The Effects of the Slavic–Balkan Contact on Lipovan Daco-Romanian

Abstract

This paper offers both a descriptive account and an analysis of the possible consequences of linguistic contact between the Daco-Romanian variety spoken by the Lipovan community and Russian (starting from a fieldwork-based corpus study) regarding (low) verb movement in neutral readings, ultimately reflected in the preference for [adv-v] word order. The situation identified in Lipovan Daco-Romanian will be compared with that of old and standard Daco-Romanian, Moldovan Daco-Romanian, and Russian.

1. Introduction

1.1. Background

Lipovan Daco-Romanian is used in numerous parts of Romania, especially by Lipovan communities. After several fieldwork sessions that took place between 2018 and 2023, I have observed that there are two tendencies in the morphosyntax of Lipovan Daco-Romanian: (i) on the one hand, it displays a series of archaic features, that can be found in old Romanian (an explanation for this situation can be the fact that the Lipovan community is a conservative community), and (ii) on the other hand, I have observed the influence of Russian on the morphosyntax of Lipovan Daco-Romanian as a consequence of the generalized Russian–Daco-Romanian bilingualism in these communities.

The immigration of Lipovan Russians took place gradually, with the first attempts of the Lipovans to come to Daco-Romanian-speaking territories taking place in the late 17th century. Therefore, there are multiple countries across Europe (besides Romania and the Republic of Moldova) and even on different continents where Lipovan Russians fled after the Russian Orthodox Church split1, i.e., Ukraine, Bulgaria, Hungary, and Turkey and even the USA, China, and Uruguay (Ipatiov 2001, pp. 30–31; Tudose 2015, pp. 129–30). In Romania, there are around 70 settlements with predominantly Lipovan population, more specifically in the north and southeastern of the country (Bukovina, Dobrudja, Wallachia). The immigration of Lipovans in Dobrudja took place in several stages, gradually making up a compact community in which traditions, language, and confessional character have been preserved and consolidated (Prigarin 2007, pp. 3–4; Tudose 2015, p. 163).

1.2. Language Contact and Its Consequences: The Case of Lipovan Daco-Romanian

It is well known that there is not only a single point of view regarding the effects of language contact. Some scholars claim that linguistic contact cannot affect all the domains of a language (Weinreich [1953] 1964, among others), while others state that there are no limitations on the influence of one linguistic system over the other (Sala 1997, p. 133; Trubetzkoy 1939, p. 82). Hickey (2010) developed a more moderate approach according to which every linguistic level could be changed as a consequence of linguistic contact, with the mention that there are notable differences in the rhythm of change. The influence of the Daco-Romanian on Slavic languages in Romania is a complex problem, and this subject has benefited the interest of the researchers in the literature, but the other side of the process, i.e., the way in which Russian influences the Daco-Romanian variety spoken by Lipovan Russians, has not been sufficiently studied. Unavoidably and involuntarily, Lipovan Russians from Dobrudja bring Russian features into their variety of Daco-Romanian, given that their mother tongue is Russian, e.g., in family settings it is always Russian that is employed. Apart from that, a considerable number of hybrid constructions can also be encountered, ultimately the result of both languages being used by the community, albeit in different settings.

1.3. Aim of the Paper, Sources, and Methodology

1.3.1. Aim of the Paper

The examination of the Romanian variety spoken in the Lipovan community from Dobrudja is approached from a linguistic and a sociolinguistic point of view; I will explain the identified particularities of Lipovan Daco-Romanian by comparing the structures that individualize this variety with those from modern Daco-Romanian, old Romanian, Russian, Romance and Balkan languages, and Moldovan Daco-Romanian (on V(erb)-movement, see Kayne 1991; Cinque 1999; Cornilescu 2000; Ledgeway and Lombardi 2005; Nicolae 2015, 2019; Schifano 2015, 2018; Costea 2019). The main objective is to give a descriptive account of the Daco-Romanian variety spoken in the Lipovan community from Dobrudja and to analyze the morphological and syntactic effects of Daco-Romanian–Russian contact regarding verb movement2.

1.3.2. Sources

My research is based on empirical data that I collected from the Lipovan community in Dobrudja, an area that was chosen given the numerous Lipovan settlements. The data are based on ethnographic and linguistic fieldwork, with the examples used being extracted from spontaneous conversations between native Lipovan Daco-Romanian speakers as well as semi-structured interviews with native Lipovan Daco-Romanian speakers. The participants are Russian–Romanian bilinguals aged 30 to 70 who speak Russian in the family, at home, and in their community and for whom Russian represents an important means of keeping the Lipovan identity alive.3 All Lipovan Daco-Romanian examples were taken from original conversations between, and with, Lipovan Daco-Romanian native speakers, and their pragmatically unmarked status was verified once again with different participants.

1.3.3. Methodology

In what follows, I adopt the current standard cartographic approach regarding the clause structure (as sketched in (1) below), according to which the inflectional domain (IP) can be split into three domains, i.e., Mood (MoodP), Tense (TP), and Aspect (AspP), with languages showing different options with respect to verb raising along the clausal spine. In what follows, the approach originally put forward by Giorgi and Pianesi (1997) and then further nuanced by Schifano (2018) will be employed, thus keeping a balance between Cinque’s (1999) cartographic approach (see the hierarchy given in (2) below; cf. also Ledgeway and Lombardi 2005) and the minimalist approach.

