Printer Friendly

Imperative in interrogatives in Estonian (Kihnu), Latvian and Livonian.

1. Introduction

The existing research on Estonian-Latvian linguistic contact in the area of syntax (e.g. Stolz 1991; The Circum-Baltic Languages 2001) has been concerned mostly with comparison of written varieties. Therefore, coinciding dialectal features which have not found their way into the standardized languages have remained unnoticed.

In this paper, we study the use of a special type of interrogative containing imperative marking (II) in the Estonian subdialect Kihnu (spoken on Kihnu Island, belonging to the Insular dialect), in standard and colloquial Latvian and in the Kuronian dialect of Livonian. Compare, for example, the following interrogative sentences in the Estonian subdialect Kihnu (1), Latvian (2), and Kuronian Livonian (3):

(1) Mioko koor rahvamajas laul-ga?

what_kind_of chorus village_hall:INE sing-3IMP

'What kind of choir is singing at the village hall?' (Saar 1980)

(2) Ko tevs no Riga-s berniem lai atnes?

what father from Riga-GEN children:PL.DAT PRTCL bring.PRS.3SG

'What will father bring the children from Riga?'

(3) Kui sieda laz tie-go?

how this:PART PRTCL do-3IMP

'How to do this? (How should this be done?)' (Krautmane 2010 : 70)

Example (1) from Kihnu is a question in which the verb is not in the indicative, as could be expected, but rather in the third person imperative mood, marked by the suffix -ga/-ga or -ka/-ka. The situation is similar in Latvian (example 2), where the only formal means used to encode third person imperatives, the hortative particle lai + finite form of the verb, is employed in the interrogative sentence. Finally, Kuronian Livonian joins Kihnu and Latvian in using its usual strategy for forming third person imperatives, the hortative particle laz + third person imperative form of the verb marked by the suffix -g(o), in questions.

Thus, what these three idioms have in common is the surprising occurrence of the imperative form of the verb in questions. This is a clear grammatical isogloss, which, interestingly, runs between the group Kihnu-Latvian-Livonian and the remaining Estonian dialects and Standard Estonian, as the latter seem to lack such imperative questions. Our aims in the present study are, first, to provide a formal, functional and distributional account of the imperative-in-interrogatives (II) phenomenon in Kihnu, second to compare our findings with the Latvian and Livonian equivalents, and third, to propose the most plausible scenario for the development of this phenomenon in Kihnu.

2. Imperative-in-Interrogatives in Kihnu dialect

The vernacular spoken on Kihnu Island is a subdialect of the Estonian Insular dialect. None of the other documented subdialects of Estonian shows imperative marking in questions, which makes the Kihnu subdialect quite distinctive among them. In Kihnu, the third person imperative marking of the verb is very productive in all possible types of questions; witness (4) containing a yes/no question (formed with the question particle kas) and (5)-(7) exemplifying different wh-questions: in (5), the wh-word functions as a subject, in (6) as an attribute, and in (7) as an oblique (adverbial):

(4) Kas puest sua-ga jahu?

Q shop:ELA get-3IMP flour:PART

'Is it possible to get flour from the shop?'

(5) Kissi oma lapsi kuolita-ga?

who own child:PL.PART educate-3IMP

'Who educates their children?'

(6) Miokost palka nad viel taht-ka?

what_kind_of:PART salary:PART they still want-3IMP

'What (kind of) a salary do they want?'

(7) Kellele Mihkel kosja tul-ga?

who:ALL Mihkel proposing_marriage:ILL come-3IMP

'To whom is Mihkel coming to propose?'

This lack of syntagmatic restrictions is reflected on the paradigmatic axis as well: the imperative form seems to show the full paradigm available for the (historical) 3rd person verb form. Thus, according to Theodor Saar (1980; 1960), such questions can be derived from impersonal sentences displaying impersonal voice marking on the verb; see example (8) and (9), and in case of past time reference, the verb may be in the past tense; cf. example (10):

(8) Kui siokso jolmaga mere min-da-ga?

how such:GEN weather:coM sea:ILL go-IMPS-3IMP

'How is one supposed to go to sea with such weather?'

