2.1 Definitions and conceptions of plagiarism

Dictionary definitions of plagiarism (e.g. ‘the practice of taking someone else's work or ideas and passing them off as one's own’, as defined by the Oxford English online dictionary) are typically brief, and perhaps for that reason definitions in academic contexts are frequently elaborated to specify the varied types of intellectual property and textual material which can be appropriated in a way that constitutes plagiarism.

Apart from varying levels of detail, university definitions tend to be quite consistent (see section 2.3); the only significant exception to this is the question of intention. Both in policy, guidelines and similar documentation, and in the common understanding of teachers and others who employ the term, the status of intention is vexed, with some sources claiming that it must be a feature of plagiarism – if the writer did not intend to copy or intend to cheat by doing so, the act cannot legitimately be called plagiarism – and others (if a minority) maintaining that intention is irrelevant (Pecorari Reference Pecorari2001; Sutherland-Smith Reference Sutherland-Smith2011).

2.2 Plagiarism as cheating or academic crime

To understand how intention came to be linked to plagiarism, it is necessary to examine the origins of the concept. The word comes from the Latin plagium (theft, kidnapping), originally referring to kidnapping a child or a slave; in the seventeenth century it emerged in the meaning of ‘literary theft’, that is, appropriation of literary texts (Mallon Reference Mallon1989). Although textual appropriation was not a new phenomenon, it was the perceptions of and attitudes to it that began to change as mass-produced texts became widely available as a result of printing innovations. Texts and ideas expressed in written form came to be seen as their authors’ intellectual property, leading to the emergence of the notions of authorship and originality, and resulting in the introduction of copyright laws in eighteenth-century EnglandFootnote ¹ (for a more detailed historical account, see Sutherland-Smith Reference Sutherland-Smith2008 and Bloch Reference Bloch2012). The notion of plagiarism in the English-speaking world is thus historically linked with the issues of authorship, originality, ownership of texts and copyright (Scollon Reference Scollon1995). It is this understanding of plagiarism with its network of related concepts that was subsequently adopted by English-speaking academia.

The historical context is important for understanding the traditional view of plagiarism as theft, and thus a crime, which is still widely accepted in academic and media debates alike. Many definitions of plagiarism found in scholarly literature rest on this conception, describing plagiarism as ‘literary theft, stealing (by copying) the words or ideas of someone else and passing them off as one's own without crediting the source’ (Park Reference Park2003: 472). As noted above, this view is also reflected in university policies about plagiarism in English-speaking academia, which consider it as an academic crime deserving punishment (Pecorari Reference Pecorari2001).

Plagiarism has long been used as an umbrella term covering various types of unacceptable behaviour, some of which, but not all, refer to textual activity. Poor referencing, inadequate paraphrase and inaccurate citation are sometimes placed in the same category as commissioning a paper from a commercial service and submitting another student's paper as one's own. Problems arising out of students’ lack of knowledge of citing conventions and lack of skills in using sources are thus placed in the context of cheating. As cheating is inherently intentional and deliberate, all the practices to which the term plagiarism traditionally refers, including poor referencing, are therefore seen as dishonest and deceitful behaviour. As a result, plagiarism is considered a moral transgression and a reflection of moral decay. Various metaphors that have been used in connection with plagiarism vividly illustrate this view: plagiarism has been described as a sin, vice, disease, cancer, plague, stealing and a crime (Howard Reference Howard1999; Park Reference Park2003; Bloch Reference Bloch2012).

Although it is commonly acknowledged (though not universally; see sections 2.1 and 2.3) that students may plagiarise inadvertently because of their lack of knowledge of citing conventions and poor paraphrasing skills rather than an intention to deceive, the traditional view nevertheless defines such textual behaviour as literary theft that needs to be punished. For instance, Park (Reference Park2004) outlines an institutional framework for dealing with plagiarism, consisting of a four-level scale of sanctions and penalties. In recognition that ‘a clear distinction must be drawn between inexperienced academic study and writing skills (especially among first-year undergraduates and international students) and wilful cheating and deception’ (p. 301), the framework treats issues such as inappropriate paraphrasing and incorrect or incomplete citations, all of which may commonly occur in the writing of inexperienced writers, as a minor first offence. However, if the problem occurs again, it is treated as the second offence, resulting in a mark of 0 for the assignment, because ‘[a]fter a first offence it is assumed that a student who plagiarizes is more likely to do so consciously, which is why the ladder of penalties . . . rises steeply’ (p. 302). The underlying assumption that mastering the use of sources is a matter of a quick fix, as we will show in section 3.2, has been criticised in recent years, particularly in reference to L2 writers.

