inflection categories using differing affix types

Discussion of natural languages, or language in general.
Post Reply
Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

inflection categories using differing affix types

Post by Vardelm »

I'm looking for languages where a given inflection or category of inflections use differing types & forms of affixes. I'm thinking of something like Arabic's perfective/past vs. imperfective/non-past, where person is basically marked via suffixes (kataba) or prefixes (yaktubu). I'm also somewhat familiar with Georgian verbs having a 1st person prefix vs. a 3rd person suffix. Is anyone familiar with more examples? For extra credit, whence comes a difference in prefixing vs. suffixing for a given inflectional group?

What I'm working towards is verbal conjugations that have significant differences between a realis paradigm & an irrealis paradigm, parallel to the differences between Arabic's past & non-past.
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Re: inflection categories using differing affix types

Post by Soap »

Prefixes can come from cliticized pronouns. Imagine a quick sketch of a language like this

so milibo "I write"
te milibe "you write"
a miliba "he/she writes"

Then assume that word-final /o/ and /e/ are lost but /a/ survives (happened in French and some other Romance languages). Suppose also that all word-initial vowels drop. This may encourages the speakers to get rid of the 3rd person pronoun /a/, which was likely optional to begin with, and come to mark 3rd person singular by the final /-a/ on the verb, which survived the change that got rid of /e/ and /o/.

This means that the pronouns that used to mark 1st and 2nd person are now mandatory, since the 1st and 2nd person forms of the verb have otherwise merged. Whether the pronouns also drop their vowels or not is up to the creator of the language, since monosyllabic words often resist changes to "final" vowels. Either way, the new verb conjugation would look something like this:

somilib
temilib
miliba

This whole setup could be analyzed as a circumfix, and that's the analysis I would go with if there is at least one form that has morphemes that appear on both sides of the root. Nobody knows how Semitic got its inflectional paradigm, but it likely goes back at least 10000 years, and maybe more than 20000 years, so there may be other languages where the original derivation is transparent. There may even be some examples of this in IE, and I'd suspect that they'd be of the type that I imagined for the example language in this post where an originally optional pronoun becomes mandatory.

Also, sorry, I thought this post was in this thread, but I think my answer can be helpful even so.
Sunàqʷa the Sea Lamprey says:
Image

Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

Re: inflection categories using differing affix types

Post by Vardelm »

Thanks. That's helpful extra credit! :)
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: inflection categories using differing affix types

Post by zompist »

You might be interested in Hindi, whose verbal system is all over the place.

First, there's a distinction between simple verbs (e.g. the perfective), and forms with auxiliaries (which are actually more common).

Then, there's the paradigms that inflect with person/number, and those which go with gender/number.

And then, the perfective has ergative-absolutive alignment, while everything else is nominative-accusative.

(Historically, many of the paradigms come from participles, which is where the second element comes from. And the alignment change seems to have come about by reanalyzing a passive form.)

Vijay
Smeric
Smeric
Posts: 2244
Joined: Sat Feb 06, 2016 3:25 pm
Location: Austin, TX, USA

Re: inflection categories using differing affix types

Post by Vijay »

zompist wrote:First, there's a distinction between simple verbs (e.g. the perfective), and forms with auxiliaries (which are actually more common).

Then, there's the paradigms that inflect with person/number, and those which go with gender/number.
Wait...aren't these the same thing? Don't the simple verbs inflect with person/number and the auxiliaries inflect with gender/number?

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: inflection categories using differing affix types

Post by zompist »

Vijay wrote:
zompist wrote:First, there's a distinction between simple verbs (e.g. the perfective), and forms with auxiliaries (which are actually more common).

Then, there's the paradigms that inflect with person/number, and those which go with gender/number.
Wait...aren't these the same thing? Don't the simple verbs inflect with person/number and the auxiliaries inflect with gender/number?
Nope. Simple verbs can go either way; so can auxiliaries.

(Plus there's other fun things, like two levels of causative, and making passives from intransitive verbs.)

Vijay
Smeric
Smeric
Posts: 2244
Joined: Sat Feb 06, 2016 3:25 pm
Location: Austin, TX, USA

Re: inflection categories using differing affix types

Post by Vijay »

I'm having trouble figuring out what you're talking about, sorry. :( Could you give me an example of a simple verb in Hindi that inflects with gender/number and an auxiliary in Hindi that inflects with person/number?

