Hannele Dufva
Introduction
Today, we see how globalization, with its cultural flows, and technology with its new developments constantly create new types of contexts and new kinds of language practices. Thus it would not be unreasonable to say that language itself is changing. Also, these large-scale changes create new environments for learning languages, and these environments, po-tentially, will influence how we conceptualize learning itself. Thus, as the contexts and usage change, it is possible that the theoretical basis of lan-guage learning needs to be rethought. Further, this gives us a reason for rethinking the pedagogical practices of language education. This paper discusses the two central concepts of second and foreign language learn-ing research -- ’language’ and ’learnlearn-ing’ -- and the potential consequences of how their reconceptualisation might influence practices of language education and pedagogy.
To redefine ‘learning’, we need to transcend the traditional dichot-omy between social and cognitive descriptions that has been typical for second language acquisition (SLA) research. In recent years, we have ob-served a movement from the strict cognitivism of the early SLA towards socially oriented arguments, some of which have turned out to be exclu-sively social in their position. As an alternative to these polarized views, language learning will be regarded here as a social-cum-cognitive process:
an activity in which the social and the cognitive are involved and in-tertwine. A holistic view is advocated, in which cognition is not placed
‘internally’ in the learner’s brain, but is extended to ‘external’ activity in
the social and physical environment. Here, I will draw on arguments from the following sources: Vygotskyan and neo-Vygotskyan views (e.g., Lantolf, 2000; Lantolf & Thorne 2006); systemic psychology (Järvilehto, 1994, 2006); distributed views on cognition (Cowley, 2004, 2006; Stef-fensen, 2009); ecological views (Gibson, 1970; van Lier, 2004, 2007);
and the Bakhtin Circle dialogism and neo-dialogism (Linell, 2009; Duf-va, 2010; Dufva et al., 2011).
It will be argued that learning is distributed cognitive activity. This is to say that the individualist notion of learning is rejected and argue that learning occurs in collaboration with and is mediated by other people and/or by different tools and artefacts of the social world1. As language or linguistic resources are being shared in the activities in which people par-ticipate, they are also constantly recycled. What is important to note is that this process is not seen as transfer of information from ‘outside’ to ‘inside’.
Learning is not regarded as an acquisition of abstract forms but as linguistic resources being appropriated by persons participating in a certain activity.
The reconceptualization of ’language’ below draws upon the recent debates in which the traditionaltwentieth-century concepts have been dismantled and deconstructed (see also, e.g., Makoni & Pennycook, eds.
2007). However, I will focus in particular on the dialogically oriented views of language and the Bakhtinian notion of heteroglossia. It will be argued that to emphasize the dynamicity and relationality of language, language learning should be regarded as appropriating different situated practices, or heteroglossic languaging.
Learning: a social-cum-cognitive and mediated process
To see learning as a social-cum-cognitive -- or distributed -- process rejects the Cartesian interpretation in which cognitive refers to ‘internal’ actions and social to ‘external’ ones. The view challenges both the cognitivism of early SLA (second language acquisition) studies but also those contempo-rary socially-based arguments that fail to give an account of the individ-ual person and his cognizing. Cognitivism that was characteristic of the
1 The view that is discussed here does not exclude the aspect of language as an embodied and material process or that this is not, strictly speaking, a social world but a material one. For the sake of brevity, the argument for the material basis will not be developed here.
traditional SLA studies was influenced to a great degree by Chomskyan thought and rationalist philosophy. It turned away from the arguments that included the social world (social interaction, societal circumstances).
The new social focus, however, has frequently resulted in the failure to consider the cognitive aspects; for a more detailed discussion, see Dufva (2010). Here, I will aim at showing that both aspects can be included to form a new, non-Cartesian and holistic viewpoint on learning.
To see mind and observable activity as inherently connected is not at all a new idea: it was a strong presence in L.S. Vygotsky’s work and the sociocultural tradition that followed. Pointing out that one needs to study the history and development of cognitive phenomena in order to understand them, Vygotsky himself aimed at showing that human mind is social in origin and that ’higher cognitive faculties’ for intellect, rea-soning, and learning are essentially collective in origin. The social world, with its artefacts, tools, and patterns of social action that have developed over time as a collective effort of mankind, is the natural environment of each infant and each child respectively develops his intellect and reason-ing in social and collaborative activity. Therefore the social world cannot accurately be described as ‘external’: it is also the cognitive world -- or cognitive workspace – into which each of us is born and in which we continue to operate.
