Analysis of an Oral Corpus of Early Russian-Italian bilinguals

INTRODUCTION

This work arises primarily from a personal interest; since my passion for languages has become more mindful and especially curious, I started being fascinated by the world of bilinguals. I have many friends and acquaintances who are bilingual, because of their family situation or country of origin, and their linguistic behavior has always aroused in me various questions. Although, to me, as I have had no choices rather than to study languages for many years and with great effort, the world of bilingual children has always caused envy in me.

What was for me, until recently, an unclear but fascinating sphere is actually a vast field, which is studied all over the world and in diffent disciplines, even though from many points of view it is still a contradictory topic, which does not make it less fascinating.

Starting from these interests, as well as my passion for the Russian language, I chose to study and analyze, within the boundless field of bilingualism, a corpus of early Russian-Italian bilinguals.

In particular, my interest was inspired by the studies about early bilingualism and second language acquisition carried out by the Canadian neurolinguist Michel Paradis (1994Paradis ( , 2004Paradis ( , 2009, by the researches on bilingual brain and language neuropedagogy of the Italian neuroscientist Franco Fabbro (1996Fabbro ( , 2004 and by the in-depths analyses concerning the application of neurolinguistic knowledge to language teaching carried out by Professors Laura Salmon and Manuela Mariani (2008).

In the sphere of early bilingualism, the most fruitful stimula for my research came, undoubtedly, from the theories of Fabbro, according to which there would be a critical threshold for the perfect acquisition of more than one language: to sum up Fabbro's thinking, -but we will investigate these hypotheses in depth further -if a child got actively in contact with more than one language within a certain age in his childhood, not only would he be able to acquire a perfect fluency, but, assumedly, it would also be a kind of guarantee against a possible erosion and loss, in the future, of the competence in these languages (Fabbro, 2004, 85-86).

Early bilingualism and child linguistic behavior are topics at the center of attention also of Russian scholars. Russia, like the rest of Europe, has been undergoing phenomena of emigration and immigration, but also and especially of inner migration. Also the XXI century Russian children, in fact, as well as their European peers, increasingly come from mixed couples and often move to live abroad. Moreover, the Russian Federation, especially the two capitals, is a destination for foreigners migration flows, especially from the former Republics of the USSR; first in the Soviet Union and in today's Republics created from its split, bilingualism has always been a tangible reality for those minorities (or often even majorities) who had been russified in the past and who, even nowadays, still maintain, in a different way, the use of Russian language, either by political constraints, or by their express wish, for economic advantages (Burykin, 2011).

For my work I have found crucial also the studies about the language of Russian speakers living abroad, in particular by the Russian linguist Elena Andreevna Zemskaja (1998, 2001a, 2001bZemskaja et al., 2001 and the North American Mary Polinsky (1995Polinsky ( , 1996Polinsky ( , 2006Polinsky ( , 2007Polinsky ( , 2010.

Zemskaja analyzed the linguistic behavior of Russian-speaking immigrants in five countries (Italy, Germany, France, the United States and Finland), dividing them into four waves of migration, starting from the Russian Revolution up to the present day, which is, indeed, the fourth wave of migration, the so-called "economic" wave, the one that left the Russian Federation after the collapse of the Soviet Union (Zemskaja, 2005). Her studies aim at highlighting the differences between the language of the different waves and especially what aspects of the Russian language are more easily subjected to erosion due to the influence and to '"meddling" by another language (Zemskaja, 2001c).

The "encounter" between the language of the so-called "Russian diaspora" and American English is definitely the kind of linguistic phenomenon which has been studied the most, given the large number of native Russian immigrants living in the United States.~ This sociolinguistic phenomenon has been studied at various levels; the analyses concern both the Russian language used by Russian speaking children who moved to the United States at an early age (Polinsky, 1996Gulida, 2004) -this kind of analyses, at least for the characteristics of the studied samples, are similar to one proposed here -and, and above all, the language of adults, but also of youth, which has developed from the contact between Russian and English, the so-called "runglish". The runglish is a mixed-language code based on Russian and enriched by loans and calques from English, used by the Russian-speaking immigrants in the United States, who settled mainly in the State of New York. This phenomenon has raised a big debate 1 within the scientific community (Perotto, 2009, 78) and can be defined as a "sociolect" (Perotto, Ambrosi, 2009, 185), or, because of some of its aspects, it has even been compared -perhaps too drastically -to pidgins (Polinsky, 1996, 75).

In Italy, a similar analysis has been conducted by the Slavist Monica Perotto, who has studied the situation of bilingualism in the Soviet Union (Perrot, 1988), the situation of the Russian language in the post-Soviet area and all the issues related to this: the language maintenance and switching, the language policies of the different States (Perotto, 1996).

Nevertheless, in particular, the researcher has been observing for years the Russian-speaking immigration in Italy, from the socio-linguistic and socio-cultural, points of view, through surveys, questionnaires, informal interviews and simple contacts and human relationships. As evidenced by the title of her work, Lingua e identità dell'immigrazione russofona in Italia (2009), Perotto's goal is therefore not only devoted to the analysis of the linguistic behavior of the Russian-speaking community in Italy, but also to a sociological analysis. Can we talk about a Russian, or Soviet, identity referring to people who are so different because of their personal, work, education history and especially bacause of their geographical origin? Russian language has always been a "glue" for the various nationalities who were living within the Soviet Union; can we say the same about those who left their homeland and, as a consequence, were forced to abandon, at least partially, this strong element of identity ? And what are they up to with this language?

The sample studied by Perotto consisted of 100 informants, who had Russian as their mother tongue (often together with another language of the former USSR) and who had lived in Italy for at least five years, which, all together, allowed them to be considered bilingual.

Starting from all these interesting data and surveys, in particular Fabbro's studies and those about adult Russian-Italian bilingualism by Perotto, I decided to analyze an oral corpus of early Russian-Italian bilinguals, i.e. young people who moved to Italy in their childhood or adolescence, and living here for at least five/six years, and who initially had only Russian as their mother tongue, to which they added Italian later. This is going to be a purely linguistic survey, although there are also some sociocultural psycholinguisitic references, which are essential to contextualize the respondents and to understand bettertheir linguistic behavior.

The goal set was to test the level of bilingualism of these young people. More precisely the interviews (although it would be more appropriate to call them just "conversations") aimed at verifying the effectivess of the Critical Threshold Theory by Franco Fabbro. The aim was to understand if indeed it is correct to speak of an age by which a person can acquire one or more languages perfectly and beyond which they can no longer be forgotten. In particular, we tried to understand if an individual who emigrated to another country at an early age, decreasing the use of his mother tongue and exposure to it, may lose the ability to express himself in this language and to make use of all its components (morphosyntax, phonetics, vocabulary, pragmatic) adequately.

Given the small sample -only ten respondents -and especially the scholar's inexperience, this work certainly does not have any scientific claims, and does not aim to publish innovative results, but it rather tries to be food for thought for future research.

The thesis is divided into three parts. The first chapter is devoted to neurological and linguistic data and to the necessary preliminary remarks about the concept of bilingualism, language acquisition, development of various languagẽ skills, their location in the brain, the possible mixing of two or more languages and all those concepts that contribute to contextualize this rich field of study.

The second chapter, instead, is entirely devoted to the research: it describes its aims, the sample of respondents and the interviews, but also the data obtained, without neglecting, as it has already been mentioned, socio-linguistic considerations, which attempt to explain the disclosed elements.

Referring to the second chapter, the third one presents the results of the research and compares them with those obtained by Monica Perotto and the studies about the situation of the Russian language abroad (especially those on runglish), but especially it compares the data which emerged from the interviews with the theories about the Critical Threshold by Fabbro.~

Chapter 1: NEUROLINGUISTIC DATA

Bilingualism is a phenomenon that encompasses various disciplines and for this reason studies about it emerge both in the experimental sciences, and in the theoretical ones. In particular the disciplines involved are: neurology, sociology, psychology, linguistics, translation theories.

Recent studies -the most important are those of the Canadian Michel Paradis (1994,2004,2009) and the Italian Franco Fabbro (1996Fabbro ( , 2004) -recognize the interdisciplinary nature of the subject and the need for a holistic approach -although, separatisms remain.

It is obvious that studies on bilingualism -in particular, in addition to Paradis and Fabbro, we mention Weinreich (1953), Cummins (1979Cummins ( , 1984Cummins ( , 1986, Grosjean (1982Grosjean ( , 1989) -have developed and continue to develop, especially in those countries where this phenomenon is more widespread and therefore more analyzed, namely in those countries which have been considered, for centuries or because of recent immigration-related issues, "bilingual countries (or often only areas)", such as the United States, Canada and the former Soviet Republics. For these countries, in fact, bilingual education is not an innovation of the XXI century, as it is today in the European world, but it is an established, accepted and institutionalized tradition.

Types of bilingualism

In neurolinguistics, as the well-known neuroscientist Fabbro states, an individual is considered bilingual if he knows, understands and speaks two languages, or two dialects or a language and a dialect, and is able, if he wants, to make a clear separation between the two linguistic systems when he expresses himself (Fabbro, 1996, 116-117 (Wartenburger et al., 2003, 160). Scholars, however, are not unanimous in determining the age of these categories; Fabbro puts in the first group the individuals who acquired a L2 within three years of age, and in the second group those who acquired a L2 within 8 years of age, other researchers establish different thresholds (ibid, Perani et al., 1998Perani et al., , 1841.

Referring to the level of competence in both languages, bilingualism can be divided into balanced and dominant. In the first case the individual masters the two languages at the same level, while in the second one the competence of either language prevails over the other (Fabbro, 1996, 119).

The Canadian researcher Wallace Lambert (1974) proposed first the differentiation between subtractive bilingualism and additive bilingualism, which has been analyzed again several times by his famous compatriot Jim Cummins (1986). In the case of subtractive bilingualism, the L2 replaces the L1, which causes negativẽ effects on the normal learning development the individual, while in the case of additive bilingualism, instead, the development of L2 proceeds together with L1, with considerable benefits for the subject (Scaglioso, 2008, 161).

Like all the attempts to classify human behavior and mind, these distinctions are in some way imperfect and changeable, and they can not be used as mere labels to classify the subjects, to sort them into groups, categories. The level of bilingualism of an individual, as we shall see, can change over the years because of several factors, and so also the given ''label". On the other hand, however, these classifications are necessary to get out of the "labyrinth" of this field of study and to focus the attention each time on a particular aspect of bilingualism .

Components of language

According to De Saussure's definitions, it is common to distinguish, within the language:

phonology, which is the set of speech sounds (phones), whose phonetics studies the production, perception and features, while the phonematics studies the distinctive function within a language system, starting from the distint units of sound, (phonemes); morphology, whose object of study are morphemes (suffixes, prefixes and infixes), which classifies words according to their belonging to different categories (name, pronoun, adjective and verb) and establishes inflection (conjugation and declension) and derivational rules;

syntax, ie the rules that determine the order of the various elements within the sentence;

semantics, ie the part of the language which studies the meaning of words (lexical semantics) (Salmon, Mariani, 2008, 49). Lebrun, 1984, 177).

These characteristics of the human brain are very important in the neurological field for the study and treatment of language disorders due to injuries to brain areas designed for its processing elaboration, aphasias 2 (Fabbro, 1996, 49).

In particular, the same lesion to the left hemisphere causes, in principle, less serious disturbances to a small child, rather than to an adult. For a child, in fact, the recovery of expressive and understanding skills will be easier than for an adult. This is because for an infant and a child both cerebral hemispheres can support the functions of language; according to some clinical research, in fact, with the maturation of the brain, particularly after 8 years, the language and its most "typical" functions (phonology and morphosyntax) are permanently represented in the left hemisphere. For this reason, in case of brain damage to the left hemisphere, for 2 As for the areas of the brain designed for the processing of language, we primarily remind (although they are not the only ones) of Broca's area, responsible for speech production, and Wernicke's area, responsible for understanding, generally located in the left hemisphere (in left-handed teople hey may also be located in the right hemisphere or in both) and connected to each other, and with other brain structures that control language, through neural circuits (Fabbro, 2004, 60).

children under the age of 8, the right hemisphere is still able to support phonological and syntactic functions, while when age increases, the ability of the right hemisphere to reorganize completely lost functions progressively decreases (Fabbro, 2004: 26-27).

Implicit memory and explicit memory, acquisition and learning

A language can be learnt in two different ways, through acquisition and through learning, corresponding to two different types of memory to which the individual resorts, the implicit and the explicit memory respectively, involving different brain circuits. According to Fabbro:

the acquisition of a language happens with natural methods, in an informal setting, with the involvement of implicit memory (Fabbro, 1996, 120).

Implicit memory, which also includes procedural memory, and responsible for the acquisition of phonology and syntax, is the only type of memory available to children up to 8-10 months, it is dominant up to 3 years, and it is the first to appear in the child and the last to disappear in the elderly (Fabbro 2004, 57).

Acquisition through implicit memory occurs in a random and unconscious way, stored capabilities are used automatically and cannot be described in a conscious way.

Awareness begins to develop in the child after 10 months, but up to 3 years the acquisition of implicit capabilities prevails on explicit one.

In particular, thanks to this type of memory children acquire the main aspects of L1, such as phonological (recognition and articulation of sounds) and morphosyntactical (inflection, words combination) features. (Ibid,59).

Learning a language, on the other hand, takes place mainly in a formal way, ie through rules, often in an institutional environment (Fabbro, 1996, 120).

The learning process involves above all explicit memories, also called declarative, which, opposite to the implict ones, acquire information through consciousness, so the information can potentially be retrieved and verbally described (Fabbro, 2004, 63 ).

The formation of explicit memory begins after two years of age and ends around puberty, but before the age of 6 it is difficult to store up informatioñ effectively (which is why it is common to start primary school just at this age) (ibid, 65 ).

Both systems contribute to the management of language skills; phonological and morphosyntactic aspects of L1 are stored in implicit memory, but words and their meanings are predominantly located in that explicit (ibid, 67). Similarly, both memories can be used to learn languages and manage these skills, even those coordinated by the explicit memory, requiring the use of consciousness, are slower and limited (ibid, 68).

First language acquisition

At the basis of language learning and many other motor skills there is the ability of man man to repeat what he observes from other individuals, the so-called imitative behavior, and it seems that mirror neurons are responsible for it (Fabbro, 2004 41).

In his research, Fabbro widely describes the different stages of language acquisition and its various components by the child during his growth and therefore his neurological development. In particular, he highlights that, according to some experimental studies, carried out most of the time through neuroimaging techniques, the acquisition of phonological aspects -more than others -of a language is interrelated to the discriminating factor of the age:

Up to six months of age, children are able to discriminate the sounds of all human languages. This ability is progressively reduced from six to twelve months. At one year of age, children are able to discriminate only the sounds of the language they are exposed to (ibid, 29-30).

Fabbro describes in detail the acquisition of L1 and the period, defined prelinguistic, during which the child, in his first year of life, goes through the production of the first vocalizations to that of the first words (ibid, 30-32), as well as the following stages, of actual language production and acquisition 3 .

The child goes through intermediate steps, from a pre-linguistic stage to the full development of grammatical and morphological skills. The latter, in which the child begins to make use of longer and more articulated sentences, of a richer vocabulary, to conjugate verbs and to agree adjectives, nouns and articles, starts after the age of 3 and ends around the age of 8, when grammatical competence is similar to an adult one.

