Cross-linguistic influence is a phenomenon in second language learning and use. It is more commonly known as language transfer, and can be described as ‘…the influence resulting from the similarities and differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired’ (Odlin 1989:208). In other words, it is the effect of various aspects of one language on another when used by speakers who have knowledge of more than one language.
The field of CLI is highly interdisciplinary as it consists of studies from a wide spectrum of academic disciplines such as Second Language Acquisition (SLA), psycholinguistics, cognitive linguistics, sociolinguistics, linguistic anthropology, cognitive sciences, etc. As such, there is no tangible detailed theory of language transfer or CLI. However, through empirical studies addressing more general questions on language and cognition, CLI has made great contributions to the reformulation of the relativity principle and thus helped advance the linguistic relativity question, both empirically and theoretically. Since the reformulation of the principle in the mid-1990s, which early studies of transfer helped develop, CLI research has become more focused on studies where researchers attempt to incorporate linguistic relativity into their analyses of transfer. The phenomenon studied in this kind of research is also known as conceptual transfer (Odlin 2008).
Originally, instances of CLI were examined in terms of particular linguistic subsystems such as syntax and semantics in the field of second language acquisition and production. However, in the course of time, studies have expanded the field by demonstrating that CLI is not limited to production and acquisition. Transfer can also be identified in a number of psycholinguistic processes such as lexical and syntactic processing, listening and reading comprehension, tip-of-the-tongue states, nonverbal communication, and conceptual representations. Especially in relation to conceptual representation, important transfer effects have been found in the interpretation of, and with reference to, emotions, many of which were previously considered to be universal (Jarvis &
Pavlenko 2008:14). This research is of particular interest to this study, as it deals with emotions as well as the relationship between language, culture, and thought. However, before going into details
with CLI and emotions, it is worth taking a look at the different types of transfer one may encounter.
Instances of CLI can be categorised mainly in two ways. Firstly, they can be categorised according to the directions forward, lateral, and reverse; forward transfer are instances of CLI where the L1 influences the L2; when later learned languages influence each other, e.g. the L2 influence the L3 or vice versa, it is called lateral transfer; and reverse transfer is when later learned languages such as L2 or L3 influence the native language. Secondly, transfer can also be categorised according to a wide range of types: lexical transfer, semantic transfer, phonological transfer, morphological transfer, pragmatic transfer, conceptual transfer, intentional vs. unintentional transfer, and covert transfer, to mention just a few (Jarvis & Pavlenko 2010: 13-15). However, here, the most important distinction is that between linguistic and conceptual transfer, i.e. the differentiation between instances of CLI studied primarily in relation to linguistic forms and structures, and instances of CLI analysed primarily in relation to the mental concepts that underlie those forms and structures (Odlin 2003). In other words, conceptual transfer is when concepts underlying words in the native language, L1, are transferred to any later learned languages, and mapped onto new linguistic labels, regardless of differences in the semantic boundaries of the corresponding words (Odlin 2008). It is in instances of conceptual transfer that SLA and linguistic relativity intersect. As conceptual transfer always contains semantic transfer, but not vice versa; it is thus necessary for this study to consider the idea of, not only semantic, but also lexical transfer.
2.7.1 Lexical and Semantic Transfer
Lexical transfer is when the formal and structural knowledge of words in one language influences a person’s formal and structural knowledge or use of words in another language. Semantic transfer is when an actual real target-language word is used with a meaning reflecting influence from the semantics of a corresponding word in another language. Alternatively it can also be when a calque is used in the target language that reflects the way a multi-word element is mapped to meaning in another language, e.g., the sentence he remained a youngman all his life, reflects semantic and compositional influence from the Swedish and Danish word “ungkarl”, meaning “bachelor”, and which is composed of the elements ung, meaning “young”, and karl, meaning “man” (Jarvis &
Pavlenko 2010: 75).
