A: ho’, si-pawas=misu
4.4.1.1 Interrogative nouns
Interrogative nouns may function as predicates of equational constructions and
precede subjects formed by a noun or a nominalized verbal clause, as exemplified by ’ima ‘who’ and ’amol ‘what’ in (4.19)-(4.20).
(4.19) a. ’ima ka’ ’ule’ hani?
who Nom child this
‘Who is this child?’
a’. ’ima (ka’) ralu’=su?
who (Nom) name=2S.Gen ‘What is your name?’
b. ’ima ka’ cyel t<um>ahuk ci ’ucyux?
who Nom Exi.Rem cook<AF>cook Acc fish ‘Who is cooking fish?’
b’. ’ima ka’ b<in>ah-en ni watan hira?
who Nom hit<Prf>hit-LF Gen Watan yesterday ‘Who was hit by Watan yesterday?’
(4.20) a. ’amol ka’ haca?
what Nom that ‘What is that?’
a’. ’amol ka’ ’ule’=su?
what Nom child=2S.Gen
‘What gender is your child?/*What is your child?’
b. ’amol ka’ cyel=su tahk-un?
what Nom Exi.Rem=2S.Nom cook-PF ‘What are you cooking?
b’. ’amol (ka’) wah-an=su sani?
what (Nom) come-LF=2S.Gen here ‘Why do you come here?’
Lit: ‘What do you come here for?’
As shown in (4.19), ’ima ‘who’ is used to question about personal names or identity of the agent or patient of an event. When ’amol ‘what’ precedes subjects referring to human beings as in (4.20a’), it is used to inquire gender; when preceding verbal clauses initiated by LF-marked motion verbs as in (4.20b’), ’amol ‘what’ is used to
seek for reasons.21
Examples of other interrogative nouns as predicates are given in (4.21)-(4.24).
(4.21) ’ima-n ka’ pila haca?
how much Stat<Prf>big Nom house=2S.Gen ‘How big is your house?’
21 In Chang (1996), it is mentioned that in Squliq Atayal, Mayrinax Atayal and Paran Seediq, when the interrogative word meaning “what” occurs before LF-marked motion verbs, it can be interpreted as
“why’ and used to seek for reasons. In Paran Seediq, both LF-marked and PF-marked verbs such as mah-an ‘drink-LF’, mah-un ‘drink-PF’ and lawa-an ‘yell-LF’, lawa-un ‘yell-PF’ can make the preceding maanu ‘what’ interpreted as “why”.
b’. kanon ka’ m<in>sa’ tayhoku (*ka’) temu?
when Nom AF<Prf>go Taipei (*Nom) Temu ‘When did Temu go to Taipei?’
Regarding the semantic analysis of interrogative nouns, three things are noted.
(i) As a nominal predicate, when ’inu refers to the subject of a verbal clause headed by a LF-marked verb as in (4.22c), it is interpreted as “where”;
otherwise, the predicate ’inu is usually translated as “which”.
(ii) Both pira’ and kito’ may refer to unknown quantity of countable and uncountable nouns, as shown in (4.23a-b). However, only kito’ can be used to inquire information about measurement.
(iii) The temporal interrogative word kanon ‘when’ may precede a demonstrative pronoun that refers to an event, as in (4.24a). While kanon ‘when’ precedes a complete verbal clause headed by ka’, it is noted that only one of the two nominative case markers ka’ may appear, as shown in (4.24b-b’).
As predicates, interrogative nouns may attract pronoun clitics that co-refer with them or indicate possessors of them,22 as shown in (4.25).
(4.25) a. ’ima=su ’isu?
who=2S.Nom 2S.Neu ‘Who are you?’
b. ’ima=su ka’ yuma?
who=2S.Gen Nom Yuma
‘Who is Yuma to you?’
When pronoun clitics neither co-refer with interrogative nouns as predicates nor indicate possessors of them, they have to be attached to the main verb of the nominalized verbal clause, as illustrated in (4.26).
