Scarica Si-Causative Passives in Italian: A Comparative Study - Prof. Belletti e più Appunti in PDF di Morfologia e Sintassi solo su Docsity!
HO.
The production of Relative clauses and intervention locality with remarks
on causatives, si-causative passive and reflexive passive
Example of a way to elicit the production of SRs and ORs: Picture description 1 Questo è l’elefante che bagna il leone This is the elephant that wets the lion 2 a. Questo è l’elefante che il leone bagna This is the elephant that the lion wets b. Questo è l’elefante che è bagnato dal leone This is the elephant that is wet by the lion In these pictures there is an elephant and a lion. In one picture the elephant wets the lion in the other picture the lion wets the elephant. Which elephant is this? This is the elephant …. Recalling Rizzi and Friedmann comprehension of relative clauses, a question comes to mind: how do we elicit production? Production in this experiment is tested through a picture description task: with no training, adults produced very few ORs hardly but many passives are produced instead. POR: passive object relative → the aim was to elicit an OR; what was obtained was an SR in the passive (what adults overwhelmingly did) Production results form adults are surprising, because they comprehended no ORs (however, we know that their computation is harder than that of SRs, even for adults) we expected them to produce ORs. Production 1. SRs, ORs, PORs in adults’ productions (Italian) (18 adults, from Belletti &Contemori 2010; comparable results reported in Contemori & Belletti 2013 with 10 adults more. Preference task, adapted from Novogrosky & Friedmann 2006) 100% 10% 88%
(Types of) PORs in children’s productions Contemori & Belletti (2013) Percentages of Passive Object Relatives produced by children in Picture description Task; Interesting comparison to be made with adult’s types of PORs (from Preference task) Children: ⎯ At the beginning we have few PORs (not surprising 3 yo); slightly more at 5 yo, which is when passives start to be mastered ⎯ At around age 8, the kind of passive that children use is stricking: out of an overall 70% of passives, most of them are of the si - causative type, i.e. questo è il leone che si fa bagnare dall’elefante (which is interesting to notice that they are also the first ones to appear). Interestingly, si - causative passives are not only the most produced, but also the first ones to appear
- POR: il leone bagnato dall’elefante..
- Reduced relative clause: il leone che è/viene bagnato dall’elefante… (NB: reduction is not just the deletion of the complementizer and of the passive auxiliary, this is a purely descriptive fact) Reduced relative clauses are very rare, even when passive is there on the part of older children. We wanted to count reduced relatives because, in observing adults’ PORs, they were mainly reduced: questo è l’elefante bagnato dal leone. Why all these PORs in adults? Because with PORs, we have no intervention in passives (assuming the smuggling derivation of passives): that is the reason why adults tended to produce passives also also in the condition aiming at eliciting ClLD productions, while children instead tended to produce: ⎯ a - topics in the match condition (exp1) ⎯ propl^ in the number mismatch condition (exp2)
Types of PORs in adults’ productions 16% - few ORs 35% - POR with copular passive 47% - POR with reduced passive 0% - si - causative passives Results from 18 adults, from B&C 2010; comparable results reported in C&B 2013 with 10 adults more. Preference task, adapted from Novogrosky & Friedmann 2006 Preference task: elicitation in a picture-description task No support, no picture, nothing fancy: of course such a task works with adults but it was unexpected that it worked with children too. Children at age 3 listen to the experimenter and answer him/her telling them what they want to be. POR with Si-causative passive: Il leone che si fa lavare dall’elefante the lion that si-makes wash by the elephant (the lion that makes the elephant wash him) POR with Copular passive: Il leone che è/viene lavato dall’elefante the lion that is/comes washed by the elephant POR with Reduced Passive: Il leone lavato dall’elefante the lion washed by the elephant No si- causative passive ( si + causative construction) A lot of reduced PORs: ll leone che è / viene lavato dall’elefante What emerged was a lack of si - causative, in this case (you were asking wich child would you rather be) is not surprising.
Types of PORs in children and adults 17% 10% 1% 0% 35% 47% Children’s data from Picture description Adults’ data from Preference task A U shape of the results emerges comparing children and adults: ⎯ the intuitively “more complex” structure is the one that children access preferably and earlier ( si - causative passive in the POR). ⎯ the intuitively “simpler” structure (POR with reduced relative) is virtually absent in children’s productions in contrast with adults. Similarly, S. Silleresi 2015, MA Thesis, University of Siena: si-causative PORs are the first ones (4;4-6;3) to appear and the preferred ones (7;3-7;11) in (early) child L2 Italian. Children behave in an opposite way, compared to adults. There is a sense, which is si- causative passives such as:
_1. Il bambino che si fa pettinare dalla mamma.