If we go on using words such as ’social’ and ’cognitive’, they are not to be understood in their Cartesian sense. ‘Social’ does not refer to ex-plicit interaction with other people or to the societal sphere as ‘external’
context, but is also a feature of human activity that is traditionally under-stood as cognitive or psychological. As Lantolf (2004, pp. 30-31) notes, sociocultural theory is not a theory of the social or cultural aspects but, actually, a theory of mind.
Therefore, it is seen as unhelpful to continue the reductionist ar-guments of either cognitivism or the radical extremist views of socially oriented paradigms. Between cognitive and social worlds there exists a reciprocal relationship that was also a central theme in Voloshinov’s phi-losophy of signs. Voloshinov (1974, pp. 33-41) argues that outer signs, inherently connected with ideology, need to be ‘engulfed’ by inner signs.
There is an interplay between the inner and the outer signs: outer signs gain their life force by becoming inner signs when appropriated by
per-sons, while inner signs are returned to the outer dialogue when uttered.
Drawing upon these arguments, language learning can be seen as process of recycling of the socially and culturally available linguistic resources; see also Dufva et al. (2011).
When analysed dialogically, linguistic signs have two aspects: while being ‘ideological’ as to their content, they need to have materiality in order to be mediated. As Voloshinov (1973, p. 26, pp. 90-91) observes, human consciousness needs ‘gesture, inner word, outcry’ to become man-ifest. Thus language needs to be spoken, written, signed, or mediated by using other potential means of expression, that is, by different mediation-al means. If we use the Vygotskyan socioculturmediation-al formulations, the lan-guage environments involve symbolic artefacts, lanlan-guage itself being the prime example, but also material artefacts: books, pens, paper, and com-puters. Regarded in this light, language learning is a mediated process in which different mediational means are at play: these include textbooks, classroom interaction, teacher-directed talk and the various resources to which language learners are exposed, such as gaming or watching televi-sion, in out-of-school contexts.
As this view of learning does not regard mediation as transfer of in-formation from ‘outside’ to inside, it is natural to continue the argument that the environment is not an ‘external’ scene but part of the learning process itself. We could say that the environment is part of the cognitive working space of the person(s) involved.
Learning: a systemic, ecological, and distributed process
Where does learning occur, then? It was commonplace to understand the cognitive processes as happening in the individual’s mind and/or lan-guage being stored and processed in its linguistic components, as the rationalist Chomskyan argumentation put it. Today, many researchers implicitly identify learning with social interaction and do not go beyond describing what happens there. Both positions base their arguments on the interpretation of social and cognitive as external and internal, and are, as I would like to argue, led astray in this. If we consider where cognizing happens, or where language learning occurs, we should not look into the black box of the (internal) mind, or seek direct equivalents in the human brain, or identify cognizing with the behaviours in social interaction.
The views expressed within early sociocultural and dialogical per-spectives by thinkers such as Bakhtin, Voloshinov, and Vygotsky do pro-vide some of the philosophical and psychological starting points. These views find support from other, more contemporary lines of thought.
These make it possible to re-examine various issues and aspects of learn-ing (e.g., how memory works) that were previously given a cognitivist analysis and remodel them in frameworks such as systemic psychology (Järvilehto, 1998), ecological psychology (Gibson, 1970; van Lier, 2004) or distributed cognition (Cowley, 2006; Steffensen, 2009) that go be-yond the individual and/or his brain.
These views suggest a need to extend the research focus beyond the individual, something that was recognised already, and importantly, by Vygotsky. The importance of other people is present in Vygotsky’s notion of learning -- ‘first external, then internal’ -- is the thought that learners have a ‘zone of proximal development’ in which they proceed, supported by others -- parents, teachers, and peers. The perspective is also present in the concept of scaffolding that draws upon Vygotskyan thinking but is developed by Jerome Bruner. As neither Vygotsky nor the contem-porary research sharing this perspective assumes a Cartesian separation between mind and activity, it is clear that we do not talk about giving ‘in-put’ to learners. Rather, we talk about ‘sharing’ resources with them. As Suni (2008) has shown in her study of conversation between native and non-native speakers, native speakers can share their linguistic resources with non-natives in the joint cognitive working space that is created in talk.