(9) Miks luba-ta-ga uese ulku?

why allow-IMPS-3IMP during_the_night[ADV] wander_about:INF

'Why is one allowed to wander about during the night?'

(10) Kas laevad ol-ga ljonnast tahakoho tuln?

Q ship:PL be-3IMP town:ELA back come:PST.PTCP

'Have the ships arrived from the town?'

The only combinatorial restrictions that could be observed from the available material concern person distinctions. Third person imperative marking is also found in interrogative sentences with 1st or 2nd person addressees, but most of the attested examples are indirect (embedded) interrogatives; cf. (11) and (12).

(11) Kusuti, kas sia ol-ga meilt lain

ask:IMPS:PST Q you be-3IMP we:ABL go:PST.PTCP

'People asked if you have left us'

(12) Tahoti tiada, kas mia min-ga tana ljonna

want:IMPS:PST know:INF Q I go-3IMP today town:ILL

'People wanted to know if I am going to town today' (Saar 1960 : 66)

The only way to check the availability of direct questions with 1st or 2nd-person subjects was to present them to native Kihnu speakers. We consulted six native speakers of the Kihnu subdialect who live permanently on the island of Kihnu. All but one of the speakers we consulted considered the constructed example in (13) to be impossible in their dialect. The only consultant whose response to (13) was positive regarded such 1st person direct interrogatives as possible in a situation where someone has the feeling that s/he is lost and is deliberating (by him/herself) which way to go.

(13) Kusso mia nud min-ga?

where I now go-3IMP

'Where should I go now?'

This brings us to the question about the semantic (or pragmatic) function of the Kihnu imperative interrogatives--this, much like the question about person restrictions, cannot be exhaustively answered by a study of published dialect texts. The only hints concerning semantics that we had before consulting native speakers were Theodor Saar's and Mari Must's remarks that the third person imperative forms in questions express doubt on the part of the speaker (Saar 1960 : 66; Must 1994). Therefore we decided to present to our Kihnu consultants twelve minimal pairs of constructed questions (based on Saar 1980; 1960)--one with a verb form in indicative and one in imperative mood. Consultants were asked first to evaluate the acceptability of the interrogative sentences containing the imperative form and then to explain the difference in meaning between the sentence with an indicative verb form and the sentence with an imperative verb form.

The majority of our consultants were familiar with the interrogative sentences with imperative marking and considered them functional in the contemporary language (only direct 1st and 2nd person questions were generally seen as ill-formed). As a response to the question about the difference in meaning in the constructed pairs, most of the consultants confirmed that the sentence with the imperative is a deliberating question, which is directed to oneself, not to the listener or anyone else. The purpose of such questions is not to elicit information from the partner in the conversation --on the contrary, the speaker's assumption is that the partner does not have the relevant information--but rather to deliberate over the situation on her own.

Questions which do not seek information but instruction or advice ("direction") have usually been labeled 'deliberating questions' or 'direction questions' (see Huddleston 1994; Metslang 1981 : 30; Palmer 2001 : 128). In the case of Kihnu, it seems that these questions do not necessarily seek an answer from the interlocutor, and therefore they bear resemblance to rhetorical questions; cf. example (14). The deliberating nature of these interrogatives is probably one of the reasons why they are often used in indirect questions in the scope of a negative expression conveying doubt; cf. (15).