Lexical transfer research has for the most part focused on either morphophonological errors or semantic errors. There are, however, differences between morphophonological transfer, also called formal transfer, and semantic transfer. The two are not always mutually exclusive, yet research suggests that the rules dictating formal transfer are different from the ones dictating semantic transfer. Some of the most recognised forms of formal transfer are the use of false cognates as seen in the utterances many offers of violence, and a crucial problem is the drinking of alcohol before sitting behind the rat, reflecting influence from the Swedish and Danish words offer and ratt /rat meaning “victim” and “steering wheel” respectively (Jarvis & Pavlenko 2010:75, Jarvis & Ringbom 2009:109); unintentional lexical borrowing involving the use of a word from a language other than the target language; or the coinage of a new word by blending two or more words from different languages, e.g. we have the same clothers, where the word clothersis a blend of the English word
“clothes” and the Swedish word for clothes: kläder(Jarvis & Pavlenko 2010:75).
Interestingly, the likelihood of formal transfer is much higher when the source and target language are closely related in form and structure, such as English and Swedish, and English and Danish, whereas the likelihood of semantic transfer is higher when the two languages are typologically distant, such as Japanese and English. Research suggests that as a consequence, formal transfer results from learners’ assumptions relating to formal similarity between source and target language, even when the similarities are not there (Jarvis & Pavlenko 2010:77, Odlin 2003:443). As an example, the Swedish and Danish languages both have many close cognates and formal similarities with English to the extent that Swedish and Danish native speakers automatically, without any hesitation, assume similarities with English even when there are none. It is, however, important to keep in mind that lexical errors do not necessarily entail transfer, and that transfer does not always result in errors; it just happens to be the case that instances of negative transfer have been much easier to detect than instances of positive transfer and have been found more compelling (Jarvis &
Pavlenko 2010:75).
Likewise, semantic transfer is an outcome of assumed similarities between source and target language, though quite different similarities from those in formal transfer. Research shows that learners tend to assume that two languages are different until they observe evidence of similarity. In contrast, when it comes to the meaning of words, they tend to assume that two languages are
semantically similar until they become aware of differences, i.e. they assume that two equivalent words in two languages are semantically similar, despite this rarely being the case. Differences in semantics are much more difficult to become aware of, as differences in meaning are not visible and thus far more difficult to detect than differences in form. As a consequence, differences in meaning are only gradually recognised and often require an introduction, extensive study, and/or exposure to how the word is used in various contexts. Until then, language users will continue to falsely assume a semantic equivalence between words in the new language and their perceived counterpart in an already known language (Jarvis & Pavlenko2010).
2.7.2 Conceptual Transfer
As previously mentioned, conceptual transfer is when concepts underlying words in the native language are transferred to any later learned languages, thereby mapping ‘old’ concepts onto new linguistic labels, regardless of differences between the two. To this may also be added influences from the semantics and pragmatics of the native language (Odlin 2003: 464). Within conceptual transfer, research has dealt mostly with instances of influences in eight fundamental areas, all of which are based on sensory-motor experiences and are encoded in most human languages. These areas are: objects, emotions, personhood, gender, number, time, space, and motion. However, the bulk of research focuses on spatial concepts and concepts relating to time and emotions (Jarvis &
Pavlenko 2010:122, Odlin 2005:10). For this study, the area of interest is that of emotions; I shall return to the details of that topic in a later chapter.
Based on Murphy’s (2002) definition of concepts, Jarvis & Pavlenko (2010) make a distinction between language-mediated and language-independent concepts. Language-mediated concepts are multi-modal mental representations that develop in the process of language-socialisation. They guide speakers of particular languages to particular conceptual distinctions, which allow them to perform naming, identification, comprehension, and inferencing tasks along similar lines. In contrast, language-independent conceptual categories are, as the term implies, conceptual categories that exist in the mind independent of language. This is consistent with Whorf’s acknowledgement of the existence of language-independent thought, i.e. thought that exists without language (Jarvis
& Pavlenko 2010:113).
Language-mediated concepts differ across languages, and as such conceptual transfer can be seen as the influence of the language-mediated conceptual categories of one language on verbal
performance in another. In other words, the patterns of conceptualisation in one’s mother tongue affects the way one speaks a second language, i.e. one speaks a second language according to the patterns of linguistic framing and organisation of information according to one’s L1 (Jarvis &
Pavlenko 2010:115). This influence of the L1 acquired patterns of thinking and organising experience, on the use of the L2 can be seen as indications of the ‘binding power of language’, and studies do indeed suggest that even highly proficient second language learners may never free themselves entirely from their native language (Odlin 2005).