model:

original language
morphemic analysis
gloss/translation

and look like this:

to      islands            preceding           mentioning-NOM-their
fem.pl.indef.gen   masc.sg.indef.gen   masc.sg.def.nom
to aforementioned islands'

OK ... so now we're down to 4 lines instead of three ...

I myself use the cm-lingmacros package, so the above would be coded up
as follows (and perhaps there's a clue to help you do something
similar with linguex?):

Notes: a. \Example is my version of cm-lingmacros' \enumsentence
b. \mc = \multicolumn

\Example{\shortex{4}
{'il\342 & \mc{2}{l}{jaz\342'ir-a mutaqaddim-i}  & dhikr-u-h\342}
{to      & islands          &  preceding         &  mentioning-NOM-their\\
& fem.pl.indef.gen &  masc.sg.indef.gen &  masc.sg.def.nom}
{to aforementioned islands'}
}

That is, there are 4 arguments to the \shortex macro:

#1 = number of columns (4, in this case)
#2 - original language (with one entry spanning 2 cols)
#3 = morphemic analysis (with a forced line break in the tabular)
#4 = gloss/translation

I just ran a test and it works out just fine.

Now, if you have to have each morphemic attribute on its own line, you
just follow the above model and use multiple line breaks within the
arg #3 (there's no reason they can't all be done within arg #2, but it
seems more logical to put all the same number of cols into one arg,
and let the spanning entry be in its own line/arg, #2, in this case).

> (I hope the tabs are not distorted too much...) I've tried groupings in the
> gloss with newlines in them, (i.e. {islands \\ fem \\ pl \\ indef \\ gen }
> etc.) but I was not surprised to see that did not work.
>
> I am now using a tabular environment to get the effect, but that puts the
> number of the example too low; not in front of the Arabic phrase, but
> before the line with pl   sg   sg'. Has anyone encountered and hopefully
> solved this problem?

If you can't use the above idea with linguex ... one trick I've found
with tabular inside \Example is to add the [t] option to it:

\Example{\begin{tabular}[t]{....}

\end{tabular}
}

Maybe that'll solve your problem with tabular?

Good luck!

> Joost Kremers

Ch.

### 29th October 1999: Re: Multi-line glosses

Subject: Re: multi-line glosses

> In some examples, I use multi-line glosses to indicate the features of the
> lexical items. However, I could not get linguex to produce such glosses. To
> give you an idea, this is what I want:
>
> (1)  'ilb  jazb'ir-a mutaqaddim-i     dhikr-u-hb
>       to      islands preceding               mentioning-NOM-their
>               fem             masc                    masc
>               pl              sg                      sg
>               indef           indef                   def
>               gen             gen                     nom
>       to aforementioned islands'

The gloss/cgloss4e package has a three-line variant, \glll, that allows you
two lines of aligned gloss instead of the usual one.  If that is not enough,
it is VERY straightforward to extend the macros to take as many lines of
gloss as you want.

Take a look at the source for cgloss4e (look at how \threesent differs
--minimally-- from \twosent, and how they are used in \gll and \glll,
respectively).  I'll even be happy to do this for you if you wish, it's really
just a few minutes' work.

A complication: Presumably you used the linguex command \bg., not \gll?
\bg internally calls \gll (from the cgloss4e package).  You can use, e.g.,
\b.\gll instead of \bg, but because of the design of linguex, if you do that
you cannot also use automatic sentence judgement facility (according to the
documentation, I did not test it).

(Alternative, not using cgloss):

>> I am now using a tabular environment to get the effect, but that puts the
>> number of the example too low;
> one trick I've found
> with tabular inside \Example is to add the [t] option to it:
>
>   \Example{\begin{tabular}[t]{....}

If Christina's hint does not for you, try enclosing the entire table in
a \vtop{...}.

Alexis

### 1st November 1999: Re: Multi-line glosses

From: doug@essex.ac.uk (Arnold D J)
Subject: Re: multi-line glosses

There are some hints about getting inter-linear glossing', as it is
sometimes called, with a couple of packages at the very slowly emerging LaTeX
for linguists' web site:

http://clwww.essex.ac.uk/latex4ling/

Also hints about how to do  other sort of thing that linguists have to
do.

Doug Arnold

>> In some examples, I use multi-line glosses to indicate the features of the
>> lexical items. However, I could not get linguex to produce such glosses. To
>> give you an idea, this is what I want:
>>
>> (1)  'ilb  jazb'ir-a mutaqaddim-i     dhikr-u-hb
>> to      islands preceding               mentioning-NOM-their
>> fem             masc                    masc
>> pl              sg                      sg
>> indef           indef                   def
>> gen             gen                     nom
>> to aforementioned islands'

### 1st November 1999: Fonts in LaTeX

From: Ashild Naess
Subject: Fonts in LaTeX

Hello people,

as a rather new and inexperienced user of LaTeX I'm having some trouble
figuring out how to use different fonts in LaTeX. I've tried to read up in
the manual and the Companion, but probably my knowledge is too basic for me
to make sense of any of it... What I need is the possibility of inserting
IPA characters into text. Would someone mind explaining *very* carefully
and thoroughly how to do this, totally from scratch? Just assume I don't
really know anything much about how LaTeX works, which is pretty close to
the truth anyway...

Also, hi to the guy who posted last week from the University of Nijmegen.
I'm going there in January, it seems, to do a PhD in linguistics. Is it a
nice place to be?

