Underlying forms

61 views
Skip to first unread message

Lena Terhart

unread,
Sep 26, 2008, 7:34:59 AM9/26/08
to Shoebox/Toolbox Field Linguist's Toolbox
Hi there,

after Toolbox support got my project settings right again (many thanks
for that!!), I started interlinearizing. Now I encountered a new
problem: there are several phonological changes in the language and I
want to deal with these using underlying forms. To exemplify this: I
have a word "nkasiaper", which should break into "ni=kasia-pa=ro". So
for the dictionary entry: "-pa" I definded:

\lx -pa
\va -per
\u -pa =ro{3SGm}

Anyway, I only get the result "ni=kasia-pa" (among other choices
including another lexeme "per"). I also tried combinations of "-pa"
and "=ro" including and excluding spaces and + in the \u-field, so
that is not the problem.
I suppose it doesn't work because the \u-function was totally deleted
from my list of MDF-markers. I included it again, but did so by only
typing the marker's abbreviation and name. I suppose that doesn't
suffice to get it work as it should...
How do I get back the original marker?

I also have a minor problem concerning clitics: I defined the = as a
morpheme break character and it works well in the text files. But in
the dictionary it seems to be interpreted as a letter, so the clitic
"=ro" appears at the end of the dictionary instead of under the letter
"R". How can I change this?

(Note that the second problem is not related to the first one, I also
tried by changing the "=" to a simple"-" and had the same problem.)

Thanks in advance!
Lena

Alex Francois

unread,
Sep 26, 2008, 8:09:34 AM9/26/08
to ShoeboxToolbox-Fiel...@googlegroups.com
Hi Lena,

I'm not sure this is going to solve the problem, but I think one part of it lies in the fact that you're using the wrong field code for coding the allomorph. Instead of \va which codes for "Variant", you should use \a (for "allomorph"):  as far as I can remember, this is the fieldname which Toolbox is designed to work with in its parsing process.  \va is no more than a text field designed to be printed in the Mdf dictionary output.

hope this helps,

Alex.


Lena Terhart a écrit :
-- 
*******
Alex François
http://alex.francois.free.fr

Lena Terhart

unread,
Sep 26, 2008, 9:20:17 AM9/26/08
to Shoebox/Toolbox Field Linguist's Toolbox
Dear Alex,

thanks for the advise, but unfortunately that doesn't help. There is
no \a-marker in the list to choose and if I add it as a new one and
use it for "-per", even the result "ni=kasia-pa" is lost. I saw that
the marker is called \a in Toolbox Help, but I thought that it would
just be the one called \va in my project (Toolbox version 1.5.3).
Maybe I should add that the project was originally created in Shoebox.
May that be why there are other markers???

Lena

On 26 Sep., 14:09, Alex Francois <alex.franc...@free.fr> wrote:
> Hi Lena,
> I'm not sure this is going to solve the problem, but I think one part of it lies in the fact that you're using the wrong field code for coding the allomorph. Instead of \va which codes for "Variant", you should use \a (for "allomorph"):  as far as I can remember, this is the fieldname which Toolbox is designed to work with in its parsing process.  \va is no more than a text field designed to be printed in the Mdf dictionary output.
> hope this helps,
> Alex.Lena Terhart a écrit :Hi there, after Toolbox support got my project settings right again (many thanks for that!!), I started interlinearizing. Now I encountered a new problem: there are several phonological changes in the language and I want to deal with these using underlying forms. To exemplify this: I have a word "nkasiaper", which should break into "ni=kasia-pa=ro". So for the dictionary entry: "-pa" I definded: \lx -pa \va -per \u -pa =ro{3SGm} Anyway, I only get the result "ni=kasia-pa" (among other choices including another lexeme "per"). I also tried combinations of "-pa" and "=ro" including and excluding spaces and + in the \u-field, so that is not the problem. I suppose it doesn't work because the \u-function was totally deleted from my list of MDF-markers. I included it again, but did so by only typing the marker's abbreviation and name. I suppose that doesn't suffice to get it work as it should... How do I get back the original marker? I also have a minor problem concerning clitics: I defined the = as a morpheme break character and it works well in the text files. But in the dictionary it seems to be interpreted as a letter, so the clitic "=ro" appears at the end of the dictionary instead of under the letter "R". How can I change this? (Note that the second problem is not related to the first one, I also tried by changing the "=" to a simple"-" and had the same problem.) Thanks in advance! Lena-- ******* Alex Françoishttp://alex.francois.free.fr

ToolboxSupport

unread,
Sep 26, 2008, 2:54:01 PM9/26/08
to Shoebox/Toolbox Field Linguist's Toolbox
Dear Lena (and Alex),

Actually, I suppose I'm part of the problem. Lena, I saw you didn't
have any \a field and for some reason I guess I assumed you didn't
want one. So your interlinear is set up to use \va as the alternate.
You didn't have a "\u" field either, so I just skipped that whole
setup.