2.3 Institutional policies

An analysis of the policies of 54 English-speaking universities in the United States, United Kingdom and Australia (Pecorari Reference Pecorari2001) found definitions of plagiarism as an intertextual relationship (material is taken from a source and transferred to another), one which is not conventional or appropriate (because it is not acknowledged) and one which is not coincidental; it has been brought about intentionally by an agent who may (or may not) intend to deceive the reader about the inappropriate intertextual relationship.

This study, like that of Sutherland-Smith (Reference Sutherland-Smith2011), found that deceptive intent is not always specified and indeed that some definitions specifically state that plagiarism can occur unintentionally. However, an orientation toward deception is implicit in how plagiarism is defined and discussed. Although in the intervening years a growing body of research has shown that much language re-use has motivations other than an attempt at cheating, both of these analyses of policy found that definitions of plagiarism were couched in legalistic terms, that it was treated and discussed analogously to theft, and that the response to the act is punitive.

A further trend in institutional policies is for them to be neutral with respect to academic discipline. This is presumably due to a sense that the playing field should be level and that rules should not vary within the same organisation; at the same time, the differences between subject areas in terms of the features of writing which are conventional extend to the use of sources, and many academics perceive university policies as a poor fit for their own realities (Jamieson Reference Jamieson2008).

3.1 The emergence of plagiarism as a theme in L2 writing research

The earliest mentions of plagiarism in the L2 writing literature moved away from equating it with cheating and gave it as an example of behaviour which was neither norm-compliant nor openly norm-defiant, but simply in need of understanding. It was often mentioned in passing, to illustrate broader points about contrastive rhetoric, culturally varied awareness, or developing discoursal and linguistic abilities and their impact on the writing of L2 users (e.g. Ballard & Clanchy Reference Ballard and Clanchy1991; Pennycook Reference Pennycook1996; Leki & Carson Reference Leki and Carson1997).

In a frequently cited landmark article, Matalene (Reference Matalene1985) argued a point, now generally accepted, that western teachers have a responsibility ‘to admit the relativity of our own rhetoric, and to realize that logics different from our own are not necessarily illogical’ (p. 807). She supported this conclusion with a number of episodes from her time spent teaching in China, in one of which her students adopted a strategy on a writing assignment which she saw as plagiarism but which they explained was a strategy they had been encouraged to use:

Accounts also appeared in the literature of L2 writers using plagiarism (a term not always used by these authors) as a way of populating their texts with accurate and idiomatic expressions:

For these writers plagiarism was a textual resource (Shaw Reference Shaw1991), an academic survival skill (Currie Reference Currie1998), or the by-product of the student's struggle, not yet won, to find her own authorial voice (Cadman Reference Cadman1997; Spack Reference Spack1997).

By charting some of the sources of textual plagiarism, these early works provided an alternative to the understanding of plagiarism as an unethical and quasi-criminal act. If plagiarism in the writing of international students can be traced back to different understandings of what is valued in a text (Gregg Reference Gregg1986) or a collectivist background predisposing toward collaboration (Leki Reference Leki1992), then alternative explanations to intentional deceit can be found.

3.2 Criticism of the traditional notion of plagiarism

Building on this broadened understanding, since the 1990s a growing number of scholars have criticised the traditional notion of plagiarism from a range of perspectives, particularly in relation to L2 writing. One strand of criticism focuses on the origin of plagiarism to show that it is not as universal a concept as the received view implies. The notion of plagiarism as literary theft emerged in response to a particular set of economic, social and technological conditions specific to eighteenth-century England, and is thus a historical and cultural phenomenon, which may not be applicable to writing cultures and rhetorical traditions marked by a different course of development (Scollon Reference Scollon1995; Pennycook Reference Pennycook1996). Fundamental to the traditional view of plagiarism is the idea of the individual ownership of discourse; in other words, the view of texts as products with commercial value, which are owned as individual property. Scollon shows how this ideological position, which served as the basis for the emergence of copyright laws, is incompatible with current theories and research on the nature of communication (for discussions of plagiarism, intellectual property and copyright in today's world, see Buranen & Roy Reference Buranen and Roy1999 and Bloch Reference Bloch2012).

