Non-finite clauses in highly-inflected languages

Discussion of natural languages, or language in general.
Post Reply
User avatar
Trebor
Lebom
Lebom
Posts: 207
Joined: Sat Apr 10, 2004 7:36 pm
Location: Canada

Non-finite clauses in highly-inflected languages

Post by Trebor »

Hi all,

For the sake of clarity, let us first define the technical term employed in the subject line. According to this page:
Non-finite clauses are subordinate clauses containing a non-finite verb phrase. Therefore, they cannot stand alone as the main clause. Non-finite clauses are typical of written English. Since non-finite clauses do not show tense or mood distinctions and there is usually no subordinating conjunction and even the subject of the verb is usually missing, the relationship of non-finite clauses to the rest of the clause may need some interpretation. Pay attention to non-finite clauses when you translate from English into Finnish! A corresponding "lauseenvastike" seldom works in Finnish.
I am very interested to know how the highly-inflected, especially agglutinating and polysynthetic, languages which you speak, are learning, or have read about handle the below sentences from the above link. Feel free to include additional examples of the same or other (perhaps-overlooked) types of non-finite clauses as you see fit.

Translations and interlinear glosses would be very helpful.
(1) to-infinitive
without S: The best thing would be to tell everybody.

with S: The best thing would be for you to tell everybody.

[An example of my own which I like better:

Our goal is to finish the project by Friday.

The boss's goal is for us to finish the project by Friday.]

(2) bare infinitive

without S: All I did was hit him on the head.

with S: Rather than you do the job, I’d prefer to finish it myself.

(3) -ing form = present participle

without S: Leaving the room, he tripped over the mat.

with S: Her aunt having left the room, I asked Ann for some help.

(4) -ed participle = past participle

without S: Covered with confusion, she hurriedly left the room.

with S: The discussion completed, the chairman adjourned the meeting for half an hour.
Thanks. :)

User avatar
Trebor
Lebom
Lebom
Posts: 207
Joined: Sat Apr 10, 2004 7:36 pm
Location: Canada

Re: Non-finite clauses in highly-inflected languages

Post by Trebor »

*crickets chirping*

Is everyone stumped? :P

User avatar
linguoboy
Sanno
Sanno
Posts: 3681
Joined: Tue Sep 17, 2002 9:00 am
Location: Rogers Park/Evanston

Re: Non-finite clauses in highly-inflected languages

Post by linguoboy »

Trebor wrote:Is everyone stumped? :P
Speaking for myself, I'm not stumped, I just can't be arsed. I don't know Osage well enough to answer the questions without a grammar refresher and it's not on my shortlist of language priorities right now.

User avatar
Whimemsz
Avisaru
Avisaru
Posts: 690
Joined: Fri Jun 20, 2003 4:56 pm
Location: Gimaamaa onibaaganing

Re: Non-finite clauses in highly-inflected languages

Post by Whimemsz »

Ojibwe uses a lot of different constructions which can be translated into English as non-finite clauses (there are no real non-finite clauses in Ojibwe in the sense of verb lacking inflection for person/number/TAM, unless you count "participles" which are basically relativized verbs used as nouns, but they take slightly different suffixes than other verbs and in any case are treated syntactically as nouns afaict). In many cases non-finite English clauses correspond to the use of the "conjunct" order in Ojibwe -- a different set of verbal inflections used primarily (but not solely) in subordinate clauses. Conjunct inflection would be used in most of the cases translatable with English participles.

Some relations like "in order to" are generally expressed using what are called "relative roots" (which can be proclitics or a part of the verb root depending on the specific verb). Often the relationship needing a non-finite clause in English can be indicated using a preverb (verbal proclitic) in Ojibwe, such as nanda- "try to...".

Finally, some of those constructions just wouldn't need any subordinate clause or other special marking in Ojibwe at all, e.g. I suspect the most natural way to say "all I did was hit him on the head" is to say something like mii eta gii-pasindibe'og = it's only PAST-slap-head-1sg>3 (though it's a little more complicated because actually this uses conjunct marking as well, but not really for the same reasons as noted above so the example is, I think, still valid).