6

Second language acquisition and critical thresholds

The study of language disorders and aphasias in children, including bilingual, has notably contributed to determine the cerebral representation of languages and critical periods in language acquisition in general, and in the L2.

The Canadian neurolinguist Michel Paradis (1992Paradis ( , 1994 notices that L1 acquisition occurs automatically and implicitly for everyone, and the same applies to the L2 (or further languages), if it is acquired in childhood, in the first decade of life or anyway within puberty, unlike a language learned in formal situationa, at school for example, in the second decade of life, which is, instead, recorded by the explicit memory (see Fabbro, 2004, 106).

For this reason, as early as the mid-twentieth century, the Canadian neurosurgeon Wilder Penfield Graves, whose cases were taken up and confirmed almost half a century later even by Paradis, affirmed that languages should be taught in the first decade of a child's life, even better if within the age of six, when the brain is still "malleable" and "specialized in language acquisition" (ibid, 102).

Fabbro analyzes the acquisition of every aspect of the L2 and its critical thresholds. Firstly he reports the results of some researches regarding the acquisition of phonological aspects of a second language: early exposure, between 9 and 10 months, to a foreign language in contexts of social interaction mantains the ability to recognize the distinctive features of the phonemes of the language (ibid, 30). Secondly, he presents the results of some researches on pronunciation and grammatical competence (the greatest difficulties that arise during the learning of a language) which demonstrate how a perfect pronunciation and grammatical competence in a foreign language are fully achevable only if it is acquired before the age of 8.

After the age of 8, in the child the phenomenon of vocal accommodation 4 get reduced and therefore also his ability to detect and perfectly reproduce the sounds of a foreign language and to imitate the prosody.

Similarly, after 8 years of age the individual slightly loses the ability to assimilate the proper use of the closed class words 5 and grammaticality judgment 6 , as well as the ability to reach a perfect grammatical competence in a language belonging to a linguistic family very distant from their own, which is, instead, a non-influential factor if you learn a L2 in the first years of your life (ibid, 85-87).

Fabbro explains that a perfect acquisition of a L2 can take place within the age of 8 as at this age phonological and morphosyntactic development in the L1, carried out by the procedural memory, is not yet finished. Before the critical age, neural systems of the procedural memory , are still very plastic and so they allow to assimilate the mechanisms for the various languages without the emergence of interferences between them (ibid, 89-90).

7

Structure of bilingual brain

For nearly a century, scholars debated about the structure bilingual brain, especially on a central question, namely whether the two languages are located in the 4 Vocal accommodationis the ability and willingness of the individual to make their own vocal expression as similar as possible to the one of their interlocutor (Fabbro, 2004, 42).

5 Closed class words, also called "function words" (articles, prepositions, conjunctions, pronouns, auxiliaries, demonstrative and possessive adjectives), are different from the open class words, also called "content words" (adjectives, nouns, verbs) whose learning does not depend on critical thresholds, but continues throughout life (ibid, 86). same brain areas or in different areas. Nowadays, the most validated hypothesis seems to be one which is intermediate between the various theories that were opposed earlier, which recognizes that languages are located partially in common areas and partially in some specific and separate areas (Paradis, 2004, 116-117).

The most recognized theory is the convergence hypothesis by David Green, which states that, especially in early bilinguals and late bilinguals with a high level of competence in L2, when the ability to communicate in L2 is gradually more and more similar to the one in L1, language computation systems of L2 tend to correspond to the ones of L1 (see. Salmon, Mariani, 2008, 64).

Through the Event-Related Potentials (ERPS) technique, deriving from electoencephalography, it was concluded that:

the age when a second language is learned can influence the type of representation of the languages in the brain (Fabbro, 2004, 95).

This question refers in particular to the location in the brain of closed class and open class words. As some structures of the memory develop at different times, if a L2 is learned before the age of 8, its most important grammatical elements, namely closed class words, are represented in the same systems of the procedural memory in which the same elements of L1 are located. Otherwise the L2 is represented in different systems of the memory with than the L1. This different location will affect the use of the two languages throughout the course of life;

phonological and grammatical competence in of L2, in fact, will always be inferior, its use will be less and less automatic and will require more effort than the use of L1

(ibid, 93-95).

However, other researches studied the cerebral representation of syntax and showed that:

In individuals who learned a second language after the age of seven, the two languages are represented in different portions of Broca's area (the area of the brain that controls language production). The age of language acquisition does not seem to affect the representation of the languages in Wernicke's area (the brain area responsible for the recognition of speech sounds) (ibid, 96).

Finally Fabbro presents the results of other researches (Wartenburger et al., 2003) on second language early acquisition (which took place between the age of 3 and 8) and very early acquisition (within 3 years of age). In both cases, the individuals were able to reach a perfect phonological and grammatical competence, but there were some differences on the neurobiological level, which cause for early bilinguals a greater difficulty in the use of the L2 (see. Fabbro, 2004, 98-99, Paradis, 2004.

Of course we must remember that brain maturation and language development, and consequently both L1 and L2 acquisition, vary from individual to individual and are influenced by external factors, such as the higher or lower socio-cultural environment in which the child is placed, which allows him to receive more or less language and sensory stimuli in general.

It was also found that, as a rule, girls acquire language more easily than boys and, above all, that motivation plays an essential role: at any age, in fact, if emotions and positive emotional situations are connected to the language learning process, then the assimilated information augments and the overall performances improve (Fabbro, 2004, 104-105).

Capabilities of bilingual individuals

In addition to a natural grammatical and phonological competence in both languages, bilingual individuals tend to have a flawless grammaticality judgment, roughly equivalent for the two languages, which enables them to determine whether a statement is compatible with the structures of the language .

This concept is connected to the inner ear, which Salmon defines as the complex device that allows you to automatically check the correctness of a statement without activating explicit reflections (Salmon, 2004, 240).

The inner ear develops thanks to an intense, frequent and lasting exposure to the language and it gets activated both when an individual speaks, and when he listens to the others monitoring language consistency in their speech, and when he reads and is able to reconstruct the presumed meaning and the prosody of sentences, with the help of punctuation and intonation memorized by the inner ear (ibid).

A bilingual is therefore predictably equipped with an inner ear developed for both languages, capable of performing the processes of language control in an immediate, unconscious and implicit way.

Another important sphere of linguistics and communication processes studies, often overlooked, is pragmatics. Pragmatics studies the ability of an individual tõ produce right utterances not only in terms of grammar, but also appropriate to the context, ie to choose, between the various options in terms of correct morphosyntactic, that one (or those ones) which expresses the purposes that communication determined in a given situation.

If morphosyntax were separated from pragmatics, there would be the risk of producing statements that are correct, but non-existent, in contrast to the use and expectations of the interlocutor in that context, which creates obvious misunderstandings.

For bilingual individuals even this selection process is unconscious and fast, both in the L1 and in theL2, and it allows them to distinguish, thanks to the inner ear, the phrases that are not marked from the one which are marked.

The criterion of functional markedness (f-markedness) is different from the concept of markedness commonly used in linguistics (the latter defines the statements with a "low occurrence"), and distinguishes marked sentences from nonmarked not according to the words used, but to the context in which they are placed (Salmon, Mariani, 2008, 106-107).

The f-marked statements show no mistake from the morphosyntactic or semantic points of view, but they violate the standard of use referred to the context.

Hence an expression must be analyzed according to the relation between who says what, to whom he addresses, how, where, when and why (wh-factors). An expression, in fact, may be common, and therefore not marked in a context, but very marked and mismatched in another (ibid) 7 .

The ability of a bilingual to understand different registers and markednesses should also allow the bilingual to understand and master the symmetries and asymmetries between terms, concepts, phrases and sentences of L1 and L2 and to be able to find, for each unit of L1, a unit of L2 which bears an equivalent fmarkedness.

It is symmetric any formal and formalisable element that, in a contrastive analysis, appears equifunctional in the transition from one language to another, which means that, in translation, it is would be "left where it is", with the same grammatical, lexical, 7 In some cases we pronounce marked sentences deliberately to express marked irony, sarcasm or detachment.

rhetorical, register etc. role: therefore we can speak of phonetic, morphological, syntactic, lexical, rhetoric, stylistic, registrer, intonational, etc. symmetry. (Salmon, 2004, 244).

Therefore, a person is bilingual if he is able to recognize and control automatically and implicitly the asymmetries of the language pair that he masters, without being subjected to interferences.

In the essay Bilingualism and Translation, Salmon and Mariani (2008) show that, however, being bilingual, or having a perfect knowledge of two languages and their cultures and being immersed in them from birth or from an early age, is not enough to become translators. Switching from one language to another is vital in order not to lose the emotional charge of a term or expression, which means not to neglect pragmatics rules and what specific realia evoke within a certain culture (Salmon, Mariani, 2008 , 81).

In thoery, though, for a bilingual trained to interlingual switching and to finding the right matches between the two languages, interlingual translation should be very similar to the intra-lingual (or endolingual) translation, ie the ability of any individual to recode "in other words" (ie, change register) a statement within the same language (ibid, 79).

To avoid interferences and errors in the switching process, however, it is also important to have the ability to keep separate the two languages, namely language tagging (ibid, 65-66).

To explain the phenomenon of tagging we should introduce the theories of Inhibitory and Control Model (ICM) by Green (1986Green ( , 1998 and the Activation Threshold Hypothesis by Paradis (2004) . The first states that When a bilingual speaks in one language inhibitory control mechanisms are called upon to suppress the irrelevant language (Schwieter, Sunderman, 2008, 215).

Among the various possible options, then, the brain selects and activates a particular element, excluding and inhibiting the other.

According to the Activation Threshold Hypothesis by Paradis, in fact, When bilingual speakers use one of their two languages, the activation threshold of the other is automatically assumed to be raised as to avoid interference (Paradis, 2004, 28).

This hypothesis is important to explain the tagging, in fact a parameter is activated when a sufficient number of pulses reaches a specific neuronal substrate. The number of pulses required to activate the parameter constitutes, therefore, its activation threshold. An important corollary of this hypothesis is that, each time a parameter is activated, its activation threshold is lowered; if, vice versa, it is not activated frequently, it becomes increasingly difficult to activate it. In case of a bilingual the selection of a language would entail, as a consequence, the simultaneous inhibition of the equivalent translating words of the other language through the elevation of their activation threshold (Salmon, Mariani, 2008, 66-67).

In case of problems in the mechanisms of control and inhibition we may witness deficiencies and defects in the switching or code-mixing situations.

Code-mixing and interferences

Code-mixing is the unintentional mixing of two languages 8 , which involves interlinguistic interference (Salmon, Mariani, 2008, 65).

As the Inhibitory and Control Model by Green and the Activation Threshold

Hypothesis by Paradis (ibid,66), thanks to the tagging in a bilingual mind there is reciprocal inhibition between the two languages. People who are used to keeping the two languages well separated develop then stronger reciprocal inhibition systems than those who are used to frequently mixing the two languages, also for business reasons (Fabbro, 1996, 147).

On a word level, interferences can be phonological, morphological and semantic, while on a sentence level we can have lexical, syntactic and prosodic interferences (Taeschner, 2005, Bernarducci, 2006.

Phonological interferences are in principle the less frequent in bilinguals, especially in the early one, who, as pointed out before, have the ability to acquire a perfect pronunciation in both languages, if they are learned within the critical threshold. Usually when phonological interferences are found in the language of children who are bilingual from their birth are due to the fact that one of the two languages is dominant over the other. Moreover, as with lexicon, it is likely that initially the two phonology systems are not perceived as separate, which means that after a period of undifferentiation it comes gradually a differentiation one.

Phonology, in fact, is closely linked to lexicon and the distinction of the two phonologies takes place simultaneously with that of the two lexicons (ibid, 76).

Morphological interferences entail the change of only one part, a morpheme, inside the lexeme (suffix, prefix, infix), so to make it similar to a possible term in the other language. This kind of interference is very common among bilinguals, especially in children who acquire the rules of two linguistic systems asynchronously, that is when they have already assimilated morphosyntactic structures in one language and not in the other, and yet the child tends to extend the morphological rules of the first language to the second (inter-lingual supra extensions) (ibid, 77).

Semantic interferences occur when the meaning of a word in a language is allocated, for generalization, even to an apparent equivalent in the other. They can be caused by phonological similarity between the words of the two languages, even though that similarity does not correspond (or it does correspond only in part), however, to a semantic equivalence (ibid, 78).

Another kind of common interferences among bilinguals are the lexical 9 ones, that occur in cases where in a sentence in one language are inserted words of another language. The sentences in which there is about the same number of words in a language and in the other are defined mixed and are found especially in the first years of life of early bilinguals, when children do not differentiate the two lexical systems yet, or when children develop two types of different vocabulary according to the situations in which they use the L1 and L2 (for example in cases where the L1 is used at home and the L2 at school), and as a consequence they have a shortage of terminology in some communication contexts, which they overcome by inserting a term from the other language (ibid, 80).

Syntactic interferences occur when the order of the words in the sentence is transferred from one language (usually the dominant one) to another (the one used less frequently). Also in this case, as for morphological interferences, we can talk about inter-lingual extensions. Both types of interferences are explained by the fact that morphosyntactic rules are, even for monolingual children, the last to be assimilated (ibid, 81).

Finally, prosodic interferences occur when intonation,inflection and rhythm of a language are transferred, partially or fully, into the other language (ibid, 75).

Almost all researches carried out in the area the bilingualism start from the assumption/prejudice that is always the L1, acquired first during childhood, to influence and disrupt the L2, and not vice versa (Salmon, 2007).

In the following chapters we will show how often, even early bilinguals are subjected to cases of code-mixing and interferences of L2 on L1.

Chapter 2: RESEARCH

In the previous chapter we considered some of the key neurolinguistic data concerning the area of early bilingualism, ie the critical thresholds that affect the process of language acquisition, the problems, more or less serious, which may arise in case of mixing of the two languages, ie mainly interferences and cases of codemixing.

Research aims

To relate with the debate on early bilingualism, the present research starts from the Critical Threshold Hypothesis by Fabbro, which states that one or more languages may be acquired in a perfect way only within the age of 8.

We enquire whether, within this age, we can actually talk about perfect acquisition, or rather whether the perfect acquisition mentioned by Fabbro may also be considered as a sort of "guarantee" that the language is acquired once and for all, or whether, instead, even though it was learned within 8 years of age, a language could, at least partially, be forgotten, which means whether morphosyntactic, lexical, phonological and pragmatic skills could be affected for an individual who is no longer massively exposed to this language and who does not make use of it anymore, or who uses it limitedly and sporadically.

It was also noticed that, talking about bilinguals, or at least about people with an almost perfect level of competence in L2, in the scientific sphere it is common to examine the interferences that the first language exerts on the second one, ignoring the possibility that also the opposite phenomenon may occur, namely that the L2 can interfere on the L1. So in addition to the question concerning the possible loss of a language acquired at an earlystage, we also investigated whether and how the second language may interfere on the first one.

