Interrogativity in Bashkir Текст научной статьи по специальности «Физика»

Maksim Fedotov

NRE HSE — ILS RAS, Saint Petersburg

INTERROGATIVITY IN BASHKIR1 1. Introduction

The general aim of this paper is to describe grammatical markers and relevant lexical elements used in different types of interrogative sentences in Bashkir, and their distribution.

The study is based on elicitation data and spontaneous texts2 collected during expeditions organized by St. Petersburg State University in the villages of Rakhmetovo and Baimovo in the Abzelilovsky district of Bashkortostan (Russian Federation) in 2011-2016.

There are three substantial grammatical descriptions of Bashkir: [Poppe 1964, Juldashev (ed.) 1981, Dmitriev 2008 (1948)]. While the treatment of interrogativity in these descriptions is far from complete, they provided a starting point for this study.

Outline of the paper. In section 2, I will make some preliminary theoretical remarks on interrogativity. In section 3, I will discuss polar (3.1, 3.2) and alternative (3.3) questions in Bashkir, which both involve a special interrogative particle, and their intonation patterns (3.4, 3.5), and then I will review the non-interrogative uses of the particle (3.6). In section 4, I will discuss content questions (" wA-questions"), including the inventory of interrogative words (4.1), their position (4.2), and the intonation patterns used in such questions (4.3). In section 5, I will

1 The study was supported by the RSF's grant No. 14-18-03406 «Грамматические категории в языках мира: иерархия и взаимодействие (типологический анализ)» ["Grammatical Categories in World Languages: Hierarchy and Interaction (A Typological Analysis)"] and by a Saint Petersburg Government's award.

1 would like to thank all my colleagues from the Bashkir expeditions and all Bashkir speakers who helped me with this study. A special thanks goes to Francis M. Tyers.

2 The overall volume of our transcribed corpus is 16 331 words, the total length of recordings 3.4 hours. Examples taken from the corpus will be marked by the name of the text in square brackets, as in (10).

review the attested peripheral markers used in interrogative sentences, i. e. tag-like and particle-like markers, which are mostly used in biased questions. Finally, in section 6, I will discuss the expression of some special types of questions, including non-standard illocutionary types ("deliberative", permissive, and hortative) (6.1), echo questions (6.2), and embedded questions (6.3).

2. Some preliminary theoretical remarks

I will define interrogative speech act as a type of speech act whose function is to make an attempt at verbally causing the addressee to perform some other speech act — most usually to cause the addressee to communicate some information (i. e. to perform a declarative speech act). Thus, the sentence/utterance produced in this act will be called an interrogative sentence (= question), and the semantic/functional domain associated with such speech acts, "interrogativity".

Three major types of interrogative sentences are usually distinguished (cf. [Siemund 2001: 1011; Konig, Siemund 2007: 291; Dixon 2012: 377, 398]):

1) polar questions ("yes/no" questions)

— cf. Did John eat my cake?,

2) alternative questions

— cf. Did JOHN3 or Mary eat my cake?,

3) content questions ("wh-questions", information/constituent questions) — cf. Who ate my cake?.

Focus4 in all types of interrogative sentences may be broad (including the whole proposition or at least its part containing the predicate) or narrow:

(1a) Did John eat my cake? (or not) (polar question; broad focus)

(1b) Did JOHN eat my cake? (or someone else did it)

(polar question; narrow focus on the subject)

3 Hereinafter I will use small caps in the examples to indicate the focus accent in English (mostly for narrow focus) and Bashkir when needed.

4 Here understood after [Lambrecht 1994: 213] as "[t]he semantic component of a pragmatically structured proposition [expressed by the sentence] whereby the assertion differs from the [pragmatic] presupposition".

(1c) Did John eat MY CAKE? (or he ate something else)

(polar question; narrow focus on the direct object)

(2a) Did JOHN eat my cake or did MARY HIDE it?

(alternative question; broad focus)

(2b) Did JOHN or Mary eat my cake?

(alternative question; narrow focus on the subject)

(3 a) What happened? (content question; broad focus)

(3b) Who ate my cake? (content question; narrow focus on the subject)

... etc.

A further important distinction in polar questions is between explicative and non-explicative5 (or "existential") polar questions [Mehlig 1991; 2011]:

(4) [What happened?] Did you lose your PASSPORT?

(explicative polar question; broad focus)

(5) [What did you lose?] Did you lose your PASSPORT?

(explicative polar question; narrow focus on the direct object)

(6) [What's that noise?] Did JOHN arrive?

(explicative polar question; broad focus)

(7) [John should've got back hours ago] Did he arrive (yet)?

(non-explicative polar question)

Finally, questions may be neutral (expressing a completely irreal situation) or biased = leading (the situation is partially real because of the speaker's expectations). Intrinsically biased are, for example, the English tag questions:

(8a) You have met him, haven't you? (expecting a positive answer) (8b) You haven't met him, have you? (expecting a negative answer)

5 Explicative questions verify the correctness of the description of a situation whose existence is presupposed by the speaker. Non-explicative questions verify the realization, the existence of a situation in the current discourse [Mehlig 2011: 178]. The distinction applies to declarative sentences, too, cf. [What's that noise?] John arrived! and [John should've got back hours ago, did he arrive yet?] Yes, John arrived.

An important formal characteristic of biased questions is their non-compatibility6 with "negative-polarity items" (NPIs) such as English ever or any (more precisely, with pronouns in the function of "question"-type indefinites) — while neutral questions, on the contrary, are compatible with NpIs (and not compatible with specific indefinite pronouns):

(9a) Have you met anyone / #someone? (neutral)

(9b) You've met someone / * anyone, haven't you?

(intrinsically biased)

3. Polar and alternative questions

Both in polar and alternative questions (but not in content questions — see section 4) Bashkir uses a special enclitic particle = m (E). It has harmonic variants =ma, =me, =mo, and =mo, and in fluent speech may be reduced to =m (cf. (16)). It attaches to entire phrases (including coordinate phrases (12)) and occupies the iconic outermost position in series of clitics7 (cf. (13), (25)).

3.1. Polar questions

polar questions involve the particle =mE and a special intonation contour (see 3.4). By default the particle is not omitted (see, however, section 3.2 below). Cf. the examples:

(10) Jawasqa-naq bidra-he-n in-der-de-q=me slops-GEN bucket-P.3-ACC enter-CAUS-PST-2SG=Q solaq-ya?

entrance.hall-DAT

'Did you take the slop bucket out to the hall [as you should have] ?'

[110711_dah_Dialog_o_byte.006]

(11) ISRAFIL-YA kil-de-q=me? Israfil-DAT come-pST-2SG=Q

'Did you come to ISRAFIL?' [110717_gaj_Dialog_o_muzhe.002]

(12) Azat ham Zoxra Azat and Zuxra

6 If the polarity of the question itself is positive.

7 Except for (other) interrogative particles and tags, see section 4.

— [ayaj menan / ham heyle] =me?

elder.brother with and younger. sister=Q 'Are Azat and Zuxra brother and sister?'8 (it is ungrammatical to attach the particle to each conjunct)

(13) Med-qa=la=mz? medical.college-DAT=also=Q

'[You applied] to the medical college too?'

[110700_rtk_rrg_Dialog_o_postuplenii.049]

(14) Ul KUPTA[N] kit-te=me? that long. ago go.away-PST=Q

'Did he leave long ago?' [110717_gaj_Dialog_o_muzhe.017]

(15) Nama, esla-gan-dar=me tege? thing work-PC. PS T-PL=Q that Hdjla-s-kan-dar=me inde sunan? speak-RECP-PC.PST-PL=Q now that.ABL

'What, did they do it? Did they come to an agreement now?'

[110717_gaj_Dialog_o_muzhe.009-010]

(16) Tsyla-0, hin asa-r-ya tela-ma-j-hey=m(e)? listen-IMP thou eat-POT-DAT want-NEG-IPFV-2SG=Q 'Listen, do you want to eat [something]?'

(17) ... "A, parazit, hin miney menan berga

oh (Russ.) parasite (Russ.) thou I.gen with together qajt-ts-y =m9 ale ? " come-PST-2SG=Q now

{Then I remembered, and so I say:} '"Oh, you vermin, [so] you've come together with me?!"'

[140719_mfsB_Domovoj .041]

(18) (Alla) JEL saula-j=m3?

or wind make.noise-IPFV=Q

{Hearing some roaring from the street:} 'Is [it] the wind making noise (, it appears)?'

8 This and other examples for which no source is indicated were obtained by elicitation.

The position of the particle =mE is determined by a set of factors, the main one being whether the sentence has an explicit predicate or not (and probably also whether this predicate is verbal or non-verbal)9.

In the first case — verbal sentences without ellipsis (examples (10)—(11), (14)-(20)) — the particle normally attaches to the finite verb (clause-final or not), irrespectively of whether the focus is broad or narrow and which constituent is in the focus domain. Thus, the particle attaches to the predicate even in the cases where the latter is not in the focus domain, cf. (11), (14) above and (19)-(20):

(19) Hin irtaga kil-a-hey=me? thou tomorrow come-lPFV-2SG=Q

{You were going to come.} 'Will you come tomorrow?'

(20) Morat hine huq-t3=ma? Murat thou.ACC hit-PS T=Q

'Did Murat hit YOU?', 'Was it YOU who(m) Murat hit?'

The alternative placement, where the particle attaches to a non-predicate word/constituent which coincides with the focus domain, is unattested in our texts and was judged as ungrammatical by most speakers:

(19') */???(Hin) irtaga=me kil-a-hey?

thou tomorrow=Q come-IPFV-2SG Intended meaning: 'Will you come tomorrow?'

(20') */?Morat hine=me huq-ts? Murat thou.ACC=Q hit-PST Intended meaning: 'Did Murat hit you?'

In the second case — non-verbal sentences proper (locational, possessive, ascriptive, etc.; examples (12), (21)-(23)) — the particle, too, normally attaches to the predicate (NP / AdjP / AdvP), even if it is not in the focus domain, cf. (22b), (23b):

9 The explicative vs. non-explicative distinction, on the contrary, does not seem to influence the particle placement, cf. the same placement in explicative (11), (17), and (18), and non-explicative (10), (15), and (16).

(21) Basqa juq=ma sigaret-tar? other neg.cop=q cigarette-PL 'There are no other cigarettes?'

[110711_bbm_Dialog_v_magazine.010]

(22a) Bulat bogon oj-ba=me? Bulat today house-LOC=Q / Bogon Bulat oj-ba=me? / Bulat oj-ba=me bogon? 'Is Bulat at home today?'

(22b) Bulat bogon oj-da=m(e)? Bulat today house-LOC=Q / Bogon Bulat oj-da=me?

{Bulat is to be at home on one of the days.} 'Is Bulat at home TODAY? [asking about the exact day]'

(23a) Oso kitap hineke=me?

this book thou. POSS. SUBS T=Q 'Is this book yours?'

(23b) Oso kitap hineke=me?

this book thou. POSS. SUBS T=Q

{Helping the addressee find a book he left in a room:} 'This book is yours?'

The alternative non-predicate placement, again, was judged as ungrammatical by most speakers (although in this case the overall grammatical acceptability is slightly higher):

(22b') ???(Bulat) bogon=md oj-da?

Bulat today=Q house-LOC

/ * Bulat oj-da bogon=mo? Intended meaning: 'Is Bulat at home today?'

(23b') ???O§o=mo kitap hineke?

this=Q book thou.POSS.SUBST / */ok[Oso kitap] =mz hineke? Intended meaning: 'This book is yours?'

Finally, in the third case — sentences with predicate ellipsis (examples (13), (24)-(28)) — the most frequent pattern is for the

particle to be attached to the last word/constituent in the clause which at the same time coincides with the focus domain (and bears the focus accent), cf. (24)-(25), (26a), (27a), (28a). An alternative pattern, with the particle attached to a non-final word/constituent which coincides with the focus domain, is frequent and grammatical too, cf. (26b), (27b), (28b). The remaining logically possible pattern — with the particle attached to the last word/constituent which does not coincide with the focus domain — was judged as acceptable by some speakers, cf. (26c).

(24) Hin oj-ga=me? thou house-DAT=Q 'You['re going] HOME?'

(25) 0s0-n0=yzna=m? this-ACC=just=Q

'[Do you need to dilute] just this one?'

[120710_hkx_Dialog_s_sosedkoj .008]

(26a) Qala-ya irtaga=me? town-DAT tomorrow=Q

(26b) ok/?Irtaga =me qala-ya?

tomorrow=Q town-DAT

(26c) OK/?Irtaga qala-ya=ma? tomorrow town-DAT=Q {When are you going to town?} '[Are you going] to town

TOMORROW?'