- Il bambino che viene pettinato dalla mamma._ Sentence 1. is not simple, because it implies a reflexive pronoun, a causative verb and also a passive. The si- causative is intuitively less simple. This kind of complexity is irrelevant, its computation must be simpler because we first compute the causative construction and then the passive one. Sentence 2. is a shorter sentence, we only have a copula and a passive, but children don’t produce it therefore we conclude it to be very hard for them. The hardest one, counter-intuitively, must be the reduced relative (it is shorter but children do not produce it). So, not always what is longer and involves more steps is harder and not always what is shorter is easier. In this case, the process leading to reduction must not be trivial at all (perhaps because we are cutting something that is embedded).
Passives in Adult Italian Si-causative passive: Quantitative comparisons and input/frequency considerations From a first counting of types of Passives used by Italian adult speakers in informal conversations in the same files (from LIP-database):
- Copular passive: 443
- Venire passive: 296
- Si-fa causative passive: 22 The causative passive is NOT a frequent structure in standard Italian (as is intuitively clear to native speakers). Early access to causative passive by children is then NOT an effect of frequency of the structure in the input, indeed the structure is only used in certain situations. (Types of) PORs in children’s comprehension Contemori & Belletti ( 2013 ) Percentages of types of PORs correctly comprehended by children in three (older) age groups Unfortunately, these data are form older children: at this age, passives are mastered so that the preference for si - causatives is only residual. Derivation of Si- causative passive Fare is a light verb, a semi functional verb, which means that its EA is the initiator of the event. → This chunk movement is shared with the derivation of passives.
The causative interpretation is here given by the causative voice: the causative head Caus is what makes fare be interpreted as a causative vs fare una torta. Si- is the initiator in this structure: Maria fa mangiare la torta a Gianni In the active causative construction, the si position is occupied by maria , which is the initiator of the event. Kayne hypothesized that si is always the EA (including the reflexive si ). This comes from Manzini’s idea that the same word must always express the same thing. Therefore, the reflexive si must be the same thing as the impersonal si. In the si - causative derivation of passive, si picks up the accusative , it moves to the accusative head: il bambino moves to the subject position, fare moves to and si behaves as a subject clitic. This is something we assume because in passives, the accusative is blocked (passive morphology has no accusaive). Interesting Iis that si - causative passives and reflexivws pick u the same morphological element si. On the nature of si
- Reflexive: in constructions such as Maria si lava 2. Intrinsic reflexive: with verbs such as pentirsi, andarsene,… 3. Impersonal, which carries plural agreement on the adjective, has a kind of generic meaning: si è felici.
- Middle si : wich is a close passive in sentences like queste cose non si dicono
queste coseIA non si dicono < _IA > → “si passivante”: si is the EA
- Ergative si: (Burzio) it is one appearing with unaccusative verbs, and which is particularly visible in pairs like: (TR) Maria ha spento la luce (UNACC) La luce si è spenta → here it is a marker of unaccusativity Ergative si is also found with the verb “to be”, like mi sono (be is the ultimate unaccusative)
The adults’ and children’s behavior in production - overwhelming production of passive in object relatives/PORs for adults and development in the same direction for children – follows from this conclusion. Development In line with the proposals discussed, formulated in terms of fRM (X ….. Z ….. Y) according to which, given the four set theoretic relations: ◦ identity ◦ Inclusion ◦ intersection ◦ disjunction ⎯ Identity in relevant features between X and Z is excluded by both children and adults, whereas intersection is properly computed by both. ⎯ There appears to be development in the proper computation of the intermediate relation, inclusion (cfr. of NP feature) ⎯ Disjunction may be seen as ‘optimal’ as no relevant features involved, which amounts to lack of intervention. Lack of intervention, as in the case of passive, optimal. It is due to a causative voice that “fare” is interpreted as causative: “fare” is in this case semi-functional because it has a rigid EA (Ramchand). SI: in this analysis, it is the initiator Cfr: Maria initiator^ farà mangiare la torta a Gianni V+DP is preposed, which is visible in the active causative too: fare pettinare il bambino. Chunk movement: movement shared with passive derivations too Because there is no intervention (it is similar to the disjunction in featural relations), so the passive seems to be optimal.