Another underlying assumption of the traditional view of plagiarism that has received criticism is the Enlightenment notion of the author as an autonomous individual who is a sole creator of original works through self-expression. The notions of unitary authorship and originality, both pillars of the received view of plagiarism, have been shown to be problematic (see, e.g., Scollon Reference Scollon1995; Pennycook Reference Pennycook1996; Ivanič Reference Ivanič1998; Myers Reference Myers1998; Howard Reference Howard1999). These scholars have argued that authorship, rather than resulting from an autonomous individual's expression of his/her original thoughts, can be seen as constructed in discourse through interaction and negotiation with other texts, and through the reuse of available linguistic and other resources. The current thinking, influenced by the Bakhtinian notion of the dialogic nature of language, is that writing is essentially intertextual: it always contains traces of other texts, either explicitly (such as through citation) or implicitly (through following a particular textual configuration), a distinction that Fairclough (Reference Fairclough1992) termed manifest and constitutive intertextuality.

Chandrasoma, Thompson & Pennycook (Reference Chandrasoma, Thompson and Pennycook2004) criticise the notion of plagiarism precisely because it oversimplifies textual relations. They consider intertextuality, which they define ‘not as a mere manifestation of texts within a text but rather as multiple strands of knowledges within texts designed to produce desired meanings’ (p. 175), to be essential in student writing. They distinguish between three intertextual modes: conceptual, which refers to appropriating concepts from sources; complementary, which ‘complements the theme or themes of a text while reinforcing the writers’ point of view’ through examples and allusions; and metalinguistic, which includes ‘linguistic resources (e.g., specific terminologies, stance markers) used in a text’ (p. 175). Their analysis of cases of student writers and their texts shows that determining what constitutes plagiarism is complex and dependent on context, with factors such as what counts as common knowledge, student development, disciplinary conventions, authorial identity and others all playing a role. They therefore question the validity of plagiarism policies designed to apply to all disciplines, levels of study, courses, and to students at different levels of writing development.

The notion of plagiarism has also been criticised on pedagogical grounds as incompatible with theories of learning. The implied emphasis on individual originality does not allow for learning through imitation and appropriation, and is therefore detrimental to writing development (Howard Reference Howard1999). The fact that extremely diverse textual practices, ranging from buying a complete paper to inexpert paraphrase, are lumped together under the plagiarism label has also been criticised as not only causing definitional confusion but also resulting in treating a stage of writing development as dishonest and ethically questionable behaviour (Chandrasoma et al. Reference Chandrasoma, Thompson and Pennycook2004; Valentine Reference Valentine2006). Finally, a growing body of research into L2 students’ source-based writing has revealed language-related problems which may lead to inadvertent plagiarism. For example, limited linguistic repertoires and reading skills lead some writers to produce what they consider to be paraphrases or summaries, by virtue of small, local changes to the source. Other readers (such as teachers) may regard the limited changes as inadequate and therefore plagiarism. This situation has lead researchers to call for extensive pedagogical support for L2 writers rather than punishment.

These criticisms have led to efforts to reframe plagiarism in ways that more accurately reflect current knowledge about discourse, writing and learning, as discussed in the next section.

4.1 Types of plagiarism and alternative terms for different source-use practices

There is now ample evidence that students may plagiarise unintentionally, as a result of uncertainty about citing conventions, what constitutes common knowledge, or limited referencing skills and/or L2 resources (e.g., Sherman Reference Sherman1992; Currie Reference Currie1998; Howard Reference Howard1999; Errey Reference Errey2002; Pecorari Reference Pecorari2003; Petrić Reference Petrić2004; Flowerdew & Li Reference Flowerdew and Li2007; Ellery Reference Ellery2008a; Shi Reference Shi2010; Li & Casanave Reference Li and Casanave2012). Although establishing intention is always problematic, the sheer volume of studies reporting that students are concerned to avoid plagiarism but lack the experience or perspective which would allow them actually to do so makes it virtually impossible to avoid the conclusion that not all plagiarism is intended to deceive. Concerns have been raised about unintentional plagiarism being treated as a form of deliberate cheating, leading to proposals for separating these types of source-use practices. Two approaches can be identified in the literature: distinguishing between different types of plagiarism and replacing the term ‘plagiarism’ by alternatives.

