Strategies of Indefiniteness Marking in Central Sicilian—Evidence from the Dialect of Delia

Abstract

This paper is meant as a contribution to the research project on variation and optionality in the determiner system in Italo-Romance, with novel data from the Sicilian dialect of Delia. The study is based on fieldwork interviews and the construction of a small corpus of 850 observations by 24 native speakers of Deliano (mean age: 36.37; age range: 19–72). The participants were asked to (i) describe in Deliano a 3 min videotape of an Italian speaking woman during her shopping session at a supermarket and (ii) talk about their own shopping routines in Deliano. These activities were designed mainly to detect the following strategies to express indefiniteness: (i) ART, i.e., a definite article with an indefinite interpretation; (ii) ZERO, a zero determiner for bare nouns; (iii) DI+ART, the so-called “partitive article”; (iv) pseudo-partitives such as ‘a bit of’; (v) the grammaticalised cardinal ‘two’; and (vi) the grammaticalised cardinal ‘four’. The data confirm (i) the preference for ZERO in negative episodic sentences in the past; (ii) the general lack of bare DI and DI+ART, and of certo ‘certain’ used as an indefinite; (iii) the use of different specialised forms of pseudo-partitive ‘a bit of’ in older speakers of Deliano; (iv) the neutralisation of this pseudo-partitive specialisation and the consequent emergence of some true optionality in younger speakers.

1. Introduction

Indefiniteness is a multifaceted, not easy to describe, phenomenon that can interact with a number of other features, such as verbal aspect, tense, telicity, scope, and negation. As a working definition, and in line with Giusti (2021, p. 262), I follow Brasoveanu and Farkas’s (2016, pp. 238–39) suggestion that we consider its counterpart (i.e., definiteness, which regards nominals referring to an individual that is already mentioned in the discourse) to define an indefinite nominal as an element that either introduces a new individual in the discourse or may not refer to any individual at all.

Irrespective of the different theoretical proposals to define the dichotomy definiteness-indefiniteness (see Molinari, this volume, for a brief overview), the following semantic types of indefinites in Italo-Romance (and in Romance, more in general) can be identified:

• Uncontroversial (or weak) indefinites, (as in (1a, b); cf. Brasoveanu and Farkas 2016);
• Quantificational (or strong) indefinites (as in (2a, b); cf. McNally 2020);1
• Pseudo-partitives of the type ‘a bit of’ (as in (3); cf. Selkirk 1977);
• Marked indefinites introduced by the complex determiner ‘a certain’ (as in (4); cf. Schwarz 2011).

(1)	a.	Bevo		Ø	/il	/del	vino.
		drink.PRS.1SG			the	of-the	wine
		‘I drink wine.’
		[Italian]
	b.	Sei	fyse		d’	aqua.
		if-there	be.SUBJ.3SG		of	water
		‘If there was water.’
		[Piedmontese; Berruto (1974)]

(2)	a.	Alcuni	ragazzi.
		some	boys
		‘Some boys.’
	b.	Alcuni	dei	ragazzi.
		some	of-the	boys
		‘Some of the boys.’
		[Italian]

(3)	Vorrei	trovare	un	po’	di	amiche.
	want.COND.1SG	find.INF	a	bit	of	friends
	‘I would like to find some friends.’
	[Italian; Cardinaletti and Giusti (2017)]

(4)	C’	è	un	certo	timore	nelle	sue	parole.
	there	is	a	certain	fear	in-the	his	words
	‘There’s a certain fear in his words.’
	[Italian]

Although indefiniteness has been covered extensively in the literature on Romance languages, southern Italo-Romance varieties, and especially Sicilian, are still quite underrepresented. In this respect, the different strategies to mark indefiniteness are also proof of a particularly rich linguistic diversity that characterises Italo-Romance in general (cf., a.o., Maiden and Parry 1997, p. 1). In this regard, the research project started by Cardinaletti and Giusti (2018) on variation and optionality in the determiner system in Italo-Romance has shown that (i) Italo-Romance can resort to a number of different strategies to express indefiniteness, (ii) among these strategies, those that are encoded in the two positions available inside the Determiner Phrase (Abney 1987) distribute along two isoglosses.2 These are represented in (5) (adapted from Cardinaletti and Giusti 2018, p. 150).

(5)			North
			Ø+Ø
			Ø+ART
	North-West	DI+Ø	DI+ART	DI/Ø+Ø	North-East
			Ø+ART
			Ø+Ø
			South

The north-south axis regards the presence of the definite article, which is a Romance innovation with respect to Latin (cf. Ledgeway 2012, p. 89ff.) and, as such, it is expected to be more present in the area from which it radiates (i.e., the centre) and less present in the extreme or lateral areas. The Ø vs. ART on the righthand side of the couples indicates whether the nominal features in D are covert (Ø) or overt (ART).3 The west-east axis regards the presence of the indefinite determiner di ‘of’ in SpecDP (in bold) along the whole axis (with the exception of some areas in north-eastern Italy), which can interact with the presence of ART in D and thus surface as the ‘partitive article’ DI+ART (cf. Cardinaletti and Giusti 2015, 2016).4

Finally, in their work Cardinaletti and Giusti (2018) also take into account another determiner, certo ‘certain’,5 that in Italian and generally in Italo-Romance varieties is used to add a specific indefinite nuance (cf. (6a); see also Baldi and Savoia 2021, pp. 37–38), but in some southern varieties with mass and plural nouns it has been documented as a basic indefinite, as in (6b) and (6c) (see also Giammarco 1979, p. 141):

(6)	a.	Per	questo	compito	ho	bisogno	di	certi	strumenti.
		for	this	task	have.PRS.1SG	need	of	certain	tools
		‘I need some specific tools for this task.’
		[Italian]

	b.	Cə	šta			cirtə	pərzonə.
		there	stay.PRS.3SG			certain	people
		‘There are some people.’
		[Abruzzese; Rohlfs (1968, p. 119)]
	c.	S’		era	corcato		mmiezo	a	ccierto	fieno.
		REFL		be.IMPF.3SG	lie.PPT		on	to	certain		hay
		‘He was lying on some hay.’
		[Neapolitan; Rohlfs (1968, p. 119)]

The aim of the present study is to contribute to this line of research by providing new data from a dialect of central Sicily (a missing point in Cardinaletti and Giusti’s 2018, 2020 documentation) in a preliminary, qualitative way. Particularly, the paper aims to (i) account for the most recurring strategies to express indefiniteness, (ii) discuss alternative strategies outside of the DP, (iii) guide the interested researchers in the Sicilian speaking areas with regard to the phrasing of the stimuli, the possible strategies to elicit indefinite nominal expressions, and the tools to use language dominance as a relevant variable to assess the extent to which Italian influence younger speakers in their indefiniteness marking choices.