(1)		(CP >)	IP MoodP	TP	AspP (> vP > VP)
(2)	a.	Higher Adverb Space (HAS)
		[sincer ‘frankly’ Moodspeech act [din nefericire ‘unfortunately’ Moodevaluative [aparent ‘apparently’ Moodevidential [probabil ‘probably’ Modepistemic [acum ‘now’ T(past/future) [poate ‘perhaps Moodirrealis [necesar ‘necessarily’ Modnecessity [de obicei ‘usually’ Asphabitual [iar ‘again’ Asprepetitive(event) [des ‘often’ Aspfrequentative(event) [intenționat ‘intentionally’ [ușor ‘slowly’ Aspcelerative(event)
	b.	Lower Adverb Space (LAS)
		[deja ‘already’ Tanterior [încă ‘still’ Aspcontinuative [mereu ‘always’ Aspperfect [doar ‘just’ Aspretrospective [curând ‘soon’ Aspproximative [rapid ‘briefly’ Aspdurative [în general ‘typically’ Aspgeneric/progressive [aproape ‘almost’ Aspprospective [de tot ‘completely’ AspCompletive(event) [bine ‘well’ Voice [rapid ‘fast’ Aspcelerative(process) [din nou/iar ‘again’ Asprepetitive(process) [des ‘often’ Aspfrequentative(process) [de tot ‘completely’ AspSgCompletive(process) [v-VP…

2. Verb Movement: Romance Languages, Balkan Languages, Modern and Old Daco-Romanian, Russian, Moldovan Daco-Romanian, and Lipovan Daco-Romanian

2.1. The View from Romance

2.1.1. Synthetic Verbal Forms: Present

In Romance languages, the present exhibits a different distribution with respect to verb raising along the clausal spine: to the mood field in French (3); to the tense field in Northern regional Italian (4); to the aspect field in European Portuguese (5); or just outside of v-vp in Spanish (6) (Schifano 2015).

(3)	Antoine	confond	probablement	(*confond)	le	poème.
	Antoine	confound	probably	counfound	art	poem
	‘Antoine probably confounds the poem.’
	(French, apud Schifano 2015, p. 59)

(4)	Nonna	conosce	già	(*conosce)	la	ricetta.
	Nonna	knows	already	knows	art	recipe
	‘Nonna already knows the recipe.’
	(Northern regional Italian, apud Schifano 2015, p. 12)

(5)	O	João	vê	sempre	(*vê)	este	tipo	de	Filmes.
	art	João	sees	always	sees	this	kind	of	movies
	‘João always watches this kind of movie.’
	(European Portuguese, apud Schifano 2015, p. 68)

(6)	Sergio	contesta	bien	(*contesta)	las	preguntas.
	Sergio	answers	well	answers	art	questions
	‘Sergio is answering well to the questions.’
	(Spanish, apud Schifano 2015, p. 63)

2.1.2. Analytic Verb Forms: Compound Past

As for the verb-movement options of [aux-v] constructions across Romance, it has been argued that auxiliaries are base-generated directly as centers of functional projections from IP (cf. Edmonds 1978). The lexical verb is generated under v, raises to v_Prt to acquire (or license) its participial morphology, then it can either move across aspect or just outside the v-vp. Therefore, the verb has a very high position in French (the auxiliary targets the mood field, but the participle does not rise above the adverb bien ‘well’ (Ledgeway 2012, p. 145) (7)), a clause-medial one in Northern regional Italian and Northern Italian dialects (the auxiliary targets the mood field and the past participle the tense field (8)), a low one in European Portuguese (the auxiliary is found between tense field and aspect field (9)), and a very low one in Spanish (where the auxiliary and the participle leave the v-vp (10)) (Schifano 2015, pp. 185–91).

(7)	a.	Il	a	bien	compris	(*bien)	la	question.
		He	aux.perf.3sg	bine	understand.pple	well	art	question
		‘He understood the question well.’
		(French, apud Roberts 2010, p. 56)
	b.	Il	a	déjà	lu	(*déjà)	le	livre.
		he	aux.perf.3sg	already	read.pple	already	art	book
		‘He has already read the book.’
	c.	Il	est	toujours	allé	(*toujours)	à	la	mer.
		he	aux.perf.3sg	always	go.pple	always	to	art	sea
		‘He always went to the sea.’
		(French)

(8)	Gianni	probabilmente	ha	(probabilmente)
	Gianni	probably	aux.perf.3sg	probably
	sbagliato.
	make.mistakes.pple
	‘Gianni was probably wrong.’
	(Northern regional Italian, apud Schifano 2015, pp. 26–27)

(9)	a.	O	Pedro		já	tem	(*ja)	estudado
		art	Pedro		already	aux.perf.3sg	already	study.pple
		muito.
		a lot
		‘Pedro has already studied a lot.’
	b.	Hoje	o	João	(*sempre)	tem	estado	sempre
		here	art	João	always	aux.perf.3sg	be	always
		a dormir.
		sleep.inf
		‘João slept all day today.’
		(European Portuguese, apud Schifano 2015, pp. 71–72)

(10)	a.	María		ya		ha		hablado		(ya)
		Maria		already		aux.perf.3sg		speak.pple		already
		de		este	asunto.
		about		this	issue
		‘Maria has already spoken about this issue.’
		(Spanish, apud Schifano 2015, p. 72)
	b.	Probablemente		él		ha			venido.
		probably		he		aux.perf.3sg			come.pple
		‘He probably came.’
	c.	Él	siempre	ha		ido	a		la	playa.
		he	always	aux.perf.3sg		go.pple	to		art	sea
		‘He always is going to the sea.’
	d.	Él	ha		comido	bien.
		he	aux.perf.3sg		eat.pple	well
		‘He ate well.’
		(Spanish)

2.2. The View from Balkan Languages

If the situation of the Romance languages is relatively organized (they show different options with respect to verb raising along the clausal spine, to the MoodP, TenseP, or Asp(ect)P field), the Balkan languages do not seem to strictly respect the classical verb-movement patterns. Although the verb—either the lexical verb or the auxiliary—is considered to rise to T/Agr (Alexiadou and Anagnostopoulou 1998, p. 501), adverbs exhibit a relatively free word order (Rivero 1994, pp. 75–76). A consequence of the absence of strict requirements regarding the placement of adverbs with respect to verbs is the (apparently) inconsistent generalizations regarding the level of the verb movement.

