Where to record Bantu noun classes ?
Ian Tupper
entry for each noun to show the pair of noun classes applying to it,
e.g. "Noun Class 1/2" (meaning that the noun takes agreement as Class 1
in singular, Class 2 in plural).
I thought I had found the right area under the heading "Grammatical
Info. Details". One option was "Inflection Class". I added to this
category the pairs of noun classes found in the language, e.g. "NC1/2",
"NC3/4", "NC5/6", etc. For every noun, I then selected the "inflection
class" that applied (where known). That worked fine for the first thirty
or so entries. However, after that, the option "Inflection class" was no
longer visible when the heading "Grammatical Info. Details" was
selected. When I've clicked on "Grammatical Info. Details", it tells me
it's "full". It seems I've chosen the wrong approach.
What is the recommended approach for recording Bantu noun classes (or
other genders)?
Ian
-- c/o SIL, B. P. 1299, Yaoundé, Cameroun
Kevin Warfel
Ian,
Andy Black has written an excellent paper on this topic:
https://lingtran.net/tiki-index.php?page_ref_id=770
If you still have problems or questions after reading that, write to me directly or post to the FLEx list and I’ll see what additional details you might need.
Kevin Warfel
Associate Dictionary and Lexicography Services Coordinator
a.k.a. Dictionary Development Coordinator
SIL International
Current technology makes it possible to provide those translating into just about any language with both a dictionary and a thesaurus in the target language, the standard tools of the trade for professional translators, so why are mother-tongue translators in minority languages still expected to do their work without these tools? Ask me about Rapid Word Collection after reading about it at rapidwords.net.
--
You are subscribed to the publicly accessible group "FLEx list".
Only members can post but anyone can view messages on the website.
To change your status, please write to flex_d...@sil.org.
You can join this group by going to http://groups.google.com/group/flex-list.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/a8ed8906-31b6-0adb-5262-6d168a6d9d53%40sil.org.
Mike Maxwell
called "classes", it's not uncommon to think that inflection classes
must be how they should be described. But they're actually agreement
classes, more like noun gender classes in Indoeuropean languages (and
Andy's analysis treats them this way). Adjectives that modify a noun,
and the verb (if the noun is in the right grammatical position, like
subject, object, or on some cases other argument positions) take the
noun class agreement affixes in Bantu languages.
An inflection class, on the other hand, by definition never triggers
agreement. (Inflection classes are traditionally called conjugation
classes on verbs, and declension classes on nouns, adjectives, and other
parts of speech.)
About the other observation--that "Grammtical Info.Details" tells you
it's "full", I have no idea why that would happen. Perhaps someone more
familiar with FLEx would have ideas on that.
On 10/12/2019 8:05 AM, Kevin Warfel wrote:
> Ian,
>
> Andy Black has written an excellent paper on this topic:
>
> https://lingtran.net/tiki-index.php?page_ref_id=770
>
> If you still have problems or questions after reading that, write to me
> directly or post to the FLEx list and I’ll see what additional details
> you might need.
>
> Kevin Warfel
>
> Associate Dictionary and Lexicography Services Coordinator
>
> a.k.a. Dictionary Development Coordinator
>
> SIL International
>
> Current technology makes it possible to provide those translating into
> just about any language with both a dictionary and a thesaurus in the
> target language, the standard tools of the trade for professional
> translators, so why are mother-tongue translators in minority languages
> still expected to do their work without these tools? Ask me about Rapid
> Word Collection after reading about it at rapidwords.net
> You can join this group by going to
> http://groups.google.com/group/flex-list.
>
> ---
>
> You received this message because you are subscribed to the Google
> Groups "FLEx list" group.
>
> To unsubscribe from this group and stop receiving emails from it, send
> an email to flex-list+...@googlegroups.com
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/flex-list/a8ed8906-31b6-0adb-5262-6d168a6d9d53%40sil.org.
>
> You are subscribed to the publicly accessible group "FLEx list".
> Only members can post but anyone can view messages on the website.
> To change your status, please write to flex_d...@sil.org.
> You can join this group by going to
> http://groups.google.com/group/flex-list.
> ---
> You received this message because you are subscribed to the Google
> Groups "FLEx list" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to flex-list+...@googlegroups.com
> <https://groups.google.com/d/msgid/flex-list/bea14ffed61d73d1a02faabc202137a5%40mail.gmail.com?utm_medium=email&utm_source=footer>.
--
Mike Maxwell
"I am, by a flood, borne back to that wondrous
period, ere time itself can be said to have begun;
for time began with man." --Herman Melville,
Moby Dick
Gray Plunkett
> Andy Black has written an excellent paper on this topic:
>
> https://lingtran.net/tiki-index.php?page_ref_id=770
>
Kevin Warfel
Thank you for this post, Gray. It brought to my attention the fact that the valuable content you referred to is no longer available there. I am actively trying to discover where it was moved to, and when I do, I will post the new link here.
Kevin Warfel
Associate Dictionary & Lexicography Services Coordinator
Rapid Word Collection workshop consultant
--
"FLEx list" messages are public. Only members can post.
flex_d...@sil.org
http://groups.google.com/group/flex-list.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/4c25cf33-06a4-43c3-9e09-cdccebbaa7e2n%40googlegroups.com.
Kevin Warfel
The missing article has been located. (Thanks, Jason Naylor!)
Here is its new “address”: https://lingtran.net/Modeling-Bantu-Noun-Classes
Best wishes to Gray and anyone else interested in modeling an African language with noun classes in FLEx!