The rest of the paper is organised as follows: Section 2 discusses the syntactic peculiarities of the Sicilian dialects that must be taken into consideration when dealing with indefinites and provides an overview of the strategies to express indefiniteness as found in Cardinaletti and Giusti’s (2018) and Giusti’s (2021) work. Section 3 presents the tools used for the present study, with the description of the corpus (and the participants in the study) and the Bilingual Language Profile (Birdsong et al. 2012). Section 4 contains the discussion of the corpus data collected. Section 5 draws the conclusions, indicates the limitations of the study, and offers some avenues for further research with related research desiderata.

2. The Expression of the Indefiniteness in Deliano

In this section, the main strategies to express indefiniteness in Deliano are discussed and compared with their Italian counterparts, when differences emerge. But before proceeding to this comparison, two relevant features of the Sicilian dialects must be highlighted that make the utterances in Deliano (and, more in general, in Sicilian) look different from those used to elicit indefiniteness in Italian and in northern Italo-Romance varieties: the use of the Ethical Dative in specific contexts and the systematic exploitation of the left periphery by verbal arguments. These features also often influence production in the regional Italian spoken in Sicily.

2.1. Ethical Dative, Focus Fronting and Syntactic Extraposition

In Deliano, the counterpart of a sentence like the Italian in (7a), featuring the consumption verb mangiare ‘eat’, is still felt as more natural when it displays the Ethical Dative (ED in the glosses):6

(7)	a.	Ieri	sera	ho	mangiato	una	pizza.
		yesterday	evening	have.PRS.1SG	eaten	a	pizza
		[Italian]
	b.	Arsira	??(mi)	mangiavu	nna	pizza.7
		last-night	ED	eat.PST.1SG	a	pizza
		[Deliano]
		‘I had pizza last night.’

The same holds for verbs such as guardare ‘watch’:

(8)

a.

Ho

guardato

la

partita

con

gli

amici.

have.PRS.1SG

watched.PPT

the

match

with

the

friends

[Italian]

b.

??(Mi)

taljavu

la

partita

ccu

l’

amici

mia.

ED

watched.PST.1SG

the

match

with

the

friends

my

[Deliano]

‘I’ve watched the match with my friends.’

b.’

… picchì

intra,

sulu,

nun

sapìa

cchi

-ffari.

because

home

alone

NEG

know.IMPF.1SG

what

do.INF

‘… because I didn’t know what to do, home alone.’

[Deliano]

More generally, the ED is obligatory (or strongly preferred)8 when the predicate describes a transitive event with a subject assuming the role of AGENT and with no indirect complements as adjuncts (cf. Masini 2008).

Besides the preference for the ED depicted in (7b) and (8b), unlike what is shown in the Italian sentence in (8a), which displays an unmarked syntactic order, the Deliano rendition in (8b) is preferably acceptable with a special intonation that presupposes a continuation in the utterer’s speech (cf. (8b’)). The unmarked order for (8b) is shown in (9):

(9)	La	partita	mi	la	taliavu	ccu	l’	amici	mia.
	the	match	ED	RC	watch.PST.1SG	with	the	friends	my
	‘I’ve watched the match with my friends.’
	[Deliano]

The example in (9) is correlated with the structural properties of information structure, which appears in different configurations among the languages of the world. Contrary to what happens in English, where prosody (and especially intonation) is the main means to express the information structure of the sentences, Romance languages have specific structural positions in which topic and focus are dislocated. Among them, Sicilian resorts to word order to mark information structure. Cruschina (2006, 2012) shows that, on the one hand, Sicilian, like Catalan and Italian, exhibits obligatory designated positions for topic constituents belonging to the presupposition/background part of the clause. On the other hand, Sicilian displays a structural position in the left periphery of the sentence which is dedicated to (non-contrastive) focus constituents. This phenomenon, known in the literature as Syntactic Extraposition (cf. Cruschina 2006, 2012), forces all Topic ([-Focus]) constituents to appear in peripheral positions, and requires clitic resumption (the RC in the glosses of (9)) for all dislocated arguments, allowing for only Focus constituents to appear within the clause.

In addition to Syntactic Extraposition, Cruschina (2012, p. 24) claims that Focus Fronting can move the remaining new information [+Focus] elements to discourse-related positions when the pragmatic and discourse-related feature, labelled as ‘relevance’ (in the sense of Sperber and Wilson 1995), comes to play. In other words, besides Contrastive Focus, Informational Focus can also be fronted when used emphatically. Cruschina (2012) resorts to the cartographic concept of discourse-configurationality (i.e., the requirement that all sentence constituents move to discourse-related positions) to claim that, because of Syntactic Extraposition and Focus Fronting, Sicilian can be identified as a language where discourse-related properties are more relevant than grammatical functions in determining the word order (cf. Kiss 1995).

This means that the Italian sentences featuring an indefinite ZERO determiner in the scope of negation like (10a) is preferably rendered as in (10b).

(10)	a.	Non	mangio	carne.
		NEG	eat.PRS.1SG	meat
		[Italian]
	b.	Carni		nun	(mi)	(n)ni	mangiu.
		meat		NEG	ED	RC	eat.PRS.1SG
		[Deliano]
		‘I don’t eat any meat.’

2.2. Different Types of Indefinite Determiners in Deliano

In this section I will discuss the different strategies of indefiniteness marking in Deliano, by taking the detailed analysis in Cardinaletti and Giusti (2018, 2020) and Giusti (2021) on Italian and some Italo-Romance varieties as a reference, and by underlining the main points of divergence with Italian. In doing so, I will refer to my interviews with the 56 participants in the Corpus of the dialect of Delia (see Section 3), whose data are to be kept separate from those of the corpus discussed in Section 3 and Section 4.

2.2.1. Indefinite ZERO and ART

Let us start by considering ZERO, i.e., the zero determiner that underlies bare nominals. In bare nominals of all Romance languages an asymmetry between the subject and the object position can be detected (Giusti 2021, p. 265; see also Delfitto and Schroten 1991; Stark 2008a, 2008b; Carlier and Lamiroy 2018), according to which bare nominals in subject position are not allowed in Romance (see Dobrovie-Sorin 2021), including Sicilian.

(11)

a.

*Vinu

jè

’ncapu

lu

tavulinu.

wine

is

on

the

table

‘Wine is on the table.’

b.

*Stannu

spuntannu

margariti

nni

lu

jardinu

mia.9

stay.PRS.3PL

emerge.GER

daisies

in

the

garden

my

‘Daisies are blooming in my garden.’

c.

*Studenti

occuparu

la

scola.

students

occupy.PST.3PL

the

school

‘Students have occupied the school.’