2.2.1. Synthetic Verbal Forms: Present

In modern Greek, the position of the verb with respect to various adverbs is ambiguous. Mavrogiorgos (2010, 198 f.n.10; 235 f.n.55; 238 f.n.58) suggests that Greek verbs move to T (for example, the placement of the verb can be either before or after idhi ‘already’ (lexicalizing Spec,T_anterior) (11)).

(11)	a.	Aftos	idhi	divazi	idhi	to	vivlio.
		he	already	reads	already	art	book.def
		‘He is already reading the book.’
		(modern Greek)
	b.	I	Daniela	idhi	ton	iksere	idhi.
		art	Daniela	already	cl.acc.3m.sg	know.pst.3sg	already
		‘Daniela already knew him.’
		(Greek, apud Mavrogiorgos 2010, p. 125f.)

However, the discussion must be nuanced: in modern Greek, there are some adverbs (e.g., ksaná ‘again’) (12) which are attached to the verbal root. They should not be analyzed along the lines of ‘typical’ adverbs, e.g., the ones given in (11) above, inasmuch as it does not indicate the absence of verb raising but rather the fact that the verb undergoes movement together with the attached constituents, i.e., phrasal movement (Rivero 1994, p. 79; cf. also Rivero 1992).

(12)	a.	Den	tha	ksana-kalo-fái	edhó.
		neg	aux.fut.3sg	again.well.eat	here
		‘He won’t eat here well again.’
		(modern Greek, apud Rivero 1994, p. 79)
	b.	Dhen	ksana-kalo-troo	edhó.
		neg	again.well.ear	here
		‘He doesn’t eat here well again.’
		(modern Greek)

Let us now turn to what we find in Bulgarian. Here, the verb can precede and also follow both high and low adverbs, such verojatno ‘probably,’ veče ‘already,’ vinagi ‘always,’ dobre ‘well,’ and često ‘often’ (13) (Krapova 1999, p. 65). Migdalski (2006, pp. 93–96, p. 210) notes that finite lexical verbs must undergo movement across VP-adverbs, reaching an Asp(ect)-related position, and may move further (presumably to a T-related (Spec) position) to prevent pronominal/auxiliary clitics from occurring clause-initially.

(13)	a.	Ivan	(verojatno)	c̆ete	verojatno	ljubovni	romani.
		Ivan	probably	reads	probably	romance	books
		‘Ivan probably reads romance novels.’
		(Bulgarian, apud Krapova 1999, p. 66)
	b.	Toi	vec̆e4	c̆ete	knigata.
		he	already	reads	book.def
		‘He is already reading the book.’
	c.	Te	(vinagi)	otivat	vinagi	na	moreto.
		they	always	go	always	to	sea.def
		‘They always go to the sea.’
	d.	Te	(dobre)	mislyat	dobre.
		they	well	think	well
		‘They think well.’
		(Bulgarian)
	e.	Ivan	(često)	celuva		(?često)	Maria.
		Ivan	often	kiss.ind.pres.3sg		often	Maria
		‘Ivan often kisses Mary.’
		(Bulgarian, apud Kallestinova 2007, p 116)

In Croatian (as well as in Serbian) (14), both high (such as vjerojatno ‘probably’) and low adverbs (such as dobro ‘good’) (14d) precede the lexical verb in contexts that are not pragmatically marked.

(14)	a.	On	(vjerojatno)	jede	vjerojatno.
		he	probably	eats	probably
		‘He probably eats.’
	b.	On	(već)	c̆ita	već	knjigu.
		he	already	reads	already	book.def
		‘He is already reading the book.’
	c.	On	(c̆esto)	idu	c̆esto	na	more.
		he	always	goes	always	to	sea
		‘He always goes to the sea.’
	d.	Oni	(dobro)	misle	dobro.
		they	well	think	well
		‘They think well.’
		(Croatian)

2.2.2. Analytic Verb Forms: Compound Past

Rivero (1994, p. 72) notes that the auxiliary rises to T/Agr in all Balkan languages (but arguably not further, as shown by, among other factors, the fact that modern Greek auxiliaries cannot precede high adverbs, e.g., malon ‘probably’), while the lexical verb occupies a lower position. Evidence for this type of movement comes from the fact that a series of constituents, such as adverbs and floating quantifiers, can be inserted in between the auxiliary and the lexical verb5 (cf. 15).

(15)

a.

Ta

pedhiá

éXun

óla

dhi

aftó

to

érgho.

art

kids

aux.pf.3pl

all

see.pple

this

art

movie

‘All the kids have seen this movie.’

(modern Greek, apud Rivero 1994, p. 82)

b.

Den

echume

pote

grapsi

s

afton.

neg

aux.pf.1sg

never

write.pple

to

him

‘I have never written to him.’

(modern Greek, apud Philippaki-Warbuton 1993, p. 55)

c.

Aftos

malon

ehi

(*malon)

erti.

he

probably

aux.pf.3sg

probably

come.pple

‘He probably came.’

d.

Aftos

ehi

idhi6

divasi

to

vivlio.

he

aux.pf.3sg

already

read.pple

art

book.def

‘He has already read the book.’

e.

Aftos

ehi

(*kala)

fai

kala.

he

aux.pf.3sg

well

eat.pple

well

‘He ate well.’

