Chapter Position of Polar Question Particles

by Matthew S. Dryer

1. Defining the values

This map shows the position of question particles in polar questions. Polar questions are ones that elicit the equivalent of a ‘yes’ or ‘no’ response. This map does not include interrogative affixes on verbs, although it should be noted that the distinction between interrogative affixes and separate interrogative particles is often hard to make. The map treats interrogative clitics that can attach to words of different categories as question particles. Map 116A shows the distribution of different ways to signal polar questions, including question particles, but does not show their position in the sentence. A number of languages shown on Map 116A as having question particles are not shown on this map because their position is not clear from available materials.

Values of Map 92A. Position of Polar Question Particles
Go to map
 ValueRepresentation
Question particle at beginning of sentence 129
Question particle at end of sentence 314
Question particle in second position in sentence 52
Question particle with other position 8
Question particle in either of two positions 26
No question particle 355
Total: 884

In some languages, the question particle occurs at the beginning of the sentence, as in (1) from !Xóõ (Southern Khoisan; Botswana).

(1) !Xóõ (Traill 1994: 18) 

|ú 

tûu 

à 

sîi 

people 

tns 

come 

‘Did the people come?’ 

The French expression est-ce que, illustrated in (2), though originally a verb plus demonstrative plus complementizer, can be considered to have become grammaticized as a question particle; French is thus treated here as a language with an initial question particle.

(2) French (Harris 1988: 237) 

Est-ce que 

le 

président 

vient? 

the 

president 

come.pres.3sg

‘Is the president coming?’ 

In other languages, the question particle occurs at the end of the sentence, as in (3) from Hatam (West Papuan; Indonesia).

(3) Hatam (Reesink 1999: 69) 

a-yai 

bi-dani 

mem 

di-ngat 

2sg-get

to-me 

for 

1sg-see

‘Would you give it to me so that I can see it?’ 

Sentence-final question particles often cliticize onto the last word in the sentence, as in (4) from Majang (Surmic, Nilo-Saharan; Ethiopia). 

(4) Majang (Unseth 1989: 126) 

dɛn-ɛ 

daaki 

tolay

see-3sg

Daaki 

Tolay=q

‘Did Daaki see Tolay?’ 

Occasionally a question particle will code other grammatical features of the clause. For example, in Aghu (Trans-New Guinea; Papua, Indonesia), the question particles also code tense (Boelaars 1950: 152). 

A third common position for question particles is second position in the sentence, following the first word or constituent, as in (5) from Yurok (Algic; California).

(5) Yurok (Robins 1958: 139)

kic 

hes 

neskʷec-okʷ 

ku 

wɹʔyɹs 

pst 

come-3sg

def 

girl 

‘Has the girl come back yet?’ 

Second position question particles very often encliticize onto the first constituent, as in the example in (6) from Mono (Uto-Aztecan; California). 

(6) Mono (Norris 1986: 44) 

Charley=w̃aʔ

mia-pɨ 

Charley=q

go-perf

‘Has Charley left?’ 

The fourth type shown on the map consists of languages in which there is a question particle with a specific position within the clause that is something other than initial position, second position, or final position. For example, in Hmong Njua (Hmong-Mien; China), the question particle occurs after the subject and before the verb, as in (7).

(7) Hmong Njua (Harriehausen 1990: 205) 

kuv 

cov 

nyaaj 

puas 

nyam 

hoob 

pee 

1sg

meas 

aunt 

like 

sing 

song 

‘Do my aunts like to sing?’ 

In Niuean (Oceanic; Niue Island, Pacific), which is VSO, the question particle immediately follows the verb, as in (8). 

(8) Niuean (Seiter 1980: 25) 

foaki 

age 

nakai 

koe 

fakamailoga 

ki 

taha 

give 

dirl 

erg 

you 

abs 

prize 

to 

pers 

someone 

‘Did you give the prize to someone?’ 

The fifth type shown on the map is languages in which question particles can occur in two of the positions represented by the first four types, without one position being dominant. This includes languages where question particles occur either at the beginning or at the end of the sentence. An example of such a language is Koyraboro Senni (Songhay; Mali), in which the question particle is the same as the word meaning ‘or’ (Heath 1999a: 225). This type also includes languages with two different question particles that occur in different positions. An example of such a language is Hunde (Bantu; Democratic Republic of Congo), in which one question particle occurs sentence-initially, as in (9a), and the other sentence-finally, as in (9b).

(9) Hunde (Kahombo 1992: 171) 

a. 

mbéni 

ámukátsí 

mu-lómbe 

 

woman 

nc-lazy

 

‘Is the woman lazy?’ 

b. 

ámukátsí 

mu-lómbe 

hé 

 

woman 

nc-lazy

 

‘Is the woman lazy?’ 

This type also includes languages in which question particles occur either in initial position or in second position, as in Ngiti (Nilo-Saharan; Democratic Republic of Congo), where there are two distinct question particles, each associated with one of these two positions (Kutsch Lojenga 1994: 379). If one position can be considered dominant (see “Determining Dominant Word Order”), then the language is coded according to that dominant position. For example, in Abun (West Papuan; Indonesia), there is an optional sentence-initial particle te  and an obligatory sentence-final particle e, as illustrated in (10).