Howard (Reference Howard1999) introduced the well-known term patchwriting, defined as ‘copying from a source text and then deleting some words, altering grammatical structures, or plugging in one synonym for another’ (p. xvii) to denote unintentional plagiarism, arguing that it should not be considered part of the plagiarism category. Although patchwriting has proved to be a useful and influential concept (see next section), distinguishing between intentional and unintentional textual behaviour is fraught with difficulty. Recognising this, Pecorari (Reference Pecorari2008) builds on Howard's work, introducing a two-level typology of plagiarism, with textual plagiarism, defined as ‘the use of words and/or ideas from another source, without appropriate attribution’ (p. 4), as a general term, further divided into two sub-types, prototypical plagiarism, where intention to deceive is present, and patchwriting, where no such intention exists. This classification allows for discussions of source use and textual analysis without reference to the writer's intention to deceive or the need to establish it, while at the same time opening up the possibility of taking intent into consideration if it can be established.

A different set of proposals, although with a similar aim, have been made by scholars such as Chandrasoma et al. (Reference Chandrasoma, Thompson and Pennycook2004) and Petrić (Reference Petrić2004), who have argued that the notion of plagiarism should be abandoned as imprecise and pedagogically detrimental, and replaced by terminology that would provide a more accurate description of source-use practices. The call for more precise and ‘ethically-neutral language’ (Petrić Reference Petrić2004) should not be seen as merely a search for a euphemism but rather as a way to shift the emphasis from crime and moral transgression to textual relations (Chandrasoma et al. Reference Chandrasoma, Thompson and Pennycook2004).

Various terms have emerged in the literature, including ‘language re-use’ (Flowerdew & Li Reference Flowerdew and Li2007), ‘mediated authorship’ (Prior Reference Prior1998) and ‘textual appropriation’ (Shi Reference Shi2006, Reference Shi2010), each of which highlights a slightly different aspect of source-based writing and relations between the cited and the citing text. The terms most commonly used are ‘textual borrowing’ (Currie Reference Currie1998; Barks & Watts Reference Barks and Watts2001; Shi Reference Shi2004; Petrić Reference Petrić2012) and ‘transgressive/nontransgressive intertextuality’ (Chandrasoma et al. Reference Chandrasoma, Thompson and Pennycook2004; Abasi & Akbari Reference Abasi and Akbari2008; Borg Reference Borg2009).

The underlying assumption behind these terms is that explicit intertextuality is one of the essential features of academic writing. Unlike plagiarism, these terms do not have negative connotations – for instance, textual borrowing is defined as ‘appropriate integration and documentation of other texts’ (Barks & Watts Reference Barks and Watts2001: 246) – but they can also be used to describe misuse of sources (e.g., illegitimate textual borrowing, transgressive intertextuality). In a recent study of faculty attitudes to student intertextuality, Pecorari & Shaw (Reference Pecorari and Shaw2012) provide a typology of intertextuality, classifying it as acceptable or unacceptable, with some types likely to be subject to interpretation in context: indirect (e.g., formulaic phrases), conventional (e.g., signalled quotation or paraphrase), unconventional (e.g., a quotation is not signalled although the source is acknowledged) to deceptive (where the writer does intend to deceive).

These efforts to change the terminology about source use also represent a conceptual change, as new terms allow us to discuss textual practices from a different perspective without invoking the stigma inseparable from plagiarism. However, the proliferation of terms may also cause confusion outside the specific research areas where they are commonly employed, which may be one of the reasons why most scholars use these terms alongside plagiarism. Ultimately, both approaches reflect the need for more accurate descriptions of the complex issue of source use.

4.2 Patchwriting as a developmental strategy

That close reliance on sources may be a necessary stage in the development of students’ academic writing was suggested as early as Campbell (Reference Campbell1990); this argument has most notably been developed in Howard's (Reference Howard1995, Reference Howard1999) patchwriting model. More recently, Abasi & Akbari (Reference Abasi and Akbari2008) have added a further distinction between localised patchwriting, defined as ‘close appropriation at the micro level of lexis and syntax’ (p. 270) and global patchwriting, referring to ‘ineffective appropriation at the level of ideas’ (p. 271).

A growing body of research has provided empirical evidence of patchwriting in L2 students’ writing (e.g., Currie Reference Currie1998; Angélil-Carter Reference Angélil-Carter2000; Pecorari Reference Pecorari2003, Reference Pecorari2008; Chandrasoma et al. Reference Chandrasoma, Thompson and Pennycook2004; Villalva Reference Villalva2006; Flowerdew & Li Reference Flowerdew and Li2007; Ouellette Reference Ouellette2008; Li & Casanave Reference Li and Casanave2012; Davis Reference Davis2013; Li Reference Li2013). Recent studies have focused on tracing the development not only in students’ source-based texts (i.e., the products of writing), but also in their conceptualisations of source use within a complex network of related issues, including their views of themselves as authors, their attitudes to knowledge, and the educational practices into which they were socialised. These studies show that L2 students at the patchwriting stage tend to be unaware of the potential problems involved in this practice, although they are concerned about committing plagiarism inadvertently. They also tend to hold naïve beliefs about source texts, which they see as authoritative containers of uncontested knowledge, while they regard citing as simply a matter of correct application of citation conventions (Abasi & Graves Reference Abasi and Graves2008). In a study of more and less experienced L2 academic writers, Abasi, Akbari & Graves (Reference Abasi, Akbari and Graves2006) found that while the more experienced writers had a sense of their own authorial identities and were able to create a representation of themselves as authors via intertextual links, the less experienced writers, whose texts exhibited numerous instances of patchwriting, did not perceive themselves as authors who could argue or disagree with other authors. Similarly, a longitudinal interview-based study of ten Chinese students at a UK university shows that developing academic literacy requires ‘a conceptual understanding of knowledge construction and conventions in the dominant academic community, rather than practice of mechanical aspects of citing and referencing’ (Gu & Brookes Reference Gu and Brooks2008: 338, emphasis in the original). However, as Abasi & Akbari (Reference Abasi and Akbari2008) have shown, faculty practices may inadvertently encourage patchwriting by placing unrealistically heavy demands on students, who then resort to it as a coping strategy.

There is an awareness arising out of this body of research that while patchwriting is not an effective use of sources, it is a useful transitional stage that helps students develop as writers both in terms of raising their rhetorical awareness and in terms of their writing practices. Although this view is becoming widespread among researchers in the area of L2 writing, it is still not widely accepted by practitioners or by researchers in other fields.

4.3 Appropriation as a way of language learning

While research suggests that patchwriting contributes to the development of source-use skills at a broader discoursal level, there is also evidence that language appropriation facilitates learning at a more micro level, resulting in the acquisition of academic phraseology, which is an essential element of academic literacy. Villalva (Reference Villalva2006) traced the literacy development of two Latina bilingual high school students in the United States during a year-long writing project the students were required to complete. Her analysis of one of the students’ multiple drafts identified extensive use of patchwriting, resulting in a mix of academic language taken from the source and informal language of the student's paraphrase. However, the final draft revealed another textual practice, consisting of original, rather than patchwritten, sentences crafted with the help of some lexical material from the source texts. For instance, the essay displayed effective topic sentences, which contained words such as ‘unrealistic’, ‘strive’ and ‘depict’, all of which occurred more than once in one of the source texts the student was drawing on for her project. In other words, the source text became a resource for extending the student's academic repertoire, which was necessary for the writing task. Interestingly, Villalva noted that the student later used the same words in an oral presentation of her project, showing that elements of academic literacy acquired from a source text became part of her own academic language. This case shows that patchwriting, followed by a more independent and creative use of building blocks acquired from source texts, is part of the process of development of both oral and written academic language.

This study thus provides empirical evidence of what Bakhtin (Reference Bakhtin, Holquist, Emerson and Holquist1981) described as the process of making unfamiliar and foreign words our own. Similar processes have been documented in the writing of students acquiring the discourses of their disciplines (Prior Reference Prior1998). In a study of more and less accomplished L2 master's students, Petrić (Reference Petrić2012) found what may be a more advanced stage in academic literacy development, in which the more successful writers used excessive quotation fragments in their writing to signal their language appropriations, a practice that can be still regarded as a learning strategy but one that reflects students’ conceptual awareness of the difference between transgressive and nontransgressive intertextuality. It has also been shown that source texts are used as resources not only for acquiring academic lexis but also for acquiring knowledge about genre features (Tardy Reference Tardy2006), which echoes Chandrasoma et al.'s (Reference Chandrasoma, Thompson and Pennycook2004) point about intertextuality playing multiple roles in student writing.

The findings of this body of research reveal that the anti-plagiarism dictum ‘say it in your own words’ oversimplifies the task for students encountering a new discourse, who do not yet possess the appropriate linguistic repertoire necessary for writing about academic or discipline-specific topics and who thus need to rely on source texts in order to acquire it.

5.1 Students’ and teachers’ views

If policy definitions are extensive at times (implying complexity) and partially contradictory (at least as far as intention is concerned) they are intended to be perceived as straightforward and definitive, their legalistic tone suggesting that, as Sutherland-Smith observes, they ‘position members of the Academy either as right or wrong’ (Reference Sutherland-Smith2011: 136). When asked about their own understandings of plagiarism, though, as a number of researchers have done, both students and staff are much more hesitant and much more varied in their constructions of the concept.

The meaning of plagiarism for individuals has been investigated by means of interviews with staff (Sutherland-Smith Reference Sutherland-Smith2005a; Borg Reference Borg2009) or students (Ashworth, Bannister & Thorne Reference Ashworth, Bannister and Thorne1997; Ashworth, Freewood & Macdonald Reference Ashworth, Freewood and Macdonald2003; Flint, Clegg & Macdonald Reference Flint, Clegg and Macdonald2006; Blum Reference Blum2010). It has also been investigated in a text-based format, with specific instances of intertextuality used to stimulate participants’ responses (Pecorari & Shaw Reference Pecorari and Shaw2012, who researched staff). Text- or scenario-based questionnaire studies have been conducted by Julliard (Reference Julliard1994; students, staff and editors); Barrett & Cox (Reference Barrett and Cox2005; staff and students); and Roig (Reference Roig1997; students and Reference Roig2001; staff).

An overarching theme of these studies is the variety of responses to, experiences of, and meanings attributed to plagiarism. Students’ understandings of plagiarism differ from those of staff, or from a received understanding of what is conventional and appropriate (Deckert Reference Deckert1993; Julliard Reference Julliard1994; Roig Reference Roig1997; Barrett & Cox Reference Barrett and Cox2005; Blum Reference Blum2010). In one survey, only 30% of students identified the act of repeating language from a source without signalling it as quotation (but with a reference to the author) as serious plagiarism (Löfström & Kupila Reference Löfström and Kupila2013), a finding confirmed by Li's (Reference Li2013) case studies. Importantly, though, two studies (Ashworth et al. Reference Ashworth, Bannister and Thorne1997; Ashworth et al. Reference Ashworth, Freewood and Macdonald2003) have shown that this is not entirely either ignorance or a flouting of accepted principles; rather, students are attuned to issues of honesty, ethics and developing as academic writers but apply those values in ways which may surprise more experienced academic writers. This is echoed in Petrić's (Reference Petrić2004) case study, which showed a broad cleft between a student's understanding of her writing and that of her teacher.

Indeed, academic staff and other gatekeepers have a variety of understandings of what plagiarism is (e.g., Crocker & Shaw Reference Crocker and Shaw2002; Shi Reference Shi2012). Sutherland-Smith's (Reference Sutherland-Smith2005a) respondents disagreed about whether intention is a necessary component. While most believed the distinction between intentional and inadvertent plagiarism was important, two thought that ‘all acts of copying are plagiarism, and intention is therefore automatically proven’ (p. 89). Roig (Reference Roig2001) asked university teachers from a variety of disciplines to comment on rewritten versions of an excerpt from an academic text, and to decide if they represented plagiarism, and found a range of responses. He then asked a smaller sample from the same group to produce their own paraphrases, and found that some re-used language from their source to an extent which ‘could be interpreted by others as possible plagiarism’ (p. 215).

The hedging in that conclusion is not surprising given the findings of Pecorari & Shaw (Reference Pecorari and Shaw2012), who showed eight university teachers from the disciplines of natural sciences, engineering and medicine five examples of source use. The teachers found it very difficult to reach a decision and ultimately disagreed with each other in most cases. Although they invoked definitional criteria in their explanations – whether there was an intertextual relationship, whether it was appropriate and whether there was intention to deceive – the individuals weighed up different pieces of evidence in different ways to arrive at their conclusions. Borg (Reference Borg2009) confirms that while the lecturers he interviewed had thought about plagiarism a great deal, ‘personal experience affected [their] perceptions’ (p. 418).

5.2 Disciplinary differences in conceptualising plagiarism

Some of this variation may be due to the disciplinary background of respondents. Several studies have investigated frequencies in plagiarism in particular academic disciplines. McCullough & Holmberg (Reference McCullough and Holmberg2005) examined master's dissertations in a range of disciplines and found that engineering students were more likely to have copied language from sources than other groups, and Selwyn's (Reference Selwyn2008) survey of self-reported student behaviour confirmed that students in engineering and technology were the most likely to have copied small amounts of text, though other groups were more likely to have copied larger amounts. Julliard (Reference Julliard1994) showed examples of source use to four groups: university teachers in medicine and in English (as well as journal editors and medical students) and found differences between the views of academic staff, with those from medicine more accepting of language repeated from the source. Rinnert & Kobayashi's (Reference Rinnert and Kobayashi2005) survey of Japanese students found that those in the arts were more likely than those in the sciences to report that their teachers would object to plagiarism.

While these differences could in principle reflect differing degrees of proclivity to cheat, it is clear that other explanations exist as well. Jamieson (Reference Jamieson2008) relates the comments of her colleagues across disciplines: ‘Most of these rules about how to use and cite sources don't actually apply in my discipline’ (p. 77); this has to do with the well known and frequently observed differences in how textual practices are valued across academic disciplines. Pecorari (Reference Pecorari2006) showed that textual plagiarism was more common in the works of students in the STEM disciplines (science, technology, engineering and mathematics), in other words, precisely those disciplines which do not ‘permit’ explicit, signalled quotation (in the sense that it is vanishingly rare and thus highly unconventional).

Borg's interview participants also revealed both differences in perspectives and a disciplinary grounding for those differences. For example, a historian observed

In the eyes of writers and their evaluators, then, plagiarism is highly contextual. Definitions with regulatory force in universities treat it as a legalistic and relatively straightforward matter: there may be a question in specific cases about whether plagiarism occurred, but when it has, it is always wrong. The literature on L1 composition and L2 writing has, however, produced descriptions of an act which appears to be plagiarism in terms of intertextual relationships but does not carry the same implication of moral failing. Studies to date have demonstrated not only that teachers and students perceive the degree of complexity in the notion of textual plagiarism which is hidden by institutional definitions, but also that because of that complexity, individuals’ understandings of plagiarism are sharply divergent.

6.1 Plagiarism and the electronic media

The spread of the Internet and the development of new information technologies have made a vast range of sources readily available to student writers. Recent research shows that both L1 and L2 students are increasingly likely to turn to the Internet for sources for their academic assignments. For instance, 91% of 186 students in Sutherland-Smith's (Reference Sutherland-Smith2008) study at an Australian university reported using the Internet as a source of information when writing assignments. Given this trend, concerns have been raised about the growth of online plagiarism due to the ease with which students can appropriate texts from the Internet by using copy and paste, as well as engaging in other types of academic dishonesty such as purchasing essays from online paper mills (e.g., Szabo & Underwood Reference Szabo and Underwood2004). In this section, we review studies dealing with the relationship between the Internet and plagiarism, focusing specifically on three issues: (i) the extent of Internet plagiarism, (ii) the role of electronic resources in the writing process of L2 students, and (iii) the changing notions of textual ownership as a result of new information technologies.

6.2 Plagiarism and culture

As noted above, the connections often asserted between culture and plagiarism were in large part responsible for the appearance of the topic on the L2 writing research agenda. A rapidly growing body of work has sought empirical evidence for the proposition that culture may explain some textual plagiarism, using approaches such as case studies; questionnaires, interviews or focus groups with teachers and/or students about understandings and attitudes; text-based interviews; and comparisons of student writing with its sources. The findings from these studies are summarised below.

• • Earlier anecdotal accounts of a tendency towards source-based writing strategies with inadequate attribution have been amply confirmed; this is a common strategy for L2 writers (Campbell Reference Campbell1990; Currie Reference Currie1998; LoCastro & Masuko Reference LoCastro and Masuko2002; Pecorari Reference Pecorari2003; Petrić Reference Petrić2004; Shi Reference Shi2004; Abasi et al. Reference Abasi, Akbari and Graves2006; Chatterjee Reference Chatterjee2006; Flowerdew & Li Reference Flowerdew and Li2007; Abasi & Akbari Reference Abasi and Akbari2008; Oullette Reference Ouellette2008; Li & Casanave Reference Li and Casanave2012).

• • L2 writers view a source text as a ‘language repository’ (Plakans & Gebril Reference Plakans and Gebril2012: 30) and copying is sometimes mediated through the practice of recording potentially useful phrases in a notebook for later use (Flowerdew & Li Reference Flowerdew and Li2007; Gu & Brooks Reference Gu and Brooks2008).

• • It may not be identified by the teacher (Currie Reference Currie1998; Pecorari Reference Pecorari2008; Li & Casanave Reference Li and Casanave2012).

• • In tasks asking them to judge the acceptability of examples, scenarios or descriptions of source use, L2 writers often give unexpected answers (Deckert Reference Deckert1993; Chandrasegaran Reference Chandrasegaran2000; Shi Reference Shi2006); in direct comparison, some source-use strategies are more frequently identified as plagiarism by L1 than by L2 writers (Hayes & Introna Reference Hayes and Introna2005; Marshall & Garry Reference Marshall and Garry2006; Sivasubramaniam Reference Sivasubramaniam2006).