(14) Koego esimene oli tama mies, kis kundorbandi viina

SPRL first be:PST.3SG this man who illegal:GEN alcohol:PART

akkas tuoma. (Kas seda voe-ga kirjuta?)

start:PST.3SG deliver:INF Q this:PART can-3IMP write:INF

'The first one who started to deliver illegal alcohol was this man. (Is it appropriate to write this down?)' (Saar 1998 : 179)

(15) Mia utlesi pulmarahva sias, et ei tia kumbal

I say:PST:1SG wedding_crowd:GEN in that NEG know which_one:ADE

enam kahju ol-ga kas kirjutajal siast voi tudrikatel Pardust?

more pity be-3IMP Q writer:ADE pig:ELA or girl:PL:ADE Part:ELA

'I said in the wedding crowd that I don't know who is feeling more sorry,--the writer, for the pig, or the girl, for Part?' (Saar 1984 [1933])

Example (14) is significant not only because it shows the similarity of deliberative and rhetorical questions, but also because it shows the imperative marking on a modal verb of possibility. We have pointed out the fact that the imperative marking in questions does not seem to be syntagmatically or paradigmatically restricted. Example (14) clearly indicates that imperative marking in Kihnu interrogative sentences shows even greater combinatorial freedom than the imperative mood in its typical use as a marker of illocutionary force. Like in other European languages (e.g. English, cf. Sadock, Zwicky 1985 : 159; Polish, Czech, cf. Besters-Dilger, Drobnjakovi?, Hansen 2009 : 171), Estonian modal verbs are not compatible with the imperative mood (cf. example (16)), which indicates that Kihnu third person imperative marking in questions no longer displays the same distribution as the imperative mood.

(16) *Voi seda kirjutada!

can.2IMP this:PART write:INF

'*Can write this!'

As already noted, the use of third person imperative forms is not attested in interrogative sentences in the other Estonian dialects. An automatic query in the Corpus of Estonian Dialects did not reveal any imperatives in interrogative sentences, although admittedly this corpus is not very rich in questions. (1) The basic function of this form in all Estonian dialects and Standard Estonian is to express indirect commands that can be addressed to any person; cf. example (17) from Standard Estonian. This generalized use of the third person imperative marker to express commands whose addressee can also be first and second person is known in Estonian descriptive tradition as jussive mood (see Erelt 2002).

(17) Isa utles, et ma/sa/ta min-gu ema juurde maale

father say:PST.3SG that I/you/s/he go-3IMP mother:GEN to[POSTP] country:ALL

'Father said that I/you/she/he should go to mother's place in the countryside'

The jussive is rather common in Estonian dialects, although not in interrogative sentences. One explanation for this would be that such imperative (or jussive) interrogatives have been in use, but have been lost. In contemporary Estonian one can actually find a very restricted use of the imperative in certain frozen rhetorical questions, which display the structure mis 'what' + see 'this' + ol-gu 'be-3IMP'; witness (18) and (19) from the Corpus of Written Estonian.

(18) Mis kuradi dekadents see ol-gu--me umber sarga tantsime?

what devil:GEN decadence this be-3IMP we around coffin:GEN dance:1PL

What the hell kind of decadence is this--we are dancing around the coffin?' (ILU1990\ilu0204)

(19) Noh, mis see siis ol-gu!?--

INTERJ what this then be-3IMP

peaaegu et rbbgatasid lugupeetavad

almost CONJ yell_out:PST:3PL honourable:PL

'Well, what is this--the honourables almost yelled out?'

Is this a relic of an earlier, wider use of the third person imperative in interrogative sentences? Unfortunately, this question has to wait for a definitive answer; as yet, both supportive and counter-evidence can be presented for this assumption. Wiedemann notes obsolete uses of imperative in interrogative sentences (like examples 20, 21 and 22) in his grammar of Estonian (Wiedemann 1875 : 468); on the other hand, the Corpus of Old Written Estonian does not reveal any examples of imperatives in questions (Kulli Habicht, personal communication).

(20) mis ta nuud seal teh-ku nii kaua

what s/he now there do-3IMP so long

'What is s/he doing there so long now?'

(21) missugune wara neil waestel ol-gu

which possessions these:ADE poor_person:PL:ADE be-3IMP

'What kind of possessions are these poor people supposed to have?'