22 In Huang (1996), piya’ ‘how much’ in Mayrinax Atayal is regarded as an interrogative verb because
(4.26) a. pira’ gal-un=su la?
how many/much take-PF=2S.Gen Part ‘How many/much did you take?’
a’. * pira’=su gal-un la?
how many/much=2S.Gen take-PF Part b. ’inu ka’ ’<in>tax-an=su?
where Nom see<Prf>see-LF=2S.Gen
‘Where did you look at?’
b’. * ’inu=su ka’ ’<in>tax-an?
where=2S.Gen Nom see<Prf>see-LF c. kanon musa’=su tayhoku?
when AF.go=2S.Nom Taipei
‘When do you go to Taipei?’
c’. * kanon=su musa’ tayhoku?
when=2S.Nom AF.go Taipei
In addition to being predicates in equational sentences, interrogative nouns referring to non-subject arguments may remain in-situ and be preceded by case markers or suffixed with -an. Some examples of interrogative nouns as arguments are given in (4.27)-(4.29) for illustration.
(4.27) a. mak-magiy laha ci ’ima ka’ watan?
want-AF.leave 3P.Neu Com who Nom Watan
‘Who does Watan want to leave with?’
b t<in>ahuk ni ’ima ka’ ’ucyux hani?
cook<Prf.PF>cook Gen who Nom fish this ‘Who cooked this fish?’
c. m<in>ahiy ’ima-n ka’ watan?
AF<Prf>hit who-Loc Nom Watan ‘Who did Watan hit?’
d. t<in>ahuk=su (ni) ’ima-n ka’ ’ucyux hani?
cook<Prf.PF>cook=2S.Gen (Ben) who-Ben Nom fish this
‘For whom did you cook this fish?’
(4.28) a. t<um>ahuk=su ci ’amol?
cook<AF>cook=2S.Nom Acc what ‘What do you cook?’
b. k<in>ac-an na ’amol ka’ walis?
bite<Prf>bite-LF Gen what Nom Walis ‘What bit Walis?’
(4.29) a. m<in>ani’ ci pira’ ka’ watan?
AF<Prf>eat Acc how many/much Nom Watan ‘How much did Watan eat?’
b. m<in>abel ø ’inu ka’ yaya’=su?
AF<Prf>sleep where Nom mother=2S.Gen
‘Where did your mother sleep?’
c. pa-kani’ ø kanon ci ’ucyux ka’ yaya’=su?
Irr.AF-eat when Acc fish Nom mother=2S.Gen ‘When will your mother eat fish?’
c’. m<in>ani’=su ci kanon (*ci) ga’iloh hira?
AF<Prf>eat=2S.Nom CI when (*Acc) banana yesterday
‘When did you eat banana yesterday?’
c’’. m<in>ani’=su (*ci) kanon ci ga’iloh hira?
AF<Prf>eat=2S.Nom (*CI) when Acc banana yesterday
‘When did you eat banana yesterday?’
Two points are noted from examples in (4.27)-(4.29).
(i) Both ’ima ‘who’ and ’amol ‘what’ may be preceded by the accusative case marker ci that marks the following noun as a comitative member as in (4.27a) or a patient as in (4.28a). However, ’ima ‘who’ behaves like a proper noun because it can be marked as locative with the suffix -an, while ’amol ‘what’ is like a common noun that is marked as genitive with na.
(ii) Interrogative nouns referring to non-core participants, such as location (’inu
‘where’) and time (kanon ‘when’), are usually not preceded by case markers.
However, as shown in (4.29c’-c’’), kanon ‘when’ can be preceded by ci when inquiring time in the past. It is noted that such a ci cannot co-occur with the
accusative case marker ci in a sentence (see also footnote 10 on p.23).
In the above discussion, interrogative nouns are all used as head nouns that may serve either as predicates or arguments. When co-occurring with other nouns, they may function as adnominal modifiers except for kanon ‘when’. Compare the following pairs of examples.
‘I am cooking ku-hua fish.’23
b’. ’ucyux ’amol ka’ cyel=su tahk-un?
fish what kind Nom Exi.Rem=2S.Gen cook-PF
‘What kind of fish are you cooking?’
c. ’alang haca ka’ cyel=mu kel-an.
village that Nom Exi.Rem=1S.Gen stay-LF ‘I live in that village.’
c’. ’alang ’inu ka’ cyel=su kel-an?
village which Nom Exi.Rem=2S.Gen stay-LF
‘Which village do you live in?’
d. tugal ngahi ka’ ni’-un=mu.
three sweet potato Nom eat-PF=1S.Gen
‘I eat three sweet potatoes.’
d’. pira’ ngahi ka’ ni’-un=su?
how many/much sweet potato Nom eat-PF=2S.Gen
‘How many sweet potatoes did you eat?’
