Postposition to (in)animate (pro)nouns

[Michel Pauw]

After a few years of absence I am returning to FLEx. The coming days I will likely bring up a number of complicated questions on how to enter/analyse data into FLEx. 

Case #1: postposition to (in)animate (pro)nouns

Aramba has two sets of postpositions. One set is applied to animate nouns and the other set is used to inanimate nouns. 

Example: the preposition 'to' has two equivalents in Aramba: 1) -fo for inanimate nouns, and 2) -mbo (or -nmbo?) for animate nouns. See the data below and the questions following below. 

This is all the data I have so far for the allative postposition 'to' in Aramba. I have more data, but they fit the data shown above. The table for the postposition 'from' would look exactly the same, but then the -o is replaced by an -a: -(n)mba. 

 

I have the following questions about how to interpret this data, when it comes to analyzing the data in FLEx. So based on this data: 

1. What should I describe as the actual suffix? 
   • I am currently inclined to consider -nmbo to be the suffix and -mbo to be an allomorph. The reason for this is that I don't want to include the -n in the auxiliary suffix, because there is a dative case marker -n that has exactly the same auxiliary suffixes. So, it is easier to describe -n- as part of the suffix
     -nmbo and consider -mbo an allomorph in a few cases, than to say the -n- belongs to the auxiliary suffix and then I need to come up with a whole set of different auxiliary suffixes that precedes the dative suffix -n. 
   • Would you agree/disagree?   
2. How many (and which) lexical entries do I enter? 
   
   • I suspect it would be best to describe two main lexical entries: one for the postpositional suffix -nmbo and another for the auxiliary suffix, but do I really have to add as many auxiliary suffix-entries into FLEx, as I see in the table? 
   • The auxiliary suffixes do indicate number (SG/PL), but only when a postpositional suffix is attached. That's why I call it an auxiliary suffix. If no postpositional suffix is attached, the nouns (as they appear in the first column) can be either SG or PL. They will not get the auxiliary suffix.   
   • It looks like there are two basic auxiliary suffixes -à (SG) and -á. All other instances can be explained morphophonemically, for example: 
     • the adding of -y can be explained by the preceding front vowel
     • the adding of -w can be explained by the preceding back vowel
     • the change of -à into -e and of -á into -é can be explained by vowel harmony (certain vowels don't like to go together with the 'a-sound')
   • How do I enter all this data according to the interpretations given above? 
     • Should I enter them all as separate lexical entries? 
     • Should I consider them allomorphs of -à and -á (and if so, how do I enter those into FLEx)? 
     • Should I stick with two lexical entries  -à and -á as the underlying morpheme of all the others and then describe some morphophonemic rules in FLEx to account for the others? If so, where and how do I describe those morphophonemic rules in FLEx? 

Thanks for advising!

Michel Pauw

Aramba Translation Advisor Papua New Guinea

[Andreas_Joswig]

Hi Michel,

For the animate postposition (I would rather call it an allative case marker, as postpositions are free forms or at least clitics) I would enter two separate entries: -nmbo for nouns and -mbo for pronouns.

The variation of the aux-suffixes looks mostly conditioned by phonological rules, but not entirely. I don't see how the change to -we and -wé for 'man' can be explained by vowel harmony, as the same stem vowels are present for the markers -wa and -wá with 'dog'. If in that case you cannot identify another phonological explanation, you will have to work with inflection classes. In any case, you can enter both morphologically and phonologically conditioned variants as affix allomorphs under the same entry. The phonologically conditioned ones need an environment to work (such as [e](C)_, whereas the morphologically conditioned ones need the inflection class with which they come. You should read up in the "Introduction to parsing" about how to work with allomorphs, environments and inflection classes.

Best wishes,

Andreas

--
"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/0d151c67-bddb-4647-8187-b91607de4bban%40googlegroups.com.

--

Andreas Joswig, PhD

Senior Linguistics Consultant Addis Ababa, Ethiopia

+49 1512 3581686 (Ger) or +251 910 497594 (Eth)

[Michel Pauw]

Thank you, Andreas, that gives me a good start!

FYI: regarding vowel harmony: the ú (central vowel) and the u (back vowel) belong to two excluding sets of vowels. The ú does not like to work with the 'a' sound (although I realize 'man' already has another 'a' in its root, which seems to be an exception that is covered by another rule). 

I am reading through the Introduction to Parsing and will go from there. 

[Michel Pauw]

After reading the FLEx documentation I'm stuck. 

This is my understanding on how to analyse the data in FLEx: 

• separate entry -fo: allative suffix for inanimate nouns
• separate entry -mbo: allative suffix for animate pronouns 
• separate entry -nmbo: allative suffix for animate nouns

• separate entry -à: auxiliary suffix SG, with -yà, -wà, -e, -we as allomorphs based on certain morphophonemic rules
• separate entry -á: auxiliary suffix PL, with -yá, -eyá, -wá, -wé as allomorphs based on certain morphophonemic rules
• (for now I leave out the auxiliary suffixes for the pronouns)

However, I struggle with the difference between Inflection Class and Inflection Feature, when and how to use each one. I've read the documentation in the Introduction to Parsing guide, as well as the help file in FLEx, but they don't take me any further. 