(27a) Irtaga qala-ya=ma? tomorrow town-DAT=Q

(27b) Qala-ya=mz irtaga? town-DAT=Q tomorrow

(27c) ??Qala-YA irtaga=me?

town-DAT tomorrow=Q {Where are you going tomorrow?} 'Tomorrow [you're going] to town?' {And then to Baimovo?}

(28a) OK/??Jasel(-e) hida=me?

green-P.3 thou.DAT=Q

(28b) OKHma=me jasel(-e)? thou.DAT=Q green-P.3

{A clown is handing out balloons on a party. — Who shall I give a green balloon to?} 'To you a green one?'

The picture presented above differs to some extent from the description and the examples in [Dmitriev 2008 (1948): 195]:

In interrogative sentences the question affix is theoretically placed on the sentence element that needs affirmation. More often it is the predicate, less often the subject (...) Hin bal kitapta aldayma? [thou this book:ACC take:PST:2sG:q] — did you take that book? Hin bal kitaptama alday? [thou this book:ACC:q take:PST:2sG] — did you take that book? Hinme bal kitapta alday? [thou:q this book:ACC take:PST:2sG] — did you take that book? The predicate (. ) attracts in some way the question affix, and only in the case of a strong logical stress on the subject the affix is placed on it. (. ) Hiney kitabay juqma? [thou.GEN book:P.2sG neg.cop:q] — do(n't) you have a book? Hiney kitabayma juq? [thou.GEN book:P.2sG:q neg.cop] — what, you don't even have a book? (or: 'you don't have a book or what?'). Saj juqma? [tea neg.cop:q] — do(n't) you have some tea? Sajmejuq? [tea:q neg.cop] — what, you don't even have tea?" 10.

In my data, such non-predicate particle placement both in verbal and non-verbal sentences is not attested in texts and was judged as ungram-matical by almost all speakers (cf. (19'), (20'), (22b'), (23b') above).

Another remark can be made on negative questions. In Bashkir they are neutral and not biased, at least not towards expecting or previously believing the answer to be positive (similar to Russian and unlike some other languages, such as English, cf. [Romero, Han 2004]) — cf. (16), (35), and (29)-(30):

(29) Sabata-na (...) bel-ma-j-ged=me?

bast.shoe-ACC know-neg-lPFV-2PL=Q

{We were wearing bast shoes on our feet.} 'Bast shoes.

you (pl.) know [what it is]?' (lit. "Don't you.")

[140710_ggs_Rabota_i_zhizn.017]

10 Translation from Russian, transliteration, glosses, bold and italic type are mine. — M. F.

(30) Hed-ga jardam karak-ma-j=me? you-DAT help necessary-neg-IPFV=Q 'Do you (pl.) need help?' (lit. "Don't you...")

3.2. Polar questions without the particle =mE In [Juldashev 1997: 215] and [Poppe 1964: 95-96], the use of the particle in polar questions in Bashkir is described as optional. Poppe provides the example: Hin bslkitapts aldsy? <thou this book:ACC take:PST:2SG> 'Did you take this book?'.

In my data, the situation is the following. Presence (non-omission) of the particle in polar questions is definitely the default option, both universally produced (and judged as grammatical by the speakers) and most frequent in the texts: 40(+3)11 tokens with the particle vs. 7(+9) without it.

However, polar questions without the particle (but with the same special intonation, see 2.4) are accepted as grammatical by some speakers (a minority in my data), and are present in the texts — the above mentioned 7(+9) tokens.

No clear difference in function was noted between questions with the particle and without the particle. Just like the former (see 2.1), the latter can be explicative (cf. (31), (36)(?)) and non-explicative (cf. (33), (34)(?), (35)(?)), can have broad (cf. (31), (33)-(35), (36)) or narrow (cf. (32)) focus. The results of negative-polarity item (NPI) tests were contradictory, but generally questions without the particle seem to be less acceptable when combined with NPIs (while questions with the particle combine with them perfectly well), cf. (37); this may indicate that questions without the particle at least tend to be biased. Also, questions without the particle were judged by some speakers as "more colloquial" or as "dialectal".

(31) Nima-kaj? PIROG bes-er-gan? what-DiM pie boil-C aus-pc . ps T '[Thinking out loud:] What's that? Did he bake a pie?'

[150707_rgg_Krtek2]

11 The first number is the number of clear cases and the number in brackets is the number of additional less certain cases.

(32) OKl???(Hin) IRTAgA kil-â-(he)q?

thou tomorrow come-lPFV-2SG {You were going to come} 'Will you come tomorrow?'

(33) OKl??Hin irtâgâ kil-A-(he)D?

thou tomorrow come-lPFV-2SG {You were going to come tomorrow} 'WILL you come tomorrow?'

(34) Basqa sigâret-târ JUQ? other cigarette-PL neg.cop 'There are no other cigarettes?'

[110711_bbm_Dialog_v_magazine.012]

(35) Anaw sertifikat kArAk-mA-J? there certificate necessary-NEG-lPFV 'And [you] (don't) need a certificate?'

[110700_rtk_rrg_Dialog_o_postuplenii.010]

(36) Sertifikat-tar-s-n bir-mâ-gân-dâr? certificate-PL-P. 3-ACC give-NEG-PC.PST-PL

'Have(n't) they given out the certificates [in your school]?' (All other schools have already given them out).

[110700_rtk_rrg_Dialog_o_postuplenii.031]

(37) OK/*Hin beraj qasan Mâskâw-ôâ bul-ds-q?

thou some when Moscow-LOC be-PST-2SG '— Have you ever been to Moscow?' {— No, never}.

3.3. Alternative questions12

In alternative questions, Bashkir uses coordination of two or more constituents together with the particle =m(E) (and a special intonation contour, see 2.5). The particle is very rarely omitted, moreover, there is usually more than one instance of it, matching the number of alternatives (see below). Coordination in alternative questions may involve simple juxtaposition or the insertion of a disjunctive conjunction âldâ~âllâ 'or' (sometimes also jâki 'or' or Russian loanword ili 'or') between the (last) two coordinated constituents. Cf. the examples:

12 I should make a disclaimer that the study of Bashkir alternative questions presented in this section is, unfortunately, based solely on elicitation data, as there were (almost) no examples of alternative questions in our texts.

(38) Azat irtaga esla-j=me alla jal it-a=me? Azat tomorrow work-IPFV=Q or rest do-IPFV=Q 'Is Azat WORKING or RESTING tomorrow?'

(39) Azat irtaga esla-j=me alla juq=mz? Azat tomorrow work-IPFV=Q or NEG.COP=Q 'Is Azat working tomorrow or not?'

(40) Hin tege sayat-te Azat-qa=ms alla Bulat-qa=mz thou that hour-ACC Azat-DAT=Q or Bulat-DAT=Q bulak it-te-y?

present do-PST-2SG

'Did you give that watch to Azat or to Bulat?'

(41) Hin bogon, irtaga, alla / alda / ili irta-nan huy thou today tomorrow or or or morning-ABL after kil-a-y=me?

come-IPFV-2SG=Q

[We know that the addressee is going to visit soon]. 'Will you come TODAY, TOMORROW, or THE DAY AFTER TOMORROW?'

The patterns of the particle placement are:

1) If predicates or full clauses are coordinated, there is almost always an instance of the particle attached to each finite verb (i. e. "Azat tomorrow [works=q] or [rests=q]?" or "Outside [wind blows=q] or [it rains=q]?"), cf. (38)-(39) above and (42); other placement patterns are dispreferred.

2) For full sentences with coordination of lesser constituents (NPs, AdvPs, PPs etc.), the preferred pattern is 2a) with particles attached to the predicate and to the second coordinated constituent, placed after the verb13 ("You that watch Azat gave=q, or Bulat=q?"), cf. (43a), (44) below. Other patterns recurrent in my data are: 2b) with particles attached to the first constituent and to the predicate, with neutral word order ("You that watch Azat=q or Bulat gave=q?"), cf. (43b) below, 2c) with an instance of the particle attached to each constituent ("You that watch Azat=q or Bulat=q gave?"), cf. (40) above

13 This is a deviation from the neutral verb-final word order of Bashkir (cf. types 2b, 2c, and 2d below). It may be the case that the structure of such examples is (or was at some point) biclausal, with predicate ellipsis in the second clause: "[you that watch Azat gave=q], or [Bulat_=q]?".

and (45) below, 2d) with an instance of the particle attached to the predicate only ("You that watch Azat or Bulat gave=q?"), cf. (41) above.

3) Finally, for sentences with ellipsis of the predicate (and also non-verbal sentences) the universally preferred pattern is the one with an instance of the particle attached to each coordinated constituent ("Azat=q or Bulat=q?"), cf. (46), (47).

(42) Tas-ta jel i6-a=me

outer. side-LOC wind blow-lPFV=Q alla jamyar jaw-a=ma? or rain rain-lPFV=Q

{Hearing some roaring from the street:} 'Is [it] the wind blowing or is it RAINING outside?'

(43a) Hin tege sayat-te Azat-qa bir-de-y=me

thou that hour-ACC Azat-DAT give-PST-2SG=Q alla Bulat-qa=ma? or Bulat-DAT=Q

'Did you give that watch to Azat or to Bulat?'

(43b) Hin tege sayat-te Azat-qa=ma alla Bulat-qa thou that hour-ACC Azat-DAT=Q or Bulat-DAT bulak it-te-y=me? present do-PST-2SG=Q

'Did you give that watch to Azat or to Bulat?'

(44) Hin bogon kit-a-hey=me alla irtaga=me? thou today go.away-lPFV-2SG=Q or tomorrow=Q 'Are you leaving today or tomorrow?'

(45) Hin (bogon=me,) irtaga=me, irta-nan huy=ma thou today=Q tomorrow=Q morning-ABL after=Q kil-a-y?

come-lPFV-2SG

'Will you come (today,) tomorrow, or the day after

TOMORROW?'

(46) Azat-qa=ma OK(alla) Bulat-qa=ma? Azat-DAT=Q or Bulat-DAT=Q

{— I gave that watch to my brother.} ' — To Azat or to Bulat?'

(47) Bdgon=mo, (alla) irtaga=me? today=Q or tomorrow=Q

{— I'm leaving.} '— Today or tomorrow?'

As for the conjunction, it is almost obligatorily inserted in case of full clause coordination (12 : 1)14, it may be sometimes omitted in case of predicate coordination (11 : 3) (in my data most such examples involved polarity alternation — "does he work or not/doesn't he", cf. (48)) or non-predicate constituent coordination (39 : 4) (most examples are of the 2c pattern type: "you that watch Azat=q, Bulat=q gave?", cf. (45) above), and it is omitted almost as often as not in the case of sentences without a verb (17 : 10), cf. (46), (47) above.

(48) Azat irtaga esla-j=me, esla-ma-j=me? / ...juq=ma? Azat tomorrow work-IPFV=Q work-NEG-IPFV=Q NEG.COP=Q 'Is Azat working tomorrow or [is he] not working? / or not?'

Examples that have some other logically possible arrangements have also been attested, including examples without particles or with a single particle attached to only one of the coordinated constituents (cf. two examples below). However, such patterns are very rare in my data, and they were judged by other speakers as ungrammatical.

(49) OK,???T3s-ta jel=me alla jamysr sawla-j? outer. side-LOC wind=Q or rain make.noise-iPFV {Hearing some noise from the street:} 'Is [it] the WIND or the RAIN making [this] noise outside?'

(50) OK/*Jel, (alla) jamysr?

wind or rain {Something is making noise outside.} 'Is it the WIND or the RAIN?'

14 The two numbers represent the number of examples (translations) in my data that have the conjunction and the number of those that do not have it, respectively. Only the translated examples are counted.

It should be noted that there is a high probability that the speakers were calquing the Russian stimuli (all of which contained the Russian conjunction ili 'or') during elicitation, so we can expect that the real ratios in actual speech would be shifted — i.e. that there would in fact be a higher number of sentences without the conjunction (in all the three cases).

3.4. Intonation in polar questions

In polar questions, the pitch rises sharply on the stressed syllable of the word that bears the focus accent (being part of the focus domain) and then usually falls (cf. [Poppe 1964: 95-96]).

Consider the diagrams15 below for different types of polar questions. The sharp rise is indicated by a thick line on the diagrams. Examples (51), (51'), (53a), (54a) were obtained by indirect elicitation to reduce the possibility of calquing the Russian intonation: the speakers were asked first to translate a declarative stimulus in Russian, and then to turn it by themselves into a question. Example (52) is taken from a spontaneous text.

