The project includes a very large amount of very different language data from Italo-Romance varieties in contact. This poses two main problems. The first one is that these data need to be recorded, handled, classified, and accurately described, but the language competence to try and decipher them is not always there.
The second problem is to identify the communities where these Italo-Romance languages are still spoken. When the project started, we had very little information about these languages in the Americas. While a lot is done on the US Italian communities (and that is one of the reasons why we included American English, especially from New York and Boston), Argentina and Brazil, and to a certain extent Canada, are almost completely untouched.
Furthermore, the fact that data need to be collected outside the EU poses an additional problem with respect to data protection protocol. So far, we managed to work on schedule, despite some inevitable delays, due to the issues just mentioned.
We projected and designed a data crowdsourcing atlas, with the help of the Digital Humanities Lab of Utrecht University. The atlas required a lot of reworking, to be made as user friendly as possible while at the same time conveying as much information as possible. The atlas is now online, has a considerable traffic and contains about 300 spontaneous speech recordings. It can be found here:
https://microcontact.hum.uu.nl/(öffnet in neuem Fenster). We also developed an app that can be downloaded from Google Play Store (
https://play.google.com/store/apps/details?id=uu.microcontact(öffnet in neuem Fenster)). So far, it counts 131 unique downloads and installations.
The atlas and the app were quite successful in capturing the attention of the Italo-Romance speakers in Italy, but less so in helping us identify the speakers’ communities in Argentina, Brazil and Canada. To attract their attention, we contacted local associations, Italian Institutes of culture, and colleagues at various universities, establishing a very large network on which we relied for our fieldwork.
Fieldwork was performed in Argentina, Brazil and Canada, yielding 51 full interviews in Brazil, 36 in Canada, and 106 in Argentina.
In the case of Argentina, where we could not find many contacts, one of our members went on a pre-fieldwork.
Fieldwork in Brazil proved also particularly difficult: the researcher often experienced problems with plugging in the pc, and reported a very strong feeling of unsafety when traveling to remote areas with relatively costly recorders and computers. Nevertheless, fieldwork was successful, also in terms of empirical findings.
We managed to achieve two important sets of information, regarding the sociolinguistics of the speaker communities in the Americas and the empirical findings.
As for the sociolinguistics aspects: we have a much more informed overall view of the status of the Italian communities in the countries under investigation. This information, unavailable before our fieldwork, can be useful not only for future fieldworkers in linguistics but also for historians and sociologists, and anyone working on migration.
From a theoretical viewpoint, we made some interesting discoveries, as the data collected show significant deviation from what is commonly assumed about heritage languages. The most important difference we found regards the expression of the grammatical subject in heritage northern varieties, in particular in Argentina. Previous studies have shown that L2 (2nd language) learners tend to pronounce the subjects in sentences that would require a null subject, independent of the contact languages.
We expected to find this in our data, but what we found is instead a spreading of the null subjects, especially for 1st person subjects (singular and plural). Not only do heritage speakers of Friulian, a northern variety, omit the full subject when they speak Friulian: they even drop it when this is not allowed by the baseline (original Friulian) grammar.
It seems that this null subject strategy is not linked to any grammatical constraint, but to discourse constraints. Null subjects appear most frequently in topic structures, i.e. in structures where “what we want to talk about” is specifically addressed (for instance, if I say “with this key, you won’t be able to open that door”, “with this key” is a topic, it’s what we are talking about). If there is a topic in the clause, the subject is omitted. This is also quite surprising, as heritage speakers are not supposed to master the correlation between topichood and null-subjecthood. In fact, this should be exactly the context in which they insert an overt subject, given that computing both grammatical subjects and topics requires a bigger cognitive effort. They don’t.
The third discovery is that change affects 1st person elements, which should be more resilient than 3rd person ones.
Last, we have an overextension of differential object marking (DOM), contra what is documented in other heritage varieties.
We are now writing up these findings in a number of joint articles.
We are also trying to find a general explanation for these unexpected data. Most likely, this difference is due to the fact that so far the contact languages considered in heritage studies were typologically rather different, whereas here very similar languages are involved. The micro vs macro level of variation could also be involved, although it is hard to see why at a macrolevel we should see phenomena that are exactly the opposite of the microlevel.