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: inflection categories using differing affix types

Post by zompist »

Vijay wrote:I'm having trouble figuring out what you're talking about, sorry. :( Could you give me an example of a simple verb in Hindi that inflects with gender/number and an auxiliary in Hindi that inflects with person/number?
Simple verb: any of them, e.g. ghumnā 'walk'.
Past is gender/number: 1m ghumā, 1f ghumī ; 2m ghume, 2f ghumĩ̄
Optative is person/number: 1s ghumũ, 2s ghume, 3s ghume, 1p ghumẽ, 2p ghumo, 3p ghumẽ

Auxiliary: honā 'be'.
Present is person/number: 1s hṹ, 2s hai, 3s hai; 1p haĩ, 2p ho, 3p haĩ
Past is gender/number: 1m thā, 1f thī ; 2m the, 2f thĩ̄

And while we're at it, the presumptive of honā is inflected by person, number, and gender, e.g. 1sm hṹgā, 1sf hṹgī, 2sm hogā...

Vijay
Smeric
Smeric
Posts: 2244
Joined: Sat Feb 06, 2016 3:25 pm
Location: Austin, TX, USA

Re: inflection categories using differing affix types

Post by Vijay »

Ohh, I somehow managed to confuse "auxiliary" with "participle" in my brain. Uy. :? And I guess I didn't realize participles counted as simple verbs (even though they're definitely used that way, too).

Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

Re: inflection categories using differing affix types

Post by Vardelm »

zompist wrote:You might be interested in Hindi, whose verbal system is all over the place.

First, there's a distinction between simple verbs (e.g. the perfective), and forms with auxiliaries (which are actually more common).

Then, there's the paradigms that inflect with person/number, and those which go with gender/number.

And then, the perfective has ergative-absolutive alignment, while everything else is nominative-accusative.

(Historically, many of the paradigms come from participles, which is where the second element comes from. And the alignment change seems to have come about by reanalyzing a passive form.)
Thanks! I'll take a closer look there. I actually looked at Hindi & several other Indian languages to redesign the phonology for this conlang.

FWIW, I think the Takelma language of Oregon comes closest to what I was thinking of. Sapir describes it as having a "verb stem" and an "aorist stem". I just need to decide on how my 2 stems would be formed, and I'd like it to be something besides just ablaut. From there, I'd like the conjugations to work differently for the 2, so Hindi & other suggestions may come in useful there.
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

User avatar
kodé
Lebom
Lebom
Posts: 175
Joined: Sun Apr 16, 2006 8:25 pm
Location: Trojan Country

Re: inflection categories using differing affix types

Post by kodé »

This seems fairly common, the more I think about it. I think the root of the phenomenon (partly) is verb forms being built off of historically non-verbal forms, like participles. For example, in Modern Hebrew, while the past and future conjugate their subjects for person, gender, and number with suffixes and prefixes, respectively (just like in your Arabic example), the present conjugates for only gender and number, and uses suffixes that are transparently derived from nominal suffixes. This is due to the MH present being derived from a participle. IIRC something similar happens in Russian.

You get a similar sort of effect in some fluid-S languages, where more "active" or volitional subjects take A(gent) agreement and more "passive" or non-volitional subjects take P(atient) agreement. For example, in Creek/Muskogee/Seminole, P-agreement is a stem-level prefix, while A-agreement is a word-level suffix. Intransitive verbs that can be used fluid-S-ly in Creek thus have suffix agreement with volitional subjects but prefix agreement with non-volitional subjects. IIRC Mayan languages do something different but sort of vaguely reminiscent (two different agreement patterns).
linguoboy wrote:
GrinningManiac wrote:Local pronunciation - /ˈtoʊ.stə/
Ah, so now I know where Towcester pastries originated! Cheers.

User avatar
Frislander
Avisaru
Avisaru
Posts: 836
Joined: Mon Feb 29, 2016 6:34 am
Location: The North

Re: inflection categories using differing affix types

Post by Frislander »

Vardelm wrote:I'm looking for languages where a given inflection or category of inflections use differing types & forms of affixes. I'm thinking of something like Arabic's perfective/past vs. imperfective/non-past, where person is basically marked via suffixes (kataba) or prefixes (yaktubu). I'm also somewhat familiar with Georgian verbs having a 1st person prefix vs. a 3rd person suffix. Is anyone familiar with more examples? For extra credit, whence comes a difference in prefixing vs. suffixing for a given inflectional group?

What I'm working towards is verbal conjugations that have significant differences between a realis paradigm & an irrealis paradigm, parallel to the differences between Arabic's past & non-past.
Probably the best example of what you appears to be getting at that I know of is the Algonquian languages.

So all of the languages (with the possible exception of Mi'kmaq, if someone has a proper grammar of the thing I'd love to have it because I can't find one) distinguish between these two modes called the independent and the conjunct, and this distinction is easily reconstructible to Proto-Algonquian. These kinda do as they say on the tin in most of the languages; the independent appears in main clauses and the conjuct appears in subordinate clauses, though a few of the languages, notably Arapaho and again Mi'kmaq, extended the conjunct to realis independent clauses as well (in Mi'kmaq I think the independent was lost completely, in Arapaho it remained in interrogative and negative sentences among others, so it's along the lines of the realis-irrealis distinction you mention). For my examples I'll use reconstructed Proto-Algonquian forms.