Any help would be appreciated, either through this group, or if anyone would be willing to have a 1 hour video call with me, that would be even better!

Thanks!

Michel 

[Michel Pauw]

For example: 

• how should I tell FLEx that a certain noun is animate or inanimate?
• how should I tell FLEx that a certain suffix only goes with an animate or inanimate noun? 

[Mike Maxwell]

I haven't seen any response, so here are my thoughts.

Inflection features are generally used for something that (more or less)
has a semantic value; whereas inflection classes are for properties that
do not have any apparent semantic value, and are therefore more or less
arbitrary. Verb classes in many Romance languages, like -ar/-er/ir
verbs in Spanish, are examples of properties without any semantic value.
(Stem classes, used where the stem of some words of a particular part
of speech undergo alternations, while other stems of that same part of
speech do no, are also properties without semantic value.)

Also, inflection classes by definition *never* cause agreement. For
example, the fact that a Spanish verb is in the -er class does not cause
its arguments to agree with it in any way. Inflection classes sometimes
(but not always) cause agreement, e.g. in Spanish an adjective agrees
with the noun it modifies in gender.

Latin (from which of course the Romance languages derive) had a
wide-ranging and complex system of both gender
(masculine/feminine/neuter) and inflection classes (often called
declensions for nouns and adjectives, and conjugation classes for
verbs). To some extent the noun declension classes cross-cut the gender
classes, for example while most second declension nouns were masculine,
some were neuter or feminine; and third declension nouns could be any
gender. Latin adjectives would agree with nouns they modified in gender
(an inflectional feature), but not in declension (an inflectional class).

Animacy usually does have some semantics behind it, so I would guess
that in the language you are working with, it is an inflectional feature.

There are gradations (of course, we're talking about language...).
Gender in IndoEuropean languages typically has a core of semantic
relevance: nouns for women are typically (but not always, as Mark Twain
pointed out) feminine, nouns for men are typically masculine. But nouns
for inanimate concepts also have gender, even though that typically
makes no sense semantically. And in some languages with animacy
distinctions, things like lightning may be animate--which may or may not
reflect a cultural belief. Or in English, 'pants', 'shorts' etc. are
plural, even if you are talking about one pair.

In sum: if this animacy distinction in your language causes agreement,
then it must be an inflectional feature, not an inflectional class (nor
a stem class). Otherwise, the semantics is probably a clue.

On 6/24/2022 10:45 PM, Michel Pauw wrote:
> After reading the FLEx documentation I'm stuck.
>
> This is my understanding on how to analyse the data in FLEx:
>

> * separate entry -fo: allative suffix for inanimate nouns
> * separate entry -mbo: allative suffix for animate pronouns
> * separate entry -nmbo: allative suffix for animate nouns
>
> * separate entry -à: auxiliary suffix SG, with -yà, -wà, -e, -we as

> allomorphs based on certain morphophonemic rules

> * separate entry -á: auxiliary suffix PL, with -yá, -eyá, -wá, -wé as

> allomorphs based on certain morphophonemic rules

> * (for now I leave out the auxiliary suffixes for the pronouns)

>> *Case #1: postposition to (in)animate (pro)nouns*

>> Aramba has two sets of postpositions. One set is applied to
>> animate nouns and the other set is used to inanimate nouns.
>> Example: the preposition 'to' has two equivalents in Aramba:

>> 1) /-fo /for inanimate nouns, and 2) /-mbo (or -nmbo?) /for

>> animate nouns. See the data below and the questions following
>> below.
>>

>> inanimate.jpg
>>
>> This is all the data I have so far for the allative
>> postposition 'to' in Aramba. I have more data, but they fit
>> the data shown above. The table for the postposition 'from'
>> would look exactly the same, but then the -o is replaced by an

>> -a: /-(n)mba/.

>> I have the following questions about how to interpret this
>> data, when it comes to analyzing the data in FLEx. So based on
>> this data:
>>

>> 1. What should I describe as the actual suffix?
>> * I am currently inclined to consider -/nmbo /to be the
>> suffix and -/mbo/ to be an allomorph. The reason for
>> this is that I don't want to include the -/n/ in the

>> auxiliary suffix, because there is a dative case

>> marker -/n/ that has exactly the same auxiliary
>> suffixes. So, it is easier to describe -/n-/ as part
>> of the suffix
>> -/nmbo/ and consider -/mbo/ an allomorph in a few

>> cases, than to say the -n- belongs to the auxiliary
>> suffix and then I need to come up with a whole set of
>> different auxiliary suffixes that precedes the dative

>> suffix -/n/.
>> * Would you agree/disagree?
>> 2. How many (and which) lexical entries do I enter?
>> * I suspect it would be best to describe two main

>> lexical entries: one for the postpositional suffix

>> -/nmbo/ and another for the auxiliary suffix, but do I

>> really have to add as many auxiliary suffix-entries
>> into FLEx, as I see in the table?