Analyzed sample

To answer these questions, we decided to analyze ten early Russian-Italian bilinguals, boys and girls born in Russia (or still the Soviet Union at the time) from

Russian-speaking parents, who initially had only Russian 10 as their mother tongue, and therefore were raised monolingual up to a certain age, and added Italian language after moving to Italy, which makes of them early bilinguals, at least according to Fabbro's hypothesis. These ten children, five male and five female, are now between 23 and 30 years of age and arrived in Italy during childhood or adolescence, between the ages of 1 and 18, have lived here for at least six years and acquired Italian implicitly, at school and in contact with other native Italian speakers, so they didn't learn it explicitly, through language courses or other meta-linguistic reflections.

As we pointed out in the first chapter, Fabbro differentiates early bilinguals, children who acquired the L2 within the age of 8, from very early bilinguals, who acquired it within the age of 3. This further classification has been omitted for the purposes of this research, considering the small number of respondents, so we decided not to further subdivide the sample and to use for all of them the generic term of " early bilinguals", without, however, omitting the appropriate remarks about those who can be defined as " very early bilinguals".

As neurolinguistic scholars are not unanimous about the Critical Threshold

Hypothesis, we decided to consider as an age limit for language acquisition sexual development (which occurs later for boys than for girls), which is usually considered last the critical threshold. For this reason we examined also subjects who moved to Italy within their teenage years.

Development of the interviews

After identifying ten subjects who fulfilled the characteristics described in the previous paragraph and after making sure that they were willing to participate in this research, we decided to carry out interviews aimed to test the oral language skills in Russian -so the L1 -of the sample.

However, instead of conducting actual interviews, and the pattern questionresponse, we opted for simple conversations, free and informal, which had an outline 12 , consisting of 23 points, about the informant personal and language history, such as their arrival in Italy, their first impressions about the country and its language, their relationship with their country of origin and their native language, their use of Russian and their meta-linguistic awareness in the use of both languages.

As they were free conversations, which aimed at reproducig as much as possible natural and spontaneous dialogues, the outline has not been followed in a 11 To respect their privacy we used fancy names, but we haven't changed all the other data. strict way (as if it was interrogation). The informants were not even aware of this written outline, but, before the interview, I explained them about the purposes of the research, without detailing, though, my will to of analyse their linguistic behavior, so that they would not excessively control their speech and would not activate conscious and unnatural metalinguistic mechanisms. They were explained the development of the interview, stating that it had to be an informal conversation in Russian about the relationship that the informant had with his native language and his country of origin.

As far as possible, the interviewee was let talk freely and in a "continuous flow", the interviewer avoided interruptions, except when some clarifications were needed and, above all, avoided the question-response mechanism, typical of interrogation, indeed, and not of spontaneous dialogues.

In cases of shy informants, and therefore less inclined to talk about themselves independently, the interviewer intervene more frequently, relying on the outline to get the conversation back to these topics. Although, nothing prevented the respondents, and the interviewer, from digressing on other issues, even the silliest:

what was important, in fact, was to get a corpus of the speech of each respondent.

These basic topics of the interview, therefore, were not meant to constrain the conversation, but to get, along with a spoken corpus, also sociological information on the sample, to be able to better contextualize them, comparing socio-and psycholinguistic data.

The dialogue being informal and spontaneous, even the exact lenght of each interview was not set, the shortest lasted for about 9 minutes, while the longest for about 39. According to the familiarity between the interviewer and the interviewee, and the intimacy that rose between the two, the loquacity or discretion of the respondent, and the desire to talk about himself that everyone showed, each interview had a different duration and development. The informant was given the freedom to speak as much as he wanted as well as to lead the course of the conversation.

The decision to conduct oral interviews, rather than written questionnaires, is motivated by the fact that the object of investigation was oral language, not the written one. Writing skills, and above correct spelling, are a secondary, optional aspect, that develops at a later stage in the individual 13 (both on the ontogenetic and phylogenetic levels), which does not affect the ability of verbal expression (a person, in fact, can still speak a language perfectly, while still making spelling mistakes).

Moreover, as we all know, any written form has the disadvantage of being potentially checked and altered, in a more or less conscious way, which interferes with the unconscious and immediate mechanisms of spontaneous orality.

Sociolinguistic overview of the sample

Conducting these interviews turned out to be very interesting, not only to understand the language behavior of the analyzed sample of early bilinguals, and then to "test" the critical threshold and maintenance/loss of L1, which will be discussed in the next section. It also revealed very important and curious information about the interest that these people have towards Russian language, their awareness and their meta-linguistic reflections on their competence in this language, but also in Italian, and the causes and reasons that led some of them to keep an alive and correct use of the Russian while others has lost it, partially or almost completely. Some of the respondents arrived in Italy after adoption and then, as they said themselves, immediately after the transfer, they had a long period of real rejection, as it often happens, towards their native language and country, which were linked to negative memories about life in orphanage (one of them claims to have completely forgotten Russian and to have started almost from scratches after enrolling at the Faculty of Languages). As they don't use Russian in their family context and everyday life anymore, they claim to speak better Italian than Russian, -which they still got back to study in high school or in college -although almost all of them developed an interest in their mother tongue, their culture and their origins. However, even those who moved to Italy with their family of origin admit to to use both languages at home, making use of the mixed-language utterances 14 and of code-switching.

The observations and examples of one informant were particularly

interesting, as he explained that often, talking with his mother in Russian, he has a hard time finding the words, because they come to his mind in Italian, so he uses Italian terms and adds a Russian ending (he makes the example of cominciarit'). They admit, moreover, that they often have doubts in Russian, they make mistakes and are corrected by their family. As for the relationship with the two languages, both the ones who were adopted, and the ones who have Russian family said they initially had a hard time learning Italian and dealing with the two languages , but today feel more confident in Italian than in Russian (only one respondent feels he masters both languages equally well). Some of them, to explain this situation, make the example of translation;

when they have to make some translation work, they are aware that they are better at translating from Russian into Italian, rather than vice versa, as they acknowledge they don't know many words in Russian or simply they have trouble finding the right words.

As for the interest in what happens in their country of origin and in general of their own culture, their opinions are different (although in some cases the informants seem to contradict themselves and to have confused ideas on this topic). About half of the respondents claimed to read books in Russian, to follow Russian news and television and to be interested in what happens at home, while the other half, above all the adopted informants and those who arrived in Italy when they were very little, confess they do not follow much the issues about their country of origin and read almost exclusively in Italian. The informants have also provided valuable information about the maintenance of contacts with their native country. Almost everyone keeps in frequent contact with relatives and friends back home, via telephone or correspondence and they often go and visit them (on average once every two years), except again for adopted children, who cut nearly all the ties with the home country.

14 Mixed-language utterances are a phenomenon thas has been largely treated by scholars on bilingualism, especially by Russians; some see it as an expression of language decay, similar to interferences, some others as a more active form of bilingualism. In her research, Perotto (2009) widely analyzes Russian-Italian mixed-language utterances, as well as the various scholars who have dealt with this matter.

For some contacts with compatriotes have endured even after their arrival in

Italy: none of them seem to be particularly looking for other young Russian speakers, but they are all happy to have to deal with people who share their same origins, to share experiences and stories of common life with. However, their accounts on the language used with these friends are different:

the majority of the boys said they use Russian when in a group of only Russian speakers, adding a few words in Italian if needed, while the girls said they are now so used to 'Italian, that they use only this language or sometimes they speak half Italian and half Russian, or even they switch completely to Russian just to be understood by the not Russian-speaking rest of the company. Another issue discussed during the conversations was the identity: only two informants claimed to feel Italian and Russian, the others, instead, said they love Italy, they feel good in this country and they feel comfortable with Italian, but they feel a strong attachment to their homeland and they feel different for some aspects from their Italian friends and acquaintances andso they are proud of their national peculiarities.

It is remarkable the fact that some of the respondents showed a certain reticence towards the interview in Russian, not only those who proved real difficulties in expressing themselves in this language, but also those who mastered it perfectly, insisting on the fact that they would feel more confident and at ease speaking in Italian. This attitude may be due not only to the fact that some of them admitted to not being used to using this language anymore, speaking it badly and rarely, and therefore to be a bit ashamde, but also to the fact that, since the interviewer's mother tongue was Italian, they perceived Italian as a language "more in common" 15 .

Being life stories, all the interviews took a different course, and every speaker insisted more on some points rather than others. What we are presenting now are therefore only the most important among all the sociological aspects that emerged, from which, although, we can portray an interesting profile -although partial, given the small number of the sample -about the relationship that young Russians, living in Italy since childhood or youth, have with their country and their language.

First classification of the sentences

Initially, the interviews were categorized according to the informant; for each interview all the statements that contained something "strange" 16 16 Appendix II presents the classification of the errore of all ten interviews.

Grammar mistakes

The most frequent errors observed in the speech of the respondents concern verbal aspect, verbs of motion, grammatical agreements, conjugation of verbs, especially irregular ones, use of the cases, of the prepositions and of the different parts of speech.

This kind of errors is typical for those who learn Russian, especially for learners whose mother tongues are not Slavic and do not have the category of aspect,

do not have such a diversified range of verbs of motion, and do not decline nouns, adjectives, pronouns , numerals and names.

Aspect

As for the category of aspect, we found 13 cases of misuse of perfective and imperfective, distributed among 5 speakers.

As the examples below show 18 , the perfective is very often used in contexts where the imperfective would be required 19 :

1. Когда я уже думал [...] изучать русский, я начал его изучать ещё (fm) и чуть чуть его *всмомню (*всмомню instead of вспоминаю) (respondent 2); 2. Прежде чем мои итальянские родители mi adottassero (cm/cs), я *ехала в *Италии и *вернулась в Россию два или три раза (*вернулась instead of возвращалась) (respondent 4); and, viceversa, the impefective when the perfective would be required:

3. Три года назад я *начинала изучать русский язык, потому что вначале поняла, что русский язык для меня это (fm) *важно, (*начинала instead of начала) (respondent 4); 4. Для ребёнка, когда он вырастет, будет его выбор *если *выбирать *один или другой язык (*выбирать instead of выбрать) (respondent 9);

Verbs of motion

Many informants, 5 out of 10, showed to confuse also the verbs of motion.

We noticed 12 mistakes in this areas, linked to the confusion between unidirectional (идти and ехать) and multidirectional (ходить and ездить) verbs:

1. Шесть лет я *ехал зимой и летом в Италию (instead of съездил) (respondent 8); 2. Сейчас, *как я работаю, больше не могу часто *ехать (*ехать instead of ездить) (respondent 7); 3. Если он начинает *идти [...] в школу в Италии...si abitua di più, cioè, (cm/cs) *привыкает лучше (fm) к итальянскому языку (*идти instead of ходить) (respondent 9), both between the verbs that carry the meaning of "walking" (ходить -идти) and "going by transports" (ездить -ехать):

4. Мне очень хотелось *идти в *России (*идти instead of ехать) (respondent 4); 5. Я думаю, что буду жить здесь в Италии, но *ходить в *России тоже (*ходить instead of ездить) (respondent 4), and issues on the use of verbs of motion preceded by a prefix: 6. Мне сразу *идёт в голову говорить по-русски (instead of приходит) 20 (respondent 2); 7. Она тоже уже давно живёт... она *ездила в Италию, разговаривает очень хорошо *по итальянскому языку, потом... её жизнь уже в Италии (*ездила instead of переехала) 21 (respondent 8).

Cases

The most copious mistakes undoubtedly concern the use of the cases. We found even 107 such errors, over 8 speakers, even among those who have demonstrated a rather fluent speech.

Most of these errors are linked to the expression of motion to a location, after which there was a tendency to use the prepositional, instead of the accusative (which means an assimilation of the motion to the state):

1. Потом я приехал с моей мамой (fm) *здесь в *Италии (*здесь instead of сюда, в *Италии instead of в Италию) (respondent 2); 2. Поеду в *Англии, в *Лондоне (в *Англии instead of в Англию, в *Лондоне instead of в Лондон) (respondent 4); 3. Хожу в *университете иногда (instead of в университет) ( respondent 7), although there are also examples of the use of the accusative instead of the prepositional: 4. Бывает …*на интернет да, я читаю (*на интернет instead of в интернете) (respondent 8).

The use of the accusative of nouns and adjectives, but also of pronouns, is also limited. In fact, to express a direct object the informants often used the nominative:

5. Я никогда не видел *русские люди здесь (instead of русских людей) (respondent 2); 6. Они *почти меня *увидят как *итальянец (*итальянец instead of итальянца) (respondent 2); 7. Мне очень хочется вернуться в *России, чтобы видеть... *мои родные люди, *мои родители (*мои родные люди instead of моих родных людей, *мои родители instead of моих родителей) (respondent 4).

However there are also opposite examples, ie proving the use of the accusative when one should simply use the nominative: 8. Они не как *меня (instead of я) (respondent 4). In this case probably the informant got interfered with the Italian form "me", which resembles the Russian form "меня" (the initial phonemes, in fact, coincide).

Even the genitive in the negative sentences is often omitted and replaced by the nominative: 9. У меня не было *возможность (instead of возможности) (respondent 4); ~ 31 ~ 10. Я должен был там сделать passaporto (cm/cs) и, *как у меня *он не было [...] (он instead of его) (respondent 2).

We found several errors in the use of the cases also after verbs or adjectives, especially if these require special cases, or together with verbs and expressions which in Russian are followed by a different case than in Italian: 11. Мы занимаемся *перевозки cargo (cm/cs) (*перевозки instead of перевозкой or перевозками) (respondent 5); 12.

[Они] Были рады *улыбки детей (instead of улыбке) (respondent 9); 13. Конечно *меня платят за эту работу (instead of мне) (respondent 9); 14. Она *меня помогает писать правильно (instead of мне) (respondent 4); 15. Они отвечают *меня «Нет» (instead of мне) (respondent 4); 16. *Мне интересует, просто времени очень... нету (instead of Меня) (respondent 7). 17. *Мне бы интересовал тоже говорить и познакомиться с кем-то, кто говорит по-русски (instead of Меня бы интересовало) (respondent 2).

The examples 13-17 may be due to the interference of the Italian pronoun "mi", which corresponds, in Russian, both the accusative "меня", and the dative "мне".

We want to highlight the recurrent wrong use of the verb "интересовать": no respondents who made use of this verb used the correct form with the accusative, but always with the dative. As we just mentioned, the confusion is probably due to the convergence of the pronoun "mi" in Italian into accusative and dative, but also to the fact that the adverb "интересно" is preceded by the dative instead. Moreover, the informants didn't show to know well when to use the verb and when the adverb, as in the following example:

18. *Мне не интересует выучить этот язык (it would be better Мне не интересно) (respondent 9).

Another frequently observed error among many respondents was the expression of the age, for example: 19. Щас *у неё одиннадцать лет (instead of ей) (respondent 2); 20. *У меня двадцать три года (instead of Мне) (respondent 4).

Probably the respondents followed the Italian construction, expressing the age through the verb "to have" -for this Russian mostly uses the construction "у + genitive" -while for the age Russian uses the dative, which can be followed by the verb "to be" in the past (было) or the future (будет).