Ashild Naess
University of Oslo

### 1st November 1999: Re: Fonts in LaTeX

From: Thomas Martin Widmann
Subject: Re: Fonts in LaTeX

Ashild Naess  writes:

> as a rather new and inexperienced user of LaTeX I'm having some trouble
> figuring out how to use different fonts in LaTeX. I've tried to read up in
> the manual and the Companion, but probably my knowledge is too basic for me
> to make sense of any of it... What I need is the possibility of inserting
> IPA characters into text. Would someone mind explaining *very* carefully
> and thoroughly how to do this, totally from scratch? Just assume I don't
> really know anything much about how LaTeX works, which is pretty close to
> the truth anyway...

1) Install TIPA:
Get it from ftp://ftp.tex.ac.uk/tex-archive/fonts/tipa
-- I don't remember what I did next, but I think it's well
described in the manual.  If not please mail again!

2) I assume that you use \usepackage[T1]{fontenc} already.
Replace that with:
\usepackage[T3,T1]{fontenc}
\usepackage[safe,T1]{tipa}

3) In text you can now use all the commands described in the
TIPA manual, e.g. \textlengthmark to get the two triangles used
to mark a long sound.  You can also enter an IPA environment
within \textipa{...}.

/Thomas
- --
Thomas M Widmann  | Master's Student | Programmer  | Uni-parken 8, 2. v.333
viralbus@daimi.au.dk|Ling. & Comp. Sci.|  Stibo DS   | DK-8000 Århus C, Danio
+45  21 67 61 27  |Aarhus Universitet|tmw@ccieurope.com |  President/DK-TUG
| +45 87 33 44 65 / T³ONF/TÅGEKAMMERET

### 1st November 1999: Re: Fonts in LaTeX

From: Christina Thiele
Subject: Re: Fonts in LaTeX

Ashild Naess writes:
> Hello people,
>
> as a rather new and inexperienced user of LaTeX I'm having some trouble
> figuring out how to use different fonts in LaTeX. I've tried to read up in
> the manual and the Companion, but probably my knowledge is too basic for me
> to make sense of any of it... What I need is the possibility of inserting
> IPA characters into text. Would someone mind explaining *very* carefully
> and thoroughly how to do this, totally from scratch? Just assume I don't
> really know anything much about how LaTeX works, which is pretty close to
> the truth anyway...
>
> Also, hi to the guy who posted last week from the University of Nijmegen.
> I'm going there in January, it seems, to do a PhD in linguistics. Is it a
> nice place to be?
>
> Ashild Naess
> University of Oslo

In addition to the comments posted by Thomas (the one that looks like
it came from Dag ... how did that happen, Dag?!) Anyways ... in
colleague has a membership in TUG, look for the TUGboat journal
issues. In issue 17,2 (1996), you'll find Fukui Rei's documentation
for TIPA on pages 102--114. As well, you'll find that article
available in .pdf format via the TUG web pages (www.tug.org) and then
click on TUGboat > Tables of Contents > and then the issue number.

That gives you a head-start on documentation at least.

Ch.

### 1st November 1999: Re: Fonts in LaTeX

From: Mike Hammond
Subject: RE: Fonts in LaTeX

Ashild:

I too am pretty new to LaTeX, but I've gotten phonetic characters to work.
As another responder indicated, get the TIPA package from whichever archive
you find easiest. installing it is unfortunately not all that obvious,
especially for a newcomer. (I got our system administrators to do it for
the mainframe, but I had to do a lot of moving around to finally get it to
work under Miktex and Oztex.) It's worth it though. It has every possible
character/diacritic you could possibly want.

to use it:

1. install it.
2. make sure your document includes \usepackage{tipa} at the beginning.
3. the codes you have to enter are not at all obvious, but the
documentation is excellent. if you're a newbie like me, just print out the
postscript doc and pay attention to the very useful charts at the end.

mike h.

### 1st November 1999: Re: Fonts in LaTeX

From: krussll@CC.UMANITOBA.CA
Subject: RE: Fonts in LaTeX

There's not a lot of reason to use anything but TIPA.  To repeat Mike
Hammond's instructions:

> to use it:
>
> 1. install it.
> 2. make sure your document includes \usepackage{tipa} at the beginning.
> 3. the codes you have to enter are not at all obvious, but the
> documentation is excellent. if you're a newbie like me, just print out the
> postscript doc and pay attention to the very useful charts at the end.

I don't recall the original poster mentioning what kind of computer they
used.  Just in case they're using Windows and MikTeX, I'll elaborate step
1 a little:

1a.  Unzip all the files pretty much anywhere in your TeX directory.
Underneath \texmf\metafont is probably the best choice.

1b.  Tell MikTeX to go find it.
From a DOS shell, run
initexmf --update-fndb
(I think more recent versions of MikTeX put an entry to do this

MikTeX should automatically take care of the rest (e.g., having Metafont
automatically generate the necessary files the first time you use the
font).

Updating the filename database (step 1b) is something you should do
every time you add a new font or a new package to MikTeX.  It took me
forever to figure it out myself, but ever since I made it a habit, I've

Good luck.

-- Kevin

### 1st November 1999: Re: Fonts in LaTeX

From: "Benjamin T. Bruening"
Subject: RE: Fonts in LaTeX

> There's not a lot of reason to use anything but TIPA.

> 3. the codes you have to enter are not at all obvious,

I've been using WSUIPA, which I downloaded from the same place you can get
TIPA.  The output seems to be comparable to TIPA, and the codes you enter
*are* obvious, at least to me.  Once I got the thing installed, WSUIPA
worked great (all the instructions I've seen mailed to the list about TIPA
apply to WSUIPA), and after a few minutes I didn't even need to refer to
the documentation to be able to get the symbols I wanted---the codes are
completely transparent.  So the statement in 3 above, if correct about
TIPA, seems to be a good reason to use something else.