However, I'm not able to translate most of the example sentences you gave, partly because of my own lack of fluency.

merijn
Lebom
Lebom
Posts: 207
Joined: Thu Dec 21, 2006 10:36 pm
Location: Utrecht Overvecht

Re: Non-finite clauses in highly-inflected languages

Post by merijn »

OK I'll bite. My specialist language is Zulu. I am not fluent in it (to my great shame), and this is not an area I have actively investigated. Zulu is not an extreme when it comes to rich morphology, but still it is much more inflected than about any West-European language.
Zulu has an infinitive, which is non-finite. It also employs a group of tenses called participials, which I wouldn't call non-finite (pariticipials are inflected for about the same categories as principal tenses (prinicipal tenses are the tenses that are indicative and not participial), and it can license subjects just the same way as principal tenses), but they are only used in subordinate clauses and they can be used without a subordinating conjunction. They are amongst others used in adverbial clauses. A second tense that is often used when English uses an non-finite verb forms is the subjunctive, which again, I would say is a finite verb form; it can be used on its own, and it can license subjects just the same way as principal and participial tenses can.
(1) to-infinitive
without S: The best thing would be to tell everybody.
I think Zulu uses an infinitive here

into enhle i-nga-ba ng-uku-tshel-a bonke
9thing 9good sm9-pot-become cop-inf-tell-fv everybody*

with S: The best thing would be for you to tell everybody.
It is possible to have a subject with an infinitive (the subject has possessive marking in that case), but my (far from native) instincts say that that sentence would probably be translated with a subjunctive

into enhle i-nga-ba ng-ukuthi u-tshel-e bonke
9thing 9good sm9-pot-become cop-conj sm2s-tell-subj everybody
[An example of my own which I like better:

Our goal is to finish the project by Friday.
I think Zulu uses an infinitive here
inhloso yethu ng-uku-yi-qed-a iphrojekthi ngoLwesihlanu
9goal 9our cop-inf-om9-finish-fv 9project by.Friday

The boss's goal is for us to finish the project by Friday.]
Again, I don't think you would use an infinitive in Zulu, but a subjunctive

inhloso yo-mphathi ng-ukuthi si-yi-qed-e iphrojekthi ngoLwesihlanu
9goal 9poss-1boss cop-conj sm1p-om9-finish-subj project by.friday
(2) bare infinitive

without S: All I did was hit him on the head.
konke engi-kw-enz-ile kw-a-ku-ng-uku-m-shay-a e-khanda
17everything rela1s-om17-do-perf sm17-past-sm17-cop-inf-om1-hit-fv loc-5head
with S: Rather than you do the job, I’d prefer to finish it myself.
I genuinely don't know how to do this in Zulu.

(3) -ing form = present participle

without S: Leaving the room, he tripped over the mat.
Here you use the aforementioned participial

e-shiy-a e-kamelw-eni u-guzuk-a o-cans-ini
prt1-leave-fv loc-5room-loc sm1-stumble-fv loc-11mat-loc

with S: Her aunt having left the room, I asked Ann for some help.
again Zulu would use the participial
ubabebekazi wakhe e-shiy-e e-kamelw-eni ng-a-m-cel-a usizo
1aunt 1her prt1-leave-perf loc-5room-loc sm1s-past-om1-request-fv 11help
(4) -ed participle = past participle

without S: Covered with confusion, she hurriedly left the room.
I am going to translate "covered with confusion" literally. Again Zulu uses the participial, this time a passive participial

e-mbes-iwe nge-sidididi a-sheshe a-shiy-e ikamelo
prt1-cover-pass.perf with-7cofusion sm1-hurry sm1-leave-sbj 5room

with S: The discussion completed, the chairman adjourned the meeting for half an hour.
again participial
ukuxoxisana ku-phel-ile, usihlalo u-wu-hlehlis-e ngo-hhafu we-hola umhlangano
15discussion prt15-complete-perf, chairperson sm1-om3-postpone-perf with-1half 1poss-hour 3meeting

*glosses: numbers are noun class or if followed by s=singular or p=plural to person sm=subject marker pot=potential cop= copular particle inf=infinitive fv=default final suffix (which is a) om=object marker conj=conjunction subj=subjunctive poss=possessive rela=relative agreement perf=perfect past=past prt=participial subject marker pass=passive

User avatar
gach
Avisaru
Avisaru
Posts: 472
Joined: Mon Feb 17, 2003 11:03 am
Location: displaced from Helsinki

Re: Non-finite clauses in highly-inflected languages

Post by gach »

Trebor wrote:According to this page
Boy had you found an old style page from my university. Though to be honest, it doesn't surprise me to find relics like this. The pages are a mess in general and have become harder and harder to navigate throughout the years.