Other recurring errors were noticed in the use of cases after prepositions. The respondents tended not to decline the nouns, leaving them at the nominative: 21. Для *ребёнок это легче выучить итальянский язык (instead of the genitive ребёнка) (respondent 4); 22. Я думаю, что *в будущее русский язык будет *полезно (*в будущее instead of the prepositional в будущем to express definite time) (respondent 4); 23. Я живу в Рекко, vicino (cm/cs)… около *Генуя (*Генуя instead of the genitive Генуи) (respondent 7), or they declined at the wrong case;

24. Я пишу *фейсбук *друзьями и подруге, которая была *со мне в детском саду (fm) (*со мне instead of со мной) (respondent 4). It is likely that the interviewee here remembered the initial phonemes of the instrumental "мной" (/ m / and / n /), which coincide with the initial ones of the dative "мне", and she confused the two forms; 25. Паспорт у меня на *белорусском переведен (instead of белорусский) (respondent 1). Probably here the respondent used the form "на + prepositional", which is the most common when talking about languages (ie "in Russian, Italian"), losing the automatism of the verb of motion "перевести" (which in this case requires the preposition "на + accusative"); 26. Больше знаю о *Путина (instead of о Путине or про Путина) (respondent 2). Here the confusion may be due to the fact that in Russian the adverbial of topic can be expressed with "про + accusative" and "prepositional о +" and then the informant may have mixed the two constructions; 27. Мы общаемся по *фейсбуке (instead of по фейсбуку or на/в фейсбуке) (respondent 8). Also in this case the informant could have used both the form "по + dative", both "+ prepositional на" or "в + prepositional". Presumably he mixed the two constructions; 28. *Во-второй... во-вторым, даже если я живу здесь в Италии уже практически половину жизни, всё равно я себя чувствую украинкой (instead of Во-вторых) (respondent 6). This example is particular because the interviewee showed she wanted to decline the unvariable idiomatic phrasẽ "во-вторых", as if it were a phrase formed by the preposition "в" followed by the adjective "второй" (in fact she was trying to find the correct ending), or they were still confusion in the agreement after prepositions: 29. Я хотела забыть *русские дети, [...] поэтому я не хочу *с ним говорить (use of the singular form *с ним instead of the plural *с ними) (respondent 4), both between adjective and noun, and among the different adjectives: 30. Только один раз я проговорил с одним (fm) *русском (*русском instead of *русским) (respondent 2). The form in -ом indicates the prepositional singular masculine of adjectives with a hard ending, but also the instrumental singular of masculine nouns with a hard eding. The adjective "русский", therefore, may have been assimilated to the noun, while being an adjective; 31.

[Interviewer]: Они говорят с акцентом?

[Respondent]: *Русском? (instead of Русским) (respondent 2). Here the respondent makes the same mistake as in the previous example, which may have been caused by the interviewer using the noun "акцент" with the instrumental ending in -ом. Linked to cases declension after prepositions, it was interesting to notice that many respondents, up to 4, showed problems in the use of the expression "speak (and read) a language / in a language" (Russian uses the adverbial form "говорить (and читать) 22 по-русски, по-итальянски, etc.", but also the form with the prepositional is widely widespread "говорить на русском, на итальянском (языке)" ): 32. С мамой *по русскому говорю, иногда *по русскому, иногда *по итальянскому (*по русскому instead of по-русски, *по итальянскому instead of по-итальянски) (respondent 7); 33. С *Анне можно разговаривать *по русскому языку (*по русскому языку instead of по-русски or на русском языке) (respondent 8).

Here the subjects probably followed "by ear" the statistical frequency of the construction "по + dative", forgetting that in this particular case the forms "порусски, по-итальянски etc." are adverbial and therefore they are invariable and idiomatic. This is definitely the most common error for this type of expression, but we detected also others, such as: 34. Я только чуть чуть говорю *русский, хорошо *итальянский, *испанский, ещё (fm) *английский (*русский instead of по-русски, *итальянский instead of по-итальянский, ecc.) (respondent 2). This can be considered as a calque from the Italian form "parlo russo, italiano, ecc.", simply translated with the accusative; 35. В семье мы говорим когда по-русски, когда *на итальянский (instead of по-итальянски) (respondent 2). Again we could refer to this as a calque, as the informant may have translated from the Italian "parlo in italiano" with "на (in) + accusative" (although it is surprising to see that in the same sentence, and very close to each other, the informant uses the first time the correct form, the second one a wrong one); 36. Я думаю, что сын должен всё время слушать... слышать два языка, *нет один день я с ним буду разговаривать по-итальянски и (fm) *один день *по-русскими (*по-русскими instead of по-русски) (respondent 9).

Another recurring phenomenon, and always related to the use of cases after prepositions, concerns the declination of numerals, as the informants tended to use invariably the nominative 23 : 37. После *три года очень очень трудно (instead of трёх лет) (respondent 9); 38. С шести лет до... *восемнадцать могут изучать итальянский язык (instead of восемнадцати) (respondent 9), or to decline them when there was no need to: 39. В *шести лет меня и моих сёстр *убрали от родителей (*шести instead of шесть) (respondent 8).

In this regard it was interesting to notice that two informants repeatedly committed an error very common in Russia: talking about dates and years they declined the ordinal numeral "две тысячи" (followed by another digit), when actually the rule states that with the ordinal adjective only the last digit of the number 23 Truly simplifying the declination of numerals is a widespread tendency in Russia, but never in cases similar to those proposed below, except for non-native speakers.~ must be declined (except for the adjective "двухтысячный", which, in fact, is not followed by anything) (GR, 2001):

40. В сентябре *двухтысячи седьмого я сдавал вступительные экзамены (instead of две тысячи) (respondent 1); 41. Я возвращалась в Россию в последний раз в *двухтысячи седьмом году (instead of две тысячи) (respondent 9).

Finally, we detected other errors regarding the use of cases, difficult to group into precise categories, such as: 42. Перед тем как начинать четвёртый год (fm), обучение *четвёртый курс (*четвёртый курс instead of на четвёртом курсе) (respondent 1); 43. Когда я хочу *более с любвиой говорить с ней (instead of с наибольшей любовью) (respondent 1). The respondent declines the word "любовь", belonging to the third declension, as a feminine noun of the first declension; 44. Первую, младшую, зовут *Веронику (instead of Вероника or Вероникой) (respondent 8); 45. Я старше *два года *Анну и *шесть лет *Веронику (the whole acceptable sentence would be: Я старше на два года Анны и на шесть лет Вероники) (respondent 8). In these last two examples the informant mistakenly uses the accusative of proper names, when the context would require the nominative for sentence 44, and the genitive for sentence 45. The confused use of aspect, verbs of motion and cases can be interpreted as a symptom of the fact that, the respondents partly lost the distinction that exists in Russian within these categories and that, influenced by Italian, they "label" in an undifferentiated and univocal way, the concept expressed by any verb, as it mostly happens in Italian, where there is no grammatical categoryof the binary aspect of verbs as in Russian.

Following the universal principle of '"economy of effort", the subjects seem to assimilate the most "economical" solutions one of the two languages to the other, unconcerned about gradulally losing the automatisms in the use of procedures and in the access to vocabulary and idioms databases.

Prepositions

Many informants, 6 out of 10, also showed difficulties in the use of prepositions, although, all in all, these cases have not been very numerous (only 19).

Errors related to prepositions concern primarily the expression of location, both in terms of state and motion to a location, or motion from a location (or origin), but also the expression of time, or other less frequent cases.

Sometimes the respondents agreed the following nouns, adjectives or pronouns the to appropriate case for the preposition that they mistakenly used:

1. Так как я работала практически всё это время *за прошлого года [...] (instead of в прошлом году) (respondent 6); 2. Есть некоторые подружки, которые *со Львова (instead of изо) 24 By switching to Italian we assume that the informant had difficulties even remembering the nouns indicating the cardinal points.

Although in this case the construction "в + prepositional" is more correct, the negative particle "ни" requires the use of the genitive (we could have accepted "ни акцента, ни говорения"), while the informant uses the nominative. 25

Grammatical gender and agreement

Another small amount of errors -21 distributed over 7 speakers -was found in the gender agreement between verbs, nouns, adjectives, pronouns and numerals.

As we know, inflection rules in Russian are for some aspects richer than in Italian, primarily because in Russian kinds are three (masculine, feminine and neuter) and because plan to accord the parts of speech not only second base to the kind and number, but also to the case. 26 1. Жизнь не *трудный (instead of трудная o трудна) (respondent 4). The informant uses the masculine rather than the feminine. This error may be due to the fact that in Russian nouns ending in -ь, can be both masculine and feminine. We also specify that, as a nominative predicate, the informant could have used both the long form of the adjective (трудная) and the short (трудна). 2. Русский язык для меня это (fm) *важно, потому что это мой *родный язык (*важно instead of важный o важен, *родный instead of родной) (respondent 4). Here too, the informant could have used either the short adjective masculine (важен), or the long one (важный), but she used the neutral form (важно) 27 . In the second case, instead, there is no gender confusion, but the informant uses the ending -ый, typical of masculine adjectives with no stress on the ending, instead of -ой, that distinguishes, instead, the stressed ending and therefore she commits an error of tonic accent, putting on the [о] in the root, instead of on the ending; 3. Это моя *родна страна (instead of родная) (respondent 4). In this case we can not know if the speaker was using the short adjective (which cannot be used for an attributive function, though) instead of the long, or she simply omitted the final morpheme [я] (the emphasis, in fact, was, correctly, on the [а]); 4. Мой родной язык *стало итальянский теперь (instead of стал) (respondent 9) 28 ; 5. [...] *мой проблема это знать что говорить и как (instead of моя) (respondent 2). Here almost certainly the respondent got the interference from Italian, where the noun "problema" is masculine, while in Russian it is feminine, as (almost) all the nouns ending in -а; 6. Я тихо (fm) теряю *русский речь (*русский instead of русскую) (respondent 5). As already said, nouns ending in -ь can be both masculine and feminine: the informant most likely perceive "речь" as masculine while it is feminine (the correct form is therefore "русскую"); 7. Лучше когда маленький изучает *две или три языка (instead of два) (respondent 2). In Russian, the word "язык" is masculine, but it is likely that the informant got confused with Italian, where "lingua" is feminine and then she used the numeral in the feminine form (but for the noun she used, correctly, the ending of the genitive masculine); 8. *Самая главная думаю это открыть среднюю [школу] или elementari (cm/cs) (*Самая главная instead of Самое главное (respondent 9). This case can be considered a calque from the Italian phrase "la cosa più importante", given that the informant uses the female form, but in Russian this phrase is expressed with the neutral; 9. Свидетельство о рождении знаешь? *Она зелёненькая (instead of Оно зелёненькое) (respondent 1). This example is particular because the interference is not due to the gender in Italian, but to the phonetics of the word "свидетельство" where the final [o] pronounced as a / a /, which makes the informant consider the noun as feminine gender. The following example can be considered as an error in the use of the number:

28 See note 19.~

Для меня это была каникула (fm) (instead of были каникулы)

(respondent 4). The noun "каникулы" in Russian is only plural 29 , but almost certainly the interviewee thought about the Italian "vacanza/e", which, instead, presents both the singular and the plural forms.

Parts of speech

Some informants, 4, got mixed up in the use of parts of speech. In fact almost all the errors of this type, 8 out of 9, concern the use of the noun indicating nationality -the respondents used, instead, the adjective -probably because in Italian there is almost never a difference between these and the corresponding nouns adjectives, but in Russian there can be two different forms (for example, the noun итальянец / -ка and the adjective итальянский / -ая / -ое):

1. Я себя чувствую *итальянской (instead of итальянкой) (respondent 7); 2. Ты не *итальянская, ты русская (instead of итальянка) (respondent 9); 3. Я думаю, что ты вообще *итальянский (instead of итальянец)

The only exception is the following example: 4. *Так жизнь! (instead of Такова) (respondent 4). Here the informant used the adverb instead of the demonstrative pronoun, probably calquing the Italian expression "Così è la vita!".

Conjugation and use of verbs

We already exposed some problems encountered in the use of verbs of motion and aspect; here we will describe the errors detected in the conjugation of verbs, regular and irregular, and the confusion that some speakers demonstrated by using some verbs rather than others, but we will analyze this problem in detail when we will talk about calques.

Overall we found 34 errors, in the speech of 5 respondents, not counting, of course, the ones which we decided to classify only as calques.

Some verbs were conjugated in the singular rather than in the plural:

1. Их родители не *хочет, чтобы они забыли язык (instead of хотят) (respondent 9); 2. У тебя спрашивает (fm) "Пожалуйста, переведи...", а если у тебя спрашивает (fm) "переведи..." (instead of спрашивают in both cases) (respondent 9); 3. Там *говорится принципиально (fm) по-русски (*говорится instead of говорят) (respondent 2). Talking about languages, this speaker uses several times the verb in the third person singular adding the reflexive ending "ся", which can only be used in some cases, but not in this. Some verbs, regular or irregular, were conjugated in a wrong way: 4. Здесь *ищат очень много *человек которые языки знают (*ищат instead of ищут) (respondent 7). The verb "искать" presents a consonant alternation (the morpheme -щ), but it belongs to the first conjugation, while the informant conjugated it as if it tobelonged the second; 5. Когда я пьяная всегда *слышут (что я русская)! (instead of слышно or слышат or слышится) (respondent 9). In this case the adverb "слышно" is more appropriate to express the impersonal form, rather than the verb in the third person plural. The respondent, however, not only did she use the verb, but, as opposed to the previous example, she conjugated it as if it belonged to the first conjugation, while it belongs to the second (the correct form of "слышать" in the third person plural would be "слышат ", and the informant used this form properly shortly later); 6. Конечно они не *сижут читать по-русски или писать на этом языке (instead of сядут) (respondent 9). This error needs to be analyzed under various aspects (in a sense we can say that it is "triple"): on the one hand the form "сижут" does not exist (the third person plural of the verb "сидеть" is "сидят"), on the other hand the informant refers to the verb "сидеть", which is imperfective, while this case requires the perfective, and indicates a state, ("be seated"), but here you need the verb indicating movement ("to sit down"), so "сесть", which is perfective and irregular (the third person plural is precisely "сядут"). Since Italian does not have all these nuances, it is likely that the informant used the first form which came to her mind, comjugating it in a more or less "regular" way (in fact the first person of the verb "сидеть" is "сижу" , so the use of "сижут" is in a certain way "justifiable"); 7. Я не знаю как *могу бы говорить (instead of мог бы) (respondent 2). The Russian conditional is formed by adding the particle "бы" to the past tense of the verb, whereas here the interviewee used the present (perhaps it is a "mistake" due to the similarity between the two forms "могу" and "мог"); 8. Не хотелось и не *будет хочется вернуться (instead of захочется) (respondent 8). The form "будет хочется" does not exist, therefore, to use an impersonal verb, in line with the previous one, the respondent should have used the perfective "захочется".

From the point of view of the verbs, some expressions from informant 8 were really interesting, as he committed several errors in the use of verbs with prefixes, sometimes even trying to correct himself, or on the other hand, he used the correct prefixes, but the wrong verbal root, albeit similar to the accurate one because of a phonetic similarity: 9. В *шести лет меня и моих сёстр *убрали от родителей (*убрали от instead of отобрали у родителей); 10. Когда Белорусь *осталась от России (instead of ушла); 11. Когда *остался... расстался советский союз (instead of расспался).