(22) ei tea, mis mull ol-gu, nagu minu suda nartsitaks

NEG know what I:ade be-3IMP as_if I:gen heart make_to_wilt:coND

'I don't know what it is with me, as if my heart were being made to wilt away.'

We will return to these sentences from Old Written Estonian in Section 4, where we will discuss the possible triggers for the rise of imperative-in-interrogatives in Kihnu.

3. Imperative-in-Interrogatives in Latvian

Latvian expresses third person imperatives with a periphrastic construction consisting of the imperative/hortative particle lai and the finite indicative form of the verb; cf. example (23) from Holvoet (2007 : 111-112).

(23) Ja vina negrib te palikt, lai iet uz virtuvi

if 3SG.F NEG:want:PRS3 here stay:INF PRTCL go:PRS3 to kitchen

'If she doesn't want to stay here, let her to go to the kitchen'

Much like in Kihnu, this periphrastic construction occurs in interrogative sentences. Holvoet has discussed such occurrences at length in connection with their modality (see Holvoet 2001 : 67-81; 2005; 2007 : 27-29, 111-118); cf. examples (24) and (25).

(24) Ko lai es tagad daru?

what PRTCL 1SG now do:PRS1SG

'What should I do now?' (Holvoet 2007 : 114)

(25) Kur lai es nemu tik daudz lidzeklu?

where PRTCL 1SG take:PRS1SG so much means:GEN.PL

'Where should I get such a sum of money from?' (Holvoet 2007 : 115)

Holvoet (2007 : 27) distinguishes between questions intended to elicit information and questions to which the expected response is directive. According to him (Holvoet 2007 : 114), it is only the second type of questions (so called 'deontic requests'), where the Latvian imperative/hortative marker lai occurs. Such questions have been called 'direction questions' or 'deliberative questions' by other researchers (e.g. Huddleston 1994). The purpose of the speaker in examples (24) and (25) is to elicit a directive rather than information from the addressee. In other words, s/he is expecting to hear something like 'Do this and that!', 'Go here or there!' from the addressee.

As in Kihnu, the Latvian imperative/hortative construction with lai occurs both in wh- and yes/no-questions; see (26) and (27).

(26) Kur lai eju?

where PRTCL go:PRS1SG

'Where should I go?' (Holvoet 2007 : 29)

(27) Vai lai eju?


'Should I go?' (Holvoet 2007 : 29)

On the other hand, unlike in Kihnu, where the imperative marker occurs predominantly in questions with third person subjects, in Latvian the construction with lai occurs in questions mostly with first person subjects (see examples 24-27), although third person subjects are also attested; cf. (28).

(28) Kur nu vina tos piragus tagad lai liek?

where then she those:ACC pastries:ACC now PRTCL put:PRS3SG

'What is she to do with [where is she to put] these pastries now?' (Holvoet 2001 : 68)

Thus, at first glance it seems that Kihnu and Latvian IIs are rather different: while the Kihnu third person imperative is used as a marker of deliberative questions without obligatory addressees apart from the speaker herself, Latvian lai (in combination with the finite verb form) expresses a deontic request to an obligatory addressee. While the Kihnu imperative marker is used in questions mostly with third person subjects, the Latvian imperative/hortative construction is used in questions mostly with first person subjects.

One should note, however, that Latvian lai is not exclusively employed to encode deontic requests. All reactions to a query posted to the Baltistics mailing list (, sent on 20.09.2010) showed that the construction lai + V-IND is used in questions to convey doubt on behalf of the speaker; cf. example (2) repeated in (29) and example (30). In terms of pragmatics, such questions coincide with Kihnu deliberative questions.

(29) Ko tevs no Riga-s bern-iem lai atnes?

what father from Riga-SG.GEN children-PL.DAT PRTCL bring.PRS.3SG

'What will father bring the children from Riga?'

(30) Kas to lai zina?

who it:ACC PRTCL know:PRS3SG

'Who knows it?'

