[Lingtyp] Identifying a type of possessive constructions

[tsukida at for.aichi-pu.ac.jp]

Dear Yuan-Lin Mickey Yang, 

My name is Naomi Tsukida and I study Truku Seediq. I am glad that you study Seediq. 
In Truku, same pairs exist. The following example is from my Ph.D thesis (Tsukida 2009, p.316). 

(1)	a.	malu    	ka  	lenglungan 	senaw 	nii.
AV.good	NOM	thought    	man   	this
Lit. This man's thought is good. = This man is honest.

      	b.	malu   	lenglungan=na    	ka  	senaw 	nii. 
AV.good	thought=3s.GEN   	NOM	man   	this  
Lit. This man is good-thought. = This man is honest.

I named such construction as “possessee demotion”. The expression for what is possessed is demoted from subject NP into the predicate. In my analysis, the possessor is the subject in the b. sentence, and the possessee is a part of the predicate. 

One may want to call such construction “possessor promotion”, but that term is reserved for such pairs as below, where possessor is promoted to Topic position. (Tsukida 2009, p.315). 

(2)	a.	wada  	geguy-un 	ka  	patas 	lawking.
is:gone 	steal-GV1 	NOM	book	Lawking
Lawking's book was stolen.

	b.	lawking 	'u,  	wada   	geguy-un 	ka  	patas=na.
Lawking	CNJ 	is:gone 	steal-GV1 	NOM	book=3s.GEN
Lawking's book was stolen.

Back to possessee demotion, the subject may be A, P or S, but must be the subject of the clause. The subject is S in (1) above, and it is P in (3), and A in (4).

(3)	a.	sehengi-an=mu       	ka  	hangan 	senaw 	gaga.
forget-GV2=1s.GEN  	NOM	name  	man   	that
I forgot that man's name. 

      	b.	sehengi-an=mu      	hangan=na     	ka  	senaw 	gaga.
forget-GV2=1s.GEN 	name=3s.GEN	NOM	man   	that
I forgot that man's name.

(4)	a.	'ini  	kerut        	baga     	ka  	renabaw	qehuni 	nii.
NEG	AV.NFIN.cut	hand.OBL 	NOM	leaf    	tree.GEN	this
The leaves of this tree do not cut hands.

    	b.	'ini 	kerut        	baga     	renabaw=na  	ka   	qehuni 	nii.
NEG	AV.NFIN.cut	hand.OBL 	leaf=3s.GEN 	NOM	tree   	this
The leaves of this tree do not cut hands.

What is allowed to demote differs in terms of possession cline (Tsunoda 1996), according to which of S, P or A the subject is. 
S: body parts, attributes, clothes, kinterms, pets, works, other possessees.
P: body parts, attributes, clothes, kinterms, pets, works, other possessees.
A: body parts, attributes, ?clothes, *kinterms, *pets, *works, *other possessees.
(Tsukida 2009, pp.612-5)

Best wishes,
Naomi Tsukida

Bibliography

Tsunoda,Tasaku. 1996. The possession cline in Japanese and other languages. Hilary Chappell and William McGregor (eds.). The Grammar of Inalienability. De Gruyter Brill. pp.565-630

Tsukida, Naomi. 2009. Sedekku-go-no Bunpou (Grammar of Seediq). Ph.D. Dissertation submitted to the University of Tokyo. (In Japanese.) (Please check ResearchGate)





________________________________________
差出人: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> が Yuan-Lin Yang via Lingtyp <lingtyp at listserv.linguistlist.org> の代理で送信
送信: 2026 年 2 月 21 日 (土曜日) 13:55
宛先: lingtyp at listserv.linguistlist.org <lingtyp at listserv.linguistlist.org>
件名: [Lingtyp] Identifying a type of possessive constructions

Dear all,I am currently working on Toda Seediq, an under-described Formosan (Austronesian) language spoken in Taiwan. Recently I just discovered a kind of possessive constructions that seem to be seldom studied by previous Formosan literature (not sure for the typology literature for the moment, tho):(1) Naqah k<n>dus-an=na                                     ka      bubu=mu.      bad     DEP<PFV>live-LOCNMLZ=3SG.GEN  NOM  mother=1SG.GEN      'My mother led a difficult life.'(2)  Ini       k-paru      pahung=na                   ka        hiya.      NEG   DEP-big   gallbalder=3SG.GEN    NOM   3SG.NEUT      '(S)he is not bold.'Toda Seediq is predicate-initial and PSA-final, and there is only one overt case marker in the language, namely the nominative ka, which can be seen as marking the boundary between the predicate and the argument the former is predicated of in a clause. In both (1) and (2) above, the predicate parts consist of complex NPs with genitive pronoun =na marking the possessor, yet there is also another nominatively marked PSA argument that is co-referential with the genitive =na. This is a less-common type of possessive cxns in the language (and perhaps also in other Formosan langauges), because (i) usually the adnominal possessive cxns only have either the genitive pronouns or the overt NP possessors that follow the possessums and are not marked by ka (as in (3) and (4) below), and (ii) the genitive pronoun and the PSA argument are not in the same NPs. (3)  rulu=mu      wheel=1SG.GEN      'my car'(4)  laqi     Emi      child    PN      'Emi's child'I can only tell syntactically (1) and (2) might involve clitic doubling, but this label is not satisfying for me. Hence, I tentatively analyze the cxns in question as hanging topic cxns, but I still wonder by what type of possession should I describe such a kind of possessive cxns. Do they exhibit a kind of external possession? Or do they belong to other types of split possession or adopt a strategy I am not familiar with? I wonder if any of you have have found the same or similar phenomena in the languages you work on, and can thus shed some lights on this issue for me.Best regards,Yuan-Lin Mickey Yang, MA Student, Department of English, National Taiwan Normal University.