Benjamin

-------------------------------------------
Benjamin Bruening
Dept. of Linguistics
MIT E39-245
77 Massachusetts Ave.
Cambridge, MA 02139

### 1st November 1999: Re: Fonts in LaTeX

Subject: RE: (IPA) Fonts in LaTeX

Could someone who's dealt with them post a quick summary of the differences/
genealogical relationships between WSUIPA, TSIPA, and TIPA?  How do they
compare in terms of features and coverage of symbols?  Is there a clear
winner that everyone should switch to?

According to the TSIPA documentation, it was written to include IPA symbols
new to the 1989 standard, which WSUIPA did not include (at that point, at
least).  Anyone know what the current situation is?

Alexis

### 1st November 1999: Re: Fonts in LaTeX

From: Christina Thiele
Subject: Re: (IPA) Fonts in LaTeX

>
> Could someone who's dealt with them post a quick summary of the differences/
> genealogical relationships between WSUIPA, TSIPA, and TIPA?  How do they
> compare in terms of features and coverage of symbols?  Is there a clear
> winner that everyone should switch to?
>
> According to the TSIPA documentation, it was written to include IPA symbols
> new to the 1989 standard, which WSUIPA did not include (at that point, at
> least).  Anyone know what the current situation is?
>
> Alexis

1. WSUIPA
---------
I helped with the development of WSUIPA back in the late 80s, early
90s. It was designed by Janene Winter at Washington State University
(WSU), and we used the Pullum and Ladusaw _Phonetic Symbol Guide_
(first ed), which had appeared shortly before Janene and I began
corresponding (thanks to intros from Dean Guenther). So WSUIPA
contains pretty much everything in P & L, along with a very few
additional oddities which I needed at the time for Algonquian
linguistics.

There's an article on WSUIPA's development in TUGboat 12,1 (1991),
pages 149--156; what I'd been using before WSUIP was ph10, shown in
Figs. 10 and 11 (bitmapped fonts, designed at Carleton University).

Anyways, that's WSUIPA. It used to be part of the TeXT1 distribution
(set of macros from WSU) but it was eventually made available as a
separate item. But it was frozen at that time, and so did not see any
upgrade following the Kiel changes of 1989. That's the biggest
difference, then, between WSUIPA and both TSIPA and TIPA. It's old ;-)

2. TSIPA/TIPA
-------------
Fukui Rei's TUGboat article (the one I'd referred to earlier
-- TUGboat 17,2 (1996), pages 102--114) says in the intro para:

TIPA is a system for processing IPA (International Phonetic
Alphabet) symbols in LaTeX. It is based on TSIPA* but both
METAFONT source codes and LaTeX macros have been thoroughly
reworked so that it can be considered as a new system.

* TSIPA was made in 1992 by Kobayashi Hajime, Fukui Rei
and Shirakawa Shun. It is available from a CTAN archive.
One problem with TSIPA was that symbols arleady included in
OT1, T1 or Math fonts are excluded, because of the limitation
of its 128 character encoding. As a result, a string of phonetic
representation had to be often composed of symbols from different
fonts, disabling the possibility of automatic inter-word
kerning. And also too many symbols had to be realised as macros.

What we don't know is what, if anything, TSIPA itself was based upon.

As for any further IPA upgrades/changes, I have no further information
as I no longer actually read any of the phonetics stuff ... I just
typeset it ;-)

But there are plenty of linguists on the list who should be able to
say whether or not the Kiel changes of 1989 are the latest ... I
haven't seen any comment here to say otherwise.

Hope this helps give a start to the archaeology ;-)

Ch.

### 1st November 1999: Re: Fonts in LaTeX

From: Johannes Reese
Subject: Re: Fonts in LaTeX

> 2) I assume that you use \usepackage[T1]{fontenc} already.
>    Replace that with:
>    \usepackage[T3,T1]{fontenc}
>    \usepackage[safe,T1]{tipa}
^^^^^^^^

What is this for? Is it the solution for a problem I have?:

I have tried to use tipa together with the ae- and times-packages. This
makes an option clash, as tipa urges T3 and T1(?)-encodings, but I want
ae, not T1, taking away the ability to use postscript-fonts.

Wishing you can help

Johannes

### 1st November 1999: Re: Fonts in LaTeX

Subject: Re: Fonts in LaTeX
From: "Todd O'Bryan"

See below for hints about TIPA...

- ----------
>From: Mike Hammond
>To: "'Ashild Naess'" , "ling-tex@ifi.uio.no"

>Subject: RE: Fonts in LaTeX
>Date: Mon, Nov 1, 1999, 8:59 AM
>

> Ashild:
>
> I too am pretty new to LaTeX, but I've gotten phonetic characters to work.
> As another responder indicated, get the TIPA package from whichever archive
> you find easiest. installing it is unfortunately not all that obvious,
> especially for a newcomer. (I got our system administrators to do it for
> the mainframe, but I had to do a lot of moving around to finally get it to
> work under Miktex and Oztex.) It's worth it though. It has every possible
> character/diacritic you could possibly want.
>
I just installed TIPA for the second time (on OzTeX) and, like Mike, had to
try several different folders for storing things to finally get things to
work out. Let's face it, (La)TeX is just not intuitive when it comes to
getting things set up.