(modern Greek)

Bulgarian distinguishes between two auxiliaries that correspond to the verb ‘to be’: sûm (a functional auxiliary, used in the formation of the present perfect; it has a lower position) and bjux (an auxiliary used in the formation of the past perfect; it has a higher position). Krapova (1999, pp. 65–67) considers that, given the structure in (16), the past perfect auxiliary rises to T1P, while the participle stands in T2Asp; this could explain why bjax can be separated from the lexical verb by various floating adverbs or quantifiers (17), unlike sûm (18) (but cf. Legendre 2000, pp. 428–29, who argues that the hypothesis proposed by Krapova is not borne out, since both auxiliaries behave similarly with respect to a class of adverbs such as vec̆e ‘already’ (19), verojatno ‘probably’ (20), and dobre ‘well’ (21)).

(16)	[AgrP Agr [T1P T1 [T2/AspP T2/Asp [VP SU [V’ [V OB]]]]]]
	(Krapova 1999, pp. 65–67)

(17)	Ivana	beše	c̆esto	razkazvala	(c̆esto)	tazi	istorija.
	Ivana	aux.pf.3sg	often	tell.pple	often	this	story
	‘Ivana often told this story.’

(18)	Ivana	(*e)	c̆esto	e	razkazvala	(c̆esto)	tazi	istorija.
	Ivana	aux.pf.3sg	often	aux.pf.3sg	tell.pple	often	this	story
	‘Ivana often told this story.’
	(Bulgarian, apud Krapova 1999, p. 62)

(19)	Ivana	vec̆e	e/beše	proc̆ela	knigata.
	Ivana	already	be.aux.pres/aux.pf.3sg	read/read.pple	book
	‘Ivana has/had read the book.’
	(Bulgarian, apud Legendre 2000, p. 429)

(20)	Toi	beše	verojatno	razkazval	(verojatno)	tazi	istorija.
	he	aux.pf.3sg	probably	read.pple	probably	this	story
	‘He probably had read this story.’

(21)	Toi	beše	dobre	mislil	(dobre).
	he	aux.pf.3sg	well	think.pple	well
	‘He had thought this through well.’
	(Bulgarian)

Croatian and Serbian have two paradigms of the auxiliary ‘to be’: one represented by clitic forms (sam, si, je, smo, ste, su) and one represented by non-clitic forms (jesam, jesi, jest(e), jesmo, jeste, jesu) (Tomić 1996, p. 837) (22). The auxiliary seems to rise higher than the lexical verb to the second position of the clause (see Migdalski 2006, p. 170), and, as a consequence, different elements can intervene between them.

(22)	a.	Ja	sam	mu	ih	dao.
		I	aux.pf.1sg	him	them	give.pple
		‘I gave them to him.’
		(Croatian, apud Tomić 1996, p. 839)
	b.	niji	su	žene		kod	kuće		c̆ekale
		they	aux.pf.3pl	wives		at	home		wait.pple
		‘and they waited for their wives at home’
		(Croatian, apud Dragomirescu and Nicolae 2016, p. 458)
	c.	On	je	vjerojatno		dosao		(vjerojatno).
		He	aux.pf.3sg	probably		come.pple		probably
		‘He probably came.’
	d.	On	je	već		proc̆itao		(već)	knjigu.
		He	aux.pf.3sg	already		read.pple		already	book
		‘He has already read the book.’
	e.	On	je	c̆esto		išao		(c̆esto)	na	more.
		He	aux.pf.3sg	always		go.pple		always	to	sea.def
		‘He always went to the sea.’
	f.	On	je	dobro		jeo		(dobro).
		He	aux.pf.3sg	well		eat.pple		well
		‘He ate well.’
		(Croatian)

2.3. The View from Modern Daco-Romanian

2.3.1. Synthetic Verb Forms: Present

In (standard) Daco-Romanian, the issue of verb movement has been discussed by several researchers, thus resulting in two main orientations: on the one hand, it is considered that the verb in Daco-Romanian rises to the highest projection in the mood–time–aspect field, namely MoodP (see, among others, Dobrovie-Sorin 1994; Cornilescu 2000; Schifano 2013, 2014, 2015, 2018; Nicolae 2015, 2019), and, on the other hand, it targets a lower position (for further references, see Ledgeway and Lombardi 2005, 2014; Ledgeway 2012, 2014; Costea 2019). One explanation for this can be the fact that there exists variation among speakers with respect to the level of V-movement in Daco-Romanian, given that speakers allow it to raise to high(er) or low(er) projections. However, it is perhaps the case that a moderate approach should be employed (cf. also Costea and Ledgeway 2024, this issue), given that, as my own experiments have shown (cf. Boioc Apintei 2023), in modern Daco-Romanian, lexical verbs are preceded in pragmatically unmarked sentences by high adverbs, such as probabil ‘probably’ (lexicalizing Spec,Mood_epistemicP) (23) or poate ’maybe’ (lexicalizing Spec,Mod_irrealisP) (cf. 23 and 24), while, in the case of lower mereu ‘always’ (Spec,Asp_perfect) (25), both pre- and postverbal placements can be unmarked, given the fact that some speakers consider its preverbal placement pragmatically neutral, while others consider its postverbal placement pragmatically neutral.

