Last modified: 2014-03-10 13:09:08 UTC

Wikimedia Bugzilla is closed!

Wikimedia migrated from Bugzilla to Phabricator. Bug reports are handled in Wikimedia Phabricator.
This static website is read-only and for historical purposes. It is not possible to log in and except for displaying bug reports and their history, links might be broken. See T26432, the corresponding Phabricator task for complete and up-to-date bug report information.
Bug 24432 - Inflection parameters
Inflection parameters
Status: NEW
Product: MediaWiki extensions
Classification: Unclassified
WikiLexicalData/OmegaWiki (Other open bugs)
unspecified
All All
: Low enhancement with 2 votes (vote)
: ---
Assigned To: Nobody - You can work on this!
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2010-07-18 14:01 UTC by Fiable.biz
Modified: 2014-03-10 13:09 UTC (History)
4 users (show)

See Also:
Web browser: ---
Mobile Platform: ---
Assignee Huggle Beta Tester: ---


Attachments
Tables of inflection parameters and of parameter modes. (11.80 KB, application/vnd.oasis.opendocument.spreadsheet)
2010-07-18 14:01 UTC, Fiable.biz
Details

Description Fiable.biz 2010-07-18 14:01:48 UTC
Created attachment 7577 [details]
Tables of inflection parameters and of parameter modes.

In order to deal with inflexions, I suggest:

1) you first build an "Inflection parameter" table, aiming at saying what inflections depend on, in any given language. The idea is a triple-entry table: language/word class/parameter whose modes would be booleans or triple-state things ("depends on" / "does not depend on" / "not known yet"). The table will say that an English verb inflects according to mode, voice, tense, person and number, but not on form (affirmative, negative, interrogative).
As far as I know (and I know little), the data management systems I know only allow double-entry tables (field/record), and not triple-entry tables with direct access to fields. A single-entry table triply indexed will do if necessary. An example of such a table is here enclosed (1st sheet). Of course, in real life, instead of names, we need numbers referring to other tables (word class table, parameter name table).
  Note that we need to fill the table from the inflection point of view, not from the semantic point of view. For instance, in French, the indicative "imparfait" (past continuous) tense has several meanings:
  past continuous TENSE: "J'ÉTAIS chez moi quand je le feu a pris. ("I WAS at home when the fire started.",
  unreal MODE: "Si j'ÉTAIS une femme, je porterais une jupe." ("If I WERE a woman, I would wear a skirt.")
while in other languages these two meanings are expressed completely differently. This doesn't matter in a dictionary, it's a grammar question. The thing that matters is the indicative past continuous of "être" ("to be") is "j'étais, tu étais, il était, elle était, nous étions, vous étiez, ils étaient". When to use the French past continuous is, to my mind, not relevant. In other words, what is called "imparfait" and for that thought to be a past tense of the indicative mode can in fact express something else (the unreal mode) but this piece of information probably doesn't need to appear in an inflection table. If it did, then the full French past continuous would appear twice. French learners are supposed to learn that, in French, the unreal mode is always expressed by the indicative mode's past continuous forms, so that, traditionally, no "unreal" mode appears in French conjugations tables.

2)   Then we'll need a table of the "parameters modes" themselves, saying that, in ancient Greek, the numbers are "singular", "dual", "plural", in Malagasy, the tenses are "past", "present", "future". Etc. Note that this table won't replace the inflection parameters table, because a given parameter can influence several word classes (for instance in French and Spanish, the number influences nouns, adjectives, verbs, demonstrative determiners, possessive determiners, demonstrative pronouns, possessive pronouns etc.) but not all. This parameters modes table should be filled by Omegawiki users, so needs a user interface. An example is also included (2nd sheet), just for Malagasy, because it's easier.

3)   We also need an "Inflection groups" table, each group being exemplified. The traditional groups are usually not enough, but necessary for human users. In French, 3 verbal inflection groups would be
+ 1st group "chanter",
+ 1st group "jeter",
+ 1st group "acheter".
We also need a user interface for readers to fill this table.

4)   Every word (couple expression + DefineMeaning) in a given language should be given a field "inflection group", which could be null (no inflection), or "unique group" if this word class in this language does inflects but in only one way.

5) Then we need a mechanism to present in 2 dimensions tables the more-than-2 parameters inflections, and enable users to fill such tables.

This will do: any reader will then be able get one or two fully inflected words of the same inflection group than the word he is looking at. By "fully inflected", I mean at the first level because, in agglutinative languages, inflections combine.

6) To go further, we'll need, for each inflection group, a function producing the tables for any word of that group.

To my opinion, the most difficult points are 5) and 6).
What do you thing of such a plan?
Comment 1 Niklas Laxström 2010-07-18 14:50:10 UTC
Does this scale? Finnish has 50 to 60 inflection groups (varies by source), and one word can have over two thousand forms (or much less, if we look strictly only at inflection forms). It's not always clear to which inflection group(s) word belongs to.

6) is impossible task to do ourselves. It's still very hard task if we try to integrate tools that can already do that.
Comment 2 Fiable.biz 2010-07-18 15:38:21 UTC
I don't know Finnish at all, and don't understand what you mean by "forms" that would not be inflexions. Could you explain? However, I didn't suggest we store all the inflexions. 
French verbs are also a few dozens of groups, and one fully inflected example per group makes a small thin book called "Bescherelle La conjugaison pour tous", including a list of about 12 000 verbs with their group number. It's used by most junior high school. The same for Spanish.

Can't we can at least go up to 5)? I think 6) is achievable too with a clever user interface.
Comment 3 Niklas Laxström 2010-07-18 15:51:05 UTC
What kind of user interface for item 6 you are talking about? I see no easy way to generate all inflection forms automatically for all languages, nor I think it makes any sense to have users to input them (because there are too many of them, we miss generalizations and there might not be one single agreed number of inflection groups).

I'm not opposed having this information available, but I think we should think little more what makes sense and what can be done in limited resources.
Comment 4 Fiable.biz 2010-08-16 09:45:50 UTC
In order not to clutter this Bugzilla, I shift to the International linguistic parlour the discussion on producing inflexions.
http://www.omegawiki.org/International_Linguists_Beer_Parlour/Inflexions#Producing_inflexions
I think we can discuss here the core of the project, which is giving the users access to a way of producing inflexions of a word, thanks to its group and to examples.
Comment 5 Mark A. Hershberger 2011-03-13 17:21:41 UTC
Is this being worked on?  Doesn't really look like anyone is showing any interest in this for at least the past four months.

Note You need to log in before you can comment on or make changes to this bug.


Navigation
Links