To come back to the subject, I'll try to give some basic description of the Finnish non-finites as I implied in the reflexivity thread. It's not really a that highly inflected language but the infinitive system is still pretty interesting and fairly rich compared to what you'll find in many related languages. I won't even attempt to go for a comprehensive description of everything there is as that would take ages. Instead I'll give some sort of a review of the more commonly used non-finites and give some remarks of their use.

The Finnish non-finites are traditionally divided into infinitives and participles. The infinitives are basically everything that isn't a participle and the three morphologically distinct forms are traditionally just marked by numbers as the first, second and third infinitive. In Estonian descriptions I've seen the corresponding forms being called infinitive, gerund and supine. The only difference here is what convention you choose to use for the naming. The infinitives can also be called the TA, TE and MA infinitives based on their underlying morphological forms. This seems to be what most researchers prefer and it really feels more appropriate than hijacking more Latin terms or just numbering everything.

1st or TA infinitive

This category has two forms with very different meanings. The short form is what you'd expect an infinitive to be and is often just called plain "infinitive" with no further attributes. It acts as the regular simple complement for other verbs

Halua-n sano-a.
want-SG1 say-INF1
"I want to say."

This form is marked by one of the allomorphs of the TA marker suffixed with the unwritten final glottal stop. The final glottal stop hasn't got any formal function in the modern language but its origin as the reflex of the ancient lative case suffix *-k is worth noting. We'll see that the process of deriving semantically bleached infinitives from ones with local semantics is still ongoing.

The long form of the 1st infinitive consists of the TA marker marked for the translative case -kse and suffixed with an obligatory possessive suffix. It's typically used to form final subordinate clauses

Soit-i-n sinu-lle sano-a-kse-ni kiitos.
call-PST-SG1 you-ALL say-INF1-TRANS-POSS.S.SG1 thank.you
"I called you to say thanks." ~ "I called you to thank"

This construction can only be used for final clauses that have the same subject as the main clause and the possessive suffix marking must agree with the finite verb of the main clause in person and number. Other types of clauses with final meanings must be formed with finite verbs (typically in the conditional mood) with the conjunction jotta ("so that")

Soit-i-n sinu-lle, jotta vo-isi-mme jutel-la.
call-PST-SG1 you-ALL so.that can-COND-PL1 chat-INF1
"I called you so that we could chat."

The long form of the 1st infinitive also occurs in some other construction, some of which are pretty lexicalised

Sa-i-t-ko syö-dä-kse-si?
get-PST-SG2-Q eat-INF1-TRANS-POSS.S.SG2
"Did you get enough to eat?"

Näh-dä-kse-ni kaikki on valmis-ta.
see-INF1-TRANS-POSS.S.SG1 all be.SG3 ready-PART
"As far as I can see everything is ready."

2nd or TE infinitive

This infinitive can be inflected for two cases, the otherwise practically lexicalised instructive and the inessive. Both are used as adverbials so calling them gerunds wouldn't be out of place at all. The 2nd infinitive instructive is used as an adverb of manner

Hän tul-i juos-te-n.
(s)he come-PST run-INF2-INSTR
"He came running."

while the inessive form describes simultaneous background action

Hän tul-i auringo-n laski-e-ssa.
(s)he come-PST sun-GEN set-INF2-INE
"He came as the sun was setting."

As you can see, this non-finite clause can be used with a subject different from the main clause. In this case the subject of the non-finite clause gets genitive marking. If the non-finite clause has the same subject as the main clause, the non-finite verb must agree in person with the verb of the finite clause similarly to the long form of the 1st infinitive

Hän laulo-i juos-te-ssa-an.
(s)he sing-PST run-INF2-INE-POSS.S.SG/PL3
"He sang as he was running."

3rd or MA infinitive

This is the largest class of infinitive inflections in Finnish and is the place of the most active development. Basically it's the most semantically loaded class of infinitive forms in modern Finnish but at the same time it's spreading towards the more purely syntactic functions of the 1st and 2nd infinitives. This infinitive is marked by the -ma morpheme, as you could have guessed from the name, and is commonly inflected for the cases inessive (sano-ma-ssa), elative (sano-ma-sta), illative (sano-ma-an), adessive (sano-ma-lla) and abessive (sano-ma-tta).

The inessive form describes ongoing activity and gives a figurative location for other activities

Istu-n syö-mä-ssä.
sit-SG1 eat-INF3-INE
"I'm sitting and eating."

The location metaphor is very prominent as this form can answer to the question missä ("where"). With the copular verb olla the 3rd infinitive inessive marks the progressive aspect

Ole-n syö-mä-ssä.
be-SG1 eat-INF3-INE
"I'm eating."