(23)	Ana	probabil	merge	la	serviciu	astăzi
	Ana	probably	goes	to	work	today
	‘Ana probably is going to work today’
	(modern Daco-Romanian, apud Boioc Apintei 2023, p. 124)

(24)	poate	alergăm	împreună
	maybe	run	together
	‘maybe we run together’
	(modern Daco-Romanian, apud Boioc Apintei 2023, p. 124)

(25)	Ana	(mereu)	mănâncă	(mereu)	prăjituri
	Ana	always	eat.ind.pres.3sg	always	cookies
	când	vine	la	noi.
	When	come.pres.3sg	to	us
	‘Ana always eats cookies when she comes our place.’
	(modern Daco-Romanian)

2.3.2. Analytic Verb Forms: Compound Past

With respect to the [aux-v] constructions in modern Daco-Romanian, there is obligatory adjacency between the auxiliary and the past participle (cf. 26a,b), i.e., adverbs and floating quantifiers cannot be placed between the auxiliary and participle7 (Dobrovie-Sorin 1994, pp. 8–9; Cornilescu 2000, pp. 90–91; Ledgeway 2012, p. 145; Nicolae 2015). In modern Daco-Romanian, it was observed (Boioc Apintei 2023) that there is a preference for the auxiliary and the lexical verb to be preceded in pragmatically unmarked sentences by high adverbs, such as probabil ‘probably’ (lexicalizing Spec,Mood_epistemicP) (26c) or poate ’maybe’ (lexicalizing Spec,Mood_irrealisP) (26d) (Boioc Apintei 2023).

(26)	a.	Ion	a	plecat	probabil/	deja/		mereu.
		Ion	aux.pf.3sg	go.pple	probably	already		always
		‘Ion has probably/already/always left.’
		(SDRo, apud Nicolae 2015, p. 81)
	b.	copiii		au	(*toţi)	citit		toţi.
		children.def		aux.pf.3pl	all	read.pple		all
		‘all the students eat’
		(modern Daco-Romanian)
	c.	Ana	probabil	a	mers	la	serviciu	astăzi.
		Ana	probably	aux.pf.3sg	go.pple	to	work	today
		‘Ana probably went to work today.’
		(modern Daco-Romanian, apud Boioc Apintei 2023, p. 125)
	d.	poate		am	înțeles	greșit.
		maybe		aux.pf.3sg	understand.pple	wrong
		‘maybe I got it wrong’
		(modern Daco-Romanian, apud Boioc Apintei 2023, p. 125)

Therefore, in the case of present Daco-Romanian, a variation regarding the level of the verb movement can be observed, since for some speakers the verb tends to stay low(er), while, for others, it can raise to high(er) projections. This variation can be explained through the Balkan character of Daco-Romanian, given the fact that a similar situation (i.e., variation among speakers with respect to verb movement) can be found in Balkan varieties, in languages such as Greek, Bulgarian, and even Serbian and Croatian.

2.4. The View from Old Romanian8

2.4.1. Synthetic Verb Forms: Present

The landing site of the present indicative verb in the texts verified from old Romanian seems to be the low aspect-related position (Nicolae 2019; Boioc Apintei 2023). As far as HAS is concerned, lexical verbs are preceded in pragmatically unmarked sentences by high adverbs, such as acum ‘now’ (lexicalizing Spec,T_{(past/future)}P) (27), neapărat ‘necessarily’ (lexicalizing Spec,Mod_necessityP) (28), and iar ‘again’ (lexicalizing Spec,Asp_{repetitive(event)}P) (29).

(27)	acum	easte	al treilea	an	de când	eram	în	casa
	now	is	third	year	since	was	in	house.def
	părinţilor		miei
	parents.dat		my
	‘now is the third year since I was in my parent’s house’
	(VS.1700, p. 145)

(28)	Dumnezeu (...)	neapărat	varsă	milostea
	God	necessarily	spreads	mercy
	‘God necessarily spreads His mercy’
	(Ev.1642, p. 119)

(29)	și	într-alt	loc	iarăş	zice…
	and	in = other	place	again	says
	‘and in other place says…’
	(VRC.1645, p. 201, 10r)

With respect to LAS, old Romanian lexical verbs are preceded in pragmatically unmarked sentences by low adverbs, such as încă ‘still’ (lexicalizing Spec,Asp_continuativeP) ((30), (31)), pururea ‘always’ (lexicalizing Spec,Asp_perfectP) (32), aproape ‘almost’ (lexicalizing Spec,Asp_prospectiveP) (33), or bine ‘well’ (lexicalizing Spec,VoiceP) (34).

(30)	încă	dăruiesc	cutăruia	pentru	mântuirea	sufletului
	still	give.ind.pres.1sg	someone	for	rescue.def	soul.gen
	meu
	my
	‘I still give [it] to someoane for the rescue of my soul’
	(ACP.1714, p. 12r–12v)

(31)	și	încăș	hotărâm	pentru	această	mare	milă
	and	still	decide.ind.pres.1pl	for	this	big	mercy
	‘and we still decide for this big mercy’
	(AAM.1713, p. 8, 35v)

(32)	iară	eu	pururea	nedejdescu	pre	tine
	And	I	always	believe.ind.pres.1sg	in	you
	‘and I always believe in you’
	(PH.1500–10, p. 10, 60v)

(33)	vreamea	mea	aproape	iaste
	time.def	my	almost	is
	‘my time is almost here’
	(CC1.1567, p. 204v)

(34)	carii	bine	vestesc	sufletelor	bucuriia	cea negrăită
	who	well	announce	souls.dat	joy.def	the unspeakable
	‘the one who well announce unspeakable joy to the souls’
	(Ev.1642, p. 227)

2.4.2. Analytic Verb Forms: Compound Past

As for the [aux-v] constructions in old Romanian, it has been argued (Nicolae 2015) that the auxiliary can be found in the IP domain (mood-filed), in contrast to the participle, which has a lower position. A further argument for the low position of verb movement is represented by the possibility of interpolating various constituents between the auxiliary and past participle.