(51) Unda JEL or-a=me?

that.LOC wind blow-IPFV=Q

{Hearing some roaring from the street:} '[Is it] the wind blowing there [it appears]?' (explicative broad-focus polar question)

300_

unda jel öräme

(51') JEL sawla-j=m9? wind blow-IPFV=Q

{Hearing some roaring from the street:} '[Is it] the wind making noise [it appears]?' (explicative broad-focus polar question)

300_

V"

jel saw laj ma

15 All intonation diagrams in this paper were made using the Praat software [Boersma, Weenink 2017].

(52) PIROG bes-er-gan?

pie boil-CAUS-PC.PST {What's that?} 'Did he bake a PIE?'

[150707_rgg_Krtek2] (= (31)) (explicative broad-focus polar question without the particle)

1

pi rok be ser gân

(the dotted line reconstructs the rising part of the contour)

(53a) Hin BOGÔN kil-a-hey=me?

thou today come-IPFV-2SG=Q

'Will you come TODAY? {or on another day}'

(explicative narrow-focus polar question)

300_

/ s.

hin bo gôn ki là heg me

(53b) Ul HINE kur-de?

that thou.ACC see-PST {Azat saw someone.} 'Did he see YOU?' (explicative narrow-focus polar question without the particle)

350

)

ul hi ne kiir Ôe

(54a) Hin irtaga KIL-A-HED=me? thou tomorrow come-lPFV-2SG=Q

{You were going to come tomorrow.} 'WILL you come tomorrow?' (non-explicative polar question)

300 __

-Mt-

hin irtaga ki la heg me

(54b) Bulat bogon OJ-DA=me? Bulat today house-LOC=Q 'IS Bulat at home today?' (= (22a))

(non-explicative polar question)

300

bulat bogon oj 6a me

(Cf. also the diagrams for examples (161a) Hin qala-ya BAR-A-D=m9? 'WILL you go to town?' and (160a) Bulat KlL-DE=me? 'Has Bulat arrived [yet]?' in section 6.2 below)

In a similar way, in declarative sentences the pitch normally falls sharply on the stressed syllable of the word bearing the focus accent (indicated by thick lines on the diagrams):

(55) Unda JEL or-a.

that.LOC wind blow-IPFV

{Hearing some roaring from the street:} '[It's] the wind blowing there', (cf. (51)) (broad-focus explicative declarative sentence)

unda jel ora

(56) Áje, ul MINÉ kür-de. yes that I.ACC see-PST 'Yes, he saw ME' (answer to (53b)).

(narrow-focus explicative declarative sentence)

350_

aje, ul mi ne kür 5e

(57) Irtaga min KIL-A-M.

tomorrow I come-IPFV-lSG

{— Will you come tomorrow?} '— I WILL come tomorrow' (cf. (54a)). (non-explicative declarative sentence)

irtaga min ki lam

Cf. also the diagram for an "out-of-the-blue" (neutral) declarative sentence in example (158a) Min QALAyA jsjsnam 'I'm gonna go to town' in section 6.2 below.

3.5. Intonation in alternative questions

Alternative questions have an intonation contour that can be partially derived from that of polar questions. The pitch rises sharply on (the stressed syllable of) the first alternative and rises (usually less prominently), or sometimes falls slightly, on the non-first alternatives. Each rise is usually immediately followed by a fall.

Consider the diagrams below for alternative questions with three (58a)-(58b) and two alternatives (59)-(60). The relevant rises and falls are indicated by thick lines.

Note that the location of these rising and falling segments does not depend on the position of the particle(s) =mE, but solely on the position of the coordinated constituents that express the alternatives.

(58a) Ssnajaq-ts AzAt, MOrAt alla BulAt wat-t9=ms? cup-ACC Azat Murat or Bulat break-PST=Q {One of three sons has broken a cup.} 'Did AZAT, MURAT, or BULAT break the cup?'

300___

/ / J

sanajaqta a zat mo rat alda bu lat wattams

(58b) Saske-ne AzAt, BULAt alda MORAT jar-ds=m9? cup-ACC Azat Bulat or Murat crack-PST=Q {One of three sons has broken a cup.} 'Did AZAT, BULAT, or MURAT break the cup?'

250_,_,_,_,

r "V

saskene a zat bu lat alda mo rat jardams

(59) Hin tege sayat-te AZAT-QA=m9 alla BULAT-QA=m9 thou that hour-ACC Azat-DAT=Q or Bulat-DAT=Q bulak it-te-y? present do-PST-2SG

'Did you give that watch to AZAT or to BULAT?' (= (40))

400

) A 1

hin tege sayatte a zat qa ma alia bu lat qa mg bulag ittei]

(60) Hin BOGON kit-a-hey=me alla IRTAGl=me?

thou today go.away-lPFV-2SG=Q or tomorrow=Q 'Are you leaving TODAY or TOMORROW?' (= (44))

300

JU / M At -^vA —

hin bo gon ki ta her) me alia ir ta ga me

3.6. Non-interrogative uses of the particle =mE There have also been attested some non-interrogative uses of the particle =m(E) in Bashkir. It is used in protases of alternative conditional-concessive constructions (in a similar way to alternative questions), cf. (61)-(62), and sometimes in hypothetical conditional protases or temporal clauses: a) together with the conditional mood suffix -hA, cf. (63)-(65a), or b) in a special type of conditional ~ temporal construction with -nd=md <-PST=Q>, which expresses, among other things, the temporal relation 'as soon as', cf. (65b)-(66).

(61) Ular-ya, 9ssn=m9 bus=m9, kaban tas that.PL-DAT true=Q vain=Q stack stone j9y9l-9p bar-yan-ya oqsa-yan. fall-CV go-PC.PST-DAT look.like-PC.PST 'True or not, it seemed to them that the hay stack is falling'.

[110716_bbm_Aslaj.004]

(62) Hin mine alda-j-hay=mz, alda-ma-j-hay=mz, thou I.acc deceive-lPFV-2SG=Q deceive-NEG-lPFV-2SG=Q min hiqa asan-ma-j-am.

I thou.DAT believe-NEG-lPFV-1SG

'Whether you're deceiving me or not, [in any case] I won't believe you'.

(63) Hin ulaj esla-ha-q(=me), thou so work-COND-2SG=Q

min kit-a-m / kit-asak-men.

I go.away-IPFV-lSG go.away-FUT-1SG {Threatening:} 'If you do so, I'll (surely) go away'. (=mE seems to add "categoricalness")

(64) Ager ul kisa kil-gan bul-ha (OK=ms), if that yesterday come-PC.PST be-C0ND=Q bal xal bul-ma-yan bul-ar ine.

this state be-NEG-PC.PST be-POT be.PST

'If he had come yesterday, this wouldn't have happened'. (here

=mE was said to add "categoricalness" too)

(65a) Kis bul-ha=ms, bota-he=la, joro-r-ga evening be-COND=Q all-P.3=also go-POT-DAT say-a ??(saq).

go.out-IPFV only.just

'As soon as the evening comes, everybody is to go out for a walk! [making terms / threatening]'

(65b) OKKis bul-dz=ma, bota-he=la joro-r-ga evening be-PST=Q all-P.3=also go-POT-DAT say-a(-lar). go.out-IPFV-PL

'[Usually] when ~ as soon as the evening comes, everybody is going out for a walk'.

(66) Bar-ap jet-te-q=me, miqa saltara-t-ar-haq.

go-CV be.enough-PST-P.2SG=Q I.DAT ring-CAUS-POT-2SG 'As soon as ~ if you arrive, you'll call me'.

The particle is also used in embedded alternative questions, even with non-finite verb forms, cf. (170)—(171) in section 6.3.

4. Content questions

Content (" wh-") questions in Bashkir are characterized by the use of interrogative words (one or several) and a special intonation pattern (see 3.3). They do not contain the particle =mE 16:

(67) Bsl qasan bul-ds (* =ms)? this when be-PST =Q

'When was it?'; 'When did it happen?' (addition of the particle is ungrammatical)

(68) Qaj-da bar-a-ysd? where-LOC go-lPFV-2PL 'Where are you going to go?'

[110700_rtk_rrg_Dialog_o_postuplenii.023]

(69) Min bslar-dsy qajnz-hs-n al-a al-a-m?

I this.PL-GEN which-P.3-ACC take-lPFV take-lPFV-1SG {Looking through some stuff.} 'Which of these may I take?'

(70) — Kupme17 hot qal-yan? — Ad.

how.much milk remain-PC.PST little '— How much milk is left? — Little'.

(71) Kem-de qaj-da ursnlas-tsr-ds-lar? who-ACC where-LOC settle-CAUS-PST-PL

{The boys were housed in several different buildings.} 'Who has been housed where?' (lit. "Whom where [they] housed?")

(72) Nima bul-ds? what be-PST

'— What happened?' {— I hurt myself}. (broad focus question)

16 This is true for plain content questions. In echo questions from content questions, however, interrogative words can co-occur with the interrogative particle (see 5.2).

17 The interrogative word kupme 'how much' contains, in fact, a fossilized instance of the particle =mE together with the adverb kup 'much' and must have originated in polar questions such as 'Is there much milk left?'. However, today such questions can only mean 'How much milk is left?' and cannot be answered 'yes' or 'no'.

4.1. Interrogative word inventory

The following interrogative words are attested in my data18:

• kem 'who (only human)'19 (cf. (71));

• nima 'what, who (inanimates and non-human animates)' (cf. (72));

• nama 'what, who (inanimates and non-human animates)' (rare; < nama 'thing, something');

• ni(j) 'what' (rare, mostly in fixed expressions);

• nisa 'how many (count quantity)' (has forms nisa-w 'how many [of them]', nisa-sar 'how many (each)', nisa-nse 'which ("how manieth")'

• (also ni saqb <what to> 'how many');

• küpme 'how much (mass quantity)' (< küp=me <much=Q>) (cf. (70));

• küpmega 'how (to what extent)' (küpme-ga <how.much-DAT>);

• nisek 'how (in what manner)';

• qalaj 'how (in what manner)' (rare) (also 'what a.!' in exclamatives);

• nislap 'why';

• niya 'why' (probably from ni-ga <what-DAT>)

• (also nima-ga <what-DAT> 'what for', ni oson 'why, what for');

• nindaj 'what [kind of]' (may be "headless", as in 'What [kind of beer] do you want?');

• qaj 'which [out of several]' (?) (rare, mostly in fixed expressions);

• qajha 'which [out of several]'

(also has a "headless" form with a possessive suffix: qajhs-hs- 'which one', plural qajhs-lar-s- 'which ones');

• qajnd(hd) 'which one [out of several]' (cf. (69));

• qaj-da 'where [at]; where [to]' (cf. (68), (71));

18 Cf. the lists in [Poppe 1964: 55-56; Juldashev (ed.) 1981: 157-158; Dmitriev 2008 (1948): 100-103].

19 The interrogative word kem can also be used to ask about personal proper names, cf. (105) Hiney isemey kem? 'What's your name?' (although nisek 'how' is possible too); (94) Kem ale ul? 'What [was] his [name]?'. See [Idiatov 2007: 67-74] on analogous uses in different languages.

• qaj-dan 'where from';

• qa(j)san 'when' (cf. (67));

• (nisla- / nimasla- 'do what').

The 'where' question root qaj- combines with only two of the three spatial case suffixes: ablative (qaj-dan) and locative (qaj-da), while dative *qaj-ya is unattested — qaj-da is used in both locative and directive meanings. It is the same pattern that is observed in demonstrative adverb(ial)s: formally locative forms bsnda 'here' and unda/sunda/anda 'there' are used in both locative and directive meanings.

The most frequent and basic word for 'what' is nima (18 tokens in our texts20), while nama is rare (1 token). The word ni(j) is rare too (4 tokens), and is used mostly in fixed combinations such as ni sayat <what hour> 'when', ni oson <what for> 'why', ni saqls <what to> 'how many'.

The 'what'-words nima and nama can refer not only to inanimates (things), but also to non-human animates (animals), while kem 'who' can refer only to humans (including small children). Cf.:

(73) Unda nima / #kem? there what who

{We hear growling (of an animal) in the street:} 'What's there?' (with kem it would be understood as referring to a human being as the source of the growling)

There is a three-way contrast of attributive interrogative words: nindaj 'what [kind of]' vs. qajhs~qajns(hs) 'which (one) [out of several]' vs. nisanse 'which ("how manieth")'.