Interestingly some verbs (mennä, "go"; tulla, "come"; lähteä, "leave") replace the 3rd infinitive inessive of the progressive construction with the inessive of an action nominal formed by the -O- affix (men-o-ssa, tul-o-ssa, lähd-ö-ssä)

Ole-n tul-o-ssa koti-in.
be-SG1 come-O-INE home-ILL
"I'm coming home."

The elative and illative forms expand the locational metaphor system of the inessive form by giving ways to say "from doing" and "into doing". Thus you can say

Tule-n syö-mä-stä.
come-SG1 eat-INF3-ELA
"I'm coming from eating." ~ "I'm coming from where I ate."

and

Mene-n syö-mä-än.
go-SG1 eat-INF3-ILL
"I'm going eating"

Similar to the inessive form, these forms can answer to the questions mistä ("where from") and mihin ("where to"). However, many verbs require these forms as their complements and in this use the forms loose much of their locational meaning. Thus you get for example

Lakkaa huuta-ma-sta!
stop.IMP shout-INF3-ELA
"Stop shouting!"

and

Hän rupes-i huuta-ma-an.
(s)he begin-PST shout-INF3-ILL
"He began to shout."

Especially the 3rd infinitive illative is a common verb complement and is replacing the 1st infinitive short form for some verbs. The classic example of this is the verb alkaa (also "begin"). In standard grammar you are supposed to use the 1st infinitive as in

Ala-n huuta-a.
begin-SG1 shout-INF1
"I begin to shout."

but colloquially many people use exclusively

Ala-n huuta-ma-an.
begin-SG1 shout-INF3-ILL
"I begin to shout."

This use is bleeding into the written language and is a common target of jokes* and grammar nazis. However, grammarians have been telling us for a while now that in something like a decade this use will be acknowledged also in the formative grammar. The development we are seeing here, where a non-finite verb form marked for a local case begins to loose its locational meaning, is exactly parallel to the development of the 1st infinitive short form where it has lost its original lative meaning.

An additional use of the 3rd infinitive illative is to mark future actions together with the verb tulla ("come")

Tule-n vielä huuta-ma-an usein.
come-SG1 still shout-INF3-ILL often
"I'll still be shouting often."

The two remaining common case forms, adessive and abessive, are manner adverbials and describe respectively the means of doing ("by doing") and the lack of doing ("without doing")

Valvo-n myöhään juo-ma-lla kahvi-a.
stay.awake-SG1 late drink-INF3-ADE coffee-PART
"I'm staying up late by drinking coffee."

Valvo-n myöhään väsy-mä-ttä.
stay.awake-SG1 late get.tired-INF3-ABE
"I'm staying up late without getting tired."

This use of the cases is parallel to their use with nouns where adessive is used to mark instruments and abessive the lack of an instrument. The 3rd infinitive adessive is very close in its meaning with the 2nd infinitive instructive and they can often be used interchangeably. There is still, however, difference in their semantics as the 3rd infinitive adessive marks more concretely the means of doing while the 2nd infinitive instructive marks plain manner.

Further stuff with MA

The -ma morpheme of the 3rd infinitive (or an allomorph -m-) also appears in many other inflections. Firstly two of the participles are based on it, I'll come to these later. Secondly it can be seen in the productive action nominal -minen. This forms general nouns referring to any action such as huutaminen ("act of shouting") and unlike the other less productive action nominals it's resistant against lexical drift. It's also used as the complement of verbs that take nominal objects and can't take an non-finite verb

Lopeta huuta-minen!
stop.IMP shout-ACT.NOM
"Stop shouting!"

The action nominal also occurs in archaic constructions denoting obligation or prohibition in which use it's traditionally labelled as the 4th infinitive. You don't see this in the modern language and I'll just note that the Wikipedia article on Finnish grammar has an example of this.

A third place where you find the -ma morpheme is in the so called 5th infinitive -mAisillA + POSS.S which has the meaning "just about to do"

Ole-n nukahta-maisilla-ni.
be-SG1 fall.asleep-just.about-POSS.S.SG1
"I'm just about to fall asleep."

This construction isn't too common because of it's restricted area of usage but it's definitely hanging around and not disappearing.

Finally the -ma morpheme appears as a plain deverbal noun derivation as in kuolema ("death") from kuolla ("to die"). This might be closest to the original use of the morpheme as there's also the word kalma ("death", archaic) which is the reflex of the same -ma ending attaching itself to the same verb root during the Proto Uralic stage.


