“Two is better than one, isn’t it?” Exploring the role of overt negation in the acquisition of Italian ‘nessuno’

In languages such as Italian, Catalan, Czech or Russian, two negative elements appearing in the same clause express a single semantic negation. This phenomenon is known as Negative Concord (NC) and it is illustrated in (1). By contrast, in other languages, the co-occurrence of two negative elements gives rise to a Double Negation (DN) reading (see, a.o., Zanuttini, 1991; Zeijlstra, 2004, building on Van der Wouden, 1994). (1) Laura non ha parlato con nessuno alla festa. Italian Laura NEG has talked with n-one at-the party NC: “Laura didn’t talk with no one” = “Laura didn’t talk with anyone” *DN: “Laura didn’t talk with no one” = “Laura talked with someone” NC readings can be accounted for assuming the establishment of an agreement relation between (multiple) elements syntactically marked as negative but crucially lacking semantic negation – called n-words, following Laka (1990) – and a single negative operator, which can remain silent or be overtly spelled out (Zeijlstra, 2004; Penka, 2007, 2011; Tubau, 2016; Deal, 2022). A DN reading is rather one in which each negative element in a sentence contributes semantic negation. In compliance with semantic compositionality, two co-occurring negative elements thus end up conveying an affirmative reading due to their two negations cancelling each other out. An example from Standard English in offered in (2). Such readings are commonly found in German and Dutch, as well. (2) Laura didn’t talk with no one at the party. Standard English Laura aux-NEG talk with n-one at the party. DN: “Laura didn’t talk with no one” = “Laura talked with someone” *NC: “Laura didn’t talk with no one” = “Laura didn’t talk with anyone” Languages that typically allow for Negative Concord readings – which henceforth will be referred to as NC languages – are further distinguished between strict and non-strict (following Giannakidou, 1998). The difference between these two classes is that, in strict NC, negative constituents must always co-occur with sentential negation, which is not the case in non-strict NC. In non-strict NC languages such as Italian, Spanish or Catalan, the sentential negative marker, when present, is considered to spell out the above-mentioned negative operator. In these languages, the co-occurrence of n-words with sentential negation depends on their position in relation to the verb. When preverbal, negative constituents stand on their own (3a). This means that they receive their negative meaning through Agree with the negative operator, which remains silent. When postverbal, negative elements rather co-occur with sentential negation (3b). Hence, in these structures, the Agree relation is established between the n-word and the negative operator spelled out as the sentential negative marker (Zeijlstra, 2004). It must be noticed that, in non-strict NC, DN readings are possible, though only in specific contexts such as fragment answers (3c) or when n-words sit in a high clausal position (3d). (3) a. Laura con nessuno (*non) ha parlato alla festa. Italian Laura with n-one (*NEG) has talked at-the party “Laura talked with no one at the party” b. Laura *(non) ha parlato con nessuno alla festa. Laura NEG has talked with n-one at-the party NC: “Laura didn’t talk with no one” = “Laura didn’t talk with anyone” c. Q: “(Laura) Con chi non ha parlato alla festa?” (Laura,) with whom NEG has talked at-the party “(Laura,) With whom didn’t she talk at the party?” A: “Ah, con nessuno!” Ah, with n-one “Oh, with no one!” NC: “[Laura didn’t talk] with no one” = “[Laura didn’t talk] with anyone” DN: “[Laura didn’t talk] with no one” = “[Laura talked] with someone” d. [Laura]TOP, [con nessuno]FOC non ha parlato alla festa! Laura, with n-one NEG has talked at-the party! DN: “Laura didn’t talk with no one” = “Laura talked with someone” All the Italian examples in (3a-d) illustrate the interaction between n-words and overt sentential negation by means of the negative indefinite nessuno (“no one”). Negative indefinites have been analysed following two main approaches. One is the quantifier approach (Barwise & Cooper 1981; Zanuttini 1991; de Swart 2000; de Swart & Sag 2002; Watanabe 2004; Bošković 2009; Zeijlstra 2011), which treats them as single semantic units, on a par with universal and existential quantifiers. The other is the decomposition approach (Bech 1955; Jacobs 1980; von Stechow 1993; Rullmann 1995; Penka 2007, 2011; Sauerland 2000), which rather considers them as being composed of different semantic units, namely, a morphological marker for negation and an indefinite. There are, to our knowledge, a few experimental works investigating the acquisition of the Italian negative indefinite nessuno which, however, focus on the mastery of its much rarer DN reading (Tagliani, 2019; Moscati, 2020; Tagliani et al., 2022). The