> to use it:
>
> 1. install it.
> 2. make sure your document includes \usepackage{tipa} at the beginning.
> 3. the codes you have to enter are not at all obvious, but the
> documentation is excellent. if you're a newbie like me, just print out the
> postscript doc and pay attention to the very useful charts at the end.
>
It's better (I think) to avoid the long codes that start with backslashes,
if possible. (These are listed as Input1: in the charts of symbols.) If you
create a TIPA environment (with the command \textipa{}) the codes you type
in the brackets are automatically converted to those obnoxious codes for
you. These codes (listed as Input2: in the charts) *are* pretty
obvious--they're what everybody uses when they try to type IPA in email: @
for schwa, N for engma, S for esh, etc. I've also found that when I'm
working with a language that has a lot of an occurrence of a character that
doesn't have a short code, I just define it. For example, I've defined \tS
to be the t-esh ligature I'd prefer to use for the alveo-palatal affricate.
To do that, you'd just type:

\newcommand{\tS}{\textteshlig}

In the rest of my document, whether in plain text or a \textipa environment,
\tS gives me the symbol I want.

Fukui Rei, if you're out there listening, thank you very much for a very
elegant and useful package. The fonts are beautiful and the interface is as
simple as it could be given how complicated the problem is. It's certainly
made the stuff I produce look far nicer than people can do with TrueType
fonts.

Todd O'Bryan

### 2nd November 1999: Re: Fonts in LaTeX

From: Christina Thiele
Subject: Re: Fonts in LaTeX

Benjamin T. Bruening writes:
> >There's not a lot of reason to use anything but TIPA.
>
> > 3. the codes you have to enter are not at all obvious,
>
> I've been using WSUIPA, which I downloaded from the same place you can get
> TIPA.  The output seems to be comparable to TIPA, and the codes you enter
> *are* obvious, at least to me.  Once I got the thing installed, WSUIPA
> worked great (all the instructions I've seen mailed to the list about TIPA
> apply to WSUIPA), and after a few minutes I didn't even need to refer to
> the documentation to be able to get the symbols I wanted---the codes are
> completely transparent.  So the statement in 3 above, if correct about
> TIPA, seems to be a good reason to use something else.
>
> Benjamin
> ...

The main reason to use TIPA is that it's consistent with the IPA's
symbols, and that WSUIPA was not upgraded to conform to the 1989 Kiel
changes.

I think there's also a case to be made for the shapes being of better
quality in TIPA than WSUIPA. A few of them aren't what I learned when
I did all those linguistics courses ;-) but still ... I think overall,
one is better off with TIPA.

Of course, I'm still using WSUIPA but soon ... soon ... Fukui, I
promise, I will switch to TIPA and 2e for the linguistics journals
;-))

Ch.

### 2nd November 1999: Re: Fonts in LaTeX

From: "J.Kremers"
Subject: RE: Fonts in LaTeX

Why do you propose to install TIPA in the \texmf\metafont directory? MikTeX
uses a \localtexmf directory and it is recommended that you put all files
you add yourself in that directory. That way they will not be overwritten
if you ever update your MikTeX. So would it not be better to put the TIPA
files in \localtexmf?

(Actually, I do not know if it is really that big a deal. At home I use
teTeX under Linux, which does not use a /localtexmf directory. So...)

By the way: I use MikTeX 1.20 and it does indeed add an option to update
the filename database in the start menu:
Start|Programs|MikTex|Maintenance|Refresh Filename Database.

Joost Kremers

>There's not a lot of reason to use anything but TIPA.  To repeat Mike
>Hammond's instructions:
>
>> to use it:
>> 1. install it.
>> 2. make sure your document includes \usepackage{tipa} at the beginning.
>> 3. the codes you have to enter are not at all obvious, but the
>> documentation is excellent. if you're a newbie like me, just print out the
>> postscript doc and pay attention to the very useful charts at the end.
>
>I don't recall the original poster mentioning what kind of computer they
>used.  Just in case they're using Windows and MikTeX, I'll elaborate step
>1 a little:
>
>  1a.  Unzip all the files pretty much anywhere in your TeX directory.
>       Underneath \texmf\metafont is probably the best choice.
>
>  1b.  Tell MikTeX to go find it.
>       From a DOS shell, run
>           initexmf --update-fndb
>       (I think more recent versions of MikTeX put an entry to do this
>
>MikTeX should automatically take care of the rest (e.g., having Metafont
>automatically generate the necessary files the first time you use the
>font).
>
>Updating the filename database (step 1b) is something you should do
>every time you add a new font or a new package to MikTeX.  It took me
>forever to figure it out myself, but ever since I made it a habit, I've
>
>Good luck.
>-- Kevin
----------------------------------------
Joost Kremers
Katholieke Universiteit Nijmegen
Talen en Culturen van het Midden-Oosten
Erasmusplein 1
6525 HT Nijmegen
tel: 024-3612996
fax: 024-3611972

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: FUKUI Rei
Subject: Re: (IPA) Fonts in LaTeX

> What we don't know is what, if anything, TSIPA itself was based upon.

> Hope this helps give a start to the archaeology ;-)

Well, let me talk about some old story.

TSIPA was written from scratch (but of course based upon cm
fonts). At that time we knew of WSUIPA. The main reason for
creating a new font was that we needed more symbols and we
found some of the WSUIPA symbols didn't look good (I mean no
harm). About the three authors of TSIPA, Kobayashi Hajime
was (and still is) working for a printing company and he was
very good at TeX and Metafont. He was trying to use to TeX
in typesetting textbooks for high school students and he
needed IPA fonts for English textbooks. Shirakwa Shun was a
graduate student of linguistics and he was working for a
publishing company as a part time job. I was a lecturer in
linguistics. I always wanted to create a new set of IPA
fonts but too busy to do so myself.