In many cases, the respondents considered as equivalent translating words pairs of verbs (Salmon, Mariani, 2008, 104) that share only a few meanings, thus applying pragmatic symmetry where there is not:

12. Не *знала разговаривать по-итальянски (instead of умела) (respondent 9).

Russian, in fact, like English, clearly distinguishes the idea of "to know", for which there is the verb "знать" (for example: я знаю русский язык or я знаю Антона), from "to be able to", "уметь" (я умею кататься на лыжах); 13. Другую, среднюю сестру *называют Анну (instead of зовут Анна) (respondent 8). Similarly, for the verb "to be called", there is the form "называться", used for all titles (эта песня называется "Миллион алых роз") and the form "звать", which also has the meaning of "bear the name of" (моего брата зовут Александр); 14. Я *учусь русский язык (instead of изучаю) (respondent 4); 15. Они где *изучают? (instead of учатся) (respondent 2); 16. Сейчас, когда я изучал много русского языка (fm) [...] (instead of много занимался русским языком) (respondent 2). In Russian there are four forms for what Italian translates commonly with "studiare": summing up we can say that "изучать", transitive, indicates the subject, the field of study tõ which we dedicate ourselves in depth (я изучаю языки), "учить", also transitive and often used instead of "изучать" in the spoken language, refers to the study of something small (он учит текст наизусть) 30 , "учиться", reflexive, refers to the "student's profession" (Мария ещё учится), the place of study (я учусь на факультете экономики), or more generally means "to learn" (take, for example, the famous saying "век живи -век учись"), while "заниматься" is used with meaning of "to do homework" (сегодня занимаюсь лингвистикой or каждый день занимаюсь в библиотеке); 17. Я *учила итальянский язык быстро […], я говорила только поитальянски (instead of выучила) (respondent 4). This class of verbs is furtherly complicated by the fact that the prefixes preceding the verb "учить", which make it perfective, change its meaning. This is the case of "выучить", followed by a noun, which takes on the meaning of "to learn" (он выучил русский язык в течение трёх лет), but also of "научиться", followed by a verb, also "to learn" (она научилась плавать в детстве). However, this can also be considered an error of aspect; 18. *Нам изучали только *русский язык (instead of Нас учили только русскому языку) (respondent 8). When on the contrary "учить" (and its perfective "научить") takes on the meaning of "to teach", in Russian it requires a construction opposite to the Italian one (the accusative for the person to whom something is taught and the dative for the subject of teaching, for example: учитель учит учеников грамматике). In this case the informant, not only does she make a mistake in the construction, using the Italian one, but she also confuses the verb "учить" with "изучать". Also all these examples can be considered calques, as well as other errors in the use of verbs such as:

19. Я не *следую политикой (instead of слежу за политикой) (respondent 4). Also in this case the Italian verb "seguire" requires two different verbs in Russian depending on the semantic relation: "следовать", which usually is followed by the dative, and "следить", followed by "+ за instrumental". Here the informant "fuses" the two structures, which in some cases are equivalent (eg "следовать моде" or "следить за модой"), in others, as in this example, have two different meanings: "следить за политикой" means to be interested in political events and keep up to date on this topic (which is what, presumably, the informant meant), while "следовать политике" means to obey the rules politics imposes; 20. В два года *неможно изучать ничего... *научить (*неможно instead of невозможно and *научить instead of научиться 31 ) (respondent 8). In this case probably the respondent thought of the Italian phrase "non si può" and he translated it literally, forgetting that in Russian the adverb "можно" can not be transforemd into negative (the negative of "можно" is "нельзя") 32 . 21. *Начали всякие проблемы (instead of Начались) (respondent 8). In Russian, the verb "to start", and "to continue" and "to finish" can be both transitive (начать), when followed by another verb (он начал говорить) or by a direct object (она начала школу) and reflexive (начаться), if, as in this case, it is the subject itself that "gets started" (первая мировая война началась в 1914 году); 22. Если я должна говорить по-английски, не думаю что *решаю сказать что-то (instead of смогу) (respondent 4). We assume that here the informant thought of the Italian verb "riesca" (or "riuscirei"), which is phonetically very similar to "решаю" and confused it with "смогу" (which corresponds, in fact, to the Italian "riesca"); 23. *Спишу почти один (fm) раз в неделю со всеми, *спишу их (instead of списываюсь. For the second example both the form списываюсь с ними and пишу им are possible) (respondent 7). Here the informant tries to use the verb "писать" with the prefix "с", but in this case it becomes reflexive and it also modifies the root (as it often happens in Russian when adding prefixes to verbs). In the first case it would be correct to use the form "списываюсь", as the respondent after uses the preposition "с" followed by the instrumental (со всеми), in the second, where the informant tries to correct herself, she could 32 The form "неможно" can be considered an expression of просторечие, ie of a low level of education, but also of child language. In ancient times it was used and considered an equivalent of "нельзя" (it was used also by Peter the Great and Pushkin).

have used both the proposed forms, either "списываюсь с ними", or simply "пишу им" (however she used the wrong case, since "писать" wants the dative, while she used the accusative).

Introduction of "indefinite articles"

As it is known, Russian language has no articles. Four speakers included in their speech either numeral, which in Russian would have a very different meaning, or the pronoun "такой".

The following examples can not be considered real errors (they have been classified as marked sentences) or просторечие (especially the phrases that contain "один"), or above all calques from Italian:

1. У нас один (fm) диктатор, *который его называют Лукашенко (instead of У нас диктатор) (respondent 8). In this case the meaning of the numeral "один" is "only one" (while most likely the respondent just wanted to say "We have a dictator") ; 2. Только один раз я проговорил с одним (fm) *русском (instead of Только один раз я проговорил с русским) (respondent 2); 3. Я знаю такую (fm) учительницу, которая просит *тридцать или *сорок Other examples, however, show the use of the numeral "один" even in cases where the indefinite pronoun is required, which in Italian is always expressed with "uno": 5. Они меня видели (fm) больше как *один из Италии (instead of как когонибудь из Италии) (respondent 2); 6. Они почти (fm) меня *увидят как *итальянец, *лучше чем как *один из Калабрии вообще (*один из Калабрии instead of кого-нибудь из Калабрии) (respondent 2); or in some cases where a demonstrative pronoun would be required: 7. Для ребёнка, когда он вырастет, будет его выбор *если *выбирать *один или другой язык (*один или другой instead of тот или другой) (respondent 9). Also these examples, as the previous ones, can be considered as calques from Italian.~

Loss of "свой" and "себя"

Half the respondents rarely used the possessive adjective "свой", that is, instead, essential in Russian.

In some cases, as in the following examples, the use of the possessive "personal" instead of "свой" is accepted, even though perceived as "not really Russian" (a native speaker, in short, would use "свой"):

1. Я люблю мою (fm) землю (it would be better saying свою) (respondent 8); 2. Мне было очень трудно разговаривать на моём (fm) языке, (it would be better saying своём) (respondent 9), in other cases the possessive adjective is redundant: 3. На итальянском я поправляю мою (fm) маму (probably a native speaker wouldn't use the possessive at all but would say На итальянском я поправляю маму) (respondent 6), in other cases, instead, it is just wrong: 4. Они разговаривают на *их родном языке (instead of своём) (respondent 9). However, a particular example is: 5. *Свои родители говорят по-русски? (instead of Ёе) (respondent 2). This case is, in a way, seems to be going "against the tide" compared to the previous examples; probably the informant, thinking for homophony of the Italian "i suoi", sought a form in Russian that corresponded to the characteristics of the nominative plural possessive adjective. She didn't remember, though, that in cases in Russian the adjective does not agree the noun that follows it, but it needs the invariable form "её" (or possibly "его" if referred to a male subject). Similarly, one informant used several times the personal pronoun "меня" instead of "себя" 6. Я считаю *меня итальянской (instead of себя итальянкой) (respondent 4); 7. Они считают себя русские, но я считаю *меня итальянской (instead of себя итальянкой) (respondent 4). In these cases we can assume that the error is due to the interference of the Italian pronoun "mi", which resembles, phonetically, to the Russian "меня". In the last instance, in fact, the informant makes a mistake only in the second sentence, while in the first one she correctly uses "себя" (perhaps aided by the fact that it resembles phonetically to the Italian "si").

In general we can assume that, the limited use of both "свой" and "себя", are due to the fact that the use of these words in Russian is different and more frequent compared to the use of the possessive adjective "proprio" and the pronoun "se (stesso)."

Other grammar mistakes

We noticed, moreover, a slight confusion in the use of the particles "так" and "как", that, depending on how they are combined, build various conjunctions and expressions:

1. Я *учила итальянский язык быстро, *так как когда я вернулась в *России, прежде чем мои итальянские родители mi adottassero (cm/cs), я говорила только по-итальянски (*так как instead of так что) (respondent 4); 2. Сейчас, *как я работаю, больше не могу часто *ехать (*как instead of так как) (respondent 7). Probably in this case the informant just forgot a part of the expression; 3. Плохая политика у нас, как в России, *как в Белоруси (*как instead of так и) (respondent 8).

Finally, we conclude the analysis of the grammatical sphere by presenting a series of grammatical errors not easily "classified" and "grouped". 6. *Заплатили никакого гида (instead of Не заплатили ни за какого гида) (respondent 9); 7. Если у тебя нет денег, они не рады (fm), если у тебя нет такой (fm) красивой машины, они не рады (fm) (respondent 9). In this sentence there is no consistency between the subjects of the various sentences: given that the topic dealt with people in general, the respondent could have said both "Если у тебя нет денег, ты не рад, если у тебя нет красивой машины, ты не рад", and "Если у них нет денег, они не рады, если у них нет красивой машины, они не рады".

In some cases, such as the following, errors are probably caused by interference from Italian: 8. Было бы тоже интересно говорить с кем-то, *что знает русский лучше чем я (instead of кто) (respondent 2). Here almost certainly the informant thought of the Italian relative pronoun "che", which, unlike in Russian, does not change, whether it refers to something animate or inanimate; 9. Потом я приехал с моей мамой (fm) *здесь в *Италии (я приехал с моей мамой instead of мы приехали с мамой) (respondent 2). This can also be considered a calque; 10. Она *некоторые раз приезжала в *Калабрии (*некоторые раз instead of несколько раз opr иногда) (respondent 2); 11. *Некоторые раз это чуть-чуть тяжело (instead of Иногда) (respondent 2).

These last two examples can also be considered calques of the Italian phrase "qualche volta"; It was noticed, moreover, the excessive omission of the subject pronoun: 12. И когда она приезжает сюда, ты ещё должна spendere... (cm/cs) платить (fm) много денег, чтобы научилась (fm) и бла бла бла (научилась instead of она научилась) (respondent 9), and some confusion in the use of the conjunctions "и" and "а", corresponding both to the Italian "e": 34 13. Я *делала первый класс, и (fm) потом приехала сюда в *Италии (instead of Я закончила первый класс, а потом приехала сюда в Италию) (respondent 4). 34 The conjunction "а" can correspond also to the Italian "ma".

Syntactic mistakes

Here briefly some syntactic errors, found in the speech of 5 speakers, but we will resume this topic in paragraph 2.9, when dealing with calques:

1. Только один раз я проговорил с одним (fm) *русском, который *он изучал языки тоже (fm); (instead of который тоже изучал языки) (respondent 2); 2. Потом мой папа приехал в *Италии […] мы поехали тоже (fm) в *Калабрии (instead of мы тоже поехали в Калабрию) (respondent 2); 3. Когда *у него было, не знаю, когда он ещё не очень был взрослый (fm) (instead of когда он ещё не был очень взрослый) (respondent 2); 4. Я не могу говорить хорошо (fm) на украинском (instead of Я не могу хорошо говорить на украинском) (respondent 6); 5. Больше нового ничего (fm) не случилось... не случается (instead of Больше ничего нового не случилось) (respondent 8).

All these examples can not be considered as real errors, but examples of просторечие or marked register. Therefore, it is hard to say if in these cases the erosion of some syntactic rules is due to the interference from Italian (as it will be for sure for the syntactic calques which will be discussed shortly) or whether it is simply the expression of a "rough" and not refined language.

Lexical mistakes

At the lexical level we found, in the speech of 6 informants, some confusion in the use of certain words, together with some "fantasy" in the invention of nouns which do not exist in Russian, (at the level of vocabulary the errors detected are 13):

1. Учусь в университете *истранных языков... *факультете истранных языков (*истранных instead of иностранных) (respondent 7); 2. В Молдавии *ещё не *говорится по-русски (*ещё instead of больше or уже) 35 (respondent 2); 35 Out of context, the use of the adverb "ещё" might seem normal. However, at that time the topic of conversation was the language situation in Moldova, where now the use of Russian is decreasing (which is why it can be assumed that almost certainly the informant meant "больше" or "уже"). свою письменную речь or своё письмо) (respondent 2); 4. *Мой проблема ни *акцент, ни *говорение […] (*говорение instead of в разговорной речи) (respondent 2). The nouns "писание" and "говорение" exist in the Russian language, but indicate different concepts than the communicative intentions of the informant in this context 36 . Perhaps he was thinking of the Italian terms "scritto" and "parlato" and associated them, from the verbs "писать" and "говорить" and the rules of formation of the Russian nouns, to "писание" and "говорение" (we could call these examples of lexical calques); 5. Это учение (fm) *главный для работы и здесь *ищат очень много *человек которые языки знают (fm) (instead of Знать русский языкважно or Этот предмет -важен) (respondent 7). Similarly, the noun "учение" is not suitable for this context (the expression is, in fact, marked); 6. Если это китайцы [...] германцы (fm) [...] (instead of немцы) (respondent 8). In this case we can consider as cause of this error the influence of Italian (as in Italian, as well as in Russian, the noun "германцы" means the population of the ancient Germans, or it denotes a strong irony). The only plausible explanation could be that the informant formed, as in some cases, the noun of nationality from the name of the State (so that he though as follows: if the inhabitants of Китай are китайцы, of Италия, итальянцы, etc., so the inhabitants of Германия will германцы); 7. Если сперва *фамилии были богатые […] (instead of семьи) (respondent 8). Here the interference of the Italian word "famiglia" is obvious (it is the only time this speaker used the word "фамилия" instead of "семья". For this example, we can speak of lexical calque); 8. В России я была в детском саду (fm), cioè orfanotrofio (cm/cs) (в детском саду instead of в детском доме) (respondent 4); 9. Это 350 километров *северо-запады Москвы... sud-est (cm/cs) (instead of на юго-востоке) (respondent 9); 10. Думаю что, almeno tu, (cm/cs) ты, точно ты (fm) должна бы говорить с ним только на родном языке (ты, точно ты instead of по крайней мере 36 They relate to the action of speaking and writing respectively. The noun "писание" also indicates the Sacred Scripture (Священное писание).~ ты oppure хоть бы ты) (respondent 9). In these three examples, thanks to the switching into Italian, we can understand that the informants wanted to express something else, in fact they switched into Italian because they realized that something is "not working" in what they are saying (their inner ear, therefore, is efficient to make at least a "self-criticism");

In the following examples the subjects "coined" nonexistent nouns, while still following the rules of formation of Russian language:

Calques

Analyzing what we called "grammar", "lexical" and "syntactic" mistakes, we repeatedly pointed out that many of them could be considered calques, as we could notice a clear "transfer" from a word, phrase construction or Italian.