(10) Abun (Berry and Berry 1999: 102) 

(te) 

nan 

nai 

nan 

bi 

suk 

it 

(q)

2sg

get 

2sg

poss 

thing 

comp 

‘Have you got your things?’ 

Because all polar questions contain a sentence-final particle, but not all contain a sentence-initial one, Abun is treated here as a language with a sentence-final question particle. 

The final type shown on the map consists of languages without question particles. This includes a number of subtypes which are not distinguished on this map but are distinguished on Map 116A. It includes languages which have interrogative affixes on verbs, languages which code polar questions by a different word order, and languages in which interrogative sentences differ from declarative sentences only in intonation. Note that in some languages with question particles, it is more common not to use these particles; a language is coded here as lacking a question particle only if there is no evidence that a question particle is ever used.

In a number of languages, the position of question particles is variable. In such languages, the position of the question particle often depends on what is the focus of the question. In most if not all such languages, however, there is one position which is associated with a more neutral question, where the truth of the entire sentence is being questioned without one constituent being the focus of the question.  In such cases the language is coded for this map on the basis of the position of the question particle in such neutral questions. In Imbabura Quechua (Ecuador), for example, there is an interrogative clitic that is attached after the constituent which is the focus of the question, as in (11a), where the focus of the question is wasi-man  ‘to the house’; if the sentence is a neutral question, with no specific constituent as the focus, the interrogative clitic occurs on the verb, as in (11b).

(11) Imbabura Quechua (Cole 1982: 15) 

a. 

wasi-man=chu

ri-ju-ngui 

 

house-to=q

go-prog-2

 

‘Are you going to the house?’ 

b. 

wasi-man 

ri-ju-ngui=chu

 

house-to 

go-prog-2=q

 

‘Are you going to the house?’ 

Since the question particle attaches to the verb in polar questions with no constituent in focus in Imbabura Quechua, and since the normal word order in this language is verb-final, it is coded on the map as a language with a sentence-final question particle. In Syrian Arabic, which is not verb-final, the neutral position for the question particle is sentence-final (Cowell 1964: 378). 

In the example in (12) from Kobon (Trans-New Guinea; Papua New Guinea), the question is signalled by an expression that involves (optionally) the word meaning ‘or’ plus a negated form of the verb of the sentence. 

(12) Kobon (Davies 1981b: 5)

ne 

kaj 

ap 

mɨd-öp 

(aka) 

mɨd-ag-öp 

2sg

pig 

indef 

be-perf.3sg

(or) 

be-neg-perf.3sg

‘Have you any meat?’ 

For the purposes of this map (and Map 116A), such negative expressions are not distinguished from question particles. Such multi-word expressions for expressing questions are restricted in the languages examined to sentence-final position.

Many languages have question particles or expressions that indicate that the question is a leading question, one where the speaker has an expectation as to what the answer will be. English, for example, employs what are called “tag-questions” for this purpose, illustrated in (13a), which differ from the neutral question form illustrated in (13b) in that its form indicates that the speaker expects a ‘yes’ answer.

(13) 

a. 

You’re planning to go to the beach, aren’t you? 

b. 

Are you planning to go to the beach? 

Similarly, Hixkaryana (Carib; Brazil) has a particle hampe, illustrated in (14), which signals doubt, and when used in sentences with interrogative intonation, suggests that the speaker expects the answer to be ‘no’.

(14) Hixkaryana (Derbyshire 1979: 5) 

ɨhme 

menahno 

hampe 

egg 

2>3.eat.pst

doubt 

‘Did you really eat the egg?’ 

A neutral question in Hixkaryana, however, differs from the corresponding declarative sentence only in intonation. The map restricts attention to the expression of neutral questions, so Hixkaryana is shown on the map as only employing intonation to signal questions. 

There are two ways in which a language can be coded as having a question particle at the beginning of a sentence. First, the particle may always occur at the beginning of a sentence, regardless of the word order elsewhere in the clause. Second, it may immediately precede the verb, and the dominant word order in the language may be verb-initial so that the normal position of the question particle is sentence-initial – even though if some other constituent precedes the verb, it will also precede the question particle. These two types of languages are not distinguished on the map. Similar comments apply to question particles at the end of sentences.

2. Geographical distribution

Apart from languages lacking question particles altogether, the most common type shown on the map is languages in which the question particle occurs in sentence-final position. Note, however, that most languages with question particles fall into three areas: (i) a belt stretching across Africa from western Africa to central eastern Africa; (ii) an area within Asia including mainland Southeast Asia and extending west into India and north through China to Japan and eastern Siberia; and (iii) New Guinea. Within these areas, sentence-final particles outnumber all other types. Outside these three areas, however, final particles are no more common than initial or second position particles. The largest area in which final particles are absent is Europe and North Africa. 

Both initial particles and second position particles are widespread, although they are uncommon within the three areas mentioned above in which final particles predominate. Initial particles are most common in (i) Europe and North Africa; and (ii) North America, including Mesoamerica.  Second position particles are most common in North America. Languages lacking question particles are fairly evenly distributed throughout the world.