In fact, when I was a student, I made a small typesetting
program and a set of fonts by myself, not knowing anything
about TeX. Some years later I became an assistant in
linguistics and used some time in typesetting a linguistic
working paper with my program. Shirakara was one of the few
students who used and maintained that program. One day, I
don't remember when, TeX appeared before me and I stopped
doing programming myself. Shirakawa and I started to learn
TeX and joined a local TeX users meeting. In one of the
application of vf fonts for commercial publishing and after
that Shirakawa and I talked with him and agreed to create a
new IPA font.

That was a pre-history of TSIPA.

Fukui Rei

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: "Harold F. Schiffman"
Subject: Re: (IPA) Fonts in LaTeX

For what it's worth, we used TSIPA for phonetic material in a book just
Spoken Tamil.   The rest was typeset with LaTeX, completely
camera-ready copy; tsipa worked just fine for the phonetics.  We used gb4e
for interlinear glossing (up to 3 lines sometimes: Tamil font, spoken
Tamil, glosses) and that worked just fine.

Hal Schiffman

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: Christina Thiele
Subject: Re: (IPA) Fonts in LaTeX

FUKUI Rei writes:
>
> > What we don't know is what, if anything, TSIPA itself was based upon.
>
> > Hope this helps give a start to the archaeology ;-)
>
> Well, let me talk about some old story.

> TSIPA was written from scratch (but of course based upon cm
> fonts). At that time we knew of WSUIPA. The main reason for
> creating a new font was that we needed more symbols and we
> found some of the WSUIPA symbols didn't look good (I mean no
> harm). About the three authors of TSIPA, Kobayashi Hajime

No offence taken at all, Fukui. No problem. Some of our strategies
were no more sophisticated than turning a character upside down ;-)
Sometimes it worked ... sometimes it doesn't. And I do think that the
TIPA shapes, in the main, are nicer ;-)

> was (and still is) working for a printing company and he was
> very good at TeX and Metafont. He was trying to use to TeX
> in typesetting textbooks for high school students and he
> needed IPA fonts for English textbooks. Shirakwa Shun was a
> graduate student of linguistics and he was working for a
> publishing company as a part time job. I was a lecturer in
> linguistics. I always wanted to create a new set of IPA
> fonts but too busy to do so myself.
>
> In fact, when I was a student, I made a small typesetting
> program and a set of fonts by myself, not knowing anything
> about TeX. Some years later I became an assistant in
> linguistics and used some time in typesetting a linguistic
> working paper with my program. Shirakara was one of the few
> students who used and maintained that program. One day, I
> don't remember when, TeX appeared before me and I stopped
> doing programming myself. Shirakawa and I started to learn
> TeX and joined a local TeX users meeting. In one of the
> application of vf fonts for commercial publishing and after
> that Shirakawa and I talked with him and agreed to create a
> new IPA font.
>
> That was a pre-history of TSIPA.
>
> Fukui Rei
>

This is great stuff! Many thanks for telling us.

Actually, the whole WSUIPA history began (for me) when I gave a paper
on TeX, Linguistics, and Journal Production' at the 1987 TUG meeting
in Seattle, Washington. I met Dean Guenther, who was coordinator or
some such for the courses before the meeting and seemed to help out a
lot during the conference as well. Anyways, I later learned that he'd
gone back to WSU and talked with Janene (she was working for him, I
and TeX and said the only thing left that needed improvement was a
real phonetic font.

Well, Janene had been playing with Metafont and however it happened,
fell on the IPA as a likely subject for MF. The back-history at WSU is
described in that TUGboat article (a conference paper, actually;
TUGboat 12,1 pages 149--150 for the background).

Anyways, she eventually got in touch with me -- that would probably be
some time in 1988, I think -- and we began corresponding via e-mail. I
across and thought was brilliant -- it showed what, for us, was an
x-height! This was perfect. She got herself a copy (it might have even
been in their library, as it had been published in 1986) and started
to really work with it. I would get samples via e-mail, run them off,
and then go into great long verbose (nothing's changed!) descriptions
of how I thought the shapes might be improved. Very hard to do via
e-mail. I knew the shapes ... she knew MF ... and we both struggled
with adjectives and verbs to get the two to mesh. I have a whole
boxfull of hardcopy from our correspondence on the subject!

As things were starting to come together, I convinced my editor that
this would be a good thing to really help support and finalise, so I
got a flight paid out to WSU, and Janene and I worked on the fonts for
3 or 4 days straight. I think we were the funniest thing around: she'd
be peering at the screen and all the MF code (Greek to me!) and then
here I'd be, flailing my hands and arms around, trying to show the up-
or downward swoop of a stroke, or the roundness of a curve, or
whatever ... and then she'd look back at the MF code, and try to take
the flailing arm movement and figure out some corresponding MF code
;-) ... then we'd both have a good laugh at our reflections in the
window!

One thing we had to deal with was: how to arrange the characters in
the grid -- how to assign them a place in a font-table array as found
in the TeXbook, Section F. Arrange by shape and look, or arrange by
phonetic similarities ... we settled on ordering them by shape: that
is, the a-like shapes were first, the b-like shapes next, and so
on. My argument was that someone keyboarding IPA material (and so
often, texts will have just a few bits here and there, not great long
paragraphs of IPA chars) would use the chart and without any phonetics
knowledge, identifying chars based on their phonetic traits would just
be too hard -- easier for everyone, linguist or not, to just look for
a-like shapes till they found the inverted print-a, or looked through
all the b-like shapes till they found beta ...