experiment we propose rather targets the acquisition of its NC reading. Moreover, it intends to compare comprehension of structures with overt mention of sentential negation followed by postverbal nessuno with that of sentences in which the preverbal indefinite is the only element contributing negation. The interest in comparing comprehension of structures with pre- and postverbal nessuno stems from recent observations from child German. In German, children up to 6 years of age have been found to struggle with production of the negative indefinite kein (“no”) in subject but not in object position – an asymmetry not observed with the universal quantifier alle (“all”) (Bill et al., 2024). Following the decomposition approach to negative indefinites, the authors argue that this asymmetry is due to structures with kein subjects to be more complex than those with kein objects. For kein subjects to be interpreted as negative, the indefinite must indeed be reconstructed under a covert negative operator sitting at the edge of vP. Such reconstruction is not needed with kein objects, since in those cases, the indefinite already sits in the scope of the silent negative operator. Hence, children’s tendency to avoid producing kein subjects but not kein objects is claimed not to denote a general difficulty with negative indefinites. Rather, it could be an attempt to avoid the costly operation of reconstructing the subject under negation. The asymmetry found by Bill et al. (2024) in German raises interesting questions that can be tested in Italian, too. One such question is whether Italian children encounter similar difficulties in production – and comprehension – of subject negative indefinites. It must be acknowledged that the expectation that preverbal subject negative indefinites might be harder than object negative indefinites in Italian, as well, follows not only from the reconstruction approach, but also from what is known as the Transparency Principle (Slobin & Bever, 1982; Jackendoff, 1990; Weist et al., 1997; Van Hout, 1998), upon which the Meaning First Approach to language acquisition (Sauerland & Alexiadou 2020; Guasti, Alexiadou, & Sauerland 2023) is built. Assuming that negative indefinites crucially lack semantic negation, and knowing that it is easier for children to learn structures encoding concepts in a one-to-one way than in a many-to-one way (van Hout, 1998), one could therefore expect that, in Italian, sentences with object negative indefinites as (4a) are easier to understand and to produce than sentences with preverbal subject negative indefinites as (4b). (4) a. Maria non ha visto nessuno. Maria NEG has seen n-one NC: “Maria didn’t see no one” = “Maria didn’t see anyone” b. Nessuno ha visto Maria. N-one has seen Maria “No one saw Maria” (4a) can indeed be held to be more transparent in that the negative operator is overtly spelled out as the sentential negative marker non – following Zeijlstra’s (2004) analysis of Negative Concord. This is not the case in (4b), though, where the operator making nessuno interpretable as negative remains silent. In (4a), all elements taking part in the agreement relation are thus overt, whereas in (4b), this relation must be retrieved based on the evidence that the preverbal indefinite bears a morphological, but semantically void, marker for negation. Notice that the hypothesis that sentences with sentential negation and object negative indefinites be more transparent and thus easier to acquire than sentences with preverbal subject negative indefinites can be tested via a more minimal comparison than the one in (4a-b) in Italian. Being a pro-drop language, Italian indeed allows for postverbal subjects, which German does not. This means that the same contrast as (4a-b) can be obtained by juxtaposing pre- and postverbal subject negative indefinites. In so doing, the constituent – i.e., the subject – remains the same; what changes is only its position in relation to the verb (5a-b). (5) a. Non l’ha mangiata nessuno. NEG it-Cl has eaten n-one b. Nessuno l’ha mangiata. N-one it-Cl has eaten NC: “No one didn’t eat it” = “No one ate it” The study we present consists of a dynamic version of the Truth Value Judgement Task (TVJT) (Crain & McKee, 1985; Crain & Thornton, 1998) in the spirit of Thornton et al. (2016), in which participants are first presented with short videos and then asked to judge whether baby squirrel Camilla, who is watching the videos, too, can correctly recall what happened in the stories. In each video, participants see three characters interacting and entertaining the possibility to do different things. At the end, the characters do some but not all the things mentioned and, crucially, one thing remains undone. This was arranged in order to meet one of the fundamental requirements for TVJTs, namely, the Condition of Plausible Dissent (Crain et al., 1996). At the end of each video, the narrator ask