However, some expressions, which we decided to consider actual "calques", seemed, more than others, a real "distortion" due to interference. Even Perotto, in her analysis Lingua e identità dell'immigrazione russofona in Italia (2009), found in the speech of her sample various calques, divided into lexical, syntactic and semantic.

Lexical calques are popular, as they involve one of the most "permeable" field of the language, ie precisely vocabulary and they occur when the use of a word or a phrase [...] gets shaped, more or less fully, from the use that another language already makes of the corresponding linguistic forms (TV, 2012) 37 Eved though she didn't remember the word "соотвечественники", then suggested to him by the interviewer, this informant proved to have a working inner ear rather: she soon realized that "одностранники" was probably not the right word, in fact she stopped and repeated the term watching the interviewer looking for a confirmation or aid. and they normally appear in compound words, for which the individual parts of a foreign compound word are transleted literally.

Similarly to lexical calques, syntactic calques occur when a syntactic construct of a language is transferred into another (ibid).

Semantic calques, instead, are produced when a word in a language, which shares one or more semantic meanings (ie one or more acceptations) with a word of another language, takes on by imitation other semantic features that it previously did not have (ibid).

These definitions are used in linguistics to define the ordinary and consolidated phenomenon of transferring terms from one language to another, which leads to the formation of new words or, in other cases, the diffusion of new meanings.

As Perotto (2009), we decided to apply this classification (which we already mentioned in the previous paragraphs before giving its definition) also to the incorrect terms, expressions and constructions encountered during the interviews.

We point out that it was often difficult to decide whether to consider a calque as lexical or semantic, since many times both interpretations would have been possible (in fact, we could have even merged the two categories). It was decided to consider as lexical calques the cases involving single words, invented from an Italian lexeme, or idioms consisting of two words, where there is clearly a literal translation of every single part. The cases in which, however, there is a clear extension of a meaning of an Italian lexeme to a Russian word or expression, which only shares in part the meaning with Italian, were considered semantic calques (although they also prove a literal translation from Italian).

Lexical calques

As examples of lexical calques we report:

1. Если хочешь есть один (fm) *сит […] (*сит instead of сайт) (respondent 8); 2. Она ещё изучает *ебрайское (instead of еврейский) (respondent 2); 3. Там *приватические учителя (instead of частные or, maximum, приватные, marked form, though) (respondent 9).; 4. Испанский difficilmente lo dimenticano, (cm/cs) очень *симилярный... no! (cm/cs) (*симилярный instead of сходный) (respondent 9) 38 ; 5. Там *говорится принципиально (fm) по-русски (принципиально instead of в основном) (respondent 2). In Russian "принципиально" means "in principal", while almost certainly the informant had in mind the Italian word "principalmente"; 6. Она приехала раньше меня, на год с половиной (fm) (instead of на полтора года) (respondent 6); 7. Но это больше раз так (fm) (instead of Но в основном это так) (respondent 2). This is a calque from the expression "la maggior parte delle volte"; 8. Один раз (fm) денег нету, во второй раз (fm) ты работаешь (instead of Иногда денег нету, иногда ты работаешь) (respondent 6). This sentence is most likely a calque from the expression "una volta… l'altra…"; 9. Спасибо Богу (fm) в этом лицее, в Деледде, я изучала русский язык (instead of Слава Богу) (respondent 9); 10. Немножко начала вспоминать мой родной язык, но всегда с трудностью (fm) (instead of с трудом) (respondent 9).

Syntactic calques

At the level of the sentence, on the other hand, we found the following syntactic calques:

1. Если ты *слышишь говорить мою маму на итальянском, ты не сразу *думаешь: "Ах это русская женщина" (instead of Если ты услышишь как моя мама говорит на итальянском, ты не сразу подумаешь: "Ах это русская женщина") (respondent 6); 2. Редко услышишь *их разговаривать по-итальянски tra di loro (cm/cs) (instead of Редко услышишь, что они разговаривают по-итальянски между собой) (respondent 9); 3. Работаю в ресторане как официантка (fm) (instead of официанткой) (respondent 7); 4. *Мой проблема ни *акцент, ни *говорение [...] (instead of Моя проблема ни в акценте, ни в разговорной речи) (respondent 2); 5. Я думаю, что сын должен всё время слушать... слышать два языка, *нет один день я с ним буду разговаривать по-итальянски, и (fm) *один день по-русскими (instead of не то что бы сегодня я с ним буду разговаривать по-итальянски, а завтра по-русски) (respondent 9). We assume that here the informant meant "Non che un giorno gli parlerò in italiano e l'altro in russo"; 6. Когда бывает, если что-нибудь по... какая (fm) новинка, я слушаю, только до *пять лет отсюда, от туда ... в эти годы не случилось *ничего не нового (instead of Когда бывает, если что-нибудь по... какая-то новинка, я слушаю, только уже пять лет... в эти годы не случилось ничего нового) (respondent 8). This sentence is not very clear; probably the informant did not know how to express the expression "da cinque anni a questa parte"; 7. Для ребёнка, когда он вырастет, будет его выбор, *если *выбирать *один или другой язык (instead of Для ребёнка, когда он вырастет, будет его выбор, выбрать тот или другой язык) (respondent 9); 8. Не знаю, *если это верно (instead of Не знаю, верно ли это) (respondent 9); 9. Потом он выберет, *если продолжать учить его или нет (instead of Потом он выберет, продолжать ли учить его или нет) (respondent 9); 10. Я не знаю, *если смогу вернуться (instead of Я не знаю, смогу ли вернуться) (respondent 6). We highlight that the error in the indirect interrogative was committed by three subjects and several times.

Semantic calques

The most examples regard undoubtedly semantic calques. When talking about verbs we noticed that some sentences calqued the pragmatics of asymmetric verbs in Italian and Russian (for example the confusion between "знать" and "уметь", between "следить" and "следовать", between "учить", "учиться "and" изучать ", which are typical errors of learners of Russian or speakers of Russian as a L2).

Although the following examples attest the tendency to consider a pair of words, that is a dictionary match, as equivalent translating words , when actually it is not always the case (Salmon, Mariani, 2008, 104):

~ 54 ~ 1. Я тихо (fm) теряю *русский речь (тихо instead of по-тихоненьку or медленно) (respondent 5); 2. Не вижу русский телевизор (fm) (instead of смотрю русского телевидения) (respondent 4). These two cases can be considered as phenomena of просторечие; 3. Я знаю тоже (fm) русских друзей, но я *с ними не вижу (я *с ними не вижу instead of я с ними не встречаюсь or я с ними не общаюсь) (respondent 4); 4. Друзья есть, но я их не вижу часто (fm) (instead of я не часто встречаюсь с ними, if we consider this as a semantic calque, or я их часто не вижу, if we consider it as syntactic) (respondent 7); 5. У меня есть ещё семья в России, но я их не слушаю (fm) (instead of я с ними не общаюсь) (respondent 4); 6. Уже (fm) удочерение (fm) очень *дорогой (instead of Усыновление по себе очень дорогое) (respondent 9). 39 Here, most likely the respondent meant "Già l'adozione è molto cara"; 7. Они *почти меня *увидят как *итальянец, *лучше чем как один из 39 In Russian there are two nouns and two verbs to indicate adoption, according to the adoption of a boy (noun: "усыновление", verb: "усыновить") or a girl ("удочерение" and "удочерить"). Generally speaking the masculine form is used. 40 We will get back to the terms and realia related to the school world in paragraph 2.11.~ The most commonly encountered calque was from the verb "fare", far more widespread in Italian than "делать" in Russian; 11. Я читаю несколько слов и скажу (fm) моим друзьям «Мы уже *делали эти *слово» (instead of «Мы уже видели эти слова») (respondent 4); 12. Я *делала первый класс, и (fm) потом приехала сюда в *Италии, *сделала здесь elementari (cm/cs) и сделала (fm) всё (instead of Я закончила первый класс, а потом приехала сюда в Италию, и здесь закончила начальную школу и закончила всё) (respondent 4); 13. Если ты хочешь *сделать вопрос (instead of задать вопрос) (respondent 4); 14. Поеду в *Англии, в *Лондоне, с подругой *два года, чтобы учить английский язык, и (fm) второй год чтобы *делать выcшее научение, второй получение (*делать выcшее научение, второй получение instead of закончить высшее образование, второе высшее образование) (respondent 4). In this case the informant also commits strictly lexical errors, as she uses nonexistent nouns (научение), whose meaning is not relevant in this context (получение), and of agreement; 15. Они не думали, что потом буду *делать языки (instead of изучать языки) (respondent 2); 16. Я *делаю теннис и хожу бегать (fm) (instead of Я играю в теннис / Я занимаюсь теннисом и бегаю) (respondent 7) 17. Я буду стараться всю мою (fm) жизнь *делать так чтобы не потерять (fm) [русский язык] (instead of Я всю жизнь буду стараться не забыть [русский язык]) (respondent 6). We assume that the expression, unnecessary, "делать так чтобы" is a calque from the expression "fare in modo di". Moreover, given the "not Russian" order of the elements of the sentence, we can consider this case also a syntactic calque. 41 41 In Russia, as early as the eighteenth century and even more in the next century, given the importance that French language had taken on an international level, especially in European courts, it began to spread what had been called "Gallicisms" (which included the misuse of the verb "делать", calquing the French "faire"). Even in the Russian Empire, nobles used to study, read, speak French and boasted of knowing this language. The same Tolstoy introduced various Gallicisms and calques from French (also syntactic ones), frequently using the word "делать" (Mal'kova, 2009).

Marked sentences

We already talked about words and expressions f-marked for the context of use. Here are more examples of f-markedness, that concern, however, whole sentences. It is not about grammar, lexical or morphosyntactic errors, but entire structures that in themselves does not have a precise meaning, which, however, can be reconstructed from the context: 1. Если она захочет ещё изучать эти языки, у неё будет большое знание (fm) (instead of она будет очень хорошо учиться) (respondent 2); 2. Много раз мне бывает что я хочу что-то сказать по-итальянски, но мне приходит в голову только по-русски […], и ещё тоже наоборот (fm) (instead of и то же самое наоборот) (respondent 2); 3. Не знаю как действительно карта, у меня плохая география России (fm) (instead of я плохо знаю географию России) (respondent 8); 4. Половина моей крови (fm) это albanese (cm/cs), мой отец был albanese (cm/cs) (Половина моей крови это albanese instead of Я наполовину албанка) (respondent 9); 5. Они не понимали, что я была madrelingua (cm/cs) (instead of носительница языка oppure что мой родной язык русский) (respondent 9). The term "носитель/-ница языка" got into Russian language only recently, so the expression "родной язык" is more common.

Examples of code-switching and code-mixing

All ten informants, even those who proved to have maintained a high level in oral expression in Russian, used mixed-language utterance, both consciously (code-

switching), or involuntary (code-mixing). 42

Given that not all scholars are unanimous about the difference between the two phenomena, we specify that here we consider as CS only the voluntary change of code, coplying with Osipova's definition, according to which CS is a real communication strategy, chosen to simplify the statement, to make it more understandable, or to establish a closer contact and greater "solidarity" in the relation with the interlocutory (Osipova, 199939, in Perotto, 2004, while we will consider CM as the unmotivated transfer from one language to another (Belikov, Krysin, 200131, in Perotto, 2009.

We thus classified as CS the cases in which the informant expressed the desire to switch somehow to Italian, mostly through a break or metalinguistic comments (eg "как сказать ...?", "Не знаю как сказать" "типа" or "how do you say ...?", etc.), or when he spoke in Italian simply because he knew to be understood by the interviewer, or also to ask for suggestions.

We considered as CM, however, the cases where the informants did not seem to have realized that they had actually "mixed" the two codes, or if they realized only afterwards (often, in fact, they repeated the same term, but in Russian, or they stopped and repeated it, smiling, in Italian, to show they were aware of their mistake).

Code-switching

After grammatical errors, CS is definitely the most common phenomenon that was found.

In most cases the respondents have resorted to Italian because they did not know a word in Russian:

1. Мои итальянские родители mi hanno adottato (cs) (in Russian удочерили меня) (respondent 4); 2. Для меня это трудно тоже (fm) писать потому что tastiera (cs) порусски... (fm) (in Russian клавиатура на русском... or русская клавиатура...) (respondent 4); 3. Мой брат более diligente (cs), переведи diligente (cs) (in Russian старательный) (respondent 5); 4. Потом один, как сказать... olandese (cs) (in Russian голландец) (respondent 8); 5. Ребёнок очень скоро начинает... come si dice "apprendere"? (in Russian учиться) (respondent 9); or they couldn't express a whole expression: 6. Если у тебя нет отсрочки, то ты считаешься, типа, не дезертир, но renitente alla leva (cs) (in Russian уклоняющимся от призыва на военную службу) (respondent 1); 7. Работал на convegno nazionale del sincro… Non so come si dice "convegno" in russo (cs). Семинар может быть (in Russian на националном симпозиуме синкро) (respondent 1); 8. Сразу было легко, потому что erano appena arrivati in Italia (cs) (in Russian они только что приехали в Италию) (respondent 9); 9. Испанский difficilmente lo dimenticano (cs) (in Russian Вряд ли они забудут испанский) (respondent 9); 10. Поэтому знаю как трудно, и для родителей... economicamente parlando (cs) (in Russian в экомическом плане) (respondent 9), or also because their inner ear suggested that the Russian expression used probably was not accurate and therefore they preferred to point it out in Italian:

11. В России я была в детском саду (fm), cioè orfanotrofio (cs) (in Russian в детском доме) (respondent 2); 12. Дети, когда они маленькие, *в неделе... come si dice "nell'arco della settimana"? (cs) не читают по-русски (in Russian в течение недели) (respondent 9) 13. Это 350 километров *северо-запады Москвы... sud-est (cs) (in Russian на юго-востоке) (respondent 9). Two informants, who didn't show any particular difficulty in speaking in Russian, made use of what Perotto called emotional CS, used to express strong emotions, conflict, curses (Perotto, 2009, 116): 14. Я проработал три месяца, culo assurdo! (cs) (in Russian we could find an expression with a similar f-markedness, for example я заебался!) (respondent 1); 15. Я проходил собесодование, но мне дали calcio in culo (cs) (in Russian пинка под зад) (respondent 5).

In cases where the informants were referring to particular realia of Italian culture, they made use of compensational CS 43 ,that is used to 43 Perotto (2009) also speaks of the necessity loans, taking the definition of Gusmani (1981, 151), and of cultural borrowings (Myers-Scotton, 1993, 169). ~ 59 ~ convey an expression that does not exist in one of two languages (ibid, 111): 16. Работал в порту operativo porto (cs) (in Russian разнорабочим) (respondent 1); 17. Сейчас у нас всех поставили в cassa integrazione (cs) (in Russian we could say Мы все были переведены на госпособие, although using an expression which doesn't fully express the semantic meaning of "cassa integrazione") (respondent 5).