Further, Järvilehto (2012) argues that the organism and its environ-ment should not be regarded in terms of two systems but one. In not sepa-rating environment from the mental activity of the organism, Järvilehto’s views provide a theoretical basis for understanding memory and, at the same time, for some aspects of language learning. In Järvilehto (1994, pp.
154-155) the metaphor of memory as an internal storage is challenged.
His argument is that the processes of remembering should be studied by regarding not only the ways in which the organism itself is organized, but should also expand the perspective to include the environments of both present and past. Memory, then, does not refer to a place, location, or storage, but rather, remembering, the ability to operate in the current
environment relying on the environments in one’s past. When we learn something new, there is a change in the organization of the organism-en-vironment-system.
If we accept Järvilehto’s (1994) argument, the metaphor of ‘inter-nal language storage’ with its ‘mental representations’ should be rejected.
Instead, ‘mental knowledge’ can be considered as action potential. This view may sound radical at first: against the classic cognitivist assump-tion of language learning as ‘internalisaassump-tion’ -- acquisiassump-tion of rules and items – learning now is seen as a process in which the persons develop in their ‘skilful linguistic action’ (Cowley, 2012) the potential to detect different linguistic resources present and their ability to act upon these as affordances (van Lier, 2004). Today, we have not developed fully ideas of how to reconceptualise the mental knowledge of language or a person’s language proficiency. Still, a tentative argument can be presented that language proficiency is not to be modelled as internal, individual, (semi) permanent knowledge of rules and items. Rather, the theoretical argu-ments seem to suggest that it might be regarded as processual knowledge, which consists of essentially situated and dynamic skills that allow learn-ers to operate across time and space.
Järvilehto’s (2006) perspective of learning extends beyond the bor-ders of the individual organism-environment systems, that is persons: ‘All efficient learning presupposes the participation of both the teacher and the pupil (or the trainer and the trainee)’. Järvilehto’s views resonate with other non-individualist, or ‘extended’, perspectives on cognition (e.g., Hutchins, 1995; Cowley, 2006). These argue that cognition is ‘spread’
among the participants, is ‘shared’ by them, or ‘emerges’ in the interac-tivity between the human agent(s) and the resources / tools present. Thus also the ability to learn language – either first or additional ones – can be understood as ways in which human agents are capable of perceiving and acting in their different linguistic environments, with other people and artefacts present.
Language: What is it that is learned?
Criticism of ‘language’ as a system
Another set of questions is concerned with how to define ‘language’. That is, what is the object that learners set out to learn? What is the object
of teaching at school and other institutions that provide instruction in languages? When one looks both at the research of language learning and the pedagogical discourses and practices, one finds several persistent met-aphors and dominant conceptualisations. These include 1) the influence of written language and literacy, 2) the impact on the national language ideologies and 3) the influential Saussurean view of language as an (ab-stract) system. These ideas have led to the idea that learners are supposed to internalize a system of abstract rules and contextless lexical entities. For a critical discussion, see Dufva et al. (2011).
The written language bias of linguistic inquiry has promoted the idea that units of ’language’ are similar to those found in written forms of language (for criticism, see, e.g., Linell, 2005; see also Voloshinov, 1973).
A literacy-based, written language bias can also be found in the ways languages are taught and language proficiencies are assessed. The written word is strongly present in classrooms where textbooks and literacy-based ideals still rule. For a survey in the context of Finland, see, e.g., Luukka et al. (2007). Also, learners’ proficiencies are still often evaluated and assessed by literacy-based standards in spite of the continuing critical dis-cussion. Thus it is almost inevitable that the written language bias is pres-ent also in language learners’ beliefs. In their studies on foreign language students’ conceptualisations -- with learners’ self-portraits, narratives and questionnaires as data -- Kalaja et al. (2008) have found a consistent pres-ence of textbooks and written materials. Their findings suggest that learn-ers see that their goal as learning the contents of textbooks, grammars, or dictionaries, that is, the decontextual descriptions of language rather than how to use language. These beliefs are no doubt advanced by the textbook-centred practices of foreign language classrooms, but they are also supported by the discourses, metaphors, and vocabularies of linguis-tic research.