In old Romanian, the lexical verb (past participle) can follow high adverbs, such as acum ‘now’ (lexicalizing Spec,T_{(past/future)}P) (35), or lower ones, such as încă ‘still’ (lexicalizing Spec,Asp_{repetitive(event)}P) (36), pururea ‘always’ (lexicalizing Spec,Asp_perfectP) (37), doar ‘just’ (lexicalizing Spec,Asp_{retrospective}P) (38), or bine ‘well’ (lexicalizing Spec,Voice) (39).

(35)	Iară	acmu	ne-am	înderetnicit	de	leage
	and	now	cl.1pl = aux.pf.1pl	depart.pple	of	law
	‘and now we stopped obeying the law’
	(VRC.1645, p. 217: 22r)

(36)	în	darul lui Dumnezeu	sau	în	păcatele	lumiei,
	in	gift	God.gen	or	in	sins.def	people.gen
	încă	nu	s-au		ales
	still	neg	cl.refl.pass = aux.pf.3pl		choose.pple
	‘in God’s gifts or in the sins of the world they hadn’t been chosen yet’
	(VRC.1645, p. 222: 26r)

(37)	pururea	l-am	sfătuit	dreptu
	always	cl.3sg.acc = aux.pf.1sg	advise.pple	well
	‘I always advised him well.’
	(Bert.1774, p. 60v)

(38)	că	doară	au	făcut	el	spăseniia
	that	just	aux.pf.3pl	do.pple	he	redemption
	‘he has just completed his redemption’
	(VRC.1645, p. 215: 2v)

(39)	și	pre	soli	bine	i-au	priimit
	and	doom	messagers	well	cl.3pl.acc = aux.pf.3pl	receive.pple
	‘and the messagers were well received’
	(ULM.~1725, p. 91v–92r)

2.5. The View from Russian

2.5.1. Synthetic Verb Forms: Present

In Russian literature, there are two different interpretations regarding verb movement: on the one side, the rich agreement of Russian verbs suggests V-to-I movement, namely to aspect (Gribanova 2013, pp. 94–95; Roberts 2019), and on the other side, the fact that manner adverbs appear to the left of Russian verbs (40) (Slioussar 2007, p. 145; Dyakonova 2009, p. 28; among others) has led some scholars to conclude that the final landing site is v (cf. Bailyn 1995; Erechko 2002; Kallestinova 2007; among others).

(40)	a.	Ja	navernoe	smogu	tebe	pomoč.’
		I.nom	probably	can.fut.1sg	you.dat	help.inf
		‘I can probably help you.’
		(Russian, apud Dyakonova 2009, p. 8)
	b.	Ivan	často	ubiraet	komnatu.
		Ivan.nom	often	clean.ind.pres.3sg	room.acc
		‘Ivan often cleans his room.’
		(Russian, apud Gribanova 2013, p. 95)
	c.	Ja	uže	em		eto.
		I	already	eat.ind.pres.1sg		this
		‘I already eat it.’
		(Russian)
	d.	Zlodei	horošo	znali	povadki	životnyx.
		malefactors	well	know.past	habits	animals.gen
		‘The malefactors knew animals’ habits well.’
		(Russian, apud Slioussar 2007, p. 146)
	e.	Petja	bystro	sdelal	uroki.
		Petya.nom	quickly	do.pst.m	homework.acc
		‘Petya quickly did his homework.’
		(Russian, apud Dyakonova 2009, p. 5)

2.5.2. Analytic Verb Forms

Russian auxiliaries surface higher manner adverbs (cf. 41), while lexical verbs follow them (Slioussar 2007, p. 146). This was explained by the fact that auxiliaries occupy a T-related position, and the verb moves somewhere between vP and T (Gribanova 2013, p. 96; Harizanov and Gribanova 2019, pp. 471–472). Furthermore, there is no adjacency requirement, such that adverbs or scrambled constituents can intervene between the auxiliary and the lexical verb (42) (Dyakonova 2009, p. 7).

(41)	Kuda	ty	ne	budeš’	bol’še	ezdit’?
	where	you	neg	aux.fut.2sg	from.now.on	go.inf
	‘Where are you not going anymore?’

(42)	Zavtra	my	skoree	vsego	budem	celuj	den’
	tomorrow	we	sooner	all	aux.fut.1pl	whole	day
	zanimat’sja.
	study.inf.refl
	‘Tomorrow we will most probably study the whole day.’
	(Russian, apud Dyakonova 2009, pp. 7, 198)

2.6. The View from Moldovan Daco-Romanian

2.6.1. Synthetic Verb Forms: Present

In Moldovan Daco-Romanian, the landing site of the indicative lexical verb seems to be a (very) low ASPECT-related position. In pragmatically unmarked sentences, the lexical verbs are preceded by both high adverbs, probabil ‘probably’ (lexicalizing Spec,Mood_epistemicP) (43) or neapărat ’necessarily’ (lexicalizing Spec,Mod_necessarilyP) (44), and low adverbs, such as deja ‘already’ (lexicalizing Spec,T_anteriorP) (45), încă ‘still’ (lexicalizing Spec,Asp_continuativeP) (46), or mereu ‘always’ (lexicalizing Spec,Asp_perfectiveP) (47).