The noun-like interrogative words — at least nima, kem, and "headless" qajhs 'which (one)' — can attach case suffixes, cf. (71) above, (81), (90) below. At least some of them can attach the plural suffix: nima-lar 'what (several things)', kem-dar 'who (several people)' (cf. (74)), nindaj-dar 'what kind of them', qajhs-lar-s-n 'which ones (ACC)'.

(74) Kem-dar unda huyss-a? who-PL that.LOC fight-IPFV 'Who's fighting there?'

20 Only the uses in questions (including embedded) were counted.

Two quantity interrogatives are contrasted: a mass one kupme 'how much' and a count one nisa 'how many'. The count quantity interrogative nisa 'how many' behaves morphologically like a numeral in that it attaches special numeral suffixes. It can attach the "headless" numeral suffix -(A)w (nisa-w 'how many [of them]'; cf. bis-aw 'five [of them]'), as well as the distributive numeral suffix -(s)Ar (nisa-sar 'how many (each)'; cf. bis-ar 'five (each)'). There is also an ordinal formnisa-nse "how manieth" (cf. 'RANK' in [Cysouw 2004: 1, 4-5, 19]) with the ordinal numeral suffix -(E)nsE (cf. ike-nse 'second', utad-ansa 'thirtieth'). Cf. an example:

(75) Hineq nisa-nse bala-q taw-da? thou.GEN how.many-ORD child-P.2SG be.born-PST ~'Which [number] child did you give birth to?' (lit. "Your how manieth child was born?")

Finally, there is a possible case of (nascent) interrogative verb: nisla- or nimasla- 'do what'21. Questions about actions ('do what?') in Bashkir frequently involve fused combinations of nima /ni 'what' and esla- 'do, work': nisla- and nimasla-. Cf. elicited examples (76)-(78) below. Forms nislarga '[don't know] what to do' andnimaslagan 'did so(?)' were also attested in the texts. Also, as mentioned above, a fossilized interrogative adverb nislap 'why' was attested, which obviously comes from a -p converb form of nisla-.

(76) (Hin) nimasld-j-[h]eq / nima (e)sla-j-[h]eq thou what.do-IPFV-2SG what do-IPFV-2SG / nisla-j-heq / ni (e)sla-j-heq? what.do-IPFV-2SG what do-IPFV-2SG

'What are you doing?'

(77) Nimasla-ne-q? / Nima esla-ne-q? what.do-PST-2SG what do-PST-2SG

{A boy comes home with a black eye. His mother:} 'What happened to you?' (lit. "What did you do?")

(78) Miqa nisla-r-ga / ni esla-r-ga / nima (e)sla-r-ga? I.DAT what.do-POT-DAT what do-POT-DAT what do-POT-DAT 'What shall I do?'

21 Cf. also a mentioning of nisla- in [Juldashev (ed.) 1981: 157].

This fusion involves not only the elision of the vowel /e/ (which is also possible in other cases, cf. irtaga (e)slaj 'Tomorrow he works'), but seemingly also the integration of the two words into a single phonological word, bearing a single stress (the stress is marked by an acute in the examples above). A similar process seems to be going on in combinations sulaj it- ~ sulajt- 'do so', ulaj it- ~ ulajt- 'do so', and seems to have finished in the case of nejt- 'do... what's its name' (< nej 'what's its name' + it- 'do').

At the same time, as can be seen from the examples, there is always the possibility to use the full alternative (ni esla- or nima esla-), and the speakers themselves supplied those full alternatives after the fused ones when slowly repeating (dictating) during the elicitation.

Interrogative words/roots are used in Bashkir as a base for derived indefinite pronominals, cf. kem-der <who-maybe> 'someone', qaj-da-lsr <where-LOC-maybe> 'somewhere', ber qaj-da(=la) <one where-LOC=also> 'anywhere'.

Also, at least some of them may be used on their own as "bare" indefinites:

(79) ... Ul nima esla-j al-ma-j inde.

that what work-IPFV take-NEG-IPFV now

'(...) then he can't do anything'.

[110717_gaj_Dialog_o_muzhe.015]

4.2. Interrogative word position

According to the elicitation data, the (single) interrogative word can be placed in several different positions in a clause, including staying in situ (cf. (83), (89), (95a), (96a), (97a)22) and occupying clause-initial (cf. (80a), (81a), (82b)) or preverbal23 (cf. (84a)) positions:

(80a) Kem-de kisa et tesla-ne ~ tesla-gan?

who-ACC yesterday dog bite-PST bite-PC.PST

(80b) / Kisa kemde et teslane ~ teslagan?24

22 Only unambiguous examples are listed.

23 The preverbal focal position is prominent in Bashkir as well as in other Turkic languages, cf. [Nikitina, this volume].

24 Here and in other similar examples, cf. (86b), what we see is probably the complex result of application of two rules: interrogative word being placed into clause-initial position and topic being placed at the left

(80c) / Kisa et kemde teslane ~ teslagan?

(80d) / *Kisa et teslagan kemde?

'Whom did the dog bite yesterday?' (the first three variants were all judged equally grammatical by two speakers)

(81a) Kem-ga Bulat huq-ta? who-DAT Bulat hit-PST

(81b) / Bulat kemga huqta? 'Whom did Bulat hit?'

(82a) Bal qasan bul-da? this when be-PST

(82b) / OKQasan bal bulda? 'When did it happen?'

(83) Bal ayas nindaj ayas? this tree what tree

'What [kind of a] tree is this tree?' (part of a non-verbal predicate)

(84a) Bal jar-da kem jad-yan?

this song-ACC who write-PC.PST

(84b) / OKKem bal jarda jadyan? 'Who wrote this song?'

Two of these three patterns were also elicited for focal constituents in declarative sentences, but seemingly not clause-initial placement (88b):

(85) Mine kisa ET tesla-ne.

I.acc yesterday dog bite-PST {Who bit you yesterday?} 'A DOG bit me yesterday'.

(86a) Mine et KISA tesla-ne.

i.acc dog yesterday bite-PST

(86b) /Mine KISA et tesla-ne.

{When did the dog bite you?} 'The dog bit me yesterday'.

periphery ("left-detachment"). Cf. [Nikitina, this volume] on left-detachment in Bashkir.

(87a) BULAT Azat-qa huq-ts. Bulat Azat-DAT hit-PST

(87b) / Azatqa BULAT huqts.

{Who hit Azat?} 'Bulat hit Azat'.

(88a) Bulat AZAT-QA huq-ts. Bulat Azat-DAT hit-PST

(88b) / */OKAZATQA Bulat huqts.

{Whom did Bulat hit?} 'Bulat hit AZAT'.

However, most of the elicited examples with interrogative words were translated with what can be analyzed as in situ placement25, so this should probably be considered the default strategy.

In our texts, too, reliable examples were only found for what can be analyzed as in situ placement (89)-(90) (and less straightforward (91)-(93)), except for one possible case of clause-initial placement (94).

(89) Asdj, hin qaj-da?.. mother thou where-LOC 'Mom, where are you?'

[150708_img_Kljuchi.054] (non-verbal predicate; in situ)

(90) Hiqa nima-ga anda? thou. DAT what-DAT there

'What do you want to go there for?' (lit. "To you what for there?") [140719_aamB_Vojna.057] (adjunct; in situ?)

(91) Taysn nima tot-qan-dar ine die u[l]? again what keep-PC.PST-PL be.PST now that 'What else did they take [in their hands]?'

[150707_rgg_Krtek2] (direct object; in situ (or preverbal))

(92) Sun [an], uns nisek tot-or-ya ikdn? that.ABL that.ACC how hold-POT-DAT be.PC.PST 'Well, how can I catch him?'

[110717_gaj_Dialog_o_muzhe.005] (adjunct; preverbal (or in situ))

25 With the reservation that in situ placement of subjects in most cases cannot be distinguished from clause-initial placement, as well as in situ placement of adverbials and objects — from preverbal placement.

(93) Kem miney... minän basqort tel-e-ndä

who I.GEN I.ABL Bashkir language-P.3-LOC imtixan al-a? ti.

exam take-lPFV say

'Who will give my. me an examination in Bashkir? (I said)'.

[110717_rmm_Ekzamen.010] (subject; in situ (or clause-initial))

(94) Kem äle ul?.. who now he

{Remembering:} 'What was his name?..' (lit. "Who [is] he now?")

[140705_zhm_Dialog.038] (non-verbal predicate; clause-initial)

My data on questioning elements of embedded clauses in Bashkir is scarce, but I can note that several speakers allowed in situ placement and at least one speaker allowed clause-initial placement of interrogative words in such sentences:

(95a) Hin [bal jar-da kem jaö-yän tip]

thou this song-ACC who write-PC.PST say.CV

ujla-j-hay?

think-IPFV-2SG

(95b) / OK * Kem hin [bal jarda jadyan tip] ujlajhay? 'Who do you think wrote this song?'

(96a) Hin [Bulat kem-de tuqma-yan-a-n] kür-de-y? thou Bulat who-ACC beat-PC.PST-P.3-ACC see-PST-2SG

(96b) / OKKemde hin [Bulat tuqmayanan] kürdey? 'Whom did you see Bulat beat up?'

(97a) Hin [Azat-ta kem tuqma-yan-a-n] kür-de-y? thou Azat-ACC who beat-PC.PST-P.3-ACC see-PST-2SG

(97b) / OKKem hin [Azatta tuqmayanan] kürdey? 'Who did you see beat up Azat?'

Finally, here are some elicited examples of sentences with multiple interrogative words (cf. also (71) above). As can be seen, there is variation with respect to position (at least clause-initial vs. preverbal), relative ordering, and employment of simple juxtaposition

vs. coordination with ham 'and'. However, at least in all the attested examples the interrogative words are adjacent.

(98) Qaj-da nima bul-ds? where-LOC what be-PST 'What was where?'

(99) Kem kem-ga nima bir-gan? who who-DAT what give-PC.PST / Kemga kem nima birgan?

{All people have given presents to each other.} 'Who gave what to whom?'

(100a) Qaj-da ham qasan bsl (xal)

where-LOC and when this state. of. affairs

bul-yan? be-PC.PST

(100b) Bsl qaj-da (ham) qasan bul-ds?

this where-LOC and when be-PST

(100c) (Bsl xal) qasan qaj-da bul-yan

this state.of.affairs when where-LOC be-PC.PST / bul-ds? be-PST

'Where and when did it happen?'

(101a) Kem nisek (ham) ni oson esla-ne? who how and what for do-PST

(101b) Bsns kem nisek ni oson esla-gan?

this.ACC who how what for do-PC.PST / OKKem nisek ni oson bsns eslagan? 'Who did (it), how and what for?'

4.3. Intonation in content questions

In content questions, the pitch rises until (roughly) the stressed syllable of the interrogative word and then falls until the end of the utterance26 (cf. [Poppe 1964: 96]). The location of the peak of this

26 The contours of content questions are often similar to those of the polar questions (see 2.4), but in general polar questions involve more sharp

rising-falling contour is tied to the interrogative word, but only loosely: this peak may also drift to the left or to the right from the interrogative word by at least one syllable (cf. (104a-b), (164a)). If the interrogative word is clause-initial or clause-final, there may be almost no rising (102) or no falling (164a), respectively.

Consider the diagrams below for content questions with different position of the interrogative word: clause-initial (102), clause-medial (103)-(104), and clause-final (105):

(102) Kem unda kil-ä?

who that.LOC come-lPFV 'Who is coming there?'

350_

^ V

kem unda kilä

(Cf. also the diagrams for examples (162a) Kem kit-te? 'Who left?' and (163a) Kem gorotqa bara? 'Who's gonna go to town?' in section 6.2 below).

(103) Bulat qasan kil-de? Bulat when come-PST 'When did Bulat come?'

400_

N

bu lat qa san kil de

rising and/or falling slopes and more precise location of the peak than content questions.

(104a) Unda kem kit-ep bar-a?

that.LOC who go.away-CV go-lPFV 'Who's walking there?'

— V- \

unda kem ki tep bara

J-

unda kem ki tep bara

(104b) Unda kem joro-j?

that.LOC who walk-IPFV 'Who's walking there [round our house]?'

350_

V— )

un da kem jö röj

450

un da kem jö röj

(105) Hineq isem-eq kem? thou.GEN name-P.2SG who 'What's your name?' (lit. "Who [is] your name?")

250__

hin [eg] isemeg kem

(Cf. also the two diagrams for example (164a) Bulat qaj-da? 'Where's Bulat?' in section 6.2 below with a clause-final interrogative word).