These (constructed) examples are absolutely functional both in Standard and colloquial Latvian (2) as well as in Latvian dialects. (3) The sentence in example (29) has a close equivalent attested in the Kihnu dialect; cf. (31).

(31) Mis aett Ponnast lastolo tuo-ga?

what dad town:ELA children:PL:ALL bring-3IMP

'What will dad bring the children from the town?'

Moreover, the epistemic nature of such questions (i.e. the sense of 'doubt' and 'deliberation') is reflected in Latvian by the positional mutability of the hortative particle. Compare example (29) in which lai occurs at the end of the sentence just before the verb with example (32), where it occurs at the beginning of the sentence just before the subject.

(32) Ko lai tevs no Riga-s bern-iem atnes?

what PRTCL father from Riga-SG.GEN children-PL.DAT bring.PRS.3SG

'What will father bring the children from Riga?'

Intuitively, the agent-oriented modality encoded in a 'deontic request' should be iconically expressed with a word order where lai occurs next to the subject agent. In its epistemic use, on the other hand, lai has a wider (propositional) scope and should be located close to the verbal predicate, which is the core of the proposition. Thus, the availability of sentences like (29) seems to show that lai is no longer only an agent-oriented modal.

4. Imperative-in-Interrogatives in Livonian

In Livonian, third person imperatives are often expressed with a structure which looks like a hybrid form of the Estonian synthetic imperative (or jussive) and Latvian analytical imperative with lai; cf. example (33).

(33) Jegayks laz vol-go sie valikstoks ala,

everyone PRTCL be-3IMP this:GEN authority:COM under

kien voimi yl tam um

who:DAT power:PART over s/he:GEN be.PRS3SG

'Everyone should obey the authority which has power (control) over him' (Krautmane 2010 : 60)

The construction laz volgo shows double hortative marking: the particle laz is a precise functional equivalent of Latvian lai, and the third person imperative morpheme -go is the etymological (and functional) equivalent of the Estonian third person imperative marker -gu/-ku.

As in Kihnu and Latvian, the structure which is used in Livonian in the formation of third person imperatives is also found in interrogative sentences. The evidence, however, is scarce: the only example we have comes from Krautmane's work on the mood system in Kuronian Livonian; (4) cf. (34).

(34) Kui sieda laz tie-go?

how this:PART PRTCL do-3IMP

'How should this be done? (How one should do this?)' (Krautmane 2010 : 70)

5. The development of the imperative from illocutionary force marker to deliberative question marker in Kihnu

The third person imperative has expanded its sphere of use in Kihnu from its original function of expressing orders or requests to a new function of conveying deliberation and doubt in questions. The imperative is essentially a deontic mood and doubt is an epistemic stance of low probability. This means that the Kihnu third person imperative marker has undergone a functional development from deontic to epistemic modality; cf. example (35) and the hypothesized reinterpretation of its modal force in Kihnu.

(35) Kas luomad seda vett juo-ga?

Q animal:PL this:PART water:PART drink-3IMP

'Would the animals drink this water?'

[Do the animals have to drink this water?] > [Is it possible/likely that the animals would drink this water?]

This assumed development is not unprecedented: on the contrary, the development of deontic markers into epistemic ones is very common in the languages of the world (see e.g. Heine, Claudi, Hunnemeyer 1991 : 175-178; van der Auwera, Plungian 1998). A more challenging question is what exactly has triggered this functional shift in Kihnu, especially in light of the fact that other Estonian dialects do not show parallel shifts.

In the search for an answer we looked into the functions of the imperative mood in the documented history of Estonian. Penjam (2004) has observed that in Old Written Estonian, in object and purpose clauses, the third person imperative (jussive) is often used instead of the conditional mood, which is the only option available in contemporary Estonian; cf. (36) in which instead of jaa-ks 'stay-coND' we find ja-ku 'stay-3IMP'.