Why do we expect adults to produce more passives than ClLDs? Structures like il libro (^) pro l’ho letto with topicalization and clitic resumption (ClLD) are widespread, especially in colloquial Italian, mainly in those with no intervention, i.e. with pro rather than with a lexical DP.and these structures do exactly the same thing as PORs, i.e. they are trying to avoid intervention of two co-occurring lexically- restricted NPs. Indeed, through smuggling, passives eliminate the intervention of the subject and it allows for an optimal satisfaction of locality in the relevant sense. Children and adults’ behaviour in production follows from this conclusion: we try to elicit structures with intervention of DPs, but what we end up with is a construction that avoids intervention. The comparison between adults and children demonstrated that revolves around the type of passive included. In the case of ClLD, when passive is produced, adults mainly produced both while the only type produced by children is the si-causative passive. This is interesting because it reflects the same difference between children and adults regarding the relative clauses. (plural → what happens to my friend the elephant: they wash him) (si-causative → le chien, is se fait laver par le chat)
Reflexive passive (si) Q: Che cosa succede al mio amico l’elefante? A: Si lava it washes itself (Olmo, 4;1 y.o.) Q: Che cosa succede ai miei amici il cane e l’orso? A: Il cane si lava e l’orso si sta asciugando (Leonardo, 4;2 y.o.) Up to 9% in 4 y.o. produced a reflexive passive when describing a transitive action, which is surprising because it seems wrong. Si-causative passive Q: Che cosa succede al mio amico il cane? A: Il cane si fa lavare (dal gatto) (Neri, 5 y.o.) Up to 11% in 5y.o. were answers with si causative passives (in fact, the only type of passive produced, as seen earlier) Virtually no reflexive passive anymore (2%) to describe a transitive action. Reflexive passive and then Si-causative passive Proposal: Reflexive passive is a first route to the passive in young children Belletti (2019) The hypothesis is that si causative is a kind of passive: the use of the reflexive is a way to get the passive in young children. Manetti decided then to manipulate the feature of animacy. As a matter of fact, the result shows that, when we manipulate the animacy, there is a complete absence of the si-causative passive. Moreover, there is an increase of the use of the reflexive passive. Going deeper: From reflexive passive to si-causative passive. Manipulating animacy of the topic Q: Che cosa succede a questa cosa, la torta? Complete absence of si - causative passives , which is in sharp contrast with the condition with animate topic discussed where the only clear passive was the si-causative one. Ample production of reflexive passive → up to 16% (1) Il libro si legge (Olivia,5;3) thebooksi-CLreads
(2) E lo taglia (exp:checosa?) Q: cosa? Il foglio si taglia (Olivia,5;3) And it.Cl cuts it.(exp:What?) The papers I.Cl cts The question was about a piece of paper that somebody was cutting: what happens to this piece of paper. And the answer was something like e lo taglia. In the first part the child was answering in the transitive (the piece of paper is the object, which means that the structure was understood), while in the second one children promoted the object to subject position with reflexive. The reflexive si is found also in older children when the animacy of the topic is manipulated. Reflexive passive and si-causative passive Animacy appears to be a relevant factor inducing the production of Si-causative passives, in contrast with reflexive passives (which really seem to be passives with a reflexive morphology - up to 16%) Comparing the derivation of reflexive passive and Si-causative passive Reflexive Si- causative Si is the EA of the lexical verb lavare , in the case of the si causative passive, we assume that the si is the EA, but it is the designator of the semi functional causative verb fare. So it gets the fix role of initiator and this means that we have all the arguments available. The causative has the property of moving the chunk so we have a smuggling operation: we have a pre posing verb and object into the internal argument.