Now onto my main point: the independent and conjunct are strikingly different with regards to the shapes of the person markers (plus a couple of other minor details which are irrelevant to my point). In the independent person marking is acheived by circumfixes, with the prefix marking the person of the sole argument/the highest ranked of the arguments (don't forget that Algonquian has an animacy/empathy hierarchy) while the suffix marks the number of the argument, with an additional suffix in transitives to mark the person of the other argument. There's more to it than that but that's pretty much how it works, e.g.

*nenepe
ne-nepe
1-die
I die

*nenepeʔmena:
ne-nep-eʔm-ena:
1-die-PRES-EXCL
We (excl.) die

*kenepeʔmena
ke-nep-eʔm-ena
2-die-PRES-INCL
We (incl.) die

*newa:pama:wa
ne-wa:pam-a:-w-a
1-see-DIR-PRES-3s
I see him

*newa:pama:wena:na
ne-wa:pam-a:-w-ena:n-a
1-see-DIR-PRES-1EXCL-3
We see him

*newa:pamekwa
ne-wa:pam-ekw-a
1-see-INV-3s
He sees me

In the conjunct, on the other hand, all inflection is acheived by suffixes. Furthermore in there is a much greater degree of fusion between the person and number markers, and in transitive verbs between the subject and object markers as well (compared to the independent mode where the different affixes are comparatively easily segmented). It has been proposed that this indicates that it is the conjunct which is the older paradigm, and the independent a later innovation.

*nepa:ne:
nep-a:n-e:
die-1s-SUBJ
If I die

*nepa:nke:
nep-a:nk-e:
die-1EXCL-SUBJ
If we (excl.) die

*nepankwe:
nep-ankw-e:
die-1INCL-SUBJ
If we (incl.) die

*wa:pamake:
wa:pam-ak-e:
see-1>3s-SUBJ
If I see him

*wa:pamakente:
wa:pam-akent-e:
see-1EXCL>3-SUBJ
If we see him

*wa:pamite:
wa:pam-i-t-e:
see-1-3s-SUBJ
If he sees me
https://frislander.tumblr.com/

First known on here as Karero

Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

Re: inflection categories using differing affix types

Post by Vardelm »

Frislander wrote:Probably the best example of what you appears to be getting at that I know of is the Algonquian languages.
Yep, this looks like a good language family to look at more. I looked up Ojibwe on Wikipedia, and specificially this chart. That's another good example of conjugations varying besides Semitic & Georgian. I think between those different language families I may be able to get enough inspiration for the conjugation. I still need to figure out how the stems might vary (looks like the Ojibwe verb stem stays fairly consistent???), but looking at both finite vs. non-finite forms might be more inspiration than only looking at finite (which is what I've been doing) will help.

I've looked at that Ojibwe Wikipedia article before in relation to one of my other langs, which will be a direct inverse language. Between that, Hindi, & others, it's surprising how often I'll look at a language for one aspect, and then end up coming back to it for something completely different and for a different conlang.
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

User avatar
Frislander
Avisaru
Avisaru
Posts: 836
Joined: Mon Feb 29, 2016 6:34 am
Location: The North

Re: inflection categories using differing affix types

Post by Frislander »

Vardelm wrote:I still need to figure out how the stems might vary (looks like the Ojibwe verb stem stays fairly consistent???
Well I can't speak for Ojibwe specifically (though given that like much of the "Central" languages it's quite conservative it'll probably be like this) but Algonquian languages as a whole tend to show stem alternation based on animacy (with the subject in intransitives and the object in transitives), but that's the bit about Algonquian animacy that people tend to overlook (this chart on the Ojibwe Wikipedia page doesn't nearly do it justice). I didn't mark the stem classes in my glosses because they weren't relevant to my point, but here's an example for you (again Proto-Algonquian):

*newa:pama:wa
ne-wa:pam-a:-w-a
1-see.TA-DIR-PRES-3s
I see him

vs.

*newa:panta:ni
ne-wa:pant-a:-n-i
1-see.TI-DIR-PRES-INA
I see it

These would come out in Ojibwe as (by my best guess) niwaabamaaw and niwaabandaan respectively.