5. Tags and other peripheral markers

In addition to the default types of interrogative marking, which have been described above, in Bashkir there is also a large inventory of peripheral interrogative markers — tag-like and particle-like — that are used to form biased or otherwise non-neutral (polar) questions.

The following markers were attested (they are presented together with their approximate meaning in questions and their etymology/ literal meaning):

1) tag-like interrogative markers:

.. .(*=mE)(,) tügelme?'isn't it? , right?'

.. .(*=mE)(,) äje M?'is(n't) it?, right? [no doubt]'

...(=mE), äjeme? 'is(n't) it?, right?'

...(=mE), dörö0mö?'is(n't) it?, right?'

2) particle-like interrogative markers:

... =mE (ä)llä? or what, really?'

... =mE ni(j)? or what, really?'

(< 'isn't?') (< 'yes, right?') (< 'yes?') (< 'true?')

(< 'or/unknown') (< 'what')

3) other particle-like elements used in questions:

...(*=mE)bit21 ? is(nt) it?, right?' ... =LEr (*=mE)? 'maybe, am I right?' .. ,(*=mE ) inde? 'probably, am I right?'

('~ right?') ('maybe') (< 'now')

... ikän ? ' [polite] ; maybe ? ' (< i . huy? or what, really?' . ul? or what, really?'

(< ikän <be.PC.PST>)

(< 'after') (< 'that')

The first group consists of elements resembling what are known as "interrogative tags" — interrogative elements that, prototypically, look like separate clauses or even utterances, have their own intonation contour and attach to whole, formally declarative clauses (cf. [Cuenca 1991: 6-8; Dixon 2012: 392-393]). In the case of Bashkir, four elements were attested that correspond (partially) to this definition.

The second and the third groups consist of particle-like elements. Elements in the second group were attested only in questions (and thus seem to be specialized interrogative markers). Elements in the third group appear both in questions and in other illocutionary sentence types, but demonstrate some unpredictable behavior and/or meaning when used in questions.

The tag tügelme (lit. "isn't?") is used in biased questions with a positive answer expectation — either when the speaker asks to simply agree with a statement he believes in (cf. (106)-(101)) or when the speaker asks to be dispelled of his sudden doubts about a statement he previously believed (cf. (108)-(109)). It is mostly used with positive-polarity clauses, but seems to allow at least constituent negation (109) (cf., however, (110))28. The clause it attaches to is pronounced with declarative intonation and cannot include the interrogative particle.

(106) Azat-ta nej-gä saqar-yan-dar tej,

Azat-ACC what's.its.name-DAT call-PC.PST-PL say.IPFV

27 The particle bit is a loanword, originating in the Russian particle ved' 'you know; right?, isn't it?'.

28 A possible explanation for the contrast attested between (109) and (110) is that in the most plausible reading of the latter negation takes a wider scope ('It is not true that you are deceiving me, right?'), than in the former ('It was true/said that you were not working, right?').

tugel=me?

NEG.AUX=Q

'Azat, it seems they invited him to what's its name [=some institute], right / didn't they?'

[110700_rtk_rrg_Dialog_o_postuplenii.052]

(107) Hin jelak jarat-a-hsr](,) tugel=me? thou berry love-lPFV-2SG neg.aux=q

'You like strawberries, right? [expecting a "yes" answer]'

(108) Asaj IRTAGA kil-a tugel=me? mother tomorrow come-lPFV neg.aux=q 'Mother is coming tomorrow, is she not? [with doubt]'

(109) Hin esla(OK-ma)-j-he] tugel=me? thou work-neg-IPFV-2SG NEG.AUX=Q

'[I thought] you were working, weren't you? [with doubt]' / '[I thought] you weren't working, were you?' (for the alternative with the negative suffix)

(110) ???/OKHin mine alda-ma-j-hs] tugel=me?

thou I.acc deceive-neg-IPFV-2SG neg.aux=q Intended meaning: 'You aren't deceiving me, are you?' ~ 'Aren't you deceiving me?'

Another characteristic of tugelme is that it is frequently pronounced without a pause before it (indicated by the absence of a comma in the examples). A similar example without a pause was attested for the next tag, aje bit. These two elements are probably in the process of turning from tags into clause-final particles.

The tag aje bit (lit. "yes, right?") is used in biased questions with a same-polarity answer expectation, mostly when asking for agreement with a statement that the speaker is sure of (cf. (111)-(114)). The clause it attaches to is pronounced not with declarative, but with polar interrogative intonation29 (although it cannot include the interrogative particle — probably due to the presence of bit, see below).

29 Which is a deviation from what we expect from (canonical) interrogative tags.

(111) OKHin mine alda-ma-j-hsq(* =ma), aje bit? thou I.acc deceive-neg-lPFV-2SG=Q yes you.know 'You aren't deceiving me, are you? [with hope; we are friends, I believe that you are not deceiving me]'

(112) Hin Maskaw-da bul-yan-haq, aje bit? thou Moscow-LOC be-PC.PST-2SG yes you.know 'You've been to Moscow, haven't you / right?'

(113) Azat kisa kit-te, aje bit? Azat yesterday go.away-PST yes you.know

'Azat left yesterday, did he / right? [we already know, we are expecting a confirmation, agreement]'

(114) Asaj IRTAGA kil-a (a)je bit? mother tomorrow come-IPFV yes you.know

'Mother is coming tomorrow, right? [there is no doubt]' (cf. (108) above)

The two remaining tags ajeme (lit. "yes?") and doroOmo (lit. "true?") are also used in biased questions that contain expectations of an answer with the same polarity which does not contradict the speaker's beliefs (cf. (115)-(117), (118)-(119)). Only positive-polarity examples were attested. The clause these tags attach to is again pronounced with polar interrogative intonation, and can include the interrogative particle.

(115) Hin jelak-te jarat-a-hsq, aje=me? thou berry-ACC love-IPFV-2SG yes=Q 'You like strawberries, right?'

(116) Azat kisa kit-te =me, aje=me? Azat yesterday go.away-PST=Q yes=Q

'Azat left yesterday, didn't he / or not? [when we do not know for sure]'

(117) Asaj IRTAGA kil-a bit, aje=me? mother tomorrow come-IPFV you.know yes=Q 'Mother is coming TOMORROW, right?'

(118) Azat kisage kit-te(=me), doroO=mo? Azat yesterday go.away-PST=Q true=Q

'Azat left yesterday, didn't he? [we already suspect that he most likely left; making sure]'

(119) Asaj IRTAGA kil-a bit, doro0=md? mother tomorrow come-IPFV you.know true=Q 'Mother is coming tomorrow, right? [we already know, just reassuring]'

Formally, all the particle-like elements are characterized by their clause-final position. The particles alla and ni(j) are always accompanied by the neutral interrogative particle =mE, while bit, =LEr, and inde, on the contrary, are incompatible with =mE, cf. (120a), (130). The particles alla and bit may occupy, aside from the clause-final, an alternative position after the first constituent — cf. (120b), (122c) and (136).

The particle bit is used in questions similarly to the four tags: it expresses expectations of an answer with the same polarity which does not contradict the speaker's beliefs. Cf. (120)—(121) (and also (117), (119) above).

(120a) Hin (mine) alda(s)-ma-j-hs] (*=ms) bit?

thou I.ACC deceive-NEG-IPFV-2SG=Q you.know

(120b) / Hin bit aldas-ma-j-hs]?

thou you.know deceive-NEG-IPFV-2SG 'You aren't deceiving (me), are you? [hoping that not; expecting a "yes" answer]'

(121) Asaj-em IRTAGA kil-a bi(t)?

mother-P. 1SG tomorrow come-IPFV you.know 'Mother is coming TOMORROW, right?'

The particles alla, ni(j), huy, and ul are, on the contrary, used in biased questions that contain expectations of an answer with the same polarity which at the same time contradicts the speaker's (prior) beliefs — i. e. the speaker asks the addressee to confirm their doubts. Cf. (122)-(126):

(122a) Bulat kit-ma-gan=me nij?

Bulat go.away-NEG-PC.PST=Q what

(122b) / Bulat kit-ma-ne *(=me) (a)lla?

Bulat go.away-NEG-PST=Q or

(122c) / Bulat alla kit-ma-ne?

Bulat or go.away-NEG-PST 'What, Bulat hasn't [really] left? [doubting, hearing a male voice from behind the wall]'

(123a) OKBulat kit-te *(=me) [a]lla? Bulat go.away-PST=Q or

(123b) OKUl kit-te=me hug?

that go.away-PST=Q after 'Did Bulat/he really leave? / What, Bulat/he has left? [unexpectedness, doubt]'

(124) Sertifkat tarat-sp jôrô-j-ôq=mô nij? certificate spread-cv go-lPFV-2SG=Q what 'What, you're handing out certificates?'

[110700_rtk_rrg_Dialog_o_postuplenii.001]

(125) OKHin bar-ma-j-hsq *(=ms) hug?

thou go-NEG-IPFV-2SG=Q after 'You what, you're not gonna go?'

(126) OKUnda jel sawla-j=ms ul? that.LOC wind make.noise-lPFV=Q that

'Is it the wind making this noise, or what? [surprised]'

The particles =LEr and inde are used in questions with the meaning 'maybe, probably', marking hypothetical or uncertain questions:

(127) Ber-ga ateO jebar-ge-geô kil-ma-j inde? one-DAT far send-ADJ-P. 2pl come-NEG-lPFV yet 'You nevertheless don't want to send [him] far away, am I right? ' [ 110700_rtk_rrg_Dialog_o_postuplenii. 05 8]

(128) Hin kisage ber-aw-ôe

thou yesterday one-NUM.suBST-ACC kur-gan-eq =der?

see-PC.PST-2SG=maybe 'Yesterday you saw somebody, right?'

(129) Hin mine alda-ma-j-hay=dar?..

thou I.ACC deceive-NEG-IPFV-2SG=maybe {Hesitant:} 'You aren't deceiving me, are you?..'

(130) OKAzat joqla-ma-j=dsr (* =ma)? Azat sleep-NEG-IPFV=maybe=Q

'Azat [shouldn't] be sleeping [I think]? [an uncertain question]'

(131) OKHin eslä-j-heq =der?

thou work-IPFV-2SG=maybe

'You're probably working? / You're working, right?'

Finally, the particle ikän30 (which elsewhere has a meaning close to '—it turns out, that') seems to be used in questions to signal an uncertain or softened — polite question, cf. (132)-(134). (In the latter function it seems to be used also in content questions, cf. (138)).

(132) Bulat bögön öj-öä=m ikän? Bulat today house-LOC=Q be.PC.PST 'Is Bulat at home today? [polite question]'

(133) Huyal-ap kit-mä6=me ikän, call.on.the.way-CV go.away-NEG.POT=Q be.PC.PST qajt-ap bar-yan-da?

come-CV go-PC.pst-loc

'Wouldn't he probably drop in, when he'll be on his way back?'

[110111_gaj_ss_DomaLiMuzh_Dialog.023]

(134) Ul aöas-ap quj-öa=ms ikän? that get.lost-CV put-PST=Q be.PC.PST

{Worried parents, about a daughter who hasn't come home yet:} 'What if she got lost?'

6. Special types of questions

In this final section, I will discuss some special types of interrogative sentences — namely, some non-standard illocutionary types (deliberative, hortative, etc.), echo questions, and embedded questions (which, strictly speaking, are not true questions, but nevertheless a closely related type).

30 See also [Say, this volume].

It should also be noted that the three types are not mutually exclusive, but their combinations would be functionally even more marginal than the pure cases I will concentrate on.

6.1. Non-standard illocutionary types of questions The best-known (and arguably the most frequent) function of interrogative utterances is to (try to) cause the addressee to communicate some information, i. e. to perform a declarative speech act. But there are types of questions that are aimed at causing the addressee to perform some other types of speech acts. Let us concentrate on three such types here. In the examples the questions will be presented (where available) together with the possible answers to them suggested by the speaker.

"Deliberative" questions (that request an imperative speech act) are based either on the "infinitive"31 form in -(E)r-GA / -mAO-KA used autonomously, cf. (135a), (136a), (137), (138a), or on the 1st person hortative form, cf. (135b), (136b), (138b), (139), together with the standard interrogative marking.

(135a) A: — Irtaga miqa kil-er-ga=me?

tomorrow I.DAT come-POT-DAT=Q

(135b) / OKIrtaga kil-aj-em=me?

tomorrow come-HORT-1 SG=Q B: — Waqst-sq bul-ha, kil-0.

time-P.2SG be-coND come-lMP 'A: Shall I come tomorrow? B: If you have time, come'.