The hypothesis could be that the young children production (both with animate and inanimate) the interpretation of si as a reflexive is not particularly triggered, it could be that due to this co-indexing with si is not at work in children’s grammar. This coindexation is not operative in children’s grammar. The reason for that is that maybe the fact that there is no agreement features so there is no intrinsic features. From reflexive passive to si-causative passive The assumed derivation of the Si-causative passive has the pronominal marker Si as the EA of the semi-functional causative verb fare. Its role is the Initiator role, given adopted assumptions. The EA of the lexical verb is thus not taken up by Si in this case. Whence the whole argument structure of the lexical verb is realized and the transitive action is correctly expressed by children through the Si-causative passive. On the animacy of the Initiator Hypothesis: It is natural to assume that the role of Initiator is most typically carried by an animate argument. This would account for why children do not resort to the Si-causative passive when animacy of the object topic is manipulated and the IA is inanimate. The inanimate derived subject could no be coindexed with the animate Si. New questions are opened by this hypothesis, in particular: Animacy may be a too coarse characterization of the relevant feature, as inanimates can be Initiators in causatives. Possibly animacy identifies the core case of a feature related to some of human or more generally biological/natural activity. This is visible in active causatives like:
- Il vento ha fatto scoperchiare i tetti delle case
- La vitamina B12 fa rinforzare le ossa 1’. I tetti delle case si sono fatti scoperchiare dal vento 2’. Le ossa si fanno rinforzare dalla vitamina B The claim is that the initiator must always be animate seems to strong, however the si - initiator must always be animate (otherwise we would have a clash of features). Inanimate subjects seem to be allowed for in the active The correspondent si - causative structures are inacceptable
On the animacy of the Initiator However, Si-causative passive would be completely impossible in these cases, suggesting that Si is always a necessarily animate Initiator. (Hence, the coindexing with the derived subject due to its reflexive status yields ungrammaticality). Children have yet to be tested n structures of this type, where both EA and IA are inanimate. However, when the IA/derived subject is inanimate, they opt for the reflexive passive, which does not implicate any (animate) initiator role ( si is a EA of the vP) nor coindexing with the derived subject, as proposed earlier (with reflexive si interpreted as a route to passive). The core idea is that reflexive passives are not interpreted as reflexives. Generic plural null subject instead of passive Generic plural null subject in active clause: Robust findings in children’s productions of ClLD, emerged with both animate and inanimate object topic: Interestingly, when manipulating animacy, children also produced a lot of active sentences with a generic plural null subject, rather than passives. Ways of being impersonal/generic and alternative routes to passive… We have proposed that in Reflexive passive: 1.The coindexation between Si and the derived subject may just not be operative in the children’s grammar (due to lack of overt agreement features on si). Si in turn may be the realization of a silent generic Agent/EA.
- Si does not count as an intervener for movement of DP/IA as it has a reduced internal structure, lacking e.g. a D layer (Holmberg and Roberts 2013). The reduced structure makes it dissimilar in the sense relevant for the computation of intervention for the moved full DP/IA
Zooming on si-causative passive and its robust appearance in children’s
productions
Outline of derivation of causatives: Fare-a ‘a’ resembles an applicative preposition/voice (cfr.Collins 2018,based on Kayne 2004; cfr.Later) ⎯ Fare is a semi functional verb selected by the causative voice ⎯ EA of the lexical verb mangiare is introduced by the preposition a (it is in dative) (connection with a - topics: affectedness) ⎯ Maria is the initiator ⎯ Causative is similar to the passive voice in the sense that it attracts a whole verbal chunk to its spec Outline of derivation of causatives: Fare - da Same derivation as fare da , but the EA the lexical verb mangiare is introduced by preposition by (different from passive because it is a kind of expletive preposition, the -role of the NP it introduces can be that of the experiencer; by doesn’t do anything: it doesn’t make the EA of the lexical verb become the agent, that is why it is considered as an expletive).
Tutti temono il terremoto → Il terremoto è temuto da tutti
*Le ultime fanno temere il terremoto da tutti However, in causative constructions the by makes the DP to be interpreted as an agent (the Caus head) in causatives, by is not an expletive preposition. Derivation of Si-causative passive Same derivation as passives and as causatives fare da Si is the animate EA And get - causative passive (EN) Get gets a causative interpretation, because it is selected by cause (similar to force) Same movement of a verbal movement. It is interesting that Caus is playing a role not only in si - causative passives, but also in get passives. Get-passive is the one accessed earlier by English speaking children which mirrors results from Italian in which the si- causative is the kind of passive accessed earlier. An extra role in Si-causative passive Through presence of the causative semi-functional v=fare verb, si-causative passive involves an extra role compared to copular and venire passive: e.g. Initiator (Ramchand 2008), taken up by SI. Through the co-indexing between SI and the derived subject, the subject of a si-causative passive is directly implicated, as the initiator, in the occurrence of the overall event.