(36) Kui minna taha / et temma ja-ku / senni kui minna tulle /

if I want:1SG CONJ s/he stay-3IMP until when I come:1SG

mea hohlit sinna seperrast?

what:PART care:2SG you about_it

'And if I want her/him to stay until I come, what do you care about it?' (Penjam 2004 : 74)

According to Penjam (2004), the authors of such texts have interchanged indicative, imperative (jussive) and conditional mood, as well as other devices (such as modal verbs), in search of an appropriate translation equivalent of the German Konjunktiv in Estonian object and purpose clauses. The occurrence of third person imperative (jussive) in object clauses in Estonian is significant because many such clauses are actually embedded questions. Recall example (22), repeated in (37), from Wiedemann, who also observed the peculiar use of the imperative mood instead of the conditional in Old Written Estonian.

(37) ei tea, mis mull olgu, nagu minu suda nartsitaks

NEG know what I:ade be-3IMP as_if I:gen heart make_to_fade:COND

'I don't know what is with me, as if somebody makes my heart fading away' (Wiedemann 1875 : 468)

Such occurrences lead us to the hypothesis that the II sentences in 19th-century Estonian (recall examples (20) and (21) from Wiedemann) are originally embedded questions with the main clause omitted. We may call this a 'desubordination scenario'. This scenario is supported by the fact that in spoken discourse the embedded interrogatives are often treated as containing more important, foregrounded information than the main clause, while the main clause functions more like an epistemic particle. (Keevallik [to appear]) Interlocutors typically answer the question posed in the embedded clause, but not necessarily. Especially the main clause ei tea 'don't know' marks uncertainty, the assumed lack of knowledge by the addressee of the projected question and can thus be left unanswered (Keevallik [to appear]; Keevallik 2006).

By the time Wiedemann was compiling his grammar (1875), such embedded questions with imperative marking were already obsolete in the Estonian spoken and written in northern Estonia; nevertheless, such embedded questions are fully productive in the contemporary Kihnu dialect (recall the indirect questions exemplified in (11) and (12)). The problem that we have to resolve if we adopt this scenario is why only the Kihnu dialect has preserved (or developed) imperative interrogatives, there is no reason to assume a stronger German influence and/or a stronger tendency to preserve archaic features of Old Written Estonian in Kihnu than in any other Estonian dialect. However, we also have to take into consideration the possible impact of Latvian and Livonian. It cannot be a coincidence that the imperative/hortative particle lai in Latvian and its Livonian counterpart laz (5) have both grammaticalized into conjunctions introducing object and purpose clauses; cf. example (38) from Latvian and (39) from Livonian, both of which contain object clauses.

(38) Es ludzu matei, lai vina noperk treso pulksteni

I beg:1SG mother:DAT CONJ she buy:3SG third watch:ACC

'I begged my mother to buy a third watch' (Holvoet 2005 : 98)

(39) Jumal nim um, kyl izientsost pyva, aga meg palam sies

God:GEN name be:3SG PRTCL by_itself holy but we beg:1PL this:INE

palandoksos, laz se sa-go pyvastot ka mad vail

prayer:iNE conj it get-3IMP celebrated also we:GEN among

'God's name is holy as it is, but we pray in this prayer that it becomes celebrated also among us' (Krautmane 2010 : 61)

It follows from the examples above that Latvian and Livonian are structurally isomorphous with the archaic Estonian as documented by Wiedemann. Geographically, the Kihnu dialect belongs to the North Estonian dialect continuum (on which Written Estonian is based), but at the same time it is within the contact area of Latvian and Livonian. Therefore, we may speculate that the desubordination process of embedded questions containing imperatives has been reinforced in Kihnu by contact with Latvian and Livonian. According to this scenario, Kihnu speakers have derived the simple interrogative sentence with a verb in the imperative mood from a complex sentence with omitted main clause. Thus, a question like Kumbal enam kahju ol-ga? 'which_one:ADE more pity be-3IMP' 'Which one is pitying more?' can be seen as a historical descendant of the complex sentence in (40), attested in Theodor Saar's texts.