Next non-MF problem ... I think it was around the 3rd day that I
suddenly realised that maybe we needed to start thinking about how to
access these new characters ... you can't really do {\ipa\char'000'}
every time you want to use an IPA character in a file!! So I went
through the P & L, and basically used the characters names found there
to build a list of macro names to eventually tie in to each char on
the chart. What was handy was that P&L also organised the IPA chars in
order of shape rather than phonetic traits (whether this is a common
enough approach, I don't know -- we came to it on our own ... or maybe
all those perusings of P&L already predisposed us to the sort-by-shape
approach ... hmmm). At any rate, Janene then took the draft macro
names and eventually, once I was back in Ottawa, we started testing
them and using them.

One goofy thing was my decision to use a short-form for the ejective:
\eject ;-))) I believe this one lasted 3 or 4 days before one of her
other respondents/testers suggested this wasn't ideal ... pages kept
being thrown out of the printer!! Oh well ... ;-)

At some point she had also began working with some other people, again
mainly via e-mail, I think, so she was having to distill all our
comments, all our probably very different ways of describing things
... and eventually the WSUIPA fonts came out. Tough job.

But it was truly a lot of fun working on the WSUIPA -- probably
because I could see that CJL would great benefit from the work, but
also because it was fun to work with someone like Janene.

When we were finally able to start using them in production for CJL
(Cdn. Jrnl. of Ling.), it was a real thrill! Up till then, we only had
a roman 10pt and 9pt IPA to work with: we had to avoid italics (not
that hard) and above all -- never any IPA in the 17pt titleblock!!
With the WSUIPA fonts, we could at last allow IPA in titles! It was
just gorgeous ... or so we thought! Well, yes! For the time, it _was_
a great step forward. That's why I'm never embarrassed by old material
-- it shows just how limited one's resources and skills and knowledge
were a long long time ago ... and hopefully one's current efforts show
a great deal more skill and knowledge and access to better resources
;-)

But the WSUIPA was a big stepping stone beyond the early bitmapped
IPA, which had to be designed quickly and used, whatever the
constraints, as we simply had to have some kind of IPA for the
linguistics journal (we used ph10 and ph9 from 1983 till probably
1988/89).

So, lots of history in these old fonts ... and we each are probably
very attached to the ones that were available when we started using
them. Nevertheless, I am looking forward to using the full set of TIPA
fonts and macros in the near future.

====

But there was/is another old IPA floating around ... I have this
notion that Emma Pease also worked on a phonetic fonts for ?CSLI
perhaps? Maybe she's reading these archival anecdotes and has some
info on that font ...

Ch.

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: "Todd O'Bryan"
Subject: Re: (IPA) Fonts in LaTeX

I know there's a 1993 revision to the IPA which is somehow different from
1989. (I don't know how it's different, of course.) TIPA includes those 1993
symbols.

Todd

- ----------
>From: Christina Thiele
>
>
> But there are plenty of linguists on the list who should be able to
> say whether or not the Kiel changes of 1989 are the latest ... I
> haven't seen any comment here to say otherwise.
>
> Hope this helps give a start to the archaeology ;-)
>
> Ch.

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: Emma Pease
Subject: Re: (IPA) Fonts in LaTeX

> But there was/is another old IPA floating around ... I have this
> notion that Emma Pease also worked on a phonetic fonts for ?CSLI
> perhaps? Maybe she's reading these archival anecdotes and has some
> info on that font ...

Yup, mine was also independent (started 1986) and haphazard (I
certainly used the rotate certain characters trick).  I started mostly
because someone here was doing a Hausa/English dictionary in TeX and
asked my help with the TeXing and pre-processing (perl would have been
very useful in those days).  She was using superimposed characters to
get the non-Roman characters and they looked awful so I insisted on
metafonting the special characters.  From there I created characters
for the local linguists.

I haven't worked on mine for years and it never had the complete IPA;
however, I suspect it is one of the few that has capital hooked D, K,
B (all used in Hausa).

Emma

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: Thomas Martin Widmann
Subject: Re: (IPA) Fonts in LaTeX

Emma Pease  writes:

> I haven't worked on mine for years and it never had the complete IPA;
> however, I suspect it is one of the few that has capital hooked D, K,
> B (all used in Hausa).

It sounds like it could be the basis of an African font, rather than
a phonetics one...

/Thomas
--
Thomas M Widmann  | Master's Student | Programmer  | Uni-parken 8, 2. v.333
viralbus@daimi.au.dk|Ling. & Comp. Sci.|  Stibo DS   | DK-8000 Århus C, Danio
+45  21 67 61 27  |Aarhus Universitet|tmw@ccieurope.com |  President/DK-TUG
| +45 87 33 44 65 / T³ONF/TÅGEKAMMERET

### 2nd November 1999: Re: (IPA) Fonts in LaTeX

From: Christina Thiele
Subject: Re: (IPA) Fonts in LaTeX

Emma Pease writes:

> ...
> Yup, mine was also independent (started 1986) and haphazard (I
> certainly used the rotate certain characters trick).  I started mostly

Aha! Actually, this rotate or flip or mirror trick was how a number of
the old ph10 characters were devised -- they were just clones of the
bitmapped CM fonts ;-)

God ... the things one does just to have a character or two ;-))

> because someone here was doing a Hausa/English dictionary in TeX and
> asked my help with the TeXing and pre-processing (perl would have been
> very useful in those days).  She was using superimposed characters to
> get the non-Roman characters and they looked awful so I insisted on
> metafonting the special characters.  From there I created characters
> for the local linguists.