23 In Plngawan, people refer to the traditional food they love to eat most as raguw cubay [vegetable real] ‘real vegetable’ or ’ucyux cubay [fish real] ‘real fish’. In Mandarin Chinese, ’ucyux cubay can be translated as 苦花魚 ku-hua fish.
As shown in (4.30), while ’ima ‘who’ has to be preceded by the genitive case marker ni to refer to possessors of the preceding noun, ’amol ‘what kind’ and ’inu ‘which’
occur right after nouns they modify. Like numerals, however, pira’ ‘how many/much’ precedes nouns it modifies.
In addition to being predicates as in (4.30), NPs composed of head nouns and interrogative nouns as modifiers may also serve as case-marked non-subject arguments in which the order between the two cannot be reversed, as shown in (4.31).
(4.31) a. m<in>ani’=su ci ’ucyux ni ’ima?
AF<Prf>eat=2S.Nom Acc fish Gen who ‘Whose fish did you eat?’
b. t<um>ahuk=su ci ’ucyux ’amol?
cook<AF>cook=2S.Nom Acc fish what kind
‘What kind of fish do you cook?’
c. cyel=su maki’ ’alang ’inu?
Exi.Rem=2S.Nom AF.live village which
‘Which village do you live in?’
c’. * cyel=su maki’ ’inu ’alang?
Exi.Rem=2S.Nom AF.live which village
d. m<in>ani’=su ci pira’ ngahi?
AF<Prf>eat=2S.Nom Acc how many/much sweet potato
‘How many sweet potatoes did you eat?’
Finally, it is found that while ’ima ‘who’, ’amol ‘what’, ’inu ‘where/which’ and kanon ‘when’ may undergo CVC-reduplication to mark plurality or emphasis, there is no such a form as *pir-pira’ or *kit-kito’ ‘how many/much’, as shown in (4.32).
(4.32) a. ’in-’ima ka’ b<in>ah-en ni yaba’?
Red-who Nom hit<Prf>hit-LF Gen father
‘Who and who did Father hit?’
b. ’an-’amol ka’ ’<in>aras=su?
Red-what Nom bring<Prf.PF>bring=2S.Gen
‘What and what have you brought?’
c. ’in-’inu ka’ ’<in>sal-an=su?
Red-where Nom go<Prf>go-LF=2S.Gen
‘Where and where have you been to?’
d. kan-kanon kaca hang?
Red-when Nom.that Part ‘When on earth did that happen?’
e. * pir-pira’ ka’ ci’uli’ hani la?
Red-how many/much Nom person this Part e’. pira’ kora ka’ ci’uli’ hani la?
how many/much all Nom person this Part
‘How many in total are these people?’
The reason for the lack of *pir-pira’24 and *kit-kito’ is obvious. While we may treat ‘’ima ‘who’ as referring to a single participant and ’in-’ima ‘who and who’
as its plural counterpart, it seems unreasonable to posit that pira’ ‘how many/much’
refers to a single entity. Thus, instead of being reduplicated to mark plurality, pira’
‘how many/much’ is modified by the quantifier kora ‘all’ when the total number of entities is inquired, as shown in (4.32e’).
In this section, we have shown that interrogative nouns may serve as predicates in equational sentences and attract pronoun clitics that co-refer with them or indicate possessors of them. Besides, they may function as case-marked non-subject arguments or as adnominal modifiers. Morphologically, ’ima ‘who’ may be affixed with the locative case suffix -an, and most interrogative nouns may undergo CVC-reduplication to mark plurality or emphasis except for pira’ ‘how many/much’
and kito’ ‘how many/much’. All these criteria help characterize the
24 As mentioned in Chapter 3, in some Formosan languages, the question word meaning “how many/much” can be reduplicated when the quantity of human beings is inquired, as shown below:
non-human human
(i) Mayrinax Atayal piya’ pa-piya’ (Huang et al. 1999:665) (ii) Amis pina pa-pina (Huang et al. 1999:665)
(iii) Seediq piya ppiya (Chang 1996:40)
Thanks to Prof. L. Huang and Prof. Joy Wu for this reminder.
above-mentioned interrogative words as nouns rather than verbs.