(40) Mia utlesi pulmarahva sias, et ei tia kumbal

I say:PST:1SG wedding_crowd:GEN in that NEG know which_one:ADE

enam kahju ol-ga kas kirjutajal siast voi tudrikatel Pardust?

more pity be-3IMP Q writer:ADE pig:ELA or girl:PL:ADE Part:ELA

'I said in the wedding crowd, that I don't know who is feeling more sorry,--the writer, for the pig, or the girl, for Part?' (Saar 1984 [1933])


The Kihnu subdialect of Estonian, Livonian and Latvian share a type of interrogative which uses the 3rd person imperative form and which functions mainly as a deliberative question. According to Wiedemann (1875), written Estonian also used to share the feature in some earlier stages. Though the context in which these interrogatives are used varies a bit from language to language, the similarities between Kihnu, Livonian and Latvian are remarkable. This interrogative type is most probably a result of the mutual contacts between Kihnu, Livonian and Latvian, but it may have originally been used as a translation of German Konjunktiv in embedded object clauses (embedded interrogatives). By the desubordination process, it has spread to main clauses. This is the most probable scenario for the Kihnu subdialect, but it is quite clear that contacts with Livonian and Latvian have reinforced the structural spread and the usage of the construction in Kihnu.


ABL--ablative case, ACC--accusative case, ADE--adessive case, ADV--adverb, ALL--allative case, COM--comitative case, COND--conditional mood, CONJ--conjunction, DAT--dative case, ELA--elative case, GEN--genitive case, ILL--illative case, IMP--imperative mood, IMPS--impersonal voice, INE--inessive case, INF--infinitive, INTERJ--interjection, NEG--negation marker, PART--partitive case, PL--plural, POSTP--postposition, PRS--present tense, PST--past tense, PTCL--particle, PTCP--participle, SG--singular, SPRL--superlative, Q--yes/no-question marker.

doi: 10.3176/lu.2011.2.01


We are very grateful to Andra Kalnaca, Dace Markus, Ilja Serzants (for their valuable information about Latvian), to Miina Norvik (whom we consulted about the situation in Salaca Livonian), to Kulli Habicht (whom we consulted about the Old Written Estonian), to Andreas Kalkun and the other participants of the Kihnu expedition (September 2010), and (last but not least) to all language guides from Kihnu.


Besters-Dilger, J., Drobnjakovic, A., Hansen B. 2009, Modals in the Slavonic Languages.--Modals in the Languages of Europe, Berlin--New York, 167-197.

Erelt, M. 2002, Does Estonian Have the Jussive.--LU XXXVIII, 110-117.

Heine, B., Claudi, U., Hunnemeyer, F. 1991, Grammaticalization. A Conceptual Framework, Chicago.

Holvoet, A. 2001, Studies in the Latvian Verb, Krakow.

--2005, Evidentialitat, Modalitat und interpretative Verwendung.--Modality in Slavonic Languages. New Perspectives, Munchen (Slavolinguistica 6), 95-105.

--2007, Mood and Modality in Baltic, Krakow.

Huddleston, R. 1994, The Contrast between Interrogatives and Questions. --Journal of Linguistics 30, 411-439.

Keevallik, L. 2006, From Discourse Pattern to Epistemic Marker. Estonian (ei) tea 'don't know'.--Nordic Journal of Linguistics 29, 173-200.

-- --[to appear] Interrogative "Complements" and Question Design in Estonian. --'Subordination' in Conversation. A Cross-Linguistic Perspective, Amsterdam-Philadelphia.

Krautmane, E. 2010, Kuramaa liivi keele koneviiside susteem eesti ja lati keele taustal, Tartu (Unpublished MA thesis. Institute of Estonian and General Linguistics of the University of Tartu).

