This paper aims at analyzing, both qualitatively and quantitively, the use of the subjunctive in subordinate clauses in social varieties of spoken Italian.
The loss of the subjunctive is a much-debated topic in Italian (socio)linguistics. Many authors (see Schena et al. 2002) have underlined that the use of the indicative instead of the subjunctive is widely attested in contemporary Italian (1). Nevertheless, it has also been observed that the subjunctive is still quite regular in some constructions (Berruto 2017: 42) and it is strikingly in expansion over the indicative in few syntactic contexts (2) (Renzi 2019).

(1) Capisco che ci sono (VS subjunctive: siano) dei periodi
understand.1SG that there are some periods
in cui non guadagno, dei periodi in cui
in which NEG earn some periods in which
va (VS subjunctive: vada) male
goes bad
“I see that there are periods in which I do not earn any money, periods in which it goes wrong”

(2) Il fatto che l’ avesse creato (VS indicative: aveva creato) con me e
the fact that (s)he had built with me and
con la mamma era un fatto davvero speciale
with the mum was a fact really special
“The fact that (s)he had built it with me and with his/her mum was a really special fact”
(ParlaTO corpus – PTA017)

The loss of the subjunctive has been traditionally related to extra-linguistic factors (Berruto 2012: 80-81 inter al.) but the topic has not been explored in detail yet due to the lack of suitable corpora.
In this paper, we will address this issue using the ParlaTO module of the KIParla corpus (Mauri et al. 2019), a recently built resource composed by almost 50 hours of semi-structured interviews collected in Turin and equipped with a large set of metadata, such as speaker’s age, L1 (e.g. Italian vs. Italo-Romance dialects), level of education and employment status.
We will discuss the use of the subjunctive in relation to both the linguistic conditioning of its variability (e.g. sentence type, lexical identity of the governor, embedded verb; see Poplack et al. 2018 and Digesto 2019) and some social factors (paying special attention to the degree of education and the age of the speaker). Firstly, we will observe the data from a qualitative point of view, also to test the possible semantic ambiguity between the indicative and the subjunctive in the scrutinized contexts. Then, we will perform a distributional analysis of the subjunctive vs. indicative morphology with the aim of identifying and ranking relevant factors that are involved in the loss/maintenance of the subjunctive. Finally, the distribution of the subjunctive in different social varieties will be discussed in relation to the well-known Labovian distinction between changes from above and from below (Labov 1994: 78), also to explain the internal and external reasons behind the ongoing change.