[Deliano]

Generally, the examples with bare subjects in (11) become grammatical with the insertion of the definite article (or ART in Cardinaletti and Giusti’s 2018, 2020 terminology) used as an indefinite determiner, that is when not referring to kind or to any element previously mentioned in the discourse (i.e., lu vinu, li margariti, li studenti). However, with plural count nouns, an alternative strategy regards the use of the pseudo-partitive ‘a bit of’, which, on the contrary, would sound a bit odd in Italian (cf. 12c’).

(12)

a

Lu

vinu

jè

’ncapu

lu

tavulinu.

the

wine

is

on

the

table.

‘Wine is on the table.’

b.

Stannu

spuntannu

li/

nna

puicu

di

Stay.PRS.3PL

emerge.GER

the

a

bit

of

margariti

nni

lu

jardinu

mia.

daisies

in

the

garden

my

‘Daisies are blooming in my garden.’

c.

Li/

Nna

puicu

di

studenti

occuparu

la

scola.

the

a

bit

of

students

occupy.PST.3PL

the

school

[Deliano]

c’.

Gli/

?Un

po’

di

studenti

hanno

occupato

la

scuola.

the

a

bit

of

students

have

occupied

the

school

‘Students have occupied the school.’

[Italian]

The reason for the oddity of the pseudo-partitive un po’ di ‘a bit of’ in (12c’) can be explained by the fact that the partitive article (or DI+ART in Cardinaletti and Giusti’s 2018, 2020 terminology)10 is preferred in Italian in this case, which is impossible in Deliano, according to my informants:11

(13)	a.	Degli	studenti	hanno	occupato	la	scuola.
		of+the	students	have	occupied	the	school
		[Italian]
	b.	*Di	li	studenti	occuparu	la	scola
		of	the	students	occupy.PST.3PL	the	school
		[Deliano]
		‘Some students have occupied the school.’

We have already seen examples in which ART is in subject position in Deliano (cf. (12)). In object position, ART appears to be possible in all Romance languages except Portuguese and Romanian (Giusti 2021, p. 266). Deliano is no exception:

(14)	a.	Ppi	oi,	mi	vivu		(l’)	acqua	’mbottigliata.
		for	today	ED	drink.PRS.1SG		the	water	bottled
		‘I’ll drink bottled water today.’
	b.	Ppaccamora,		cugliemmu		(li)	margariti
		For-now		pick.PRS.1PL		the	daisies
		‘We’re picking daises for now.’
		[Deliano]

If we consider episodic sentences in the past, the scenario looks quite complex. Deliano and Italian behave differently with respect to the sentence polarity. In (15), in the presence of a positive episodic sentence, the nominal expression in object position must occur as ART.12 On the contrary, in a negative episodic sentence, ZERO is obligatory (cf. (16)).13

(15)	a.	Cci	misi	#(l’)	acqua	nni	lu	bbicchjiri.
		RC	put.PST.1SG	the	water	in	the	glass
		‘I poured water in my glass.’
	b.	Cci	misi	#(li)	nuci	nni	lu	saccu.
		RC	put.PST.1SG	the	walnuts	in	the	glass
		‘I put walnuts in my bag.’

(16)	a.	(*L’)	acqua	nni	lu	bbicchjiri
		the	water	in	the	glass
		nun	ci	nni	misi.
		NEG	RC	RC	put.PST.1SG
		‘I didn’t pour any water in my glass.’
	b.	(*Li)	nuci	nni	lu	sacchettu
		the	walnuts	in	the	bag
		nun	ci	nni	misi.
		NEG	RC	RC	put.PST.1SG
		‘I didn’t put any walnuts in my bag.’
		[Deliano]

Irrespective of the polarity, Italian can resort to ZERO and DI+ART with episodic sentences in the past:14

(17)

a.

(Non)

ho

versato

acqua

nel

bicchiere.

NEG

have.PRS.1SG

poured

water

in+the

glass

b.

(Non)

ho

versato

dell’

acqua

nel

bicchiere.

NEG

have.PRS.1SG

poured

of+the

water

in+the

glass

‘I poured/didn’t pour any water in my glass.’

(18)

a.

(Non)

ho

messo

noci

nel

sacco.

NEG

have.PRS.1SG

put

walnuts

in+the

bag

b.

(Non)

ho

messo

delle

noci

nel

sacco.

NEG

have.PRS.1SG

Put

of+the

walnuts

in+the

bag

‘I put/didn’t put any walnuts in my bag.’

[Italian]

The same scenario is found in episodic ‘there be’ sentences in the present, where Deliano displays the division of labour between ART (in positive sentences; cf. (19)) and ZERO (in negative sentences; cf. (20)), whereas Italian allows for either ZERO or DI+ART, irrespective of the sentence polarity. Note that turning the Deliano examples in (19) and (20) into the past (e.g., cc’era l’acqua ‘there was water’ and cc’èranu li zanzari ‘there were mosquitoes’) does not alter the already said division.15

(19)	a.	Nni	stu	bbiduni	cc’	è	#(l’)	acqua.
		in	this	jerrycan	there	is	the	water
		‘In this jerrycan, there is water.’
	b.	Nni	sta	stanza	cci	su	#(li)	zanzari.
		in	this	room	there	are	the	mosquitos
		‘In this room, there are mosquitoes.’

(20)	a.	Nni	stu	bbiduni	acqua	nun	ci	nn’	è.
		in	this	jerrycan	water	NEG	there	RC	is
		‘In this jerrycan, there is no water.’
	b.	Nni	sta	stanza	zanzari	nun	ci	nn’	è.
		in	this	room	mosquitoes	NEG	there	RC	is
		‘In this room, there are no mosquitoes.’
		[Deliano]

(21)	a.	In	questa	stanza	(non)	ci	sono	zanzare.
		in	this	room	NEG	there	are	mosquitoes
	b.	In	questa	stanza	(non)	ci	sono	delle	zanzare.
		in	this	room	NEG	there	are	of+the	mosquitoes
		‘In this room, there are (no) mosquitoes.’
		[Italian]

Besides ZERO and DI+ART, Italian can also resort to ART in cases like those shown in (21).

2.2.2. Pseudo-Partitives and Grammaticalised Low Cardinals

Another relevant distinction in the recent literature regards noun classes. Besides the particular behaviour of abstract nouns that sets them apart from both mass nouns and plural count nouns, Giusti (2021, p. 279) introduces another distinction that is internal to mass nouns and has to do with the possibility to resort to other types of indefinites, namely the grammaticalised low cardinals ‘two’ and ‘four’.16 Singular mass nouns such as pasta behave differently from plural mass nouns, i.e., plural nouns that can be thought of as mass nouns, such as spaghetti and spinaci ‘spinach’. Only the latter can be marked by means of grammaticalised low cardinals. But there is a further distinction to make within plural mass nouns. On the one hand, nouns such as spinaci ‘spinach’ are always semantically mass, so we cannot extract a singular such as *Uno spinacio ‘a piece of spinach’.17 On the other, nouns such as spaghetti can be interpreted either as semantic mass or as regular plural count nouns (so that Uno spaghetto meaning ‘a string of spaghetti’ is possible). In this case, high cardinal numerals, which are always quantificational, force the plural count reading and, thus, can be used as diagnostics to detect genuine plural count nouns. Giusti (2021, p. 280) claims that this fact is relevant in those Italo-Romance varieties that do not display DI+ART, since they can resort to grammaticalised ‘two’ and ‘four’ as an alternative strategy to ART. Deliano, lacking DI+ART, fits this prediction, but, contrary to what happens in Italian, the grammaticalised quattru ‘four’ is preferred over du (or dui)18 ‘two’ (cf. Fare due chiacchiere lit. ‘Make two small-talks’ vs. Scangiari quattru paroli lit. ‘Exchange four words’, both meaning ‘Have a small talk’).

However, grammaticalised quattru in Deliano takes on more functions than its Italian counterpart quattro, which is in competition with other strategies of indefiniteness (i.e., DI+ART, strong indefinites such as qualche (singular) and alcuni/e (plural) ‘some’, and pseudo-partitive un po’ di ‘a bit of’). In Deliano, pseudo-partitives tanticchja di ‘a bit of’ (also contracted in tantì di, and with the alternative form tecchja di) and nna puicu di ‘a bit of’ are used, respectively, with mass (cf. (22a)) and plural count nouns (cf. (22b)):

(22)	a.	Ppi	stasira	(nn’)	accattammu
		for	tonight	ED	buy.PRS.1PL
		tanticchja	di	murtatella.
		a-bit	of	mortadella
		‘We can buy some mortadella for tonight’s dinner.’
	b.	Ppi	stasira	(nn’)	ordinammu
		for	tonight	ED	order.PRS.1PL
		nna	puicu	di	panina.
		a	bit	of	sandwiches
		‘We can have some sandwiches for tonight’s dinner.’
		[Deliano]

In this respect, plural mass nouns such as spinaci behave like singular mass noun such as pasta and select tanticchja di (cf. (23a)). With indefinite plural count nouns that can be interpreted as mass, grammaticalised ‘four’ must be used (cf. (23b)).19 In both cases, the use of the definite article would be interpreted as referring to either kind or something already introduced in the discourse.

(23)	a.	Ppi	stasira	nni	mangiammu
		for	tonight	ED	eat.PRS.1PL
		tanticchja	di	spinaci/pasta.
		a-bit	of	spinach/pasta
		‘We can have some spinach/pasta for tonight’s dinner.’
	b.	Ppi	stasira	nni	mangiammu
		for	tonight	ED	eat.PRS.1PL
		quattru	italeddra/	maccarruna.
		four	ditali	macaroni
		‘We can have some ditali/macaroni pasta for tonight’s dinner.’
		[Deliano]

Generally, quattru is preferred to nna puicu di as the actual indefinite for small quantities with plural count nouns in positive sentences, cardinal ‘two’ being limited to very small quantities, and nna puicu di (which usually has a positive nuance) being preferred for greater indefinite quantities. Thus, Deliano pseudo-partitive nna puicu di does not fully correspond to its Italian counterpart un po’ di. Consider the examples in (24). The question in (24a) can be uttered by a speaker with a bag containing a relatively small quantity of oranges, so that the hearer knows that he or she cannot expect to receive a lot of them. The question in (24b), on the other hand, is likely to be uttered by a speaker whose bag is full of oranges, so that the hearer can actually expect to receive a lot of them. Using grammaticalised ‘two’, as in (24c), implies that the speaker can only give the hearer a very small quantity of oranges (say, two or three). On the contrary, for most Italian speakers un po’ di is neutral with respect to the number of oranges the speaker is willing to/can give the hearer (cf. (25a)). Using the complex indefinite un bel po’ di ‘a big deal of’ as in (25b) is a clear indication for the hearer that he or she can expect to receive a great quantity of oranges. Finally, and again differently from Deliano, grammaticalised ‘two’ in Italian can be used as a real indefinite, thus not necessarily implying a small quantity, although some variation with this respect may be expected.

(24)	a.	L’	â-bbuliri	quattru			aranci?
		RC	have-to+want.INF	four			oranges
		‘Would you like some oranges?’ (implying: I can give you only few of them)
	b.	L’	â-bbuliri	nna	puicu	d’	aranci?
		RC	have-to+want.INF	a	bit	of	oranges
		‘Would you like some oranges?’ (implying: I’ve got plenty of them)
	c.	L’	â-bbuliri	du			aranci?
		RC	have-to+want.INF	two			oranges
		‘Would you like some oranges?’ (implying: I have very few oranges)
		[Deliano]

(25)	a.	Le	vuoi	un	po’		di		arance?
		RC	want.PRS.2SG	a	bit		of		oranges
		‘Would you like some oranges?’ (no implication on the quantity of oranges)
	b.	Le	vuoi	un	bel	po’		di	arance?
		RC	want.PRS.2SG	a	nice	bit		of	oranges
		‘Would you like some oranges?’ (implying: I’ve got plenty of them)
	c.	Le	vuoi		due				arance?
		RC	want.PRS.2SG		two				oranges
		‘Would you like some oranges?’ (possible no implication on the quantity of oranges)20
		[Italian]

Note also that, when a person has already asked for an indefinite quantity of, say, oranges (to stick to the current examples) by using nna puicu di and they want more, nantri du/dui ‘another two’, nantri quattru ‘another four’ and nantra puicu ‘another bit’ can be used to modify their previous request, again depending on the quantity desired (as shown in (26)). But if they have used quattru in the previous request, the new one containing nantra puicu (as in 26c) would sound semantically odd. This confirms the following hierarchy for plural count nouns, from least to greatest quantity: du/dui > quattru > nna puicu di.

(26)	a.	Dammilli	nantri	du/dui.
		give.IMP.2SG+to-me+them	another.PL	two
		‘Give me just a couple more.’
	b.	Dammilli	nantri	quattru.
		give.IMP.2SG+to-me+them	another.PL	four
		‘Give me a few more of them.’
	c.	Dammilli	nantra	puicu.
		give.IMP.2SG+to-me+them	another.F.SG	bit
		‘Give me more of them.’
		[Deliano]

2.2.3. Bare DI

Finally, bare DI (following Cardinaletti and Giusti’s 2018, 2020 terminology) is not allowed in Deliano. This is in line with Giusti’s (2021, p. 294) data on different Italo-Romance varieties, according to which bare DI is restricted to some areas of northern Italy (such as Piedmontese; cf. Berruto 1974; Cerruti and Regis 2020), where it can covary with DI+ART, as in Piacentino. In Italian, bare DI can indeed be found in some restricted contexts, where it can be either obligatory (i.e., with right dislocated bare nouns) or optional (i.e., with left dislocated bare nouns). These cases, however, regard quantified nouns. Consider the examples in (27), adapted from Giusti (2021, fn. 13):

(27)	a.	Ne	ho	viste	molte,	*(di)	ragazze!
		RC	have.PRS.1SG	seen	many	of	girls
		‘I saw many girls!’
	b.	(Di)	ragazze,	ne	ho	viste	molte.
		of	girls	RC	have.PRS.1SG	seen	many
		‘Girls, I saw many.’
		[Italian]

The ‘unpopularity’ of bare DI in Deliano is such that even in the specific contexts in which it is obligatory in Italian, it is neither used in Deliano nor in the local regional Italian:21

(28)	a.	Nn’	aju	vistu	assà	(*di)	carusi!22
		RC	have.PRS.1SG	seen	many	of	girls
		[Deliano]
	b.	Ne	ho	viste	assai	(*di)	ragazze!
		RC	have.PRS.1SG	seen	many	of	girls
		[Regional Italian]
		‘I saw many girls!’

Note that a relic of bare DI can be detected in the expression aviri di bbisuignu ‘need’ (lit. have of need), which is being progressively replaced by the more Italianised, bare DI-less aviri bbisuignu, especially in younger speakers.

Now that all the different strategies of indefiniteness marking in Deliano have been presented, we can turn to discussing the data that make up the small corpus.

3. The Corpus

The data collected for this study are of a different nature. On the one hand, a small corpus of 850 observations was extracted from one of the text genres that make up the Corpus of the dialect of Delia (or CorDel; Di Caro in prep.), whose data were collected between 2020 and 2021. On the other hand, continuous fieldwork in Delia in the period 2020–2022 was conducted to (i) control for the presence of the strategies to express indefiniteness discussed in the literature that did not emerge from the corpus, (ii) verify that the indefinite marking strategies found in the corpus were shared by the community of speakers, and (iii) assess the extent to which some of those strategies could be the result of the pressure exerted by Italian on Deliano.

3.1. How the Corpus Data Were Collected and Organised

The participants were asked to accomplish two tasks:

• Describe in Deliano a 3 min videotape excerpt depicting an Italian-speaking young woman during a shopping session at a supermarket (I will refer to this task as ‘elicited production’);
• Talk about their own shopping routines in Deliano, first in general and then during a specific past event, i.e., the last time they went to do the shopping (I will refer to this task as ‘semi-spontaneous production’).

In the video, the young woman is addressing her audience, showing in detail what she is buying. Considering the length of the excerpt, the participants were provided with a list of questions such as “What does the young woman buy in the fruit and vegetable aisle?” or “Does she buy any meat?”. It was pointed out to the participants that they were not obliged to answer those questions singularly or follow that particular order, since the list was meant to just support them in the recollection of the content of the video. They were also encouraged to add personal comments if they felt like doing so. During the semi-spontaneous production session, participants were free to follow a random order in describing their shopping routines. They were only suggested to be as detailed as possible and not to focus only on the main products such as water, meat, fruits, and vegetables.

All the oral productions of the 24 participants were transcribed. The corpus for the present study only took into account the sentences containing any occurrence of the following strategies to express indefiniteness. These were tagged as follows:

• ZERO for a zero indefinite determiner;
• ART for a definite article with an indefinite meaning;
• PART for the combination DI+ART (i.e., the ‘partitive article’);
• DU for the grammaticalised cardinal ‘two’ to express indefiniteness;
• QUATTRU for the grammaticalised cardinal ‘four’ to express indefiniteness;
• QUARCHI meaning ‘some’;
• NNA PARA meaning ‘a pair (of)’. Instances of un paru di ‘a pair of’ are also coded as NNA PARA;23
• NNA PUICU meaning ‘a bit (of)’. Instances of the Italianised un po’ di ‘a bit of’ associated with plural count nouns are also coded as NNA PUICU;
• TANTICCHJA meaning ‘a bit (of)’. Instances of tantì di, tecchja di ‘a bit of’, tanticchjeddra di ‘a little bit of’, and the Italianised un po’ di ‘a bit of’ associated with mass nouns are also coded as TANTICCHJA;
• NUMBER for any cardinal number, including non-grammaticalised ‘two’ and ‘four’, to express a specific number for plural count nouns;
• QUANTIF for any other quantification such as ‘200 gr. of’ and for pseudo-partitives other than ‘a bit of’, such as ‘a bag of’ and ‘a box of’. For the purpose of the present study, the occurrences tagged as NUMBER and QUANTIF were not taken into account. The rest of the occurrences amounts to 850 observations.

For each observation containing any form of indefiniteness marking, two relating lexical items were indicated: the nominal expression (e.g., ‘bread’, ‘bell peppers’, ‘oranges’) and the predicate (‘to pick’, ‘to buy’, ‘to prefer’). Moreover, the following variables were considered:

• The class of the nominal expression (CLASS) with three levels (singular, plural, and mass);
• The syntactic position of the nominal expression (SYNTAX) with four levels (subject, object, dislocated subject, dislocated object);
• The polarity (POLAR) of the sentence (either positive or negative);
• The type (TYPE) of the sentence (either episodic or generic);
• The tense of the predicate (TENSE) with five levels (present indicative, imperfect indicative, preterite indicative, present perfect indicative, and imperfect subjunctive);
• LIST, i.e., whether the nominal expression occurs in a list (e.g., ‘I bought apples, oranges, and pears) or as isolated (e.g., ‘I didn’t pick any apples.’, ‘I love oranges.’);
• VIDEO, i.e., whether the observation belongs to the elicited or the semi-spontaneous production.

3.2. An Overview of the Participants

A group of 24 participants (12 male and 12 female) was selected with ages ranging from 19 to 72 years. Each participant was given a pseudonym that is shown in this paper in the relevant sentences reported, together with their gender, age, and index of language dominance (see

Table 1. An overview of the participants with their sociolinguistic data.

Participant	Pseudonym	Gender	Age	Index of Language Dominance
1	Ilenia	F	27	+49
2	Calogero	M	26	+38
3	Marianna	F	32	+75
4	Fiorella	F	21	+71
5	Filomena	F	52	+27
6	Rosario	M	23	+22
7	Maria	F	72	+54
8	Stella	F	60	+23
9	Nicoletta	F	39	+74
10	Gaspare	M	54	+95
11	Palmina	F	62	−10
12	Raffaele	M	33	−28
13	Linda	F	19	+82
14	Gioacchino	M	65	+79
15	Teresa	F	22	+52
16	Pasquale	M	21	+60
17	Geraldina	F	21	+74
18	Davide	M	37	+96
19	Vincenzo	M	26	+103
20	Alberto	M	25	+11
21	Federico	M	28	+64
22	Nicola	M	66	+20
23	Francesco	M	24	−80
24	Paola	F	25	+12

).

3.3. The Bilingual Language Profile and the Index of Language Dominance

The Bilingual Language Profile (henceforth, BLP) was developed by Birdsong et al. (2012) as a tool to draw up a linguistic profile of those people who can speak two languages to different degrees (for different reasons and in different contexts).24 The document consists of a free access self-assessment questionnaire that can be completed either on paper (with or without assistance by the researcher) or online and generally takes 10–15 min to complete.25 It can be compiled in either of the two languages whose dominance it aims to assess. Four separate modules, focusing respectively on the participant’s language history, use, proficiency, and attitudes, contribute to calculating the index of language dominance, but they can also be used as separate research tools. An additional biographic section which can be expanded by the researchers according to the specific needs of their research (see Gertken et al. 2014) is generally included. However, it does not contribute to the final index of language dominance.

Language dominance is to be intended as a relative and continuous construct (see Grosjean 2001; Gertken et al. 2014) that goes from variety B to variety A. The index of language dominance as calculated by the BLP is the result of the subtraction of the indexes as individually calculated for the two varieties and ranges from −218 (total dominance of B) to +218 (total dominance of A), with 0 indicating a perfectly balanced bilingual. It is important to clarify that dominance does not correspond to proficiency (Gertken et al. 2014), in the sense that one can be more dominant in A than B for whatever reason but without being proficient in A (or in either language). To make this statement straightforward, Birdsong (2015) claims that measuring language dominance in a given subject who can speak two languages is very similar to measuring how dominant they are in the use of their right or left hand.

Crucially for the present study, the linguistic scenario characterising most Italo-Romance speaking areas in Italy can be described as ‘bilectal’ (in the sense of Rowe and Grohmann 2013) or ‘dilalic’ (in the sense of Berruto 1987, 1989). This means that the two languages in the speakers’ repertoire are not sociolinguistically equal but, contrary to what happens in cases of diglossia, the high variety (or A, in this case Italian) can be used (and, actually, is more and more used) in informal contexts too, which are the only contexts where the low variety (or B, in this case Deliano) is spoken. Thus, in a dilalic scenario, the high variety (i) can be one of the speakers’ mother tongues, and (ii) is the only variety used in official written communication, the other variety hardly ever enjoying a widespread commonly accepted written code. This has led to two consequences. First, the BLP could only be administered in Italian.26 Second, some of the items of specific modules of the BLP, which was originally conceived of to assess language dominance in, e.g., people of the United States speaking English and Spanish or people of Canada speaking English and French, are not equally suitable for a dilalic situation, thus forcing the final score of the index towards Italian. This latter fact will be taken into account in the analysis of the corpus.

4. Data and Discussion

The data collected point to many different directions, in line with the multi-faceted nature of indefiniteness marking. In this paper I will mainly focus on those findings that could inform future research.

4.1. A Brief Note on the Listing Effect

It is well-known that, cross-linguistically, coordination of two or more nominal expressions can license ZERO indefiniteness marking, even in subject position (cf. Heycock and Zamparelli 2003). Compare (29a) with (11):

(29)	a.	Vinu	e	acqua	su	’ncapu	lu	tavulinu.
		wine	and	water	are	on	the	table
		[Deliano]
	b.	Vino	e	acqua	sono	a	tavola.
		wine	and	water	are	at	table
		[Italian]
		‘Wine and water are on the table.’

This tendency is confirmed by the data from the corpus. The occurrence reported in (30) shows a case of coordinated ZERO. If the occurrence had only contained either ‘apples’ or ‘pears’, ZERO would have been very unlikely to occur in both Deliano and the local regional Italian.27

(30)	Però	a	-mmìa	mele	e	pere
	But	to	me	apples	and	pears
	nun	mi	pjàcinu.
	NEG	RC	appeal.PRS.3PL
	‘But I don’t like apples and pears.’
	[Teresa, F, 22, +52]

Finally, the preference for ZERO in coordinations and lists is further confirmed by Rohlfs (1968, p. 118), who reports that in Italian ZERO is normally used in lists, while DI+ART, although possible, would add a specific reading (see Rohlfs 1968, p. 118, fn. 3).

4.2. ZERO vs. ART in Negative Sentences

The corpus contains 27 occurrences of negative sentences divided into 13 episodic and 14 generic sentences. Of the 13 negative episodic sentences (that can be either in the past or in the historical present), one displays a mismatch between the determiner used (i.e., the singular feminine ART la) and its resumptive clitic (i.e., quantitative nni, which is only compatible with ZERO, instead of accusative la). For this reason, this occurrence (shown in (31)) is taken out of the lot. Of the remaining 12 occurrences (10 of which display a left dislocated object) 10 display ZERO and only 2 display ART.

(31)	… e	-bbitti	ca	la	carni
	and	see.PST.3SG	that	the	meat
	mancu	nn’	avìa	a	-ppigliari.
	NEG	RC	have.IMPF.3SG	to	take.INF
	‘… and she realised that it was not a good idea to buy meat.’
	[Nicola, M, 66, +20]

For the negative generic side, the following preliminary observations must be made:

• Two occurrences of ZERO are coordinated in the same utterance (the case is presented above in (30)) and are thus excluded from the count;
• One occurrence of ART (i.e., lu pani patati e nuci ‘the bread potatoes and walnuts’) is not counted because the mass noun ‘bread’ is specified by ‘potatoes and walnuts’;
• One occurrence of ZERO is out because the bare noun is modified by a restrictive relative clause (i.e., frutta ca nun è di staggione ‘fruits that are not seasonal’);
• One occurrence of ART (i.e., Li bbruìcculi ’mmeci nun nn’accattu ‘I don’t buy broccoli instead.’) is out because it displays the same mismatch discussed in (31);
• Two occurrences of ZERO are out because the bare nouns are followed by an adjective (i.e., respectively, verdura confezzjonata ‘prepackaged vegetables’ and cosi sicchi ‘dried food’ lit. ‘dried things’).

We are then left with only eight available observations: one ZERO and seven ART.

The data from the corpus on negative sentences in Deliano are in line with the results of Cardinaletti and Giusti’s (2020) questionnaire on the expression of indefiniteness in colloquial Italian, according to which ART is the preferred strategy in generic negative sentences in the present. On the contrary, in episodic negative sentences ZERO is much more frequently used than ART.