Lindstrom, L. 2001, Eesti murrete korpuse iseloomustus argivestlustega vorrelduna.--Keele kannul. Puhendusteos Mati Erelti 60. sunnipaevaks, Tartu (Tartu Ulikooli eesti keele oppetooli toimetised 17), 212-221.

Metslang, H. 1981, Kusilause eesti keeles, Tallinn.

Must, M. 1994, Lectures on Kihnu dialect (Manuscript).

Palmer, F. R. 2001, Mood and Modality, Cambridge.

Penjam, P. 2004, Jussiivi kasutusest eesti vanemas kirjakeeles.--ESA 50 (2004), 68-92.

Saar, T. 1980, Markmeid Kihnu grammatika kohta (Manuscript. Institute of Estonian Language, KT 291).

-- --1984 [1933], Murdetekst Kihnust (Manuscript. Archives of Estonian and Kindred Languages, University of Tartu, T 493).

-- --1960, Ulevaade Kihnu verbist (Manuscript. Institute of Estonian Language, KT 184).

-- --1998, Kihnu raamat, Tallinn (Eesti Teaduste Akadeemia Emakeele Seltsi toimetised 68).

Sadock, J. M., Zwicky, A. M. 1985, Speech Act Distinctions in Syntax. --Language Typology and Syntactic Description. Vol. 1. Clause Structure, Cambridge, 155-196.

Sjogren, A. J. 1861, Joh. Andreas Sjogren's Livische Grammatik nebst Sprachproben / im Auftrage der Kaiserlichen Akademie der Wissenschaften bearbeitet und mit einer historisch-ethnographischen Einleitung versehen von Ferdinand Joh. Wiedemann, St. Petersburg.

Stolz, T. 1991, Sprachbund im Baltikum? Estnisch und Lettisch im Zentrum einer sprachlichen Konvergenzlandschaft, Bochum (Bochum-Essener Beitrage zur Sprachwandelforschung 13).

The Circum-Baltic Languages. Typology and Contact. Vol. 2. Grammar and Typology, Amsterdam--Philadelphia 2001.

van der Auwera, J., Plungian, V. 1998, Modality's Semantic Map. --Linguistic Typology 2, 79-124.

Wiedemann, F. J. 1875 (2005), Grammatik der ehstnischen Sprache, zunachst wie sie in Mittelehstland gesprochen wird, mit Berucksichtigung der anderen Dialekte, St. Petersbourg.


Petar Kehayov

University of Tartu, Institute of Estonian and General Linguistics

tel. (+372) 737 6127

Liina Lindstrom

University of Tartu, Institute of Estonian and General Linguistics

tel. (+372) 737 6127

Ellen Niit

University of Tartu, Institute of Estonian and General Linguistics

tel. (+372) 737 6127

* The work reported on here was supported by the Estonian Science Foundation grant no. 7464 and grant no. 7006.

(1) This is due to the sharp division of the roles of the interviewer and the interviewee during lingustic fieldwork. Also, the interviewer is rarely fluent in the dialect from which linguistic material is collected (see Lindstrom 2001).

(2) Andra Kalnaca, personal communication.

(3) Dace Markus, personal communication, concerning so-called "deep" Latvian idioms of North-East Vidzeme.

(4) Our search for imperative interrogatives in Sjogren's texts of moribund Salaca Livonian (Sjogren 1986) was fruitless.

(5) In Livonian, it is obligatory to use the third person imperative form after the conjunction laz (Krautmane 2010 : 60), which actually means that the whole construction laz + V-3IMP serves to encode the subordination, not the particle alone.
COPYRIGHT 2011 Estonian Academy Publishers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kehayov, Petar; Lindstrom, Liina; Niit, Ellen
Publication:Linguistica Uralica
Article Type:Report
Geographic Code:4EXLA
Date:Jun 1, 2011
Previous Article:Valts Ernstreits, Liivi kirjakeele kujunemine.
Next Article:On voiced stops in Finnish.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters