Behind the concepts of multiliteracies and media literacy in the renewed Finnish core curriculum: A systematic literature review of peer-reviewed research

Finnish basic education faces a significant change with the 2016 commissioning of the renewed core curriculum, which introduces a new transversal competence, termed multiliteracies—a concept closely related to media literacy. This systematic literature review examines the research literature on media literacy and multiliteracies, analysing and comparing the nature of knowledge constructed and the varying definitions of the two concepts. Previous review articles (Marten 2010; Potter 2010) found little consensus among scholars regarding the definition of media literacy. This review examines the research literature published in international peerreviewed academic journals between 2010 and 2014 to investigate whether a mutual understanding of the concept has since been established. The article argues that significant differences exist between the concepts of media literacy and multiliteracies and, further, that Finnish core curriculum defines multiliteracies differently than the research literature defines the term. In line with previous research, this article finds no consensus on the definition of media literacy in the research literature. Based on the multifaceted nature of the concept, this article rejects attempts to establish a universal definition of media literacy and presents a theoretical framework for conceptualising media literacies based on their abstraction levels. The article aims to facilitate understanding of the concept and its operationalisation in research and practice and discusses future opportunities for research on media literacy and multiliteracies.


Introduction
Finnish basic education will face many changes with the introduction of the new national core curriculum in 2016. The renewed core curriculum for basic education was completed and published at the end of the 2014 by the Finnish National Board of Education (FNBoE). The new transversal competences addressed in several subject areas are likely to have a significant impact on media education. One of these competences is termed multiliteracies, defined as "the skills to interpret, to produce and to evaluate different kind of texts. These skills help students to understand diverse cultural forms of communication and to build their identity" (FNBoE, 2014, p. 22). The curriculum defines texts as information presented through various symbol systems (linguistic, visual, auditive, numeric, kinesthetic or a combination of these). The definition of multiliteracies is closely related to the concept of media literacy, which is traditionally defined as an ability to access, analyse, evaluate and communicate messages in a variety of forms (Aufderheide 1993).
Media education can promote both multiliteracies (FNBoE, 2014, p. 86) and media literacy (Kupiainen & Sintonen, 2009). Definitions of these concepts clearly overlap. It is important to distinguish between multiliteracies and media literacy, because unclear definitions create challenges in operationalising the terms in practice and in research. This raises research questions: How are the concepts of multiliteracies and media literacy related? How are these concepts defined in the recent research literature, and what kind of knowledge is constructed in the research literature? Clarifying concepts on a national level to define the line between media literacy and multiliteracies is important for scholars of media education, education planners and education practitioners. The national core curricula are critical for basic education in Finland due to their role in research, policy and practice. Curricula lead and guide the planning and implementation of educational practice.
The concept of media literacy has been under discussion for several decades, but a consensus regarding its definition has yet to be reached. The concept is multifaceted, and scholars tend to add and subtract others' ideas in seeking to define it (Potter, 2013). This study aims to produce new information and construct new knowledge by analysing the various definitions presented in the research on media literacy and multiliteracies. The article further contributes to the work of education scholars, planners, practitioners and policy makers by presenting a theoretical framework for defining media literacy. By examining the research and exploring definitions of the concepts, the article contributes to the media literacy discussion both nationally and internationally.
The following section will contextualise the research within the discussion of national Finnish education and will provide an overview of discussion of media literacy definitions. Next, the methodology of the systematic literature review will be presented and discussed. The findings section will address the research literature itself, focusing first on the nature of the knowledge constructed and next on the various definitions of media literacy and multiliteracies. In the final sections, the findings, implications and limitations of the study are discussed and conclusions are drawn.

Background of the study Media education and media literacy in the Finnish core curriculum
Finland's core curriculum is renewed approximately once a decade. The current curriculum is from 2004; the new curriculum will be introduced in the fall of 2016. Kauppinen (2010) studied the concepts of literature and reading literacies in Finnish basic education curricula. According to her research, the concept of media literacy has appeared in a variety of core curricula since the 1970s. The definitions, roles and perceptions of media literacy, however, have followed the development of wider discussions concerning media culture, meaning that varying aspects of media literacy have been included and thus promoted in core curricula. In the 1970s, for example, the focus was on mass communication, whereas curricula in the 1980s and 1990s used a concept of communication education (Kauppinen, 2010, pp. 230-232.) In the 2004 core curricula, media literacy relates to the cross-curricular theme communication and media skills (FNBoE, 2004). This development stems from a wider discussion on media education and media literacy in Finland (Kupiainen, Sintonen, & Suoranta, 2008).
Media literacy is a multi-dimensional concept discussed by scholars, professional educators, activists, parents and others worldwide. It is, however, a relatively new concept; most of the discussions began just a few decades ago (Potter, 2010). There is as of yet no consensus about the definition of media literacy (Martens, 2010;Potter, 2010;Hobbs, 2011a;Potter, 2013). At the national level, media education and media literacy have long roots in the context of Finnish education (Ruokamo, 2005;Kupiainen, Sintonen, & Suoranta, 2008).
The new core curricula states that students need multiliteracies to interpret the world around them and perceive its cultural diversity. According to the curriculum (FNBoE, 2014), multiliteracies refers to the skills to acquire, mix, edit, produce, express and evaluate information in various forms, environments and situations, with the help of a variety of tools. The acquisition of multiliteracies supports the development of critical thinking and learning skills. Ethical and aesthetic questions are also related to multiliteracies, as are information technology and communication technology skills. Cultural multiliteracies can be promoted through media education and by taking into account students' and families' media culture (FNBoE, 2014, p. 88). These notions make it interesting to scrutinise the relationship between the concepts of media literacy and multiliteracies. However, core curricula are abstract in the sense that they are contextualised later at the local level. As conceptual clarity is presumed in the field of research, where concepts must be defined with a high degree of precision, this article focuses on the research literature and the ways these concepts are defined in research articles.

Debate about the concept of media literacy
One of the most cited definitions of media literacy was presented in 1992 at the National Leadership Conference on Media Literacy, where media literacy was defined as "the ability to access, analyse, evaluate and communicate messages in a variety of forms" (Aufderheide 1993). However, this definition has been criticised for its ignorance of the themes central to media literacy (Potter, 2013), and many definitions have been presented since the conference (Martens, 2010;Potter, 2010;Potter, 2013).
One important motivation behind the conceptual discussion is the demand for a mutual understanding of the concept's content. The development of science is based on the cumulative nature of knowledge construction; scholars and others in the field need to know what is meant by the field's core concepts. Attempts to define media literacy have used various methods. A number of researchers have analysed the concept in relation to other closely related concepts. For example, Koltay (2011) andFantin (2010) analysed media literacy in relation to information literacy and digital literacy. Another method analyses how other scholars have defined and used the term. This article combines these two methods to offer a comparative view of media literacy and multiliteracies based on a systematic literature review.
James Potter (2010) presented a synthesis of various media literacy definitions and noted that the term can have multiple meanings. Potter argued that there is no consensus about the concept and that authors "continue to add and subtract ideas from other definitions when constructing their own, and continue to struggle with the questions 'what is media literacy?'" (Potter, 2010, p. 679; see also Potter, 2013, p. 420).
According to Potter (2010), four common themes are generally agreed upon in the field of media literacy: (1) mass media has potentially negative effects, (2) media literacy has a protectionist purpose against those potentially negative effects, (3) media literacy needs to be developed and (4) media literacy has many dimensions. Renee Hobbs (2011aHobbs ( , 2011b critiqued this vision, stating that Potter's ideas are too narrow in scope and that valuable perspectives are left out of the definition. "Most importantly, in conceptualizing media literacy primarily as a response to counteract the negative effects of mass media and popular culture, Potter's vision of media literacy mischaracterizes the depth and complexity of the field" (Hobbs, 2011a, p. 420). According to Hobbs, it is important to understand media literacy as more than just an antidote for mass media exposure, because media literacy education has a much wider range of objectives.
Potter (2013) deepened his analysis in a later review article, noting that mass media has a wide range of potential effects on individuals and that the purpose of media literacy is not just protectionist but also empowering in the sense that people can use media to achieve their goals. Potter also introduced three new themes upon which the field of media literacy seems to agree: (1) mass media also has an influence upon larger social structures, (2) people are more susceptible to media influence when they are passive and (3) media literacy involves skills besides knowledge (Potter, 2013). According to Hans Martens (2010), the scholarly literature defines media literacy mainly in relation to the knowledge and skills needed for analysing, evaluating and producing media messages. According to Buckingham (2007, p. 48), four broad conceptual aspects are generally understood to be essential components of media literacy: representation, language, production and audience.
One aim of the paper is to point out the multifaceted nature of the concept of media literacy and clarify its meaning. This aim is in line with Potter's recommendation to identify the most useful parts of the existing definitions and combine them into a concept that is "broad enough to apply to all media and all cultures but also detailed enough to be useful to researchers and instructors" (Potter, 2013, pp. 429-30). Potter (2013, p. 430) has emphasised a need for more carefully designed applied studies and for research built on a stronger conceptualisation of media literacy. Conceptual clarity is important in practice as well as theory. Mihailidis and Diggs (2010) noted that the interdisciplinary nature and broad definition of media literacy have influenced utilisation of the concept in practice. A concept must be broad enough to include a variety of aspects, but it cannot be too wide or it loses its definitive power. If a concept is defined too narrowly, however, it may not be versatile enough.
The Finnish core curriculum defines the concept of multiliteracies very broadly and includes a variety of literacies within the concept (FNBoE, 2014). The risk is that this umbrella concept becomes too broad to be utilised effectively in practice or in research. The broadness of the concept raises questions: What has been left out? Which literacies are not included?
Regarding media literacy, Teurlings (2010) noted that it would be hard to find another concept that connects such a variety of approaches and perspectives (p. 360). Although the broadness of a concept might explain its popularity, the risk in conceptual stretching-adding multiple ideas and things under a single concept (Sartori, 1970)-is that it can represent a deliberate attempt to lessen the conceptual value. To illustrate this idea, Sartori (1970) refers to the "Hegelian night in which all the cows look black (and eventually the milkman is taken for a cow)" (p. 1040). This also applies conversely. The higher the discriminating power of a concept, the better the information (Sartori, 1970(Sartori, , p. 1039.

Methodology
This study applied the methodology of systematic literature review (SLR). A common methodology in various disciplines, SLR is suitable for the search and analysis of large datasets and a wide range of literature. The methodology of SLR is employed in disciplinary fields such as medicine, psychology and education and also in interdisciplinary fields such as engineering education and female entrepreneurship (Borrego, Foster, & Floyd, 2014;Strech & Sofaer, 2011).
SLR is by nature a secondary research method; it uses primary studies as the data for analysis (Finfgeld-Connett, 2014). SLR can be used to promote evidence-based activity (Horvath & Pewsner, 2004), to improve policy making and decision making, to enable more objective critiques of past literature, to answer empirical questions and to evaluate methodological approaches (Borrego, Foster & Froyd, 2014;Finfgeld-Connett, 2014;Strech & Sofaer, 2011). SLR can also be used to construct an overview of a specific field to find trends or gaps in the research (Henry, Foss & Ahl, 2015). In this study, the review method was applied to gather research articles on media literacy and multiliteracies to analyse the definitions of the concepts and to scrutinise how these concepts have been utilised in the research.
One important reason for conducting SLR is to glean ideas from researchers studying the same topic (Fraenkel & Wallen, 2006, p. 67). This review gleaned ideas about defining the concepts of media literacy and multiliteracies. According to Onwuegbuzie et al. (2010), the benefits of SLR include identifying relationships between concepts and practice and identifying research methodologies and designs. This study focused especially on the relationship between the concepts of media literacy and multiliteracies and on the methodologies of the research using and researching the concepts. The review seeks to analyse the multifaceted nature of these concepts, clarify their meanings and evaluate the relationship between them.
SLR is used in multiple disciplines and thus may have various applications, but some basic principles are generally shared. SLR consists of two interconnected phases: the search of the relevant literature and the analysis of the data. To ensure the transparency of the process and to enable others to evaluate the adequacy of the method and to reproduce the process, it is important to explain both of these methodological phases thoroughly.

Search of the studies
Relevant peer-reviewed articles published from 2010 through 2014 were searched using the scientific databases ERIC (ProQuest), i Academic Search Elite (Ebsco), ii ScienceDirect (Elsevier), iii SpringerLink, iv SAGE Journals v and Emerald Journals. vi These databases were chosen because of their multidisciplinary scope and their relevance to media literacy and multiliteracies research. The search terms used were media literacy, media literacies, multiliteracy and multiliteracies. The search terms focused on the articles' author-supplied keywords, presuming that if a concept were one of the keywords, the study would define it. Both of the concepts-media literacy and multiliteracies-were searched differentially to compare the search results. Relevant research articles were included in the review based on the following inclusion criteria.

Inclusion criteria and sample size
To be included in the review the article had to be While contributions on media literacy and multiliteracies can be found in a wide variety of journals and other documents, for this review we focused only on research articles, based in part on the presumption that the relevant concepts would be defined in the research with a high degree of precision; the cumulative nature of science requires that concepts utilised in research be precisely defined. In addition, empirical and theoretical research articles are valuable sources for conceptual analysis because of their highly contextual nature and pursuit of conceptual clarity. Thus, other documents were excluded, including book reviews, online books, editorials, book chapters, opinion articles and lesson plans. This study focused particularly on the multiliteracies and media literacy research written in English, because of the historical and the widest language group bases in this research field. Authors are aware that literature published in languages other than English could give wider perspective to examine these concepts. To ensure that the articles were relevant for the discussion of media literacy and multiliteracies, only articles which listed either or both of these concepts as keywords were included.
This study focuses particularly on the recent research literature, including only articles published between 2010 and 2014 in the review. This decision was made based on previous discussion and review articles in the field of media literacy (Martens, 2010;Potter, 2010;Potter, 2013).

Analysis of the data
Content analysis was used as a methodological framework for the data analysis. Research articles are valuable data for analysis in the sense that they are highly contextual texts. They must be analysed with a high level of precision to maintain their original meaning. This is especially important when focusing on the definitions used in the articles. Undoubtedly, the complexity of qualitative content analysis increases in relation to the amount of data. Scientific precision demands that close attention be paid to sample size. According to Finfgeld-Connett (2014), the risks of a too-large sample size include the possible loss of important nuances and overwhelming redundancy (p. 349). Hannes and Macaitis (2012) noted that the median number of research reports analysed in systematic reviews of health care literature was 14. According to Finfgeld-Connett (2014), this sample size can be analysed using content analysis. For these reasons, the sample size for this review is 14 articles for each dataset.
The phase focusing on the analysis of the studies consisted of two separate steps, data extraction and content analysis.

Data extraction
In line with the research questions, the analysis focused specifically on definitions used in the literature and how the research was performed. The data relevant to answering these questions were extracted from the articles into a separate document. Extracted information included articles' reference details (authors, publishing year, the name of the article), methodological information (methodological approach, the nature of the data, the data analysis method) and concept information (the definitions of media literacy/media literacies, multiliteracy/multiliteracies, reference details if external definitions).

Content analysis
A variety of qualitative analysis methods for literature reviews are available (Onwuegbuzie et al., 2012). To systematically analyse the relevant extracted data, this study used the content analysis method. Content analysis is a widely used method in various disciplines. It is not, however, a single analysis technique but rather a methodological framework that includes different approaches (Hsieh & Shannon, 2005). Although content analysis has been used in primary studies, there is also a growing trend toward its use in systematic literature reviews. Finfgeld-Connett (2014) noted that the key differences between using the method for primary studies and systematic literature reviews stem from the notion that the data used in SLR are highly systematised and contextualised. "Due to this difference, qualitative systematic reviewers should avoid overmanipulating and processing the data (e.g. deconstructing and abstracting) for purposes of analysis" (Finfgeld-Connett, 2014, p. 350).

Findings
The first section will present basic information and provide an overview of the dataset. It describes the number of articles and divides them by year. The findings from media literacy research and multiliteracies research are presented separately. The second section focuses on the nature of the knowledge constructed in the research literature and the analysis of the definitions of the concepts. The final section presents a theoretical framework for media literacy.

Data description
Searches of the databases yielded 711 references. Of these, 619 were references from searches on media literacy/literacies, and 92 were from searches on multiliteracy/multiliteracies. Duplicates, document types other than research articles and articles without the defined keywords were excluded from the research based on evaluation of the studies' abstracts. After the exclusions, 237 articles were included in the study. Of these, 188 addressed media literacy, and 49 addressed multiliteracies. The statistical software SPSS was used to collect a random sample from each dataset. Based on the reliability reasoning described in the Methodology section, 14 media literacy articles and 14 multiliteracies articles were included in the final qualitative analysis.

Nature of the knowledge produced
One objective of this review was to analyse the nature of the knowledge constructed in these studies. This was done by focusing on the data used in the research and the analysis methods of the research. Based on the data used in the research, three of the analysed media literacy articles used a qualitative approach (Chu & Lee, 2014;Vega & Barranquilla, 2013;Teurlings, 2010). Four of the analysed media literacy articles used a quantitative approach (Chang, Miao, Lee, Chen, Chiu, & Lee, 2014;Ashley, Maksl, & Craft, 2013;Espinoza, Penelo, & Raich, 2013;Mizuno, Narimatsu, Kishi, Kodama, Murashige, Yuji, & Kami, 2010). Two of the analysed articles used a mixed-method approach (Del-Moral & Villalustre, 2013;Sidekli, 2013). Five of the analysed media literacy articles were theoretical in nature and thus not so easily classified on a qualitative and quantitative scale (Andrist, Chepp, Dean, & Miller, 2014;Holladay & Coombs, 2014;Lin, Li, Deng, & Lee, 2013;Radigales, 2013;Tejedor & Pulido, 2012).

Media literacy as an educational outcome
The rigor of media literacy definitions in the articles varied. In some cases, media literacy was defined thoroughly and with great precision, whereas in other cases media literacy was referred to without a proper definition; instead, the meaning of the concept was implied. This variance can originate partly from the research topic and how closely media literacy is related to it. Articles closely related to media literacy focused on the development of the concept (Holladay & Coombs, 2013;Lin et al., 2013;Chu & Lee, 2014;Teurlings, 2010) or on measuring levels of media literacy (Ashley et al., 2013;Espinoza et al., 2013;Del-Moral & Villalustre, 2013;Sidekli, 2013;Chang et al., 2014). In all of these articles, the term media literacy was conceptualised thoroughly and with precision. For instance, topics covered include how media literacy is understood and what the aims of the concept are, what the origins of the concept are and what kind of discussion is related to it-for example, different approaches and point of views. Lin et al. (2013), for example, defined media literacy in detail, because their conceptual article presented a theoretical framework for the concept "new media literacy". The concept of media literacy was, in a sense, at the core of the article. Lin et al. (2013) also described different approaches in the discussion of media literacy and the arguments for promoting it.
In other articles, although media literacy was included in the keywords, the concept played a less significant role. Andrist et al. (2014) analysed the use of videos in the classroom, and the authors saw media literacy as a possible learning goal for some of the video genres; Vega and Barranquilla (2013) described an educational media project related to television viewing in which the aims were related to media literacy; Mizumo et al. (2010) described a study that analysed news reporting practices; Tejedor and Pulido (2012) analysed literature from the perspective of internet risks and Radigales (2013) described the development of opera in relation to the change in media culture. Radigales (2013) discussed the genre from the perspective of media literacy: "Theories on media literacy promote the idea that communications media inform, entertain and teach. When an artistic genre becomes an agent or subject of media literacy, it adopts the prototypical drive that is characteristic of a communications medium" (p. 163). These articles all referred to media literacy to some extent but did not explicitly define the concept. When a concept is not explicitly defined, its contents cannot be analysed. Andrist et al. (2014) used the notion of media literacy as a learning goal for video use in the social studies classroom and clearly related it to critical thinking: "Films can also be used to build media literacy and sharpen critical analysis skills" (p. 197). In introducing the concept of media literacy, Andrist et al. made reference to other media literacy sources. They used the definition provided in these references but did not define the concept explicitly in their article. The authors implied its meaning by giving examples of how media literacy can be built through the analysis of videos from various genres. According to Andrist et al., pop fiction films can be used to identify messages, assumptions and meanings; propaganda films can be analysed from the perspective of persuasion and détournement videos can be useful for analysing the meanings of videos and changing them. These notions imply that media literacy is the ability to analyse and understand the messages, assumptions and meanings of media texts and the ways persuasion is used in the texts. The problem here is that a clear and explicit definition was not given; rather, it is assumed that there is consensus about the meaning of media literacy and that this meaning is obvious to the reader. For example, the boundaries of media literacy are not clearly defined. This raises the question: Do the authors think that these ideas cover all media literacy, or are these just examples of singular aspects of it? For example, these ideas do not include the production or creation aspects, which are relatively common in discussions on media literacy. Mizuno et al. (2010) included media literacy as one of their article's keywords, but the concept was not defined or even mentioned in the actual research text.
By using the keyword, the authors connected the article to the discussion of media literacy, but they did not make any explicit effort to explain the relationship between the contents of their research and media literacy. This connection was left for the reader to make.

Multiliteracies as a pedagogical approach
In most of the analysed articles, the origins of the multiliteracies concept were consistently traced back to 1994, when a group of scholars met in New London, New Hampshire, and 1996, when their article, "A pedagogy of multiliteracies: Designing social futures", was published. The concept's theoretical basis appears to be relatively homogenous, since the origins of the concept can be traced back so clearly. For media literacy, the case is different. Some of the analysed articles traced the first media literacy discussions back even to the 1930s (Lin et al., 2013). Although some references were cited more often than others, a single theoretical basis for media literacy definitions could not be found in the analysed articles. Based on this observation, it can be assumed that the theoretical basis of media literacy is broader than that of multiliteracies.
In line with the conceptualisations presented in the curriculum in Finland, in some articles, multiliteracies or multiliteracy was seen as a set of communication abilities. However, in most of the articles, multiliteracies was also conceptualised and analysed as a pedagogical approach. This differentiates the concept of multiliteracies in research from the concept of multiliteracies in the Finnish core curriculum.
Compared with the media literacy articles, clear definitions of multiliteracies as an educational outcome or ability were hard to find in the multiliteracies articles, but teaching practices and pedagogical questions and content were addressed more thoroughly. In this way, the differences between the concepts of media literacy and multiliteracies became clarified. For example, Tan and Guo (2014) explicated and analysed the use of multiliteracies pedagogy in the context of secondary education. "We implemented the New London Group's (Cope & Kalantzis, 2000) pedagogy of multiliteracies in 2 year two (14-yearold) English language classrooms, in collaboration with their language arts teacher (Tan & Guo, 2009;Tan et al., 2010). The New London Group's (Cope & Kalantzis, 2000) pedagogy of multiliteracies was suitable for the school as it offered a framework for the collaborating teacher to include a range of semiotic modes of meaning-making in a wide array of multimodal texts that the students were likely to encounter in their everyday lives" (Tan & Guo, 2014, p. 31). The concept of multiliteracies was not used to describe just the educational outcomes, but it was also used to describe a pedagogical approach in education.
Besides a pedagogical approach, multiliteracies was also defined as an educational outcome. According to Westby (2010), "Students not only need to be able to communicate effectively in oral and written language, but they also need to communicate effectively in multimodal ways-they need to become skilled in multiliteracies" (p. 64). Multiliteracies is seen as a set of communication skills or abilities which students develop. In line with this is the definition offered by Ajayi (2011): "In this study, multiliteracies is used to refer to the ability to interpret and construct different possibilities of meanings made available by differing text types associated with digital technologies and multimodal texts such as the Internet, video games, digital video, visual images, graphics and layouts" (p. 398). This definition differs in focus from the original writings of the New London Group (NLG). If multiliteracies are presented as a set of skills or communication abilities, the role of a specific pedagogical approach loses its value. In this sense, the outcome is important, not the way it is achieved.
In analysing the articles, it became clear that most of the multiliteracies articles shared a common theoretical basis. The citation patterns reveal a number of references to the writings of the NLG. In 10 of the analysed articles, references were made either to NLG's original multiliteracies article from 1996, in which the authors present the concept for the first time in article form, or to their later writings (Cope & Kalantzis, 2000;Cope & Kalatzis, 2009).
In the original multiliteracies articles (NLG, 1996;Cope & Kalantzis, 2009), the concept's founders did not intend for it to represent a strictly defined set of skills; rather, they intended it to represent a pedagogical approach for English literacy teaching. This approach takes into account the different modes of meaning, which are "dynamic representational resources, constantly being remade by their users as they work to achieve their various cultural purposes" (NLG, 1996). This thinking continued in the later writings of the NLG (Cope & Kalantzis, 2009). Their multiliteracies approach is based on two separate but interconnected societal developments: globalisation and technological development. According to the NLG, the notion of multiliteracies supplements traditional literacy pedagogy with these developments, which also creates the basis for the "multi" in multiliteracies. The first "multi" is multilingualism. Through globalisation, the role of the diversity of languages is increased in world. With regard to different languages, multiliteracies pedagogy also takes into account professions, ethnicity, subcultures and interest and affinity groups. Differences in the use of the same language-in this case, English-in different countries is also taken into account in the pedagogical approach of multiliteracies. Literacy is not, however, merely the formal literacy traditionally promoted in schools; it also covers various forms of literacies outside the school context. Literacies are not limited to a single language but are rather related to a number of different languages. The broadened understanding of literacy also covers ideas about formal and informal literacies. In their article, Marshall, Hayashi and Yeung (2012) analysed how people use different literacies in digital contexts. In their informal literacy practices, in a digital context, they can combine not only different forms of communication but also different languages. "Unlike Julia's sample of academic writing [...], Amy's communication via Facebook has many language forms associated with informal, digital literacies as well as a range of complex multilingual communication strategies" (Marshall et al., 2012). Another "multi" in multiliteracies is multimodality, which involves different modes of meanings, such as linguistic, visual, audio, gestural and spatial. As Ajayi (2011) notes: "Literacy has hitherto been defined as the ability to read and write print-based materials. However, this is increasingly becoming inadequate in the face of digital, multimodal and hybrid textual forms made possible by new media technologies" (p. 398).
An important aspect of multiliteracies pedagogy is the idea of metalanguage, which helps learners and teachers discuss and describe forms of meaning such as language, image, texts and meaning-making interactions (NLG, 1996, p. 77). The purpose of metalanguage should be to describe the differences between texts and connect those to the relevant social and cultural contexts. According to the NLG (1996), metalanguage should be flexible and open-ended in the sense that it can be used as a tool kit for "working on semiotic activities, not a formalism to be applied to them". At the core of metalanguage is the idea of Design. "The metalanguage of multiliteracies describes the elements of Design, not as rules, but as an heuristic that accounts for the infinite variability of different forms of meaning-making in relation to the cultures, the subcultures, or the layers of an individual's identity that these forms serve" (NLG, 1996, p. 88).
NLG authors see teachers as designers of learning processes and environments rather than dictators of doing and thinking. In this way, the pedagogy of multiliteracies emphasises transformability. According to the NLG (1996), meaning-making is an active and dynamic process that is not governed by static rules. The NLG understands any semiotic activity as a matter of Design, which includes three elements: Available Designs, Designing and The Redesigned (NLG, 1996, pp. 74-76). The multiliteracies pedagogy includes four interconnected components: Situated Practice, Overt Instruction, Critical Framing and Transformed Practice.
Although multiliteracies is not defined in the original NLG articles as a teaching and learning outcome as such, but rather as a pedagogical approach for English literacy, skills are presented which are related to the broadened view of literacy. For example, changes in working life have brought a need to develop skills for access to new forms of work. According to the NLG (1996), teachers need to help students develop skills to "speak up, to negotiate, and to be able to engage critically with the conditions of their working lives" (p. 67). In a globalised world and society, students must also be prepared to negotiate "regional, ethnic, or class-based dialectics; variations in register that occur according to social context; hybrid cross-cultural discourses; the code switching often to be found within a text among different languages, dialects, or registers; different visual and iconic meanings; and variations in the gestural relationships among people, language, and material objects" (NLG, 1996, p. 69). The authors also noted that learners can substantively gain metacognitive and meta-linguistic abilities when they juxtapose different languages, discourses, styles and approaches (NLG, 1996, p. 69).

Theoretical framework for media literacies
Despite the active research tradition and substantial discussion of media literacy, consensus about its definition has not been reached. Based on the multifaceted nature of media literacy, this article rejects attempts to establish a universal definition and instead highlights the need for a more nuanced understanding of the concept, presenting a theoretical framework for conceptualising media literacy. This framework is based on abstraction levels and aims to clarify the concept and facilitate its operationalisation in practice.
As we have seen, the theoretical basis of media literacy is wide and multifaceted. Teurlings (2010) described the media literacy discussion as "a movement in which radically opposed paradigms and methodologies meet: exponents of rhetorical criticism find themselves in bed with political economists, media psychologists, cultural studies scholars and mainstream communication researchers; textually oriented scholars work together with audience scholars; and both quantitative and qualitative methods are used" (pp. 359-360). Ashley et al. (2013) noted that inconsistent operationalisation of the concept of media literacy can increase reliance and the possibility of invalid inferences. These ideas call for a better understanding of the concept in relation to its levels of abstraction.
In his classic article, Sartori (1970) presented the idea of a ladder of abstraction divided into three levels: high, medium and low. A concept can be moved along the ladder in relation to its intension and extension-in other words, by widening and narrowing the scope of the concept by reducing or increasing its attributes, or the properties, which determine what belongs under the concept. The challenge is to make extensional gains without losing precision or empirical testability. If a concept stretches too wide in its scope, it loses the ability to be clearly defined; instead of meaning something, the concept means everything (Sartori, 1970). Potter (2013) noted that definitions of media literacy vary in their levels of detail. This was also noted in the analysis of the articles. Some authors may define media literacy with high precision, and others may define it more generally-in other words, with different levels of abstraction. Next, a theoretical framework for conceptualisations of media literacies based on these abstraction levels is presented.
High level of abstraction

Media literacies
Medium level of abstraction Different subsets of media literacy Low level of abstraction Contextualised media literacy definitions Table 3. Theoretical framework of media literacies The analysed articles revealed that media literacy is defined and understood in various ways. Taking this into account, we present a notion of various media literacies, and a theoretical framework for illustrating their differences is presented. The framework divides media literacy definitions into (1) media literacies with a high level of abstraction, (2) different subsets of media literacy with a medium level of abstraction and (3) contextualised media literacy definitions with a low level of abstraction.

Media literacies
The notion of media literacies is in the "high level of abstraction" category. This is the broad umbrella concept, which encompasses the different conceptualisations of media literacy-the subsets of media literacies. Different approaches affect how the media literacies' goals are viewed. For example, Teurlings (2010) used a political criterion to demonstrate that media literacy can have conservative, feminist, liberal and radical approaches. Each of these approaches has its own ideas about which aspects of media literacy to highlight and how the concept should be defined. In the liberal approach, media literacy is seen primarily as an individual ability; in this sense, media literacy is a matter of one's own personal transformation. The radical approach focuses on the transformation of the media system. Clearly, the chosen perspective and theoretical approach will influence how the concepts are understood. This notion highlights the importance of clear conceptualisations. This study's aim in conceptualising the media literacies concept is not to provide a definition for use in empirical studies, but rather to create a space for theoretical discussion and gather the different theoretical perspectives together. The risk here is that if the definition used in the research is too broad, it can result in confusion and a lack of precision. This relates to the ability to discern what is not included under the label of the concept-to define what media literacies are not. When a concept in the "high level of abstraction" category is defined by its negation, the concept can move closer to the concrete level. In this case, the level of abstraction decreases. This is the case in more precisely defined media literacies.
Media literacy is a normative concept in the sense that it is a desired state-an outcome of education (Teurlings, 2010), especially media education (Chu & Lee, 2014). Within the field of media literacy, different goals regarding its promotion have been suggested-for example, the protectionist and empowerment rationales noted earlier (Potter, 2010;Hobbs, 2011;Potter, 2013). Based on the analysis, both of these approaches are still evident in the field. Chang et al. (2014) examined the association of media exposure and media literacy with alcohol and tobacco use, and Espinoza et al. (2013) focused on media literacy's association with body image and eating disorders.
According to a traditional definition, a media literate person can "decode, evaluate, analyze and produce both print and electronic media." (Aufderheide, 1993). This definition forms the basis for media literacy definitions. Teurlings (2010) referred to Firestone (1993), who stated in the forewords of the report of the National Leadership Conference on Media Literacy that media literacy is "the ability of a citizen to access, analyse, and produce information for specific outcomes". Firestone's definition differs somewhat from the previous definition: "access" is added and "evaluate" and "decode" are omitted.
Although these may present a basic framework for definitions, different emphases are placed upon different aspects of media literacy. In a similar way, as Potter (2013) has noted, certain characteristics are added, changed or left out of the definitions. Sidekli (2013) illustrated this, noting that media literacy is a life skill for citizens and it focuses on access to, analysis of and evaluation of written and unwritten messages. Chu and Lee (2014) stated that in Hong Kong, "media literacy is defined as a life skill that enables young people to critically understand, analyse, use, and influence the media" (p. 130). Access, evaluation and produce are not mentioned in this definition, but rather critical understanding and the use and influence of media are emphasised.

The subsets of media literacy
The different subsets of media literacies form the "medium level of abstraction" in this framework. The idea is to emphasise the differences in the definitions and perspectives. This view rejects attempts to establish one universal "fit-for-all" definition, but rather highlights the importance of future discussion on the relationships between the different definitions and subsets. Compared with media literacies, concepts at the medium level of abstraction are defined with greater precision and higher intension. Media literacy can be understood in very abstract or more concrete ways. Based on the articles analysed in this study, the subsets of media literacies include news media literacy (Ashley et al., 2013), public relations literacy (Holladay & Coombs, 2013) and alcohol and tobacco literacy (Chang et al., 2014). These literacies are explicitly defined in relation to media literacy.