Kevin Warfel
Associate Dictionary & Lexicography Services Coordinator
Rapid Word Collection workshop consultant
From: flex...@googlegroups.com <> On Behalf Of Gray Plunkett
Sent: Monday, May 30, 2022 1:00 PM
To: FLEx list <flex...@googlegroups.com>
--
"FLEx list" messages are public. Only members can post.
flex_d...@sil.org
http://groups.google.com/group/flex-list.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/4c25cf33-06a4-43c3-9e09-cdccebbaa7e2n%40googlegroups.com.
jeff_...@sil.org
I am glad you all found the article. The most important concept you need for modeling Bantu in Flex is that gender-noun prefixes are inflectional features and not parts of speech. It is recommended that you have a custom field to enter the plural forms in the entries. There should not be separated entries for plural and singular. If you are going to use the parser then the allomorphs for all the prefixes should be entered along with their respective environments.
For each noun you will need to specify the inflectional features. The dialogue looks like this:
Jeff Shrum
SIL International
Language Technology Consultant.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/8427cf111f300e425e00d647e589e47d%40mail.gmail.com.
Gray Plunkett
1. You said it is recommended to have a custom field to enter the plural forms in the entries. I set that up, but then found that there is a variant type plural already in FLEx. Why is it better to create a custom field for plural forms?
2. You also said: "There should not be separated entries for plural and singular." I was wondering why you wrote that? Doesn't it depend on the language? I definitely want to have an entry for all the plural forms since there are noun class prefixes, and so looking up a plural form such as anyɩm would not be obvious when the singular is ɔnyɩm. I would want to have an entry "anyɩm nA2 pl. of ɔnyɩm". And I would also want that entry to show the agreement class for anyɩm (nA2) since the a- prefix can be either agreement class 2 or 4. So for example "ágbaa nA4 pl. of dɩ́gbalɩ".Or do you think I should avoid doing this?
Gray
jeff_...@sil.org
Gray,
You ask some very interesting questions. I am happy to try and answer them.
- The variant plurals are for plurals that do not fit the normal patterns for a language. If you look at the possible choice in that area one is “irregular plurals”. It is extremely unlikely that you will find many irregular plurals in a Bantu language. The plurals are very regular. In some cases, the singular is a null. The most come is for some noun class 1 nouns, and that is typically designated as noun class 1a. It is good to have the plurals recorded to since gender-noun prefixes are in pairs. Having the singular as the headword and the plural as a custom field documents the noun class of a noun.
- This question is dependent on the purpose of your lexicon/dictionary. If you want a dictionary for people who are not aware of the grammar of the language, then you might want to have separate entries for singular and plurals. The general rule that I have heard about dictionaries is that if it is predictable based on the grammar, then a separate entry is not called for. Think of English, Spanish or other European language dictionary, the plural is listed somewhere with in the main entry for the singular. In the case of Flex if you have the root as the headword, the headword is marked with the correct inflectional class, and the singular and plural prefixes have entries, then Flex will be able to parse both the singular and plural forms of nouns. Some dictionaries for low-resource languages will have a basic phonology and grammar to help non-native speakers to be able to use it.
I hope I have answered your questions. If not ask again, and maybe others would have good things to add to your questions.
Jeff Shrum
SIL Language Technology Consultant
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/5064f908-6d81-4a80-8e76-2a0cefd09c0bn%40googlegroups.com.
Mike Maxwell
> The general rule that I have heard about dictionaries is that if it is
> predictable based on the grammar, then a separate entry is not called
> for. Think of English, Spanish or other European language dictionary,
> the plural is listed somewhere with in the main entry for the singular.
Athabaskan languages, and a few others are prefixing, and of course many
languages have both prefixes and suffixes (Mayan, Nahuatl, Semitic for
example).
Unfortunately, the alphabetization algorithm we use starts at the
beginning of the word, so one would naively look for a word to be sorted
starting with the prefix. That's especially true for languages where at
least one prefix is obligatory: people need special training to search
for roots if those roots are not actually words (in the common sense) of
the language.
If a prefix is obligatory for some major word class (roots are not valid
wordforms by themselves), there are no perfect solutions. One approach
is to alphabetize by the root, whether or not an obligatory prefix is
shown on the headword. This is what's typically done in Arabic
dictionaries. Another is to alphabetize starting with the prefix, but
since the prefix is likely to be the same (apart maybe from allomorphy)
for large classes of words, this means there will be pages and pages of
entries whose headwords all start with the same letters.
If prefixes are optional (or there is a null prefixed form), then
typically one alphabetizes by a non-prefixed form, even if that form may
be semantically anomalous. As an example of a semantic anomaly, imagine
a language where singular nouns are un-affixed but plurals bear a
prefix. Some nouns might be pluralia tantum (obligatorily plural), like
English 'pants', 'scissors'; these would either need to be represented
by a non-existent singular form, or listed in their plural prefixed
form, unlike all other nouns in the dictionary.
Of course if you are creating an electronic dictionary for users, rather
than a paper dictionary, this problem more or less goes away. Users can
type in the obligatorily prefixed form, and the dictionary app will
bring its entry up just fine. You still have to train them which
prefixed form of the word to type in, though, unless you have a
morphological parser as a front end to the dictionary.
Mike Maxwell
University of Maryland
--
This email has been checked for viruses by AVG.
https://www.avg.com