* rektio.png, a badge telling that "you can talk about rection disorder"

User avatar
gach
Avisaru
Avisaru
Posts: 472
Joined: Mon Feb 17, 2003 11:03 am
Location: displaced from Helsinki

Re: Non-finite clauses in highly-inflected languages

Post by gach »

Splitting the Finnish non-finites into two posts.


Participles

Let's turn to the participles. The basic participle system consists of a two by two grid of present/perfect and active/passive participles

sano-va, say-PRES.A.P, "saying" ~ "sth that says"
sano-ttava, say-PRES.P.P, "sth that is said"
sano-nut, say-PRF.A.P, "said" ~ "sth that has said"
sano-ttu, say-PRF.P.P, "said" ~ "sth that has been said"

These are used in a straight forward manner to build relative clauses alongside finite relative clauses using relative pronouns and I won't be going further into that here. Moreover the perfect participles are used with the copula to form the perfect and past perfect tenses like in English and much of western Europe.

The last two participles use the -ma morpheme and are the so called agent participle and the negative participle

sano-ma, say-AG.P, "sth said by someone"
sano-maton, say-NEG.P, "sth that is not said"

The negative participle should again be pretty self explanatory. The agent participle is basically a passive participle which indicates the subject of the relativised verb by a possessive construction

sinu-n sano-ma-si sana
you-GEN say-AG.P-POSS.S.SG2 word
"the word said by you" ~ "the word you said"

Two less obvious types of non-finite clauses involving participles are non-finite temporal clauses and complex non-finite complement clauses. We've already seen the temporal clause denoting ongoing actions which uses the 2nd infinitive inessive

Hän tul-i auringo-n laski-e-ssa.
(s)he come-PST sun-GEN set-INF2-INE
"He came as the sun was setting."

There's a corresponding construction for preceding actions that is based on completely different morphology and uses the partitive of the passive perfect participle (perhaps retaining the original ablative/departing meaning of the partitive case)

Hän tul-i auringo-n leske-ttu-a.
(s)he come-PST sun-GEN set-PRF.P.P-PART
"He came after the sun had set."

This is a nice example that semantically symmetric constructions don't have to be symmetric in their morphology or syntax. To go further, clauses expressing following background actions must be formed as finite subordinate clauses or be build with derived deverbal nouns

Hän tul-i ennen, kun aurinko ol-i laske-nut.
(s)he come-PST before than sun be-PST set-PRF.A.P
"He came before the sun had set."

Hän tul-i ennen auringonlasku-a.
(s)he come-PST before sun.set-PART
"He come before the sunset."

The complex non-finite complement clauses use the four core present/perfect participles. The present participles correspond to present tense and the perfect participles to past tense in the complement. Passive participles may be used to indicate indefinite subject person. The subject of the complements is marked by genitive as in the temporal clauses above. These clauses act as objects of the main clause and so the participles get accusative marking (formally identical to the genitive)

Tiedä-n auringo-n laske-va-n
know-SG1 sun-GEN set-PRES.A.P-ACC
"I know that the sun is setting."

Tiedä-n auringo-n laske-nee-n
know-SG1 sun-GEN set-PRF.A.P-ACC
"I know that the sun has set."

These clauses are always replaceable by finite subordinate clauses beginning with the conjunction että ("that")

Tiedä-n, että aurinko laske-e.
know-SG1 that sun set-SG3
"I know that the sun is setting."

Tiedä-n, että aurinko on laske-nut.
know-SG1 that sun be.SG3 set-PRF.A.P
"I know that the sun has set."


That's about it for the basic system. There are some additional lesser used infinitive forms and loads of minor constructions that keep researchers busy or would give a conlanger stuff to consider for the rest of the life.

I'll see if I can find enough coherent material from any other significantly different language to make another useful contribution.


Some glosses:
INF1 - 1st infinitive
INF2 - 2nd infinitive
INF3 - 3rd infinitive
PRES.A.P - present active participle
PRES.P.P - present passive participle
PRF.A.P - perfect active participle
PRF.P.P - perfect passive participle
AG.P - agent participle
NEG.P - negative participle
PST - past
COND - conditional
IMP - imperative
PASS - passive
ACC - accusative
GEN - genitive
PART - partitive
INE - inessive
ELA - elative
ILL - illative
ALL - allative
ADE - adessive
ABE - abessive
TRANS - translative
INSTR - instructive
POSS.S - possessive suffix
Q - Interrogative particle

Post Reply