Anyway, there is nothing "magic" about the choice of marker, except
when you go to print your dictionary, the MDF assumes certain field
markers have certain uses.

So maybe the best thing would be to change your \va fields to \a
fields. You can use \va when it is more appropriate. How to do this:
1) First, click on the dictionary (either of yours will do, since they
are the same database type). This will bring the dictionary into focus
and notify Toolbox that it's *this* database you want to modify.
2) Do Database, Properties. This will bring up the list of markers
which are currently available to the database.
3) Locate the va marker in the list of markers. Select it and choose
"Copy". Change the marker and the Field Name by adding a couple x's
after them. (This is only temporary.) So you will have vaxx with a
Field Name of Variant form(s)xx. ( This will let you preserve the
description, etc, of the original va marker. This will make more sense
as we go along.)
4) Click OK and return to the list.
5) Choose the original va marker again and click Modify.
6) Change the "va" in the first box to "a". This should bring up a
question box asking if you want to change the marker everywhere.
7) Choose YES. You do want to change it everywhere. (This will change
it in your database and also in the interlinear specifications!)
8) Modify the rest of the dialog box, specifying "alternate" as the
Field Name, under "lx" in the hierarchy, "u" for the "following
field" (this means if you hit an Enter at the end of the \a field, it
will add a \u field), and "vernacular" for the Language Encoding. In
the description you can say it is "the alternate surface forms for
interlinear."
9) Click OK.
10) Select the copy of va -- vaxx or whatever you modified the marker
to become.
11) Remove the x's that were added. Again, Toolbox will ask if you
want to change the marker everywhere.
12) Choose YES. You want to change it everywhere.
13) Remove the x's from the Field Name also.
14) Click OK.
15) Click OK again to return to the main window.

16) Now click on one of the text databases.
17) Do Database, Properties, Interlinear tab.
18) Select the Parse process and click on Modify.
19) Click on Lexicons. This brings up a dialog box with four large
rectangular windows and at the very bottom a small one labelled
"Marker to Output".
20) Change the Marker to Output to "u".

That should solve part of the problem anyway...

Toolbox Support
> > Alex.Lena Terhart a écrit :Hi there, after Toolbox support got my project settings right again (many thanks for that!!), I started interlinearizing. Now I encountered a new problem: there are several phonological changes in the language and I want to deal with these using underlying forms. To exemplify this: I have a word "nkasiaper", which should break into "ni=kasia-pa=ro". So for the dictionary entry: "-pa" I definded: \lx -pa \va -per \u -pa =ro{3SGm} Anyway, I only get the result "ni=kasia-pa" (among other choices including another lexeme "per"). I also tried combinations of "-pa" and "=ro" including and excluding spaces and + in the \u-field, so that is not the problem. I suppose it doesn't work because the \u-function was totally deleted from my list of MDF-markers. I included it again, but did so by only typing the marker's abbreviation and name. I suppose that doesn't suffice to get it work as it should... How do I get back the original marker? I also have a minor problem concerning clitics: I defined the = as a morpheme break character and it works well in the text files. But in the dictionary it seems to be interpreted as a letter, so the clitic "=ro" appears at the end of the dictionary instead of under the letter "R". How can I change this? (Note that the second problem is not related to the first one, I also tried by changing the "=" to a simple"-" and had the same problem.) Thanks in advance! Lena-- ******* Alex Françoishttp://alex.francois.free.fr- Hide quoted text -
>
> - Show quoted text -

ToolboxSupport

unread,
Sep 26, 2008, 2:58:07 PM9/26/08
to Shoebox/Toolbox Field Linguist's Toolbox
I should have explained at the end of my message below. Having the
"\u" marker as the "Marker to Output" for the interlinear does not
mean that you always have to have a \u. It means that if there is a \u
it will be used. Otherwise it uses the first item in the list of
"Markers to Find" -- which is the lx marker. As explained in the
Training materials, you can use \a with or without \u.