(43)	Probabil	fac	exerciții	la	mall.
	probably	do.ind.pres.3pl	drills	to	mall
	‘They probably do some drills at the mall.’
	(MDRo, apud Costea 2019, p. 36)

(44)	Neapărat	îl	sun	acum.
	necessarily	cl.acc.m.3sg	call.pres.ind.1sg	now
	‘I’m obligatorily calling him right now.’
	(MDRo, apud Costea 2019, p. 36)

(45)	Încerc		să-i	lămuresc	că
	try.ind.pres.1sg		să.subj = cl.dat.3g	explain.subj.pres.1sg	that
	deja	e	timpul	să	facem
	already	is	time.nom.def	să.subj	make.subj.pres.1pl
	un copilaș.
	a little.baby
	‘I’m trying to explain to him that it’s time to have a baby together.’
	(MDRo, apud Costea 2019, p. 37)

(46)	Afară	încă	e	zăpadă,	dar	feciorașii
	outside	still	is	snow	but	sons.nom.def
	m-au	bucurat	cu	primii
	cl.acc.1sg = aux.perf.3pl	be.happy.pple	with	first.adj.pl.def
	ghiocei.
	snowdrops
	‘There’s still snow outside, but my sons brought me snowdrops, and they made me happy.’
	(MDRo, apud Costea 2019, p. 37)

(47)	Moldovenii	noștri		mereu	caută
	moldovans.nom.def	ours		always	search.ind.pres.3pl
	să	mai	ia	bani	de	undeva.
	să.subj	more	get.subj.pres.3pl	money	from	somewhere
	‘Our Moldovans always want to get some extra money from different places.’
	(MDRo, apud Costea 2019, p. 37)

2.6.2. Analytic Verb Forms: Compound Past

In the case of the compound past, the [aux-v] cluster raises to T_anterior in Moldovan Daco-Romanian, but not further. Therefore the [aux-v] should be preceded by mereu ‘always’ (Asp_perfective) (48) in pragmatically unmarked sentences. Regarding the placement of deja ‘already’ (T_anterior) (49), both the preverbal and postverbal placement of it are accepted, but only the preverbal placement one is regarded by native speakers to be pragmatically unmarked (cf. Costea 2019, pp. 38–41).

(48)	Avem	niște	părinți	de aur,	care	mereu
	have.ind.pres.1pl	some	parents	of gold	who	always
	ne-au		ajutat.
	cl.acc.1pl = aux.perf.1pl		help.pple
	‘We have some wonderful parents, who have always helped us.’
	(MDRo, apud Costea 2019, p. 39)

(49)	Zi-i	că	tu	ai	făcut-o
	tell.imper = cl.dat.3sg	that	you	aux.perf.2sg	do.pple = cl.acc.f.3sg
	deja	cu	un	amic	de-al lui9
	already	with	one	friend	of = him
	‘Tell him you have already slept with one of his friends…’
	(MDRo, apud Costea 2019, pp. 38–39)

2.7. V-Movement in Lipovan Daco-Romanian10

2.7.1. Synthetic Verb Forms: Compound Past

In contrast to standard Daco-Romanian (where verb targets a high position within the sentence), the Lipovan Daco-Romanian verb seems to have a lower Asp(ect)-related position, between Asp_perfective and Voice. The verb is consistently preceded in pragmatically unmarked contexts by lower adverbs, such as deja ‘already’ (Spec,T_anteriorP) (50), încă ‘still’ (Spec,Asp_continuativeP) (51)11, or mereu ‘always’ (Spec,Asp_perfectP) (52), while strictly preceding bine ‘well’ (Spec,VoiceP).

(50)

a.

noi

deja

mâncăm

we

already

eat.ind.pres.1pl

‘We have already eaten.’

b.

noi

deja

avem

40

și

ceva

de

ani

we

already

have. ind.pres.1pl

40

and

something

prep

years

de

când

n-am

făcut

chestia

asta

prep

when

neg.aux = pf.1sg

do.pple

thing.def

this

‘We already are 40 years old since we did not do this thing…’

c.

au

zis

că

acolo

deja

umblă

aux.pf.3pl

say.pple

that

there

already

go

pe jos

cu

picioarele

descălțați

on.foot

with

legs

barefoot

‘they said that they already are walking barefoot’

d.

cu

miru

nu

dă,

că

el

deja

este

with

ointment.def

neg

give

because

he

already

is

dat

și

nu

botează

given

and

neg

baptize

‘he doesn’t apply ointment because it is already applied, and he is not baptized’

(LDRo)

(51)	Tot	părinții	ajută...	Ajutorul	încă	merge
	continuously	parents.art	help	help	still	goes
	înainte.
	forward
	‘It is the parents that continue to help… the help keeps coming.’
	(LDRo)

(52)	a.	Ei	mereu	își	schimbă	denumirea.
		they	always	cl.refl.3pl	change.ind.pres.3pl	name.def
		‘They are always changing their name.’
	b.	mereu	vorbește,	are	prieteni
		always	speak.ind.pres.3sg	has.ind.pres.3sg	friends
		care	sunt	veniți	din	Republica Moldova
		that	are ind.pres.3sg	come	from	Republic of Moldova
		‘she is always speaking, she has friends from Republic of Moldova’
	c.	mereu	învață	limba	rusă	ca	să
		always	learn.ind.pres.3sg	language.def	Russian	to	să.sub
		vorbească	cât mai bine
		speek	better
		‘she always learns Russian to speak as well as possible’
		(LDRo)

Unlike standard Daco-Romanian (53), where the verb shows a high position, in both old Romanian (54) and Lipovan Daco-Romanian (55), the verb undergoes movement to a lower position, i.e., below low adverbs such as deja ‘already’ or mereu ‘always.’ Indeed, as already shown above, a similar position of the verb can be found in Moldovan Daco-Romanian (56).