(136a) A: — Miqa alda/alla kit-er-ga=me?

I.DAT or go.away-POT-DAT=Q

(136b) / OKMin nima, kit-aj-em=me?

I what go. away-HORT-1 SG=Q

31 This complex non-finite verbal form, traditionally called "infinitive" in Bashkir studies, is composed of the potential suffix -(E)r (or its negative counterpart -mA0) plus the dative case suffix -GA. Its basic function is to form purpose clauses — '(in order) to P'.

B: — Kit-ke-ged kil-ha, kit-$-eged.

go.away-ADJ-P.2PL come-C0ND go.away-IMP-2PL {During a quarrel, an offended person:} 'A: Shall I leave or what?! B: Leave, if you want to'.

(137) A: — Mi]a basqa kil-ma0-ka=me?

I.DAT other come-NEG.POT-DAT=Q B: — Juq, kil-maO-ka, min ud-em-da

NEG.COP come-NEG.POT-DAT I self-P. 1SG-L0C esla-j al-a-m.

work-IPFV take-IPF V-1SG

'A: I shouldn't come anymore?' B: No, don't come, I'll cope with it by myself.

(138a) A: — Maset-ka nisek bar-ar-ya (ikan)?

mosque-DAT how go-POT-DAT be.PC.PST

(138b) / OKMaset-ka nisek bar-aj-am?

mosque-DAT how go-H0RT-1SG

B: — Uram bujla-p bar-sr-ya ham u]-ya

street pass-CV go-POT-DAT and right-DAT borol-or-ya. turn-PASS-POT-DAT

'A: Can you tell me the way to the mosque?' (lit. "How can I go to the mosque?") B: Go straight down the street, then turn right'.

(139) Agursa tura-p ultsr-t-aj[-am]? cucumber chop-CV sit-CAUS-HORT- 1sg=q

'Shall I chop the cucumbers and put them [on the table]?'

[110711_dah_Dialog_o_byte. 001 ]

Apart from 1st person, 3rd person deliberative questions are possible too. In Bashkir, such questions can be formed with jussive forms:

(140) A: — Ts]la-0, Azat irtaga hine

listen-IMP Azat tomorrow thou.ACC qarss al-han=ma? opposite take-JUSS=Q

B: — Aje, qarsa al-han.

yes opposite take-JUSS 'A: Listen, shall Azat meet you tomorrow? B: Yes, let him meet [me]'.

Permissive questions (that request permission)32 involve a number of constructions: several similar constructions based on the "infinitive" form in -(E)r-GA (/ -mAO-KA) in combination with predicates jaraj 'all right, okay' (cf. (141a), (143a)), momkin 'it's possible, it's allowed' (cf. (143c)), or bul- 'be' (cf. (141b), (142b), (143b)); one based on the 1st person hortative form (cf. (144b)); one based on the imperfective form in -A/-j combined with the predicate al- 'take' (has the meaning 'be able, be permitted') (cf. (69) above and (142a) below); and one based on the conditional form in -hA combined with the predicate bul- 'be' (cf. (144a)). All of them are accompanied by standard interrogative marking.

(141a) A: — Miqa ultar-ar-ya jaraj=mz?

I.DAT sit-POT-DAT all.right=Q

(141b) / Miqa ultar-ar-ya bul-a=mz? I. DAT sit-POT-DAT be-IPFV=Q B: — Jaraj. all.right

'A: May I sit down? B: Yes, [sit down]'.

(142a) A: — Min irtaga qala-ya kit-a

I tomorrow town-DAT go.away-IPFV al-a-m=mz? take-IPF V-1 sg=q

(142b) / Miqa irtaga qala-ya bar-ar-ya bul-a=ma? I.DAT tomorrow town-DAT go-POT-DAT be-IPFV=Q

B: — Es-eq-de esla-p bot-ha-q,

work-P.2SG-ACC work-CV end-COND-2SG

32 Permissive questions, again, can refer not only to the 1st person, but also to the 3rd person: May John come today?. However, I have no data on such questions in Bashkir; below I discuss only 1st person permissive questions.

kit-a-he]. go.away-IPFV-2SG

'A: [Asking for permission:] May I go to town tomorrow? B: If you finish your work, you'll go'.

(143 a) A: — Qasan hi]a kil-er-ga jaraj?

when thou. DAT come-POT-DAT all.right

(143b) / Hi]a qasan in-er-ga bul-a?

thou.DAT when enter-POT-DAT be-IPFV

(143c) / Hi]a qasan in-er-ga momkin? thou.DAT when enter-POT-DAT possible B: — Tos-tan hu] kil-0.

noon-ABL after come-IMP {Arranging a visit:} 'A: When may I visit you? B: Come after lunch'.

(144a) Mi]a irtaga kil-ha-m bul-a=ma?

I.DAT tomorrow come-C0ND-1SG be-IPFV=Q

(144b) / Irtaga kil-aj-em=me?

tomorrow come-HORT-1SG=Q 'May I come tomorrow?'

Hortative questions — that arguably request for (acceptance of) an offer to do something together — are based on the 1pl hortative form together with the standard interrogative marking:

(145) A: — Nima, kino-ya bar-aj-aq=ma?

what cinema-DAT go-H0RT-1PL=Q

B: — OK Aje, (ajda) bar-aj-sq.

yes come.on go-HORT-1pl=q 'A: What, let's go to the cinema? B: Yeah, let's go'.

6.2. Echo questions

With an echo question the speaker seeks to receive clarification or confirmation of what the interlocutor has just said, usually repeating ("echoing") at least some part of the original utterance. Thus, echo questions should be classified not only according to their own form, but also according to the form of the "echoed" utterance.

Below I exemplify the coding of different subtypes of echo questions in Bashkir:

1) echoing a declarative sentence by a broad-focus polar question:

(146) A: — Min qala-ya bar-sr-ya jsj-sn-a-m.

I town-DAT go-POT-DAT collect-REFL-IPFV- 1SG

B: — (A?) Hin qala-ya bar-sr-ya

eh thou town-DAT go-POT-DAT jsj-sn-a-(hs)q =m9 ? collect-REFL-IPFV-2SG=Q

/ A, ni ma? Qala-ya bar-sr-ya jsj-sn-a-m

eh what town-DAT go-POT-DAT collect-REFL-IPFV-1SG

tej-(h)eq=me?

say.IPFV-2SG=Q

/ Nima? (/ Aaa?) Qala-ya kit-er-ga what eh town-DAT go.away-POT-DAT

jsj-sn-a-m tip ajt-a-heq=me?

collect-REFL-IPFV- 1SG say.CV say-IPFV-2SG=Q

'A: I'm gonna go to town...

B: [Eh?] (Did you say) you're gonna go to town?'

{A: Yes, I'm gonna go to town.}

(147) A: — Hiqa kem-der saltsra-t-ts.

thou.DAT who-maybe ring-CAUS-PST B: — (A?) (Miqa) kem-der saltsra-t-ts(=ms)?

eh thou. DAT who-maybe ring-C AUS -PS T=Q

'A: Someone called you.

B: Eh? (Did you say) someone called (me)?'

{A: Yes, someone called.}

Cf. also (160b-c) below.

2) echoing a declarative sentence by a narrow-focus polar question (with ellipsis):

(148) A: — Min qala-ya jsj-sn-a-m.

I town-DAT collect-REFL-IPFV-1SG

B: — A? Qala-ya tej-eq=me? / Qala-ya=m3?

eh town-DAT say.IPFV-2SG=Q town-DAT=Q / OK Qala-ya? town-DAT

'A: I'm gonna go to town...

B: Eh? [You're gonna go] to town (you said)?'

{A: Yes, to town.}

3) echoing a declarative sentence by a question with a narrow-focus interrogative-word substitution:

(149) A: — (Irtaga) min qala-ya bar-sr-ya

tomorrow I town-DAT go-POT-DAT jsj-sn-a-m.

collect-REFL-IPF V-1SG

B: — Qaj-da bar-sr-ya jsj-sn-a-hsq?

where-LOC go-POT-DAT collect-REFL-IPFV-2SG=Q / Aa? Qaj-da bar-sr-ya jsj-sn-a-m eh where-LOC go-POT-DAT collect-REFL-IPFV-1SG

(ti-heq)? say.IPFV-2SG=Q 'A: I'm gonna go to town. B: Eh? You're gonna go where (you said)?' {A: To town, I said!}

(150) A: — Aftobus un ike=la jegerme-la kil-a.

bus ten two=also twenty-LOC come-IPFV B: — Un ike=la nisa-la (* =me) ? ten two=also how.many-LOC=Q

/ Un ike=la nisa-la kil-a

ten two=also how.many-LOC come-IPFV ti-heq (*=me)? say.IPFV-2SG

{B: At what time does the bus arrive?}

'A: The bus arrives at twenty past twelve.

B: At how many past twelve (did you say it arrives)?'

{A: At twenty past twelve.}

Cf. also (162b) below.

4) echoing a declarative sentence by a question with a broad-focus33 interrogative-word substitution:

(151) A: — Irtaga mine (sayat) jete-la ujat-0.

tomorrow I.ACC hour seven-LOC wake.up-IMP B: — Aa? Nima (e)sla-r-ya (tip ajt-te-q)?

eh what do-POT-DAT say.CV say-PST-2SG / Nima irtaga? what tomorrow

A: — (Irtaga) sayat jete-la ujat-0 tip

tomorrow hour seven-LOC wake.up-IMP say.CV ajt-te-m. say-PST-1SG

'A: Wake me up at seven tomorrow.

B: Eh... What (did you say) (I should do) tomorrow?

A: Wake me up tomorrow at seven, I said'.

5) echoing a polar question by a polar question:

(152) A: — Hin qala-ya bar-a-(ha)q=ma?

thou town-DAT go-IPFV-2SG=Q

B: — Min qala-ya bar-a-m=mz?

I town-DAT go-IPFV-1SG=Q 'A: Will you go to town? B: Eh? Will I go to town (you said)?' {A: Yes, will you go to town? B: Probably yes.}

(153) A: — Bulat kil-gan=me?

Bulat come-PC.PST=Q

B: — Nima ti-heq? Bulat kil-gan=me?

what say.IPFV-2SG Bulat come-PC.PST=Q

33 Echoing with a substitution for the whole original utterance (apart from the regular broad-focus questioning means nima esla- 'do what?) can also be done with Nima? 'What?', Nisek? 'How?', or A? 'Eh?'.

Bulat kil-gan Bulat come-PC.PST

/ Nima ti-he]? what say.IPFV-2SG 'A: Has Bulat come? B: What? Has Bulat come (you said)?' {A: Yes, has he come? B: Yes, he has.}

ti-he]=me? say.IPFV-2SG=Q

(154) A:

Azat, Azat B: — Nima?

what / Nima, what

ataj(-s]) father-P.2SG

Ataj father ataj-s] father-P.2SG

oj-da=me? home-LOC=Q

oj-da=me? home-LOC=Q

oj-da(=me) home-LOC=Q

tip

say.CV

tej-e]=me?

say.IPFV-2SG=Q

hora-j-hs] =ma ? ask-IPFV-2SG=Q

oj-da(=me) home-LOC=Q

/(Nima,) ataj oj-da(=me) what father home-LOC=Q

/ (Nima tip hora-j-hs]), ataj what say.CV ask-IPFV-2SG father tip=me hora-j-hs]? say.CV=Q ask-IPFV-2SG {Shouting at a distance:} 'A: Azat!.. Is your father at home?..

B: What('re you asking)?.. Is my father at home (you're asking)?..' {A: Yes, is your father at home?.. B: Yes, he's at home!..}

6) echoing a polar question by a question with an interrogativeword substitution:

(155) A:

B:

Hin thou Nima? what

asa-r-ya eat-POT-DAT

tela-ma-j-he] =me? want-NEG-IPFV-2SG=Q

/ Min nima-ne tela(-ma)-j-em?

I what-ACC want-NEG-IPF V-1SG {Person B is absorbed in reading. Suddenly person A asks:} 'A: Look, do[n't] you want to eat [something]? B: What? / Er... Do I want what [you asked]?' {A: Do you want to EAT something?

B: No, I don't.}

7) echoing a content question:

(156) A: — Kem qala-ya bar-a?

who town-DAT go-IPFV B: — Kem qala-ya bara=ma? who town-DAT go-IPFV=Q

/ Kem qala-ya bar-a ti-heq=me? who town-DAT go-IPFV say.IPFV-2SG 'A: Who's gonna go to town? B: Who's gonna go to town (you said)?' {A: Yes, who's gonna go? B: Azat and I.}

(157) A: — Bulat qaj-ôa?