Another idea that has been much criticized during recent years is how we have regarded languages as internally homogeneous entities, still cat-egorically different from others. This idea of language, influenced by the ideologies of nation states, not only conceptualizes languages as boundar-ied entities (Finnish, French, German) but also promotes a monolingual bias, an ideology that still often dominates the educational discourses and language classrooms where borrowing, hybridity, and mixing are ‘wrong’
and where use of more languages than one may be judged as pedagogically unfavourable. Further, the assumed stability and singularity of norms and the entailing policy of ‘one correct answer’ is maintained in classrooms, exams, and language tests. The alternative views advocate subjecting the norms and language use to negotiation, and not only for tolerating but also promoting ‘translanguaging’ in the classrooms (see, e.g., Blackledge
& Creese, 2010).
The third notion that needs a rediscussion is whether language as (an abstract) system consisting of, e.g., syntax, morphology, phonology, and lexis, should actually be seen as the goal of the language learner. It has been commonplace in the study of language learning as a process in which a language system is internalized. However, as has been pointed out by many authors since Voloshinov (1973), a system of this kind is neces-sarily an artefact produced by the linguist’s analysis: a selective description of the formal properties of language use. Valuable as they may be, these artefacts are not to be confused with the actual reality of language use or
‘first-order languaging’ (see, e.g., Cowley, 2005; Steffensen, 2009); gram-mars -- whether linguistic or pedagogical -- inevitably select, summarise, and reduce the material they choose to describe and systematize.
It should be also pointed out that the conventional linguistic and grammatical descriptions may not be adequate at all to describe the pro-cesses by which language users actually operate. Although it has been exceedingly popular in (psycho)linguistic research to speak about mental grammars and internal lexicons, the metaphor may be faulty in many senses: as both the early dialogical and sociocultural arguments (see, e.g., Voloshinov, 1973, p. 38) and recent research seem to indicate, the nature of mental language knowledge is very much an open question. To this point, Steffensen (2009) argues that ‘there is no reason to posit internal representations of linguistic units’. With Cowley (2011, p. 21) we can say that language is to be found not in one’s internal storage, but with ‘the resources of the world’s language stores’.
Finally, if language proficiency is seen in terms of decontextualised formal knowledge, the repercussions involve a decontextual approach in language teaching. It is at the very core of the conservative tradition of language teaching to focus the classroom practices and homework rou-tines on decontextual practices, to focus on memorising grammatical
rules, lexical items, and formal translation equivalents. Instead of seeing situated and contextual practices as their target, the learners grow to dis-associate the ‘knowledge of language’ from its use.
The viewpoint of heteroglossic languaging
The contemporary discussion around the notion of ‘language’ often stress-es its dynamic qualitistress-es, and also, many scholars point to its relational char-acter. The dynamicity -- the flow-like character -- of language is present in the formulations of language as languaging (Maturana, 1995; Becker, 1991). For a closer discussion, see Dufva and Pietikäinen, (forthcoming);
as communicative activity (Thorne and Lantolf, 2007); as doing (van Lier, 2004); and as practices (Pennycook, 2010). Many new formulations also frequently embed a notion of language use (and learning) as collaborative or systemic activity. If these qualities of language are highlighted, it seems to follow that, implicitly, the views also highlight functional and mean-ingful elements rather than formal and structural ones. In all, language is regarded as a purposeful rather than a mechanical process -- and it may well be regarded as ‘the game rather than the building blocks’.
Here, I will draw particularly upon the linguistic arguments of the Bakhtin Circle and the notion of languaging. I will suggest that the goal of the learners is to appropriate language practices that are heteroglossic in nature. The implication of the notion of heteroglossic languaging is to see the learners’ goal not in learning a ‘language’ (as a singular entity), but learning situated usages (practices). In this, both the quality of doing/
action and the essential diversity of language usages is highlighted. This
action and the essential diversity of language usages is highlighted. This