homograph numbers and multiple writing systems

[Josh Jensen]

I'm working on a database that uses two vernacular writing systems, OsageNew [practical] and OsagePhon [phonetic]. The default is OsageNew. Most of the entries are currently entered only in OsagePhon (the non-primary writing system). The result is that every lexical entry where the OsageNew field is empty has a homograph number assigned to the string in the OsagePhon field.

I have run Reassign Homographs to no avail.

[[Attached Screenshot: shows a portion of Entries pane. Left column is Headword (OsageNew). Next column is Headword (OsagePhon). Note the homograph numbers!]]

Thanks.

Josh Jensen

UT Arlington

[Beth (work) Bryson]

This is because "empty" matches "empty"....    All empty strings are homographs of each other!

In what ways is it causing a problem for you?  The homograph number displays in the dictionary; that can be configured to be turned off.

I do remember that at one point FLEx would crash if the homograph number was greater than 256.  I know they made it not crash, but I can't tell what FLEx does if there are more than that many homographs.

-Beth

--
You received this message because you are subscribed to the discussion group "FLEx list". This group is hosted by Google Groups and is open for anyone to browse.
To post to this group, send email to flex...@googlegroups.com
To unsubscribe from this group, send email to flex-list+...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/flex-list
<Screen Shot 2012-05-05 at 11.01.04 AM.png>

[Joshua and Amy]

Hi, Beth. Thanks for the quick reply.

 

In what ways is it causing a problem for you? 

Supposing Osage has honest-to-goodness homographs, then turning off the homograph number display means the real ones won't get marked. Not apocalyptic, to be sure, but homograph marking (when marking actual homographs) is nice.

I think another solution is making the phonetic writing system the primary one. But the help files on multiple writing systems frighten me: they warn about bad things happening if you're not careful. But honestly, I didn't understand the help files on this topic.

Josh

[Jonathan]

You can switch which writing system is primary without problem.  I have done it before.  (Back up before doing it so that you can revert if it does something funky. )  Then see if it automatically redoes the homographs...

Is the practical system a phonetically based system?  i.e. could you map character to character between the two?  If so then you could map a converter to take the data in one and populate the other.

Hope it goes well,

Jonathan 

[John Thomson]

It should be pretty easy to program it so that empty string is treated as a special case and not assigned a homograph number. It seems unlikely it would ever be useful to assign HNs to all entries with empty forms in the default orthography.

 

Another option would be to consider the "best vernacular" writing system when assigning HNs. This would assign them to words that are HNs in the practical orthography if both have empty strings in the (default) phonetic one. It might produce somewhat surprising results if more than two writing systems are in use, though.

 

Even with just two writing systems, we'd have to think about edge cases like whether, say, an entry with form "a+diacritic" in the default phonetic orthography)and form "a" in the practical orthography is or is not a homograph of another entry with form "a" in the practical orthography but no form at all in the phonetic orthography.

 

On the whole, I think it may be best to continue to consider just one writing system when assigning homograph numbers, but make a special case for empty forms.

 

JohnT

[J V C]

Personally, I'm inclined to agree. Otherwise, the software might seem to be doing some black-box "magic" that could be confusing until the user read the right help file. Of course, if there are users who come forward with a clear use case for the "best vernacular" idea, then it's by no means a bad idea.

Jon

[J V C]

Yeah, I've never had a problem either, although usually it's the primary
*analysis* writing system that I'm switching, rather than the vernacular.

I think you'll find that things are friendlier once your primary
vernacular WS is set to something for which the bulk of your entries
have actually filled in Lexeme Form. I'm pretty sure that any dialog
that searches based on wordform won't work well otherwise. That includes
the Find dialog (Ctrl + F), and the chooser dialog for most link fields
(Components, Cross Reference, Lexical Relations, etc.) Likewise, the New
Entry dialog can't warn you that you might be creating a duplicate entry
if it can't find any matches. (It can, however, warn you based on an
identical Gloss, if you fill that in before pressing Ok.)

BTW, I believe a recent discussion on this list concluded that a single
vernacular WS should be used for all interlinear baselines, so that's
something to keep in mind as well.

Jon

[Joshua and Amy]

Thanks. And I'll check out the discussion on WS and baseline.

To unsubscribe from this group, send email to flex-list+unsubscribe@googlegroups.com

For more options, visit this group at http://groups.google.com/group/flex-list

--
You received this message because you are subscribed to the discussion group "FLEx list". This group is hosted by Google Groups and is open for anyone to browse.
To post to this group, send email to flex...@googlegroups.com

To unsubscribe from this group, send email to flex-list+unsubscribe@googlegroups.com