Bulat where-LOC

B: — (Nima?) Bulat qaj-ôa(=md)?

what Bulat where-LOC=Q {Shouting at a distance:} 'A: Where's Bulat?.. B: What?.. Where's Bulat [you said]?..' {A: Yes, where's Bulat?.. B: He's gone!..}

As can be seen from the examples, there are two general strategies that can be used in all these types:

a) the original utterance is echoed as an autonomous clause, which normally looks like indirect speech (i. e. the deictic elements are chosen according to the current speech act);

b) the original utterance is echoed as a finite34 embedded clause introduced by a verb of speech and normally looking like direct speech (i. e. the deictic elements are chosen according to the original speech act).

As for the interrogative particle =mE, its use in echo questions is guided by the same basic rules as in plain questions (the particle is

34 With a possible exception of a uniquely attested non-finite alternative to (155): Asa-r-da hora-na-q=ma? <eat-POT-ACC ask-PST-2sG=Q> {What did you say?} 'Did you ask [if I would] eat?' (lit. "You asked [about] going to eat?"). Cf. example (166c) in the next section.

present in polar questions and absent in content questions), with two special cases.

In the case of echo questions from polar questions using the embedded strategy, two instances of the interrogative particle may be present simultaneously: one for the (embedded) original question35 and one for the echo question itself, cf. (154) above.

In the case of echo questions from content questions (cf. (156)-(157) above), both with the embedded and the autonomous strategies, Bashkir allows the co-occurrence of an interrogative word and the interrogative particle =mE in the same sentence (which is impossible in plain content questions, cf. (67) in section 4 above)36. The presence of the particle in such questions can be understood as reflecting the (polar) interrogativity of the echo question itself, while the interrogative word is "inherited" from the original utterance.

If we compare this with echo questions that involve interrogativeword substitution (cf. (149)-(151), (155) above) we will see that here the particle =mE is not used (similarly to plain content questions). It conforms to the fact that in this case the interrogative word is not inherited but is a part of the echo question itself, both in those questions that use the embedded strategy and those that use the autonomous strategy.

All types of echo questions involve a special intonation contour: the pitch is gradually rising through the whole utterance. This contour may additionally be realized with a higher overall pitch and/or a steeper pitch rise at the end than the corresponding contours of plain polar and content questions. Cf. the diagrams for different types of echo questions without interrogative words (158)-(161) and with interrogative words (162)-(164) (all examples are given together with the corresponding plain questions / declarative sentences for comparison).

While for most types of echo questions this rising contour seems to be the only option, echo questions from declarative sentences

35 Cf. section 5.3 on embedded questions and examples (165a), (166a) ff.

36 Similar behavior had been described for Turkish in [Kornfilt 2013 (1997): 32-36] — with the difference that in Turkish in such echo questions the particle attaches to the interrogative word itself (it can attach to non-predicate focal constituents in plain polar questions too).

(without interrogative words) constitute a special case. They may, too, have the special rising contour (cf. (158b), (159), second pronunciations, and (160c)), but they may also have a contour with a local rise and fall (cf. (158b), (159), first pronunciations) — similar to that of the plain polar questions, see section 3.4.

(158a) MinQALAyAjojonam.

'I'm gonna go to town' (= 1st line in (148)).

(out-of-the-blue declarative sentence)

(158b) Qalaya=mo?

{— I'm gonna go to town.} '— [Eh? You're gonna go] TO TOWN [you said]?' (= 2nd line in (148))

(narrow-focus polar echo question from declarative sentence)

(159) Hin QALAyA barsrya jsjsndq=ms?

/ QALAyA barsrya jsjsnam tejeq=me? {— I'm gonna go to town} '— [Eh?] (Did you say) you're gonna go to town?' (= 2nd and 3rd lines in (146))

(broad-focus polar echo question from declarative sentence)

300,

u

hin qalaya Ьагэгуа jsjsnagms

300

qalaya Ьагэгуа jsjsnam tejegme

(160a) A: — Bulat KIL-DE=me?

Bulat come-PST=Q 'A: Has Bulat arrived [yet]? ...'

(plain non-explicative polar question)

550_

bu lat kil de me

(160b) B: — Bulat KIL-DE.

Bulat come-PST .. B: [Yes,] Bulat arrived ...'

(non-explicative declarative sentence)

550_

) —___

bulat kilde

(160c) A: — Bulat KIL-DE=me nij?

Bulat come-PST=Q what '... A: [What,] Bulat has [already] arrived?'

(broad-focus polar echo question from declarative sentence)

550__

bu lat kil de me nij

(161a) Hin qalaya BARAD=ms ? 'Will you go to town?'

(= 1st line in (152)) (plain non-explicative polar question)

300„___

wv

hin qalaya ba rag ms

(161b) Min qalaya BARAM=ms ?

{— Will you go to town?} '— [Eh?] Will I go to town?' (= 2nd line in (152)) (echo question from polar question)

300

)

min qalaya ba ram ma

Even in echo questions with interrogative words there is a gradual pitch rise spanning over the whole utterance. This is observed both in echo questions with interrogative-word substitution

(162b) and in echo questions from content questions (163b), (164b) (all examples are given together with the corresponding plain questions for comparison):

(162a) Kem kit-te?

who go.away-PST

'Who left?' (plain content question)

450_

kem kit te

(162b) Kem kit-te?

who go.away-PST {— Bulat left.} '— [What?] WHO left?'

(echo question from declarative sentence with interrogativeword substitution)

400

) X- -——

kem kit te

(163a) Kem gorotqa bara?

'Who's gonna go to town?' (= 1st line in (156)) (plain content question)

450___

—

kem gorotqa bara

(163b) Kem gorotqa bara tejey =me? / Kem gorotqa bara=ms? {— Who's gonna go to town?}

'— Who's gonna go to town (you said)?' (= 3rd line in (156) (echo question from content question)

(a rising intonation contour was produced for the second alternative too)

r

kem gorotqa bara tejer) me

(164a) Bulat qajda?

'Where's Bulat?' (= 1st line in (157)) (plain content question)

350

bu lat qaj Öa

>

bu lat qaj Öa

(164b) Bulat qajda(=m9)?

{— Where's Bulat?} '— [What?] Where's Bulat?'

(= 2nd line in (157)) (echo question from content question)

350,

bu lat qaj ôa

550

>

bu lat qaj 6a ma

In the last pair of examples (164a-b), diagrams of two recordings (of the same two speakers) are given for each example. As can be seen, despite the considerable variation of the pitch contour in the plain content questions and even though in this case (with a clause-final interrogative word) their contours may also involve a rise towards the end (164a), the contours of the corresponding echo questions (164b) still differ from them noticeably. In general, they seem to be realized with an overall higher pitch and/or to involve a steeper pitch rise at the end.

6.3. Embedded questions

Embedded, or "indirect", questions provide a "retelling" of some interrogative speech act in an embedded clause. As mentioned above, they are thus not, strictly speaking, true questions (interrogative utterances). But, as in many other languages, in Bashkir they share many formal characteristics with true, non-embedded questions.

Several general constructions were attested in all types of embedded questions (polar, alternative, and content): the finite construction with (quasi-)conjunction tip and several non-finite constructions (nominalizations) based on the forms with suffixes -GAn

<pc.pst> / -GAn-LEQ <pc.pst-nmlz> / -(E)w <nmlz> / -(E)r <pot> combined with possessive suffixes.

In the case of embedded polar questions, all five constructions may be used. In the construction with tip (cf. (165a), (166a), (167a), (168)), the embedded clause looks like a quotation of the presumed original polar question (although on closer inspection it shows some deviations from it, see below) and in all the attested examples includes an instance of the particle =mE.

In the constructions with -GAn (cf. (167c)), -GAn-LEQ (cf. (167d)), -(E) w (cf. (165b), (166b) (167e)), and -(E)r (cf. (166c)) the particle = mE normally is not and cannot be used (cf., however, a seemingly special case in (169)).

Of special interest is the fact that, among the translations of embedded polar questions, a certain number of " A-not-A"-type37 cases were attested (which is not the case for non-embedded polar questions). They contained "A-not-A"-modifications of both the finite (167b) and non-finite (165c), (167f) constructions.

(165a)

(165b)

(165c)

Zöxrä, Zuxra tip

say.cv

Azat minän Azat I.ABL hora-m. ask-PST

Zöxrä, Azat minän Zuxra Azat I.abl kil-ew-eq-de] come-NMLZ-P. 2SG-ACC

Zöxrä, Azat minän Zuxra Azat I.abl kil-ew-eq-de come-NMLZ-P. 2SG-ACC

[hin kil-de (OKl*-y)=me] thou come-PST-2SG=Q

[hineq thou. GEN beles-te. inquire-PST

[hineq thou. GEN kil-mä-w-eq-de] come-NEG-NMLZ -P. 2 S G-ACC

37 "A-not-A", or "disjunctive-negative", is a strategy of forming polar questions that involves a disjunction of the predicate and its negated copy (lit. "You'll go (or) not go?" for 'Will you go?'). It is used as a default strategy in some languages (especially in East Asia and Papua New Guinea), cf. [König, Siemund 2007: 297-298; Dixon 2012: 398-400; Luo 2013: 32 ff., 185-187].

beles-te. inquire-PST

'Zuxra, Azat asked me if you [=Zuxra] had come'.

'I

(166a) (166b) (166c)

i

(167a) (167b)

(167c)

(167d)

(167e)

(167f)

Min Azat-tan I Azat-ABL hora-ns-m. ask-PST-1SG

[(hin) kil-ä-hey=me] tip thou come-lPFV-2SG=Q say.cv

[unsy kil-ew-e-n]

that. GEN come-NMLZ-P.3-ACC

Min Azat-tan I Azat-ABL hora-ns-m. ask-PST-1SG

ok!??Min Azat-tan [unsy kil-er-e-n] hora-ns-m.

I Azat-ABL that.GEN come-POT-P.3-ACC ask-PST-1SG asked Azat if he [=Azat] would come'.

Azat Bulat-tan [min Azat Bulat-ABL I

[min I

Azat Bulat-tan Azat Bulat-ABL tip beles-kän. say.CV inquire-PC.PST

Azat Bulat-tan Azat Bulat-ABL beles-te. inquire-PST

Azat Bulat-tan Azat Bulat-ABL hora-ns. ask-PST

OKAzat Bulat-tan Azat Bulat-ABL beles-te. inquire-PST

OKAzat Bulat-tan Azat Bulat-ABL

kil-de=me] come-PST=Q

kil-de=me come-PST=Q

tip hora-ns. say.CV ask-PST

kil-mä-ne=me] come-NEG-PST=Q

[miney I. GEN

[miney I. GEN

kil-gän-em-de] come-PC.PST-P. 1SG-ACC

kil-gän-leg-em-de] come-PC.PST-NMLZ-P. 1SG-ACC

[miney kil-ew-em-de]

I. GEN come-NMLZ-P. 1 SG-ACC

[miner [kil-ew kil-ma-w\-em-de\

I.GEN Come-NMLZ COme-NEG-NMLZ-P.lSG-ACC

beles-kan / hora-ns.

inquire-PC.PST ask-PST

'Azat asked / inquired of Bulat if I had come'.

(168) Azat minan [OJ-GA qajt-sp bar-a-hsj=m3\ Azat I.ABL house-DAT come-cv go-lPFV-2SG=Q tip hora-ns.

say.cv ask-PST

'Azat asked me if I was going home (or elsewhere)'.

(169) Min Azat-tan [unsj kil-a=me ikan-e-n\

I Azat-ABL that.GEN come-IPFV=Q be.PC.PST-P.3-ACC

hora-ns-m.

ask-PST-lSG

'I asked Azati if heij would come' (cf. (171) below).

Although the construction with tip often looks like direct speech quotation, it has in fact mixed direct-indirect speech properties, in embedded questions as well as in other embedded speech contexts (see [Knyazev, this volume]). For example, in (166a) both the subject pronoun and the verb agreement suffix are chosen according to the reported speech act (as in quotation). In (165 a) and (167a), zero (3rd person) verb agreement is chosen according to the reported or the current speech act, and the subject pronoun is chosen according to the current speech act.