>> * The auxiliary suffixes do indicate number (SG/PL), but

>> only when a postpositional suffix is attached. That's
>> why I call it an auxiliary suffix. If no
>> postpositional suffix is attached, the nouns (as they
>> appear in the first column) can be either SG or PL.
>> They will not get the auxiliary suffix.

>> * It looks like there are two basic auxiliary suffixes
>> -/à/ (SG) and /-á/. All other instances can be
>> explained morphophonemically, for example:
>> o the adding of -/y/ can be explained by the
>> preceding front vowel
>> o the adding of -/w/ can be explained by the
>> preceding back vowel
>> o the change of -/à /into -/e /and of -/á /into
>> -/é/ can be explained by vowel harmony (certain

>> vowels don't like to go together with the 'a-sound')

>> * How do I enter all this data according to the
>> interpretations given above?
>> o Should I enter them all as separate lexical entries?
>> o Should I consider them allomorphs of -/à/ and -/á
>> /(and if so, how do I enter those into FLEx)?
>> o Should I stick with two lexical entries -/à/ and
>> -/á /as the underlying morpheme of all the others

>> and then describe some morphophonemic rules in
>> FLEx to account for the others? If so, where and
>> how do I describe those morphophonemic rules in FLEx?
>>
>> Thanks for advising!
>>
>> Michel Pauw

>> /Aramba Translation Advisor Papua New Guinea/

>> --
>> "FLEx list" messages are public. Only members can post.
>> flex_d...@sil.org
>> http://groups.google.com/group/flex-list

>> <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/0d151c67-bddb-4647-8187-b91607de4bban%40googlegroups.com

>> <https://groups.google.com/d/msgid/flex-list/0d151c67-bddb-4647-8187-b91607de4bban%40googlegroups.com?utm_medium=email&utm_source=footer>.
>
>
> --
>
> Andreas Joswig, PhD
>
> Senior Linguistics Consultant Addis Ababa, Ethiopia
>
> +49 1512 3581686 <tel:+49%201512%203581686> (Ger) or +251 910
> 497594 (Eth)
>

> --
> "FLEx list" messages are public. Only members can post.
> flex_d...@sil.org
> http://groups.google.com/group/flex-list

> <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

> <mailto:flex-list+...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/flex-list/53723b35-09a0-4428-ad37-995fa42224d9n%40googlegroups.com
> <https://groups.google.com/d/msgid/flex-list/53723b35-09a0-4428-ad37-995fa42224d9n%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
This email has been checked for viruses by AVG.
https://www.avg.com

[Andy Black]

On 6/24/2022 8:01 PM, Michel Pauw wrote:

For example: 

• how should I tell FLEx that a certain noun is animate or inanimate?

First, go to Grammar / Inflection Features and use Insert menu item / Feature...  Open up noun-related / noun agreement and noun class.  Click on animate class and inanimate class:



Click on Add.

You will now have the needed animate and inanimate inflection features.

To tell FLEx that a certain noun is animate or inanimate, you need to add the appropriate inflection feature to each noun under its Grammatical Info. Details section.  You'll probably need to click the Box before Show Hidden Fields at the top:



Scroll down to the bottom and you should see an Inflection Features field (be sure to get this one and not the Inflection Class field).  Click in it and then on the chooser button.  You should see a place showing radio buttons before "animate class," "inanimate class," and "None of the above."  Choose animate or inanimate as appropriate for this noun.  The Grammatical Info. Details portion may then look something like this:



To avoid having to always show all hidden fields, you can click on the blue circle with a triangle in it before "Inflection Features and choose Field Visibility / Always visible.

This, of course, just does one noun at a time.  You can use Lexicon / Bulk Edit Entries to do many nouns at one operation.  You will need to show Grammatical Category and Inflection Features. (Click on the icon in the upper right hand corner under the blue bar near the top.) Filter the entries by Noun.  (Click on on the chooser down-arrow button underneath Grammatical Category; use Choose and click on Noun.)

Now use the List Choice tab option (at the bottom).  Set the Target Field to Inflection Features.  Set the Change To field to animate.  Check every noun that is animate and uncheck every noun that is inanimate.  Click on Preview to make sure things are right and then click on Apply.  Do the same for inanimate.

• how should I tell FLEx that a certain suffix only goes with an animate or inanimate noun?
  

In Lexicon / Lexicon Edit, find such a suffix.  Show hidden fields and look at the bottom under Grammatical Info. Details.  If the suffix is inflectional, then there will an Inflection Features field.  Set it just like you did for a noun lexical entry.  If the suffix is derivational, there will be a From Inflection Features field and a To Inflection Features field (among others).  If this derivational suffix can only attach to animate nouns, then set the From Inflection Features field to animate by clicking in the field and then on the chooser button.  The To Inflection Features field is for any derivational affix that results in a noun stem; you use this to set the animacy of that derived stem.

--Andy

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/6ad49eb3-828b-44ed-aeda-0ca0a88572acn%40googlegroups.com.