Ah ... that would be Roxana Ma Newman's book, no? I have it listed in
the first (and so far only) TUGboat TeX and the Humanities column. I
really really should get the next part done ... ;-( Well, once these
proceedings are finished, maybe I'll have more time ... yeah, right!!

One can dream, though ...

> I haven't worked on mine for years and it never had the complete IPA;
> however, I suspect it is one of the few that has capital hooked D, K,
> B (all used in Hausa).
>
> Emma

Ch.

### 3rd November 1999: Re: (IPA) Fonts in LaTeX

From: Joerg Knappen
Subject: Re: (IPA) Fonts in LaTeX

Viralbus scripsit:

>> however, I suspect it is one of the few that has capital hooked D, K,
>> B (all used in Hausa).

> It sounds like it could be the basis of an African font, rather than
> a phonetics one...

In fact it is, I borrowed from this font while making the fc fonts. The fc
fonts aren't full ipa fonts, on the other hand they contain upper case letters
and italics for african languages (plus maltese and sami).

--J"org Knappen

### 9th November 1999: Speech Recognition Theory

From: Batzorigt Rentsen
Subject: from Batzorigt (student at National University of Mongolia)

Hi, My name's Batzorigt. I study at National
University of Mongolia  . I'm working at the project
named Speech Recognition because I very interest in
this topic. I want do it very much, but I don't know
Speech Recognition Theory (books, handbooks, methods,
techniques) and technology. I think that I'll write
program which associates this project in C++ language.
Send please materials (sample programms, source code
and SDK Help) which associate this project by
following address: National University of Mongolia,
Electronic-4, Batzorigt Rentsen.

### 23rd November 1999: Aligning glosses

From: "Todd O'Bryan"
Subject: Aligning glosses

Hey all!

I'm having a problem getting what I want to happen to happen with multi-line
glosses. Maybe someone has a suggestion...

I'm using gb4e.sty and trying to line up the following lines:

1) a line of running text in a language (in this case Malagasy)
2) a morphological analysis of that text
3) glosses for line 2
4) a free translation

I'm using the \glll command, and let me demonstrate my problem (use a
monospaced font to get the effect):

This is what I'd like:
sasan-dRasoa            ny  zaza
sasa- ana-     dRasoa   ny  zaza
wash- PASSIVE- by.Rasoa the child
'The child was washed by Rasoa.'

This is what I get:
sasan-dRasoa                   ny  zaza
sasa-        ana-     dRasoa   ny  zaza
wash-        PASSIVE- by.Rasoa the child
'The child was washed by Rasoa.'

Does anybody have a cute trick for keeping the automatic alignment feature
from pushing over the glosses to after the first word of line 1? I tried
putting everything on lines 2 and 3 that corresponded to a single word in
line 1 in brackets, but then I lose the alignment of the morphemes and
glosses. I suppose what I get isn't awful, but you'd think there would be a
way to do this.

Todd

### 23rd November 1999: Re: Aligning glosses

From: Christina Thiele
Subject: Re: Aligning glosses

Todd O'Bryan writes:
> ...
> I'm using gb4e.sty and trying to line up the following lines:

I don't use gb4e.sty but ... I may have a trick for you ...

> ...
>
> This is what I'd like:
> sasan-dRasoa            ny  zaza
> sasa- ana-     dRasoa   ny  zaza
> wash- PASSIVE- by.Rasoa the child
> 'The child was washed by Rasoa.'
>
> This is what I get:
> sasan-dRasoa                   ny  zaza
> sasa-        ana-     dRasoa   ny  zaza
> wash-        PASSIVE- by.Rasoa the child
> 'The child was washed by Rasoa.'

To avoid having TeX measure the length of the 1st line element
sasan-dRasoa', hide most of it ;-)

\rlap is what I use all the time in order to have TeX print but don't
measure'; \llap's the same but for stuff on the left side that you
won't want to have included in column width measurements.

Sooo ... I'd say, try this:

sasan-\rlap{dRasoa}                   ny  zaza

... and I'm assuming that to set up your alignment columns, you use
something like & to separate them, right?

> Does anybody have a cute trick for keeping the automatic alignment feature
> from pushing over the glosses to after the first word of line 1? I tried
> putting everything on lines 2 and 3 that corresponded to a single word in
> line 1 in brackets, but then I lose the alignment of the morphemes and
> glosses. I suppose what I get isn't awful, but you'd think there would be a
> way to do this.

My mantras:

a. \llap and \rlap mean print but don't measure'

b. \phantom means measure but don't print'

Without 'em, I'd never get tree diagrams to fit on a page! Or do the
same thing you're trying with alignments that need to be under
something else but at the width of that something else ...

> Todd

Send word that it's worked! Or that it hasn't ;-(

Ch.

### 23rd November 1999: Re: Aligning glosses

From: Emma Pease
Subject: Re: Aligning glosses

I also find that defining the following

% center lap creates a box of 0 width centered on the point
\def\clap#1{\hbox to 0pt{\hss#1\hss}}

creates a useful command.  It is like \rlap and \llap except it
centers the text on the current location.

Emma

ps. It is defined in the Stanford lingmacros.sty file.

### 23rd November 1999: Re: Aligning glosses

From: Christina Thiele
Subject: Re: Aligning glosses

Emma Pease writes:
> I also find that defining the following
>
> % center lap creates a box of 0 width centered on the point
> \def\clap#1{\hbox to 0pt{\hss#1\hss}}
>
> creates a useful command.  It is like \rlap and \llap except it
> centers the text on the current location.
>
> Emma
>
> ps. It is defined in the Stanford lingmacros.sty file.