In the case of embedded alternative questions, the construction with tip and two of the non-finite constructions were attested. Here the particle =mE is used in a way similar to true (non-embedded) alternative questions (see 2.3)38 — i. e. several instances of it may be (and usually are) used, both in finite and non-finite constructions:

(170a) OKAzat minan [kisa=me alla bogon=mo Azat I .abl yesterday=Q or today=Q

38 With one possible difference: the pattern 2a (lit. "You that watch Azat gave=Q, or Bulat=Q?"), which is preferred for non-clausal coordination in non-embedded alternative questions, was not observed in the translations of embedded alternative questions.

kil-de-y] tip hora-ns.

come-PST-2SG say.CV ask-PST

/ OKAzat minän [kisä=me ällä bögön kildey=me] tip horans. / OKAzat minän [hin kisä ällä bögön qajttsy=ms] tip horans.

(170b) Azat minän [(miney) kisä=me ällä bögön=mö

Azat I.abl I.GEN yesterday=Q or today=Q qajt-zw-zm-ds] hora-ns. come-NMLZ-P. 1 SG-ACC ask-PST

(170c) Azat minän [(miney) kisä(=me) ällä

Azat I.abl I.GEN yesterday=Q or

bögön(=mö) qajt-qan-zm-ds] horas-ts.

today=Q come-PC.PST-P.1 SG-ACC ask-PST

(170d) Azat minän [kisä=me bögön=mö Azat I.ABL yesterday=Q today=Q qajt-qan-zm-ds] horas-ts.

come-PC.PST-P. 1sg-acc ask-PST 'Azat asked me whether I arrived yesterday or today'.

(171) Min Azat-tan [asa-r-ya telä-j=me juq=ms

I Azat-ABL eat-POT-DAT want-IPFV=Q neg.cop=q ikän-e-n] hora-ns-m.

be.PC.PST-P.3-ACC ask-PST-1sg

'I asked Azat whether he wants to eat or not' (cf. (169) above).

In the case of embedded content questions examples were attested with the tip-construction and with one of the non-finite constructions. As in non-embedded content questions, only the interrogative words are used, but not the particle:

(172a) Min Azat-tan [qasan kil-de-y] tip

I Azat-ABL when come-PST-2SG say.CV

hora-ns-m.

ask-PST-1sg

(172b) Min Azat-tan [unsy qasan kil-gän-e-n]

I Azat-ABL he.GEN when come-PC.PST-P.3-ACC

hora-ns-m.

ask-PST-1sg

'I asked Azat when he arrived'.

(173) Alia minan [kem (hija) huq-ts\ tip hora-ns. Alia I.abl who thou.DAT hit-PST say.CV ask-PST 'Alia asked me who hit me'.

(174) Ir-dar-dan [ular ni oson eske-ne man-PL-ABL that. PL what for drinking-ACC tasla-yan-dar-3\ tura-hs-nda hora-ns. throw-PC.PSt-pl-p.3 about-P.3-LOC ask-PST

'[She] asked the men (about) why they gave up drinking'.

[140708_rmm_Gazeta.007\

7. Synopsis

In this paper I have described the grammatical markers and relevant lexical elements used in different types of interrogative sentences in Bashkir. The main points of the paper are summarized below.

In polar questions (3.1), Bashkir uses an enclitic interrogative particle =m(E) and a special intonation contour. The default position of the particle is after the predicate in full verbal and non-verbal sentences, irrespectively of the information structure. However, in sentences with predicate ellipsis the particle by default attaches to the word/constituent that coincides with the focus domain (which is also most often, but not always, the last word of the sentence). Negative polar questions in Bashkir are neutral and do not presume expectation or previous belief that the answer would be positive.

The use of the particle =mE is definitely the default option for polar questions in Bashkir, but polar questions without the particle (3.2) are accepted by some speakers and are also present in spontaneous texts. No obvious difference, semantic or pragmatic, was noted between the questions with and without the particle.

In alternative questions (3.3), the same particle =mE is used. There is usually more than one instance of the particle, matching the number of coordinated constituents that express the alternatives, or a single instance, attached to the predicate. Coordination involves the conjunction alda~alla 'or' (or a similar one), or sometimes simple juxtaposition. Alternative questions without a single instance of the particle are considered ungrammatical by most speakers.

The particle =mE has some non-interrogative uses (3.6): in alternative conditional-concessive constructions ('Whether you're deceiving

me or not, ...'), in hypothetical conditional protases with -hA ('If you do so, ...'), and in conditional ~ temporal clauses with -n3=m3 ('As soon as / if you come, ...').

Content questions (" wh-questions") are characterized by the use of interrogative words (one or several) and a special intonation pattern, and do not contain the particle =mE (section 4).

The attested inventory of interrogative words consists of at least 13 (synchronically) simplex items plus a number of derived ones (see the full list and commentaries in 4.1).

The position of the interrogative word (4.2) is normally in situ (including interrogative words that question elements of embedded clauses), although two other positions are possible: clause-initial and preverbal positions (the same two can be occupied by focal constituents in declaratives). In multiple-interrogative-word questions the interrogative words are normally adjacent and sometimes coordinated by ham 'and'.

Among the peripheral interrogative markers (section 5) are four tag-like elements and two clause-final particle-like elements; all of them are used to form biased questions. Additionally, there are a number of particle-like elements that are not confined to interrogativity, but demonstrate some unpredictable behavior/meaning in questions.

The three non-standard illocutionary types which were examined (6.1) were "deliberative", permissive, and hortative. "Deliberative" questions (requesting an imperative speech act: 'Shall I come?') may be formed on the base of the "infinitive" form in -(E)r-GA used finitely or on the base of the 1st person hortative form. Permissive questions (requesting permission: 'May I sit down?') may involve the same 1st person hortative form, constructions with the form in -(E)r-GA in combination with the predicates jaraj 'all right, okay', momkin 'it's possible, it's allowed', or bul- 'be', and constructions "-A/-j+al-" ('be able, be permitted') or "-hA + bul-" ('~ be permitted'). In hortative questions ('What, let's go to the cinema?'), Bashkir uses the 1pl hortative form.

Echo questions (6.2) employ two general strategies: a) the original utterance is echoed as an autonomous clause looking like indirect speech (the deictic elements are chosen according to the current speech act); b) the original utterance is echoed as a finite embedded clause introduced by a verb of speech and looking like direct speech (the deictic elements are chosen according to the original speech act).

In echo questions from polar questions (using the embedded strategy), two instances of the interrogative particle =mE may be present (one from the original question and one for the echo question itself).

In echo questions from content questions (both with the embedded and the autonomous strategies), Bashkir allows the cooccurrence of an interrogative word and the interrogative particle =mE in the same sentence (which is impossible in plain content questions and in echo questions with interrogative word substitution).

Embedded ("indirect") questions (6.3) involve two types of constructions: a) the finite construction with (quasi-)conjunction tip; b) several non-finite constructions based on forms with suffixes -GAn <pc.pst> / -GAn-LEQ <pc.pst-nmlz> / -(E)w <nmlz> / -(E)r <pot> combined with possessive suffixes.

Among the translations of embedded polar questions given by speakers some "A-not-A"-type (cf. footnote 37) modifications of finite and non-finite constructions were prominent (lit. "He asked if I came-didn't come" or "He asked of my coming-not coming") — which was not the case for non-embedded polar questions.

Not only the finite construction with tip , but also at least some of the non-finite constructions were attested in all the three major types: embedded polar, embedded alternative, and embedded content questions. The latter two are of special interest. In the case of the embedded polar questions it is possible to interpret them as cases of imprecise periphrasis and not as true indirect questions (as in English I asked him about his coming is not equal to I asked him if he would come). But the possibility of the non-finite constructions in embedded alternative and especially embedded content questions shows that these constructions, too, form true embedded questions (cf. the impossibility of analogous verbal-noun periphrases in English: #I asked him about [his coming yesterday or today]; *I asked him about [his coming when]).

Intonation patterns were studied for all three major types of questions, and also for echo questions.

In polar questions (3.4), the pitch rises sharply on (the stressed syllable of) the word bearing the focus accent and then usually falls. (It is similar to the sharp fall in declarative sentences).

In alternative questions (3.5), the pitch rises sharply on (the stressed syllable of) the first alternative and rises (usually less prominently), or sometimes falls slightly, on the non-first alternatives.

In content questions (4.3), the pitch rises until (roughly) the stressed syllable of the interrogative word and then falls until the end of the utterance.

In echo questions (6.2), the pitch is gradually rising through the whole utterance (and there may be a higher overall pitch and/or a steeper pitch rise at the end than in plain polar and content questions).

Bibliography

Boersma, Weenink 2017 — P. Boersma, D. Weenink. Praat: doing phonetics by computer [Computer program]. Version 6.0.26 (2 March 2017). (http: //www. praat. or g) Cuenca 1997 — M. J. Cuenca. Form-use mappings for tag questions // W.-A. Liebert, G. Redeker, L. R. Waugh (eds.). Discourse and Perspective in Cognitive Linguistics. Amsterdam — Philadelphia: John Benjamins, 1997. P. 3-19. Cysouw 2004 — M. Cysouw. Interrogative words: an exercise in lexical typology. A handout of a talk given at the Bantu Grammar: Description and Theory Workshop 3 (13 February 2004, ZAS Berlin). Manuscript. 2004.

Dixon 2012 — R. M. W. Dixon. Basic Linguistic Theory. Vol. 3: Further

Grammatical Topics. New York: Oxford University Press, 2012. Dmitriev 2008 (1948) — N. K. Dmitriev. Grammatika bashkirskogo jazyka [A Grammar of Bashkir]. Moscow: Nauka, 2008. (First publication: Moscow — Leningrad: AS USSR Press, 1948). Idiatov 2007 — D. Idiatov. A typology of non-selective interrogative pronominals.

PhD thesis. University of Antwerp, Antwerp, 2007. Juldashev (ed.) 1981 — A. A. Juldashev (ed.). Grammatika sovremennogo bashkirskogo literaturnogo jazyka [A Grammar of Modern Standard Bashkir]. Moscow: Nauka, 1981. Juldashev 1997 — A. A. Juldashev. Bashkirskij jazyk [Bashkir language] // Jazyki mira. Tjurkskie jazyki [Languages of the World. Turkic Languages]. Moscow — Bishkek: Kyrgyzstan, 1996. P. 206-216. Knyazev, this volume — M. Ju. Knyazev. Konstrukcii peredachi chuzhoj rechi v bashkirskom jazyke [Reported speech constructions in Bashkir] // This volume.

König, Siemund 2007 — E. König, P. Siemund. Speech act distinctions in grammar // T. Shopen (ed.). Language Typology and Syntactic Description (Second edition). Vol. I: Clause Structure. Cambridge: Cambridge University Press, 2007. P. 276-324.

Kornfilt 2013 (1997) — J. Kornfilt. Turkish [Descriptive Grammars Series]. London — New York: Routledge, 2013. (First publication: London: Routledge, 1997).

Lambrecht 1994 — K. Lambrecht. Information Structure and Sentence Form: Topic, Focus, and the Mental Representations of Discourse Referents. Cambridge: Cambridge University Press, 1994.

Luo 2013 — T. Luo. Interrogative strategies: An areal typology of the languages of China. PhD thesis. Universität Konstanz, Konstanz, 2013.

Mehlig 1991 — H. R. Mehlig. Ekzistencial'nyje i eksplikativnyje voprosy [Existential and explicative questions] // Russian Linguistics 15, 1991. P. 117-125.

Mehlig 2011 — H. R. Mehlig. General yes-no questions and verbal aspect in Russian // Scando-Slavica 57(2), 2011. P. 177-200.

Nikitina, this volume — T. V. Nikitina. Otklonenija ot kanonicheskogo porjadka slov v ustnykh bashkirskikh tekstakh [Deviations from the canonical word order in oral Bashkir texts] // This volume.

Poppe 1964 — N. N. Poppe. Bashkir Manual. Descriptive Grammar and Texts with a Bashkir-English Glossary. Bloomington — The Hague: Indiana University Press, 1964.

Romero, Han 2004 — M. Romero, Chung-Hye Han. On negative yes/no questions // Linguistics and Philosophy 27, 2004. P. 609-658.

Say, this volume — S. S. Say. Lichno-chislovoe soglasovanie nezavisimykh glagol'nykh skazuemykh v bashkirskom jazyke [Person-number agreement of autonomous verbal predicates in Bashkir] // This volume.

Siemund 2001 — P. Siemund. Interrogative constructions // M. Haspelmath, E. König, W. Oesterreicher, W. Raible (eds.). Language Typology and Language Universals. Vol. 2. Berlin — New York: de Gruyter, 2001. P. 1010-1028.