In these cases there is no symmetry between the two languages, as some concepts, situations, phenomena (Italian realia, indeed) are not present in Russian reality, and therefore there are not even equivalents translating words; the informants probably came in contact with these terms in Italy and therefore they must have used them almost exclusively in Italian (except when explaining their meaning to a compatriot) 44 . The subjects didn't make an effort to find a similar expression in Russian, not only because they were aware that the interviewer would have understood them anyway, and maybe even better, but also probably because they realized, even if unconsciously, to have to deal with " fully Italian concepts." ~ 60 ~ 23. Год спустя я уже сдавал экзамен terza media (cs), пошёл в колледж (fm), superiori (cs) (in Russian средней школы and в лицей, respectively, or в училище) (respondent 10); 24. Я даже получил там trenta e lode (cs) по английскому (in Russian тридцать с плюсом) (respondent 1); 25. Моя оценка финального экзамена была «buono» (cs) (in Russian «хорошо») (respondent 6). The Italian school system is organized differently from the Russian one, whose compulsory education consists of a single cycle, going from the first to the eleventh class.

In Russian there are of course terms indicating the various Italian and European education institutions (начальная школа, средняя школа and лицей, or училище, depending on whether it's a high school or a vocational school). However, it is likely that the informants, even those who knew these words, didn't automate their use in Russian, having used them almost certainly only in Italian.

Similarly, the evaluation system of schools and universities in Russia is from 0 to 5, so marks like "trenta e lode" or "buono", although they can be translated as "тридцать с плюсом" and "хорошо", do not belong the imagery of a Russianspeaker.

In some cases the terms that have a certain morphological similarity with Russian were integrated to its language system, and they were processed as Russians nouns and declined as such (Perotto, 2009, 111): 45 26. Получил borsu (cs) Эрасмус (in Russian стипендию Эрасмус) (respondent 1); 27. В mediach (cs) мне сначало не нравилось (in Russian В средней школе) (respondent 3).

Finally, it was curious to notice the frequent use of English words by an informant recently returned from his Erasmus programme in England:

28. Там государственные finance (cs) (in Russian денежные средства) (respondent 1); 29. Я могу поехать работать в Москву [...] чисто для experience (cs) (in Russian для опыта) (respondent 1).

Code-mixing

The involuntary transfer to Italian was certainly less frequent than CS and it concerned only single words, but never whole phrases or sentences:

1. Они чуть чуть хуже говорят по-итальянски, vero (cm)? (in Russian правильно or верно, an adverb which, phonetically, resembles to the Italian "vero") (respondent 2); 2. Я живу в Рекко, vicino (cm) ... *Генуя (in Russian около) (respondent 7); 3. Не хотелось и не *будет хочется вернуться, потому что gover... (cm) как сказать, плохая политика (in Russian правительство) (respondent 8); 4. Мой русский очень плохой... E poi, (cm) я помню что […] (in Russian А потом) (respondent 8).

Often CM was found in words, very frequent in Italian, which, though, do not have a precise equivalent translating word in Russian:

Начала comunque (cm) [университет], чтобы не забыть этот язык (in

Russian we could say на самом деле) (respondent 9); 6. Если я пишу и magari (cm) не помню какое-нибудь слово […] (in Russian we could say может быть) (respondent 9); 7. Надо знать английский язык очень, ma (cm) очень хорошо, ma (cm) я это (fm) не знаю (in Russian we would say Надо знать английский язык очень, очень хорошо, но я его не знаю) (respondent 4). In this case, the first "ma" is an intensifier that is used in Italian, but not in Russian, while the latter corresponds to the adversative conjunction "но".

A particular and repeated display of CM was the inclusion of the so-called слова-паразиты in Italian, ie those words, superfluous and lacking of a precise meaning, considered by Russian linguistics as "parasites", because they are inserted in the speech in an uncontrolled and unconscious way, and often obstructing thẽ comprehension and worsening the quality, and that unlikely the subject will be able to eliminate 46 (ŠŽ, 2007): 8. Она, va bè, (cm) я ходила в школу (in Russian we could insert the word, паразитное too, ну) (respondent 6); 9. Mah (cm), мы общаемся по *фейсбуке (in Russian we could insert the паразитные words Ну вот) (respondent 8); 10. Только два года была в университете e bom (cm) (in Russian и всё) (respondent 9); 11. Эти учителя очень дорогие, cioè (cm), дорогие, много денег просят (in Russian we could say в смысле) (respondent 9).

Doubts, self-correcting and hypercorrection

In general, during the interviews many informants expressed some understandable hesitation in their speech: very often they had breaks, repeating the sentences said to here how they "sounded", they asked for suggestions to the interviewer and they corrected themselves.

Despite their mistakes and the phenomena of CS and CM, they proved to have preserved, as we have already pointed out several times, in some ways, their inner ear, which can recognize some of the expressions used incorrectly 47 .

In some cases, the informants corrected themselves, understanding that they made a mistake: 48 1. Я решила учить русский язык чтобы написать (fm) *моих… моим родителям (respondent 4); 2. Мне очень хочется вернуться в *России, чтобы видеть... *мои родные люди, *мои родители... Если они уже (fm) живы... ещё живы (respondent 4); 3. Мне очень хочется вернуться *там... туда (respondent 4); 4. *Пойду... Поеду в *Англии (respondent 4); 5. Она тоже […] иногда не помнит (fm), не вспоминает слово (respondent 7); 6. Немножко начала *вспомнить... вспоминать мой родной язык (respondent 9); 7. *Две... два языка всегда нужны (respondent 9); 8. Всё время разговаривают с итальянскими *детями... детьми (respondent 9); 9. Из Южной Америки не удоч... не удочеряют (fm), не усиновляют (respondent 9); 10. Потом они в *фамилии... в семье разговаривают на *их родном языке (respondent 9).

In some examples the subjects had a doubt, even though they had expressed the idea in a totally correct way: 49 11. Меня позвали щас быть свидетелем на свадьбе моего лучшего друга, он 48 In these cases we do not present the correct form, as it is already expressed by the informants themselves. We will highlight it in bold italics, while the initial incorrect form will be presented, as usual, in bold, preceded by an asterisk or followed by (fm).

49 Here we will highlight in bold the wrong form, preceded with an asterisk, that informants have added at a later time, and then repeat the right one.

50 Here we highlight in bold italics both correct forms pronounced dall'informante.~ 15. Я думаю, что сын должен всё время слушать... слышать два языка (respondent 9); 16. Только с мамой разговаривала на русском... по-русски (respondent 9); 17. И в Италии очень мало учителей по русскому языку... *на русском языке... русского языка (respondent 9); 18. Эти учители очень дорогие, cioè (cm), дорогие… много денег *прошают… просят (respondent 9). Here the informant unnecessarily precise the meaning of the adjective "дорогой": as in Italian, also in Russian, in fact, this adjective bears the meaning of "dear, beloved" and "expensive", and it is also used to indicate something or someone "who charges a lot". In the second case, however, the informant corrects the conjugation of the verb.

Prosody

An important distinctive element of each language is represented by prosody, as:

Every verbal production is characterized by a certain prosodic structure, focused on intonation and rhythmic patterns produced by individual words and their concatenation (Marini, Nocentini, 2003, 24).

The prosodic processing can be divided into emotional prosody, which refers to the pragmatic function of a sentence, and into linguistic prosody, which is focused on demarcation function of language, which is responsible for determining the linguistic mechanisms underlying the allocation of prosodemes, 51 both at a lexical level (lexical prosody) and at a phrasal level (phrasal prosody) (ibid).

The basis of both concepts are the notions of duration of emission, accent, intonation, rhythm and pause (ibid).

As pointed out in chapter 1.9, the prosody of a language may cause interference to the prosody of the other one.

Here we decided to examine, on the one hand, the errors regarding the word stress, on the other, intonation and pronunciation, which are after all the spheres in which prosodic interference occurs more explicitly.

51 Prosodemes are minimal prosodic units that have distinctive meaning (Ferranti, 2011, 1).

Although the branch of linguistics that studies pronunciation is not prosody, but phonetics, we decided to analyze these two aspects together, since the two disciplines are closely related to each other (pronunciation, in fact, deals with the articulation of sounds, while phonetics with the peculiarities of a language).

Word stress

In the speech of six respondents we found a total of 16 errors in the use of word stress within individual words.

Where a word is the same (or very similar) in Russian and Italian, especially when it comes to foreign loans, we can read the error as interference from Italian:

1. Бывает *на Интернет, да, я читаю (instead of в интернЕте) (respondent 8); 2. Мы общаемся по *фЕйсбуке (instead of по фейсбУку or на/в фейсбУке) (respondent 8); 3. У нас всегда дерутся в ПарламЕнте (instead of в ПарлАменте) (respondent 6).

In other cases, however, some errors were found, most likely due to the fact that in Russian the word accent is movable, so it often moves within the same word, depending on the grammatical cases, on the number, gender, person, etc.:

[Interviewer]: На ДнепрЕ?

[Informant]: Да, на ДнЕпре (instead of на ДнепрЕ) (respondent 2); 5. Как будто я там никогда не бЫл (instead of нЕ был) (respondent 2); 6. «Мы уже *делали эти слOва» (instead of словА) (respondent 4). This can also be an agreement mistake, as the informant may have thought of the word "слово" in the singular, which is pronounced, as she did, putting the accent on the first [o]; 7. Они имеют выгодУ (instead of вЫгоду) (respondent 6); 8. Мне не нравится Учиться (instead of учИться) (respondent 7); 9. ОбнОвлю мой русский паспорт (instead of обновлЮ) (respondent 9).

Given the complexity, and often irregularity, of the location of the word stress, errors of this kind are fairly common in Russia, especially among less educated people, and they are interpreted as an expression of просторечие. 52

Pronunciation and intonation

In the first chapter we referred to Fabbro's theories regarding the acquisition of a language perfect pronunciation, according to which this would be possible only when this language is acquired by the critical threshold, which the neuroscientist fixed at eight years, because after this age in the child the phenomenon of vocal accommodation would decrease (Fabbro, 2004, 86).

Although here we are analysing only the informants' speech in Russian, the interviewer was able to converse with them also in Italian and to see that only two maintained a foreign accent (shortly, it is understandable that they are Russians when speaking Italian), while all the others reached an almost flawless pronounciation, even those who arrived in this country after the critical age, which does not seem to confirm Fabbro's theory on the acquisition of a perfect pronunciation in L2.

About their speech in Russian, however, it was noticed that some respondents maintained a "native" pronounciation, while others seem to have been subjected to Italian interference, as they speak like "foreigners" and with an Italian accent.

The same goes for the prosodic level; basically those who maintained a perfect pronunciation, also kept the correct phrasal prosody and therefore they use the correct intonation (or actually the different intonations) in Russian.

We decided to give to each speaker, a "judgement", positive or negative, for their pronunciation and intonation in Russian, on the basis of the entire development of the interview (see table below).

Those who received a positive judgement proved to have kept native pronunciation and intonation, while a negative judgement was given to those who proved to speak with Italian accent and intonation. 53 53 Only informant 4 was given an "intermediate" judgement (the sign ±), as initially this informant spoke with a thick Italian accent, but in the course of the conversation her pronunciation and intonation aligned with those of a Russian annd she proved to speak "like a native".

As the table shows, it seems appropriate, and very curious, to make some remarks according to their "gender"; all male respondents proved to have maintained native pronunciation and intonation, while the female respondents, with the exception of informant 3, showed that they had lost some of these important language components because of Italian language.

Ĩ

Levels of bilingualism

While all the informants belonged to the category of early bilinguals 54 (the informant N°7, arrived in Italy when she was only one year old, so she can be defined as "very early bilingual "), the data presented in the previous chapter show that the maintenance of the L1 and language levels are different.

In this analysis we have not tested the competence in L2, but in L1, and it was interesting to notice that in most cases it was observe a '"erosion" of this language, and in some cases also remarkable.

To go back to the definitions of additive and subtractive bilingualism 55 presented in Chapter 1, some individuals have shown not only that one can acquire perfectly a L2 even after the critical threshold established by Fabbro, but above all that the acquisition of L2 has hindered the command of L1. In fact, only for two informant we have not detected any erosion at all due to the L2, while another two have maintained a good language level, although with various "imperfections". In general, however, Italian seems to have almost always "had the upper hand" over Russian, negatively affecting the speech in L1, both on the morphosyntactic and lexical and on the phonological levels.

Based on the fluency in Russian demonstrated during the interviews, the number of errors detected and the more or less maintained Russian pronunciation and intonation, for each informant it was decided to assess whether the level of 54 Remember that in this research we considered as critical threshold for the acquisition of L2 the generic age of puberty, unlike Fabbro, who established the critical threshold at 8 years.

55 It was decided to consider only the opposition additive / subtractive bilingualism and leave apart the others presented in Chapter 1, as this is the most appropriate to deal with the phenomenon of maintenance or loss of L1 (other classifications refer mostly to the acquisition of L2 and the level achieved in both languages).

bilingualism reached seems to be more of an additive or subtractive (see table below). 56 However, quite often he inserts into his speech "typically Russian" terms (eg "прямо", "вообще") or unusual (as "матёк" 57 ), which usually indicate a high fluency.

Informant 3: moved to Italy with her mother and said she has maintained strong contacts with her family and friends at home; perhaps this is the reason why her Russian was flawless, while speaking Italian she has got a foreign accent. This aspect seems to confirm the Fabbro's theory (2004, 83), which states that one of the most difficult aspect to acquire in L2 is a perfect pronunciation. Inforamant 7: lives in Italy with her Russian mother and an Italian father. It is the only case of very early bilingual among those analyzed, which shows, however, that Italian is really dominant over Russian, as she says, both on the morphosyntactic (various errors in Russian), and lexical (she does not many know words) and

phonological (her accent is Italian) levels.

57 The term "матёк" is rare and has the same meaning as the most common verb "материться". The dictionary definition is as follows: загибать / загнуть (согнуть) матька. Сквернословить (Akademik, 2010).

58 It was difficult to assess whether bilingualism of this informant can be considered additive or subtractive; for this subject, more than for the others, there would be the need of a more detailed analysis.

~ 71 ~

Informant 8: in Belarus he used to live in an orphanage and now he lives in

Italy with his adoptive parents. Like informant N°3, he has maintained a foreign accent when speaking in Italian and therefore he seems to confirm the Fabbro's theory of on the acquisition of a perfect pronunciation in L2. However he showed various lexical gaps and morphosyntactic imperfections in Russian, together with the use of some constructions typical of Belarusian (belarusisms).

Informant 9: fluent in Russian (perhaps due to her exuberant character, more than to her language skills), but she made several errors, mostly due to interferences with Italian. She constantly relied on the switching, since her vocabulary in Russian was not very rich. However, very often she knew that she had committed an error and, where she was able to, she was correcting herself.

Informant 10: moved to Italy with his mother and brother and his wife is

Russian. The level of oral expression in Russian, as in Italian, is perfect, (as a "native") from all points of view. About this informant we could definitely talk of balanced bilingual.

Russian language abroad

The language of Russian-speaking emigration living abroad, namely the contact between Russian and the national languages of these different countries, has been studied mainly in the countries where there is a large Russian community, or at least from one of the former Soviet Union republics.

In particular, the contact between Russian and French was analyzed in France and Canada 59 (Golubeva-Monatkina, 2001a, as well as the language of the Russian-speaking community in Finland, in Israel (Najdič, 2005, Ben-Rafael et al. , 2006 and especially in the United States and Germany. About the mixed-language code used by the Russian-speaking community in the United States, by far the most numerous and most investigated (Andrews, 1997Osipova, 1999Zemskaja, 2000, 2001a, 2001bZemskaja et al., 2001 In Canada, since the end of the nineteenth century there has been a large community of Doukhobory, an opposition movement to the Orthodox church, developed in the Russian Empire in the eighteenth century.