Toolbox Support
> > > Alex.Lena Terhart a écrit :Hi there, after Toolbox support got my project settings right again (many thanks for that!!), I started interlinearizing. Now I encountered a new problem: there are several phonological changes in the language and I want to deal with these using underlying forms. To exemplify this: I have a word "nkasiaper", which should break into "ni=kasia-pa=ro". So for the dictionary entry: "-pa" I definded: \lx -pa \va -per \u -pa =ro{3SGm} Anyway, I only get the result "ni=kasia-pa" (among other choices including another lexeme "per"). I also tried combinations of "-pa" and "=ro" including and excluding spaces and + in the \u-field, so that is not the problem. I suppose it doesn't work because the \u-function was totally deleted from my list of MDF-markers. I included it again, but did so by only typing the marker's abbreviation and name. I suppose that doesn't suffice to get it work as it should... How do I get back the original marker? I also have a minor problem concerning clitics: I defined the = as a morpheme break character and it works well in the text files. But in the dictionary it seems to be interpreted as a letter, so the clitic "=ro" appears at the end of the dictionary instead of under the letter "R". How can I change this? (Note that the second problem is not related to the first one, I also tried by changing the "=" to a simple"-" and had the same problem.) Thanks in advance! Lena-- ******* Alex Françoishttp://alex.francois.free.fr-Hide quoted text -
>
> > - Show quoted text -- Hide quoted text -

ToolboxSupport

unread,
Sep 26, 2008, 3:24:38 PM9/26/08
to Shoebox/Toolbox Field Linguist's Toolbox
Dear Lena,

I didn't answer the second question you asked:

> I also have a minor problem concerning clitics: I defined the = as a
> morpheme break character and it works well in the text files. But in
> the dictionary it seems to be interpreted as a letter, so the clitic
> "=ro" appears at the end of the dictionary instead of under the letter
> "R". How can I change this?

The problem is that the = needs to be in the Sort Order, but ignored
for sorting.

To solve this,

1) bring your mouse over the \mb marker in the text database. (Yes, in
the marker pane itself.) Do a Right Click.
2) This will bring up the Marker Properties dialog box. This is a way
to be sure you get to the right language encoding.
3) On the right side of the dialog box is a button labelled "Language
Properties". Click that button.
4) This will bring up the Language Properties dialog box. The first
tab is the Sort Order.
5) Click Modify.
6) The Sort Order Properties is a complex dialog box. But what you
need is in the lower right corner. That is the list of characters or
combinations to be ignored.
7) Place your cursor in the "Ignore characters" box. Type a space and
then type the =.

It appears that hyphen is already in the box, in which case I don't
know why using hyphen didn't solve your problem. It's possible that
the apparent hyphen there and the hyphen you typed are different
characters. Anyway, type the equals, then click OK all the way back to
the main Toolbox window.

Check if the sorting now places =ro under "r". If so, check the end of
the database. Are there words listed at the end beginning with hyphen
(like -pa)? If so, go back into the Sort Order Properties and add the
hyphen that you type. (Who knows what's already there...) Be sure to
type a space between the equals and the hyphen.

(It worked when I did it, and the suffixes are already sorting
properly. I don't know what happened when you tried changing the = to
-.)

Toolbox Support

Lena Terhart

unread,
Sep 27, 2008, 5:45:20 AM9/27/08
to Shoebox/Toolbox Field Linguist's Toolbox
Dear Toolbox Support,

thanks for the answer. I put the = in the "Ignore characters box" and
that works fine, now.
Anyway, my first problem remains, because when I come to "6) Change
the "va" in the first box to "a". This should bring up a question box
asking if you want to change the marker everywhere." I don't get the
question box, but a dialogue box telling me "Cannot merge marker". I
had that problem before, when I tried to change some of the markers
that had been added to Toolbox into existing MDF compatible ones. (It
worked just fine with others.) I think I dealt with it, opening the
text file in a text editor and do a simple replace command. But I
guess, this time that won't work, because that doesn't change any
interlinear specifications. I unlocked all my files, so that can't be
the problem either.

Lena

ToolboxSupport

unread,
Sep 27, 2008, 7:13:04 AM9/27/08
to Shoebox/Toolbox Field Linguist's Toolbox
Dear Lena,

There are a couple things going on here. One is that Toolbox won't
"merge" a marker that is part of an interlinear process -- but it will
change one. (I discovered this by reading the Help file --
occasionally it *is* helpful! -- and by experimenting as I was writing
the instructions to you.)

So, my question -- did you sometime along the line already *add* the
\a marker? That is, is \a in the list of markers when you do Database,
Properties for a dictionary? If so, you need to take a different
approach.

Specifically, you need to
1) click on a TEXT file
2) Do Database, Properties, Interlinear
3) Select the parse process and click on Modify
4) Click on Lexicons
5) There, in the lower half of the dialog box, you will see the va
included in the list of Markers to Find.
6) Select the "a" marker from the list of Available Markers on the
left
7) Click on Insert. It should be placed at the bottom of the "Markers
to Find" list.
8) Select "va" in the "Markers to Find" list and click on Remove
9) Since you're there already, change the Marker to Output to "u".
10) Close by clicking OK until you're back to the main Toolbox window.

11) Now click on a DICTIONARY file, do Database, Properties, and try
changing the "va" to "a". Toolbox should ask
"Merge marker?"
12) Choose "yes".

This should work if you already have added the \a.

Toolbox Support
Reply all
Reply to author
Forward
0 new messages