(53)	Ana	merge	des	la	cumpărături.
	Anna	go.ind.pres.3sg	often	prep	shopping
	‘Anna often goes shopping.’
	(SDRo)

(54)	ce	se	pururea	pomeneaște
	which	cl.refl.pass	always	mention.ind.pres.3sg
	‘which is always mentioned’
	(Prav.1581, p. 258r)

(55)	ei	mereu	își	schimbă	denumirea
	they	always	cl.refl.dat.3pl	change.ind.pres.3pl	name.def
	‘they are always changing their name’
	(LDRo)

(56)	Moldovenii	noștri	mereu	caută
	moldovans.nom.def	ours	always	search.ind.pres.3pl
	să	mai	ia	bani	de	undeva.
	să.subj	more	get.subj.pres.3pl	money	from	somewhere
	‘Our Moldovans always want to get some extra money from different places.’
	(MDRo, apud Costea 2019, p. 37)

Therefore, the situation seems to be similar to that found in Russian (57), where the verb is generally preceded by adverbs lexicalizing Spec,Mod_epistemicP (navernoe ‘probably’), Spec,T_anteriorP (uže ‘already’), Spec,VoiceP (xorošo ‘well’), and Spec,Asp_celerativeP (process/process) (bystro ‘fast’) (see Section 2.5) in pragmatically neutral contexts (Bailyn 1995; Harves 2002; Kallestinova 2007).

(57)	Saša	bystro	ljubit	Veru.
	Sasha	quickly	love ind.pres.3sg	Vera
	‘Sasha loves Vera quickly.’
	(Russian, apud Dyakonova 2009, p. 33)

2.7.2. Analytic Verbal Forms: Compound Past

Unlike native speakers of standard Daco-Romanian, Lipovan Daco-Romanian speakers tend to place adverbs such as deja ‘already’ (Spec,T_anteriorP) (58), încă ‘still’ (Spec,Asp_continuativeP) (59), or mereu ‘always’ (Spec,Asp_perfectiveP) (60) before the lexical verb in pragmatically unmarked contexts.

(58)	a.	noi	deja	ne-am		pregătit
		we	already	cl.acc.1pl = aux.pf		ready
		‘we are already ready’
	b.	deja	toți	episcopii	care	au	fost
		already	all	bishop.def	that	aux.pf.3pl	be.pple
		au	murit
		aux.pf.3pl	die.pple
		‘all the bishops that had been around died’
	c.	harbuz	deja	s-a		copt	la	noi,
		watermelon	already	cl.refl = aux.pf.3sg		ripe.pple	at	us
		așa de frumos	și	bun
		so.beautiful	and	tasty
		‘the watermelon is already ripe here, and it is so nice and tasty’
		(LDRo)

(59)	încă	numerele	nu	am	pus...
	still	numbers	neg	aux.pf.1sg	put.pple
	‘I haven’t put the numbers yet…’
	(LDRo)

(60)	a.	mereu	a	mers	cu	mine	la pescuit
		always	aux.pf.1sg	go.pple	with	me	fishing
		‘(he) always went fishing with me’
	b.	ea (...)	tot	încă	nu	a	văzut
		she	also	yet	neg	aux.pf.3sg	see.pple
		‘she is yet to see [it]’
	c.	când	mi-a	spus	că	vrea	să
		when	cl.1sg = aux.pf.3sg	say.pple	that	wants	să.subj
		mănânce,	eu	mereu	am	cumpărat	și
		eat	I	always	aux.pf.1sg	buy.pple	and
		am	adus
		aux.pf.1sg	bring.pple
		‘When she told me that she wants to eat, I always buy something and bring to her.’
		(LDRo)

Therefore, Lipovan Daco-Romanian undoubtedly displays similarities to old Romanian (61), Moldovan Daco-Romanian (62), and Russian (63), inasmuch as the lexical verb consistently surfaces to the right of low adverbs. The presence of the constructions currently found in Lipovan Daco-Romanian can be explained not only through Romanian–Russian contact, but also through the fact that a similar phenomenon existed in old Romanian. In other words, Lipovan Daco-Romanian has preserved a feature from old Romanian under the influence of Russian (cf. also Dindelegan and Dragomirescu 2016, pp. 636–37). On the other hand, this situation is not to be found in standard Daco-Romanian (64), where the verbal complex including the perfective auxiliary targets a high position, i.e., at the bottom of the mood field, between Mod_necessarily and Asp_habitual.

(61)	acela	curând	iase	den	întunearecul	morției
	that	quickly	emerges	from	dark	death.gen
	‘that quickly emerges from the darkness of death’
	(Ev.1642, p. 147)

(62)	Unde	deja	a	nins?
	where	already	aux.perf.3sg	snow.pple
	’Where has it already snowed?’
	(MDRo, apud Costea 2019, p. 39)

(63)	Ivan	budet	(haraso)	pet	(haraso)
	Ivan	aux.viit.1pl	well	sing	well
	’Ivan will sing well.’
	(Russian)

(64)	(De obicei)	am	încercat	(de obicei)	să
	usually	aux.perf.1sg	try.pple	usually	să.subj
	fim	împreună	de	Paște.
	be.subj.pres.1pl	together	of	Easter
	‘We usually tried to spend the Easter together.’
	(SDRo, apud Costea 2019, p. 31)

3. Conclusions

In this paper, I took the placement of lexical verbs with respect to the adverbs stated in Cinque’s hierarchy as an indicator for the level of V-movement in Lipovan Daco-Romanian. I present empirical evidence that the placement of the present and present perfect forms of the verb in Lipovan Daco-Romanian is similar to that found in old Romanian, Russian, and Moldovan Daco-Romanian. More precisely, the Lipovan Daco-Romanian verb seems to appear to the right of low adverbs, such as deja ‘already,’ încă ‘still’ or mereu ‘always,’ given the corpus studied (Appendix A).

The difference between Lipovan Daco-Romanian and standard Daco-Romanian (where the verb targets a high position) regarding the different level of displacement of the verb can be explained through the fact that low verb movement was also an option of old Romanian, and, under the Russian influence, this phenomenon strengthened an already available syntactic option. In other words, the Romanian–Russian contact helped to preserve and consolidate a phenomenon like the one that appears in old Romanian, which is different from the one found in standard Daco-Romanian.