~ 72 ~ 1996,1998,2006,2007,2010), the term Runglish (example of Franglais and Spanglish) was coined.

Similarly, the language used by Russian immigrants in Germany 60 -reserched especially by Meng (2001), Goldbach (2005), Belenčikov (2006Belenčikov ( , 2007

-is called

Nemrus or Deutschrussisch or Quelia.

In Italy, similar analyses were conducted by the aforementioned Slavist Monica Perotto (2003Perotto ( , 2004Perotto ( , 2009Perotto ( , 2010.

Although all of these studies were carried out on samples of adults, so they do not belong to the category of early bilingual -with a few exceptions, mostly concerning the children of Russian-speaking immigrants raised in the United States (Gulida, 2004Polinsky, 1996 -it is possible to compare some of the results of this research both with the ones obtained by Perotto, and with numerous analyses on runglish.

Aspects in common with Perotto's studies

Also Perotto, in her research, found some of the calques from Italian which we found in this analysis, for example:

-The use of the noun *фамилия instead of семья; -The use of the verb "делать" in expressions which, in Russian, require other verbs, such as *делать душ instead of принять душ (similar to the case we presented of *сделать вопрос instead of задать вопрос, *делать языки instead of изучать языки, *делать теннис instead of играть в теннис, etc.); -The calque from the verb "sentirsi" (in its acceptation of "communicating with someone"), come *услышимся instead of созвонимся (similar to the sentence я их не слушаю instead of я с ними не общаюсь) (Perotto, 2009, 106).

Similarly Perotto found the syntactic calque of the indirect question, for example: он меня спросил, *если я завтра поеду к нему instead of он меня спросил, поеду ли завтра поеду к нему (ivi, 107).

There are many aspects in common between the informants of both researches and their use of CS by, which can be:

-emotional, namely the use of curse language and swearings, for example:

[...] Я по крайней мере выкрикну это vaffanculo! (ivi, 116-117) (similar to the sentence pronounced by our respondent 1: Я проработал три месяца, culo assurdo!); 61 -compensatory, to introduce concepts of Italian reality that are not present in Russian (the same way as terms referring to the Italian school system we mentioned before), as in the sentence: я уже шесть лет работаю в regole, contributi у меня оплачены (ivi, 111). In many cases Perotto's informants resorted to integrated compensatory CS, ie they declined Italian nouns assimilating them to Russian nouns (lexical adjustment) for example: дай ребёнку какую-нибудь merendinu or дай мне milju ("milja" meaning the old "one thousand lire", following the example of the noun двушка, the Russian currency worth two kopecks) (ibid, 113-114).

-expressive, which conveys irony, a desire for language game (ibid).

Perotto found such expressions as: давай пофумарим (from the Italian verb "fumare") or давай поманджарим (from "mangiare"). During the interviews presented in this research similar expressions were not detected, however one informant said to make use of similar constructions when speaking with his mother, adding a Russian ending to the root of an Italian lexeme (for example cominciarit').

-of complicity, used in order not to be understood by others (ibid, 118), There were no examples in this research, but the informants of both surveys talked about it.

On a grammatical level, however, the errors, albeit very rare, detected by Perotto concerned the nouns declension (for example the genitive plural: * врачов instead of врачей and * местов instead of мест) and the numerals (ibid, 89), errors that, instead, were found abundantly in the present research.

Aspects in common with runglish

Similarly, also the studies on runglish revealed language erosion detected in the speech of Russian speakers living in the United States, due to the contact with English.

In Polinsky's analysis, American Russian: An Endangered Language (1996), on second-generation immigrants 62 , arrived in the United States between the age 5 and 14, the researcher noticed a general erosion in the nouns declension, especially in the oblique cases, and reduced to the nominative and accusative (for example: Мой дедушка был *на мировая война instead of на мировой войне, Polynski, 1996, 34, or У неё нет *муж instead of мужа, ivi, 39).

Verb mistakes were also common (В Чикаго все *боится *преступники, instead of боятся преступников, ivi, 37), including those concerning the use of the aspect (Я никогда не *прочитал *та книга, instead of Я никогда не читал ту книгу, ivi, 53) and of the verbs of motion (В *август я *иду в Сиэтл instead of В августе я еду/поеду в Сиэтл, ivi, 56, or Мой дядя, он часто *приехал к нам в Бруклин instead of Мой дядя часто приезжал к нам в Бруклин, ivi, 57).

Again on a grammatical level, Polinsky noticed errors in gender and number agreement (В *университет книги *будет *дорого, instead of В университете книги будут дорогие/дорогими, ivi, 33), and a rare use of the possessive pronoun "свой" (Он говорит о *его делах только, instead of Он говорит только о своих делах).

The researcher also noticed various syntactical errors (*Дима и я, мы были вместе в школе, instead of Мы с Димой были вместе в школе, ivi 59), due to the influence of sentence structure in English. 62 Polinsky defines the language of the emigrants of the last wave of migration as "American Russian", while for the language of the speakers belonging to the first wave she uses the term "Emigré Russian".

Interferences and calques from English, similarly to the cases presented in this study, are undoubtedly evident in the following examples:

1. Я не имею машину instead of У меня нет машины (interference from the verb "to have") (ivi, 40); 2. *Они были в любви instead of Они любили друг друга (calque from the expression "to be in love") (ivi, 26); 3. Сегодня моя машина, она не *начиналась invece di У меня сегодня не заводилась машина (calque from the sentence "My car wouldn't start today") (ibidem); 4. Он не *умывал его лицо instead of Он не умывался (calque from the sentence "He didn't wash his face") (ivi, 45). Moreover, as the example 3 shows, as well as the following one and one of the sentences presented above about verbs of motion, the speakers of American Russian overuse and misuse the subject pronoun, which is almost always expressed in English, while our informants, instead, omitted it even too frequently, because of the influence of Italianlanguage.

As expected, Polinsky also found several cases of CS and CM, which the scholar considered mostly a lack of vocabulary rather than language strategies.

It is interesting to notice that also her informants inserted English terms in their speech to talk about their experiences at school and made mistakes in the use of verbs that carry the meaning of "to study", for example: -Моя сестра, она *учит в elementary *школа instead of Моя сестра учится в начальной школе (ivi, 24); -Всегда я получала Аs *от тот профессор instead of Я всегда получала пятёрки у того профессора (ibidem).

Finally, Polinsky noticed a phenomenon, opposite to the one we found: her informants had a tendency to translate literally some English discourse markers (what we called "слова-паразиты") into Russian, in particular "well" (То будет, *хорошо, *неудобное 63 , ivi, 27).

Also other studies about runglish, mostly on adults, showed a similar language erosion: together with a simplification in the declensions, Zemskaja (in 63 In this case the correct form of the adjective would be the short one (неудобно). ~ 76 ~ Perotto, 2009, 84) also points out errors in the use of numerals (mostly collective), in the conjugation of irregular verbs, in verb rection.

Mixed-language utterances in the first generation speakers, in contrast to those of the second generation studied by Polinsky, seems to be more often voluntary and integrated (for example: Поставь посуду в dishwasherku, Osipova, 1999, in Perotto, 2009 or Я всегда покупаю в sejle, Zemskaja et al., 2001, in Perotto, 2009.~

CONCLUSION

Through informal interviews in Russian, this research analyzed the oral language of a sample of ten early Russian-Italian bilinguals, who moved to Italy in their childhood or before their puberty age, in order to detect their maintenance or loss of their mother tongue, Russian, indeed.

The profile resulting from the interviews and the examples presented is not homogeneous. Together with speakers who showed to have maintained a high competence in their L1, the majority of the presented data demonstrates a language erosion on many aspects, due to the preeminence of use of Italian. The loss detected, in fact, involves lexical, syntactic, morphological and phonological levels.

For some speakers their move to Italy caused a gradual, often voluntary, abandon of their L1, and from an everyday language it has been relegated to a secondary one. This caused, on the lexical level, a number of gaps, which the informants tried to solve resorting to CS, or "inventing" words, on the basis of Russian lexical system or even more frequently, calquing from Italian, the most familiar language to them now.

On a syntactic level we noticed the tendency to tie together the various elements of the sentence following very often Italian rules rather than Russian ones.

As a consequence, this resulted either in completely incorrect (asteriskable) sentences or in acceptable (marked) sentences, but "not really Russian", which can be compared to the verbal productions of subjects with a low educational level.

Errors on a morphological and grammatical level are the most numerous and varied: the most frequent concern nouns, adjectives, pronouns and numerals declension, the category of the aspect of the verb, the use of verbs of motion and prepositions, the agreement of gender, number and case, the use and conjugation of verbs, especially irregular ones. This kind of error is typical for learners of Russian as a L2, especially for those whose mother tongue, like Italian for example, presents a very different language system compared to the Russian one, and Slavic languages in general.~ From the phonological perspective, it was noticed that some informants have maintained a perfect pronunciation and intonation, while others have shown a certain Italian accent when speaking in Russian and their speech follows the prosodic rules of Italian.

In particular, we highlight once again one thing which, more than others, seems curious and deserves to be examined more in depth: all the male speakers, regardless of the language level shown, maintained their Russian intonation and pronunciation, while all the female informants (except for one, whose competence in Russian remained impeccable on all aspects) proved to speak Russian with a strong Italian accent.

Interference from Italian can be found in all the aspects presented and it is undeniable especially in the calques, which prove an obvious transfer of a form, be it syntactical, be it semantic, be it lexical, from Italian into Russian, as well as the tendency to consider as symmetrical certain words or structures that are not.

The informants made an abundant use of CS strategies and sometimes they inadvertently incurred CM.

CS was often used to face recurring lack of vocabulary, to clarify key concepts the informants were not sure about, to say strong expressions, but above all to refer to realia of Italian culture and society, which are difficult to express through Russian words (or with which the informants were not familiar).

CM, instead, less frequent than CS, was noticed mostly at the level of individual words, often conjunctions, hesitations or short words (one or two syllables) in Italian, which we called "слова-паразиты" and that show a strong presence in the speech of the informants. Similar cases of erosion and language interference were noticed in other studies about the Russian language abroad, both in Italy and in various European and non-European coutries (whose results were compared to the ones of the present study), as well as in many studies on early and late bilingualism, conducted on different language pairs (we could mention an endless quantity).

The most alarming data on this language erosion is that the influence of Italian seems to be so strong in the speech of the informants, that it has affected their language automatisms in L1. In fact, only four out of ten informants can possibly bẽ defined as additive bilinguals, while for the remaining six the L2, Italian, seems to have replaced the L1, Russian (subtractive bilingualism).

The data presented and the observations that emerged can be analyzed in relation to Fabbro's theories about early acquisition of L2 (and in general foreign languages), and in particular about the Critical Threshold Hypothesis (Fabbro, 2004, 89) , which was, after all, the starting point of this research.

We remind that, according to Fabbro, L2 can be acquired perfectly only if it is learned implicitly within 8 years of age, since after this age the ability of an individual to reproduce perfectly the characteristic sounds of a language and to assimilate its morphosyntactic rules (the two most characteristic and most difficult aspects of all languages, Fabbro, 2004, 86) decreases, especially if L2 belongs to a language family which is far from L1 (ibid, 87). Moreover Fabbro, as well Paradis and other neuroscientists, agree that very early bilinguals, those who acquire the L2 within 3 years of age, show a lower difficulty using the L2 than early bilinguals (ibid, 99).

The results emerged from this research seem, partially, not to confirm Fabbro's theories. Most of the respondents, in fact, acquired Italian after the age of 8, yet their competence in this language now seems to be comparable to that of a native, both from the morphosyntactic and phonological points of view (only two informants have maintained an accent when speaking in Italian). Furthermore, regarding the competence in L2, although this is not the object of our research, we did not notice any difference between the informants arrived in Italy at only one year (the only case of very early bilingualism analyzed by us) and all the others.

The most curious aspect, however, is that language proficiency in L1, the object of study of this research, of the majority of informants seems to have suffered a sharp deterioration due to the influence of L2. These data are very interesting because, we remind that all the informants, for the first years of their life, grew up in a Russian monolingual environment, except for the one case of very early bilingual.

Definitely it is necessary to consider other extralinguistic factors (which we already mentioned), which contributed to language maintainance or caused a gradual erosion, primarily adoption or move to Italy with their families. Almost all speakers who live with their Russian-speaking family expressed their willingness to valuẽ their roots and therefore to maintain their native language, although not all of them showed a particularly high level language. On the other hand, those who were adopted, not only did they show more difficulties when expressing in Russian, but also they admitted to not being particularly interested in preserving this language, or that this interest has arisen only recently.

Consequently, the fact that they acquired a language, and above all their mother tongue within the critical threshold, does not seem to have been, for all ten informants who participated in this research, a guarantee against the loss of many morphosyntactic and phonological aspects of their L1 (we voluntarily don't mention vocabulary because this is one of the most "open" components of language. Also Fabbro, in fact, states that vocabulary, unlike morphosyntax and phonology, is located in the explicit memory and the acquisition of "open class words" does not depend on the critical threshold, but it continues throughout all life).

As stated from the beginning, given the small number of the sample, this work has no pretensions of scientific nature. It is, in fact, solely a linguistic analysis, although it does not lack of references to sociolinguistic factors and neurolinguistic data upon which and from which it was inspired, but above all it aims to provide food for thought and a starting point for any future and more in-depth researches in the spheres of early bilingualism and contact between Russian and Italian, especially in this period when the Russian-speaking community living in Italy is expanding.

~ 81 ~

ABSTRACT

According to neurolinguistic studies, in particular the theories of the Italian neurophysiologist Franco Fabbro, anyone can perfectly acquire one, two or more languages (including their morphosyntactical and phonetical peculiarities) on one condition: it must be acquired in childhood, within the age of 8, in a procedural way.

Can the early age of acquisition be considered as a guarantee that the acquired language will not be exposed to "dissolution" and/or oblivion? And in general, when a person lives overseas, can he forget his mother tongue because of the interference of the new language?

The aim of this analysis is to look for the answers to these questions through experimental data.

We conducted an experiment with ten early Russian-Italian bilinguals, who were born in Russian or in one of the former USSR countries and moved to Italy in their childhood or adolescence. First Russian was their only mother tongue, but after moving to Italy, Italian became their second language and, day after day, the dominant one.

The speech of the informants was analyzed through interviews, which showed that the informants' first language became actually their second one. In their speech we observed grammar, lexical and syntactical mistakes, which are typical for foreigners and which can be explained as linguistic interference of Italian language, together with different calques and cases of code-switching (CS) and code-mixing (CM).

Moreover, also their pronounciation and intonation, the most characteristic aspects of a language, seemed to have got deteriorated because of the interference of Italian language.

If we compare the results of this analysis with the researches on the speech of Russian-speaking migrants, living in Italy and abroad, above all in the USA, we notice that the widespread phenomena of CS and CS, borrowed words and calques are not the choice of the speaker's language behaviour, but a symptom of language loss.~

APPENDIX I OUTLINE OF THE INTERVIEW