Contextualised media literacy definitions
Research always takes place in a specific context, and concepts must be defined precisely. Media literacies should be defined in a way that not only relates to the relevant theoretical discussion but also takes into account the context in which the research takes place and its social and cultural characteristics. Contextualising the definition can also help other researchers and practitioners better understand the concept's meaning. This allows planners and practitioners to utilise media literacy in practice, and researchers can construct knowledge and build on other studies more coherently. Contextualized media literacy definitions relate to the "low level of abstraction", where media literacy needs to be defined in relation to the concrete context in which the actual study takes place.
Media literacy is so multifaceted and multidimensional that in empirical research it must be well contextualised by definition. This makes the contents of the concept more understandable and avoids potential confusion from using the same concept with different meanings. Based on a critical theory of literacy, Chu and Lee (2014) argued that the meaning of media literacy cannot be separated from the social context. This view underscores the importance of definitions for concepts that fall into the "low level abstraction" category. The question is this: What is meant by "media literacy" in the specific research context? Chang et al. (2014) illustrated the importance of contextualised definitions of media literacy by noting that caution should be exercised when comparing media literacy across studies and countries, noting how alcohol and tobacco advertising policies vary from country to country. This suggests that media literacy studies should be placed within social and cultural contexts by defining the concepts with high precision and intension. In discussions and interpretations, studies must also take into account the differing contexts in which the research is conducted.

Discussion and conclusion
To conclude, we raise possibilities for further research, noting important aspects which should be considered more closely and discussed among researchers in the fields of multiliteracies and media literacy. We also point out limitations of this research, which should be taken into account when interpreting the findings.
This study emphasised the differences in media literacy definitions and rejected attempts to establish a universal definition that can be applied in all cases, suggesting instead that media literacy should be understood as multiple media literacies. The study designed a theoretical framework-based on the abstraction levels of the concepts-to facilitate media literacy conceptualisations.
The analysis of the English definitions showed that media literacy is defined in various ways and with different levels of precision. If these differences are recognised and acknowledged, the definitions could more easily be put under critical evaluation. Contextualised media literacy definitions would help to operationalise media literacy in research and practice and facilitate theoretical discussion and cumulative knowledge construction.
In this article, we presented a framework for understanding media literacy. The framework was based on the idea of different abstraction levels. The framework presented in this article is not intended as a tool for establishing a conclusive definition of the concept. Its purpose is not to end the discussionquite the opposite: Its purpose is to emphasise the importance of conceptual clarity and raise discussion. The challenge is to find balance between the scope and precision of the concept.
In this research, several limitations are worth considering. The inclusion criteria omitted a significant amount of relevant media literacy literature; the roots of media literacy research reach back many decades. The decision to limit the study to articles published between 2010 and 2014 was based on the cumulative nature of research. Other reviews have taken into account the earlier research (Martens, 2010;Potter, 2010), and this study supported the findings of these reviews in that there is still variance in the ways media literacy has been defined. The inclusion criteria also omitted literature outside the research realm and research published in languages other than English. For example, much media literacy research is conducted and published in Finland (Pekkala, Pääjärvi, Palsa, Korva, & Löfgren, 2013) and in other European countries (Livingstone, Papaioannou, Grandío Pérez, & Wijnen, 2012), not to mention the other language areas in the world. However, the role of English in international scientific communication is important to take into account, and thus the inclusion criteria used in this study is arguable.
This study argued for a stronger focus on the relationship between the concepts of multiliteracies and media literacy. The contents of these concepts clearly overlap; for example, both recognise multimodality. This raises the question: Do the theoretical discussions share a common frame of reference and, if so, to what extent? Are any shared sources relevant to both discussion traditions? For reasons of synergy, it would be useful to find ways to explicate these mutual ideas for bridge building between these thus-far differing fields. Creating common references-for example, through bibliographical analysesis one possible starting point. Interviews with influential researchers in both fields could provide insight into the relationships between the two concepts. More theoretical and practical research is needed to discover the mutual connections between the research traditions. This could open up possibilities for cooperation between the scholars of multiliteracies and media literacy.
This study demonstrated that media literacy and multiliteracies research have used a number of methodological approaches. This theoretical variance suggests that more discussion on theoretical questions is needed. What possible challenges do the differing theoretical bases pose? What kinds of theoretical-epistemological and ontological-presumptions are related to the research traditions? What does the variance in theoretical presumptions mean for the nature of constructed knowledge and the foundation of cumulative knowledge?
Further discussion is needed on the pedagogical approach to multiliteracies in Finnish education, since multiliteracies is defined as an outcome in the core curricula. Due to the overlapping contents of the concepts, research and conceptualisations of media literacy can offer useful support for education planners and educators in implementing outcome-driven multiliteracies in the Finnish school system. The theoretical framework presented in the article facilitates a better understanding of the various conceptualisations of media literacy. The framework also highlights the importance of contextualisation, which can lessen confusion about the concepts' meanings. In media literacy research, contextualised definitions are also essential for the construction of cumulative knowledge.