The data also confirm that, in a variety like Deliano in which DI+ART is not used, in the negative domain the only strategies allowed in object position are ZERO and ART, as shown in

Table 2. Number of the occurrences for each level of INDEF by the sentence polarity.

	Polarity of the Sentence
Strategy of Indefiniteness	Negative	Positive
ZERO	16	246
ART	11	429
DU	-	17
QUATTRU	-	10
QUARCHI	-	26
NNA PUICU	-	33
TANTICCHJA	-	52
PART	-	6

(where also the six occurrences of DI+ART discussed in Section 4.3 are included).28

4.3. Some Considerations on the Partitive Article

As anticipated in Section 1, there are only six occurrences of indefinite DI+ART in the corpus, by two participants. These are summarised in

Table 3. Occurrences of DI+ART in the corpus.

Participant		Observation
Filomena [F, 52, +27]	a.	Sta carusa […] vidi di li mandarina. ‘This girl sees some tangerines.’
	b.	[…] s’accatta […] di li peperuna. ‘She buys (herself) some bell peppers.’
Nicola [M, 66, +20]	c.	[…] picchì piglià di la frutta. ‘because she picked some fruits.’
	d.	[…] ppi -gghjiri a -ppigliari di la pasta. ‘to go and fetch some pasta.’
	e.	[…] cci sunnu propriu di… dei reparti di l’affettati. ‘there are specific aisles for deli meats.’
	f.	Accattu molte delle volte anche di li… degli… di l’hamburger. ‘I often buy some hamburgers.’

.

Besides the low number of both the occurrences of DI+ART (only 6 observations) and the participants that produced them (only 2 out of a sample of 24 participants), some other remarks are noteworthy. First, except for (e) in Table 2, all the occurrences of DI+ART are in object position. This primacy of the object position for the partitive article is in line with Rohlfs’s (1968, p. 116) claim that the partitive article was originally used in object position and that its presence in subject position can be considered as a ‘further mechanisation’. Moreover, the two participants involved are both over 50 and quite well balanced as bilectals (with an index of language dominance of, respectively, +27 and +20). This could suggest an interference effect caused by Italian. This is especially clear in (e) and (f) of Table 2 where the utterances are mixed with Italian. On the one hand the Italian partitive form dei (i.e., di ‘of’ + i ‘the’) emerges; on the other, the Italian partitive form degli (i.e., di ‘of’ + gli ‘the’) is preceded by its Deliano counterpart di li, which is replicated and chosen as the final determiner form (i.e., di l’hamburger). Crucially, in both cases it is possible to detect some hesitation from the speaker, graphically rendered with the ellipsis points (‘…’), which could hint at some interference between the two languages of his repertoire.29

4.4. The Case of Nna Puicu di ‘a Bit of’ vs. Tanticchja di ‘a Bit of’. A Possible Shift to Optionality

In Section 1 we saw how grammaticalised quattru ‘four’ and the pseudo-partitive nna puicu di ‘a bit of’ in combination with plural count nouns are not used as synonyms, since the former implies a lower quantity with respect to the latter (which, in turn, is not synonymous with Italian un po’ di ‘a bit of’). We have also seen that Deliano displays separate indefinite markers for mass (i.e., tanticchja di) and for plural count nouns (i.e., quattru, nna puicu di).

However, two facts from the data collected during my fieldwork that regard the younger speakers point to a possible neutralisation of the fine-grained distinctions discussed so far:

• Nna puicu di used as synonymous to grammaticalised quattru, i.e., not implying a greater quantity with respect to the latter;
• An increasing use of nna puicu di as an indistinct indefinite marker that can thus be used also with mass nouns (as in nna puicu di latti ‘a bit of milk’ or nna puicu di zzùccaru ‘a bit of sugar’).

Since nna puicu di is the exact lexical counterpart of the Italian un po’ di (the contracted form of un poco di), which is an indefinite marker that is used for both mass and plural count nouns and does not imply a greater quantity with respect to the grammaticalised quattro ‘four’, it is no surprise that younger bilectal speakers are transferring the properties of Italian un po’ di to the Deliano nna puicu di. In this sense, it is reasonable to state that some true optionality in the expression of indefiniteness is emerging in Deliano. Note that, as a consequence of the neutralisation of the differences between nna puicu di and tanticchja di (and all its variants presented in Section 3.1), it is also possible to find tanticchja di with plural count nouns. This is the case of the occurrences shown in (32) from the corpus:

(32)

a.

… e

ddicidi

d’

accattari

and

decide.PRS.3SG

to

buy.INF

tanticchjeddra

di

mandarina.

a-little-bit

of

tangerines

‘… and she decides to buy a little bit of tangerines.’

[Vincenzo, M, 26, +103]

b.

Poi

si

piglià

tanticchja

di

cucuzzi.

then

ED

take.PST.3SG

a-bit

of

courgettes

‘Then, she bought (herself) some courgettes.

[Raffaele, M, 33, −28]

c.

o

tanticchjeddra

di…

cosi

tipu

pumadoru.

or

a-little-bit

of

things

like

tomatoes

‘… or a little bit of… stuff like tomatoes.’

[Nicoletta, F, 39, +74]

Interestingly, of the six participants that produced no occurrences of TANTICCHJA, only one is over 32 and four of them are under 27. The latter have also produced no occurrences coded as NNA PUICU. Let us now consider some other noteworthy cases that may concern the interference of Italian in the production in Deliano:

• The participant pseudonymised as Pasquale [M, 21, +60] recorded seven occurrences of NNA PUICU, of which six are associated with plural count nouns, as expected, and one with a plural mass noun (i.e., Nna puicu di spezie ‘A bit of spices’). He also produced two occurrences of NNA PUICU with singular mass noun but surfacing as the Italian un po’ di ‘a bit of’ (i.e., Un bel po’ di carne ‘a lot of meat’, and Un po’ di latti ‘a bit of milk’);
• Raffaele [M, 33, −28] recorded only one occurrence of tanticchja di but associated with the plural count cucuzzi ‘courgettes’;
• Vincenzo [M, 26, +103] recorded six occurrences of TANTICCHJA, of which five, all surfacing as tecchja di ‘a bit of’, are associated with mass nouns as expected, while that surfacing as tanticchjeddra di ‘a little bit of’ is associated with plural count mandarina ‘tangerines’. This participant also recorded seven occurrences coded as QUARCHI. Like its Italian counterpart qualche, Deliano quarchi ‘some’ is always associated with singular count nouns. In one of those occurrences, however, quarchi is associated with plural count passiluna ‘pickled olives’, probably overgeneralising its use in a variety that, unlike Italian, lacks plural alcuni/e ‘some’ and DI+ART;
• Federico [M, 28, +64] recorded 12 occurrences coded as NNA PUICU. In five cases nna puicu di is associated with singular mass nouns (i.e., virdura ‘vegetables’, lattuca ‘lettuce’ twice, salami ‘salami’, and pisci ‘fish’).30 Like the aforementioned case of Vincenzo, this participant also produced one case of tanticchjeddra di ‘a little bit of’, but here associated with the plural count (Italianised) olive ‘olives’, interestingly followed by an instance of nna puicu di with ulivi ‘olives’.31

The difficulty to produce tanticchja di consistently associated with mass nouns—even in cases of slight dominance of Italian as in Nicoletta [+74] (cf. (32b)), Pasquale [+60], Vincenzo [+103], and Federico [+64]—may indicate that this indefiniteness strategy is being replaced by nna puicu di in its neutralised version (i.e., not implying a great quantity) or directly by the Italian un po’ di,32 with a transitional period in which nna puicu di and tanticchja di can be used as synonymous and thus be associated indifferently with mass and plural count nouns, giving rise to true optionality.