Sounds like a handy little tool to have -- esp. if the other glosses
are centre-aligned.

Ch.

### 24th November 1999: Re: Aligning glosses

From: "Todd O'Bryan"
Subject: RE: Aligning glosses

Thanks to Christina Thiele, this problem is solved...

I'm using gb4e.sty and trying to line up the following lines:

This is what I'd like:
sasan-dRasoa            ny  zaza
sasa- ana-     dRasoa   ny  zaza
wash- PASSIVE- by.Rasoa the child
'The child was washed by Rasoa.'

To get this, use an \rlap{} command around the first word of the first line,
and then use empty brackets to save space for the other morphemes, like so:

\glll \rlap{sasan-dRasoa} {} {} ny zaza\\
sasa- ana- dRasoa ny zaza\\
etc.

Thanks to everyone who responded. Hopefully, Christina's insight will help
other people, too.

Todd

### 24th November 1999: Re: Aligning glosses

From: Daniel Glatz
Subject: RE: Aligning glosses

Hi everybody,

I didn't keep track of the discussion, sorry. But did anybody mention
covingto.sty? I guess it is much better than gb4e.sty ...

Greetings from  Heidelberg,
Dan

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Daniel Glatz
Center for Computational Linguistics
University of Heidelberg
Karlstrasse 2
D-69117 Heidelberg
Germany
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

### 24th November 1999: Re: Aligning glosses

From: Christina Thiele
Subject: Re: Aligning glosses

Daniel Glatz writes:
>
> Hi everybody,
>
> I didn't keep track of the discussion, sorry. But did anybody mention
> covingto.sty? I guess it is much better than gb4e.sty ...
>
> Greetings from  Heidelberg,
> Dan
>
> ...

Now, now ... don't go dumping on someone else's preferences ;-))

I myself use neither ;-)

I think I just looked at all the descriptions in that inventory of so
long ago (lingmac.tex in tex-archive/info/ on CTAN) and chose
cm-lingmacros.sty, which is Chris Manning's variant of
covingto.sty. But I still need to make use of \llap and \rlap in order
to have overlong material print while not being used to determine
column width. I don't think this is a typesetting problem that's
specific to any given package -- it's a fact of basic TeX life ;-)

And, as I mentioned yesterday, I also employ both lap's' when
devising tree diagrams (using Emma Pease' tree-dvips.sty) to balance
out branches to nodes where one node has more text than the other. For
example, if I have

C

Comp    V

I'll eyeball the text and figure that the middle lowercase om' are
only slightly wider than cap-V and cap-C, so I'll do this

\llap{C}om\rlap{p}

so that the two diagonal lines are almost symmetric as they go down
from C to Comp and C to V.

I still prefer the {tabular} approach to setting trees, but some day
I'll have to give Avery Andrews' trees' preprocessor a try. It's like
a front end to Emmas's tree-dvips'.

Ch.

### 26th November 1999: Re: Aligning glosses

From: Avery Andrews
Subject: Re: Aligning glosses

My package gets has commands for modifying the default mother daughter and
sister sister spacing when you want to, one thing it doesn't support
is putting all the terminals on one line at the bottom.  But since it's
based on Emma's package, you can use {tabulars} whenever you want to.

The most important thing the preprocessor does is generate the line-drawing
commands, I found that to be extremely tedious.  Anyway all the
trees in Andrews & Manning 1999 (Complex Preds and Info Spreading) were
done with it, and the pikkies on pg 78-79 illustrate combined operation of
the trees & Chris' avm.sty.  I'll confess however that the circles around
the nodes on pg 5 were done with great difficulty, & I couldn't be
bothered to do more of them.

> > Christina Wrote:
> >
> > > I'll eyeball the text and figure that the middle lowercase om' are
> > > only slightly wider than cap-V and cap-C, so I'll do this
> > >
> > >      \llap{C}om\rlap{p}
> > >
> > > so that the two diagonal lines are almost symmetric as they go down
> > > from C to Comp and C to V.
> > >
> > > I still prefer the {tabular} approach to setting trees, but some day
> > > I'll have to give Avery Andrews' trees' preprocessor a try. It's like
> > > a front end to Emmas's `tree-dvips'.
> > >
> >
> > Actually it's a front end to both tree-dvips and my trees.sty, the latter
> > does the real work of working out the spacing, while tree-dvips manages the
> > line-drawing.
> >
> >   - Avery.Andrews@anu.edu.au
>
> Hi, Avery! Well -- whatever it's used as a front end to, it's still
> something pretty neat and something that I haven't yet tried. I think
> I like the control I think I have by doing this directly into
> {tabular} but ... I'd have to try it out before saying anything more
> ;-)
>
> Ch.
>
> P.S. Your reply only came to me -- maybe it should also go to
>      ling-tex, so that people have the more correct info ...

### 7th December 1999: NELS proceedings

From: Dag Langmyhr
Subject: NELS proceedings

Hi all.  Has any of you ever made or come across a LaTeX style file
for a paper for the NELS proceedings?  I imagine it won't be too much
work to put one together, but if someone has already done the work....

Thanks,
Bob Frank

###################################################################
Bob Frank
Department of Cognitive Science		410-516-8699
Johns Hopkins University		       -8020 (fax)
3400 N. Charles Street
Baltimore, MD 21218
http://www.cog.jhu.edu/faculty/rfrank.html

Last updated 1999/12/17 by Dag Langmyhr.