Probably the best thing I could recommend short of emailing you some PDFs (which I could do for you, just send me a PM) is to recommend you read Leonard Bloomfield's 1946 sketch of Proto-Algonquian and read through the sections on verb finals.
https://frislander.tumblr.com/

First known on here as Karero

Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

Re: inflection categories using differing affix types

Post by Vardelm »

Frislander wrote: Probably the best thing I could recommend short of emailing you some PDFs (which I could do for you, just send me a PM) is to recommend you read Leonard Bloomfield's 1946 sketch of Proto-Algonquian and read through the sections on verb finals.
A very quick Google search got me [urlhttps://home.cc.umanitoba.ca/~oxfordwr/bloomfield1946/contents.html]the sketch, complete with a table of contents[/url]. :)

Frislander wrote:Algonquian languages as a whole tend to show stem alternation based on animacy...
I'm still VERY new to Algonquian languages (actually North American aboriginal languages in general), so I may well be missing some info. I scanned through the section on verb finals, and also the one on verb orders & modes. From what I see, all of the finals, etc. are prefixes or suffixes. After reading your last 2 posts, I realized that I may have a very narrow interpretation of what constitutes a "stem" or a "change to a stem". This is due to still being an amateur, linguistics hobbyist, and probably because my initial foray into linguistics was with Tolkien's Khuzdul and Semitic languages. When I say "how verb stems might vary", I'm thinking of non-concatenative changes, like ablaut, consonant mutation, metathesis, etc. It's only now that I see how narrow my focus is regarding stems. I might need to expand that a bit for the language I'm working on.
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

User avatar
Frislander
Avisaru
Avisaru
Posts: 836
Joined: Mon Feb 29, 2016 6:34 am
Location: The North

Re: inflection categories using differing affix types

Post by Frislander »

Vardelm wrote:
Frislander wrote: Probably the best thing I could recommend short of emailing you some PDFs (which I could do for you, just send me a PM) is to recommend you read Leonard Bloomfield's 1946 sketch of Proto-Algonquian and read through the sections on verb finals.
A very quick Google search got me [urlhttps://home.cc.umanitoba.ca/~oxfordwr/bloomfield1946/contents.html]the sketch, complete with a table of contents[/url]. :)
Oh yeah I meant to put the link to that I thought I had I must have forgot.
Frislander wrote:Algonquian languages as a whole tend to show stem alternation based on animacy...
I'm still VERY new to Algonquian languages (actually North American aboriginal languages in general), so I may well be missing some info. I scanned through the section on verb finals, and also the one on verb orders & modes. From what I see, all of the finals, etc. are prefixes or suffixes. After reading your last 2 posts, I realized that I may have a very narrow interpretation of what constitutes a "stem" or a "change to a stem". This is due to still being an amateur, linguistics hobbyist, and probably because my initial foray into linguistics was with Tolkien's Khuzdul and Semitic languages. When I say "how verb stems might vary", I'm thinking of non-concatenative changes, like ablaut, consonant mutation, metathesis, etc. It's only now that I see how narrow my focus is regarding stems. I might need to expand that a bit for the language I'm working on.
I think you might be confusing "stem" with "root". Exact definitions differ but essentially the root is completely indivisible morphologically speaking. The stem on the other hand is what you add inflection to. So it can be the bare root but it can also be a root plus derivational affixes or a compound. So Algonquian doesn't show alternation within the root (with the exception of initial change, but that's irrelevant to animacy) but it does show alternations in the stem, as you have seen. This can all come in one piece as in Algonquian, however this is not always the case, e.g. in Athabaskan languages where a verb root may also come with an obligatory prefix to complete its meaning but several inflectional prefixes may intervene between those parts.

You might want to take a look at Goddard 1990 for more on Algonquian stem derivation (tell me if the link doesn't work).
https://frislander.tumblr.com/

First known on here as Karero

Vardelm
Avisaru
Avisaru
Posts: 329
Joined: Sat Sep 20, 2008 2:37 pm
Contact:

Re: inflection categories using differing affix types

Post by Vardelm »

Frislander wrote:I think you might be confusing "stem" with "root".
Yep. I'm using them interchangeably. I've now learned something new today!

Frislander wrote:You might want to take a look at Goddard 1990 for more on Algonquian stem derivation (tell me if the link doesn't work).
Link works. Eyes going crossed. Good stuff. Thanks!
Tibetan Dwarvish - My own ergative "dwarf-lang"

Quasi-Khuzdul - An expansion of J.R.R. Tolkien's Dwarvish language from The Lord of the Rings

hwhatting
Smeric
Smeric
Posts: 2315
Joined: Fri Sep 13, 2002 2:49 am
Location: Bonn, Germany

Re: inflection categories using differing affix types

Post by hwhatting »

zompist wrote:Then, there's the paradigms that inflect with person/number, and those which go with gender/number.
This is also the case in Russian - the past tense and conditional mood inflect for gender and number, the present / future tense for person and number. The reasons are similar to those for Hindio as well, the present / future tense continue Indo-European finite paradigms, while the past tense and conditional mood go back to a construction with a participle.

Post Reply