4.5. Certo ‘Certain’

My preliminary fieldwork had already confirmed that Deliano does not fall under the group of southern Italo-Romance varieties in which certo ‘certain’ can be used as a generic indefinite devoid of any semantic specialisation (cf. Rohlfs 1968; Giammarco 1979; Ledgeway 2009). The data from the small corpus do not provide any evidence of this generic use of certo either. However, the fact that no occurrences of certo, even in its specific reading, have been detected in the corpus suggests that this might not be a good way to elicit that kind of data in general.

5. Concluding Remarks

Given the preliminary nature of the present contribution, it is expected that the open questions that arise are more numerous than the results found. In this section, I will first summarise the main findings, I will then discuss the limitations of this study and, finally, I will propose some avenues for future research.

5.1. What the Study Has Shown

The main results of this contribution can be summarised as follows:

• The results confirm Cardinaletti and Giusti’s (2018) distribution of ART and ZERO in Italo-Romance varieties along a vertical axis, an isogloss that radiates from the centre to the extreme North and South (including Sicily and Sardinia);
• The results of Cardinaletti and Giusti’s (2020) quantitative study on informal Italian indicating that ZERO is preferred in negative generic sentences in the present, whereas ART is preferred in positive episodic sentences in the past hold true for Deliano as well;
• Certo ‘certain’ is never used as indefinite determiner in Deliano;
• The use of the pseudo-partitive nna puicu di ‘a bit of’ does not fully correspond to that of the Italian counterpart un po’ di. To express generic indefiniteness with subject plural count nouns, Deliano prefers nna puicu di (in competition with ART), whereas Italian has DI+ART. Moreover, while Italian un po’ di is used to express small quantity indefiniteness, Deliano nna puicu di expresses a greater indefinite quantity and prefers grammaticalised low cardinals du/dui ‘two’ and quattru ‘four’ for small quantities;
• Bilectal speakers under 30 are progressively using nna puicu di also with (singular and plural) mass nouns and as synonymous to quattru, thus transferring to it the properties of the Italian counterpart un po’ di and leading to some true optionality and, probably in the future, to the exclusive use of nna puicu di or its possible replacement by Italian un po’ di.

5.2. Limitations of the Research

The preliminary nature of the present contribution has already been underlined. Two main limitations can be detected: (i) the small dimensions of the sample, (ii) the heterogeneous nature of the data collected.

In regard to the sample, although it is well balanced (with 12 male and 12 female participants) and describes a large section of the population (with the youngest participant being 19 years old and the oldest 72 years old), it only contains a subset of the sample used for the CorDel (56 participants), since each participant only covered part of the many genres included in the corpus.

In regard to the characteristics of the data collected, both the elicited and the semi-spontaneous productions can vary a lot from participant to participant. Since they were not timed and were free to add personal comments, the amount of oral production was inevitably influenced by the participant’s nature as more or less talkative and prone to provide detailed descriptions. In any case, it was not possible to get enough material to control for all the features that interact with indefiniteness discussed in Cardinaletti and Giusti’s (2018) work, namely Scope (wide vs. narrow) and Aspect (telic vs. atelic). Although I could check them with my informants (see Section 1), a dedicated quantitative study clearly falls under the desiderata.

5.3. Open Questions for Future Research and Desiderata

In this paper I have highlighted the main aspects that make the expression of indefiniteness in Deliano quite a complex matter that interacts with the bilectal nature of its speakers’ community, i.e., the fact that, as most Italo-Romance varieties throughout Italy, Deliano is losing ground as the dominant variety (in the sense of Birdsong 2015) with respect to Italian. As a consequence, a hybrid system is emerging that has a Deliano base (cf., e.g., the lack of bare DI and DI+ART) influenced by Italian (cf., e.g., the vanishing specialisation of tanticchja di for mass nouns and nna puicu di for count plural nouns).

Considering the scarcity of data available from Sicily as regards the expression of indefiniteness in the research line started by Cardinaletti and Giusti (2018) (see Section 1), this study has aimed to provide a preliminary set of data that could inform future research by means of dedicated studies. Among the related research desiderata, I will list the following:

• A quantitative questionnaire on the indefinite determiners in Sicilian, based on acceptability judgements, in which (i) the stimuli are worded taking into account the syntactic peculiarities discussed in Section 2.2 (i.e., Ethical Dative, Focus Fronting and Syntactic Extraposition); (ii) ZERO and ART are tested together with grammaticalised low cardinals and pseudo-partitives (replacing DI, DI+ART and certo);
• A study to assess whether bilectal speakers of Italo-Romance varieties that do not have DI and DI+ART among their indefiniteness marking strategies (also outside Sicily) make use of DI (in the specific syntactic contexts exemplified in (27b)) and DI+ART when speaking their regional Italian in informal contexts and Standard Italian in formal contexts in similar ways as the speakers from dialectal areas that make use of those strategies do. This could be done by means of some corpora analyses;
• A study to assess the extent to which younger bilectal speakers have neutralised the opposition tanticchja di ‘a bit of’ vs. nna puicu di ‘a bit of’ (as discussed in Section 4.4), in Delia and the surrounding area;
• A study to assess the presence of Italian un po’ di and of Italianised un pocu di ‘a bit of’ in the Deliano production of younger bilectal speakers;
• The elicited production task used in the present study to get data from Deliano could be reproposed to get data from the Italian spoken in the same community. This could favour a comparison between the two indefinite marking systems and help detect the points of convergence.

Besides all the possible avenues for future research proposed in this section, the data collected in the small corpus at issue need further analysis (and are indeed currently being analysed in a quantitative way) and can hopefully help to unveil other trends in the development of the strategies of indefiniteness marking in Deliano, as a representative of central Sicilian, in a bilectal contexts increasingly affected by the pressure of Italian.