• • Teachers express a belief that cultural differences explain some plagiarism (Bretag Reference Bretag2004) and some L2 writers perceive differences between the rules and practices that are acceptable in their new educational context and those of their home countries (Shi Reference Shi2006; Abasi & Graves Reference Abasi and Graves2008).

Explanations offered for differences in source-use practices with special applicability to L2 writers include

• • limitations on students’ linguistic and discoursal skills (Currie Reference Currie1998; Hayes & Introna Reference Hayes and Introna2005), on their prior experience of assessment writing tasks (Hayes & Introna Reference Hayes and Introna2005, Reference Hayes and Introna2006; Chatterjee Reference Chatterjee2006), or on their exposure to explicit instruction in writing and/or source use (Rinnert & Kobayashi Reference Rinnert and Kobayashi2005; Shi Reference Shi2006)

• • experience of and facility with memorisation, because of an educational background which has placed emphasis on memorisation and rote learning

• • delays in developing an authorial voice (Currie Reference Currie1998; Ouellette Reference Ouellette2008), possibly due to an educational background which did not involve extensive writing for assessment and/or did not place an emphasis on self-expression in writing

• • confusion as to what is required (Evans & Youmans, Reference Evans and Youmans2000; Gu & Brooks Reference Gu and Brooks2008) or a belief that inappropriate strategies were in fact legitimate (Pecorari Reference Pecorari2003; Lankamp Reference Lankamp2008; Li Reference Li2013), or declarative knowledge about what university rules say about plagiarism combined with insufficient procedural knowledge to avoid it (Pecorari Reference Pecorari2003; Li & Casanave Reference Li and Casanave2012)

• • a need for survival or coping skills (Currie Reference Currie1998; Rinnert & Kobayashi Reference Rinnert and Kobayashi2005) when confronted with the challenges of adapting to a new academic context and negotiating it through the medium of an L2

• • uncertainty about distinguishing between one's own and others’ ideas (Rinnert & Kobayashi Reference Rinnert and Kobayashi2005; Shi Reference Shi2006).

Yet despite this seemingly clear indication that source use is a problematic area for L2 writers, and the temptingly plausible explanations for it grounded in their linguistic and educational backgrounds, there is evidence to challenge the idea that L2 writers are different in this respect from their L1 peers. Three studies which looked at the source-use behaviour of L1 and L2 writers found no significant differences between the two groups (Hyland Reference Hyland2009; Martin Reference Martin2012; Weigle & Parker Reference Weigle and Parker2012), and some studies of attitudes and ability to recognise problematic source use have similarly found L2 and L1 users broadly in line (Evans & Youmans Reference Evans and Youmans2000; Maxwell, Curtis & Vardenega Reference Maxwell, Curtis and Vardanega2008).

What is the source of this disagreement? Part of it is likely to be traceable to differences in methodologies and approach. Many of the studies addressing this question have compared L2 writers not with L1 writers but with what the researchers assume to be the received view of good practice. However, as noted above, there is ample evidence that experienced academic gatekeepers differ in their view of what is and is not appropriate, so studies using a nominal received view as a baseline for comparison run into an immediate methodological difficulty.

Others have not attempted such a contrast directly, but have simply described the L2 writer situated in new environments. It may be, therefore, that the issues above are pertinent but not unique to the L2 writer. Indications that this may be the case come, for example, from Shi's (Reference Shi2006) interview study in which both L1 and L2 respondents expressed uncertainty about defining the boundaries of intellectual property, and Howard et al.'s (Reference Howard, Serviss and Rodrigue2010) analysis of L1 student texts, which showed patchwriting to be a common strategy among that group as well.

Even if the evidence showed unanimously that L2 and L1 writers use sources in clearly different ways, making the connection to culturally grounded explanations would be much more difficult. If our home culture predisposes us to think, and therefore to act, in a particular way – an assumption which is itself contested – it is not the only influence on our actions. Furthermore, our home culture consists of a set of predispositions which do not always pull in the same direction. These observations – which are essentially about the notoriously slippery nature of any culturally grounded explanations for behaviour – may account for the fact that the body of informed commentary on plagiarism and culture is as fragmented and contradictory as the body of empirical evidence.

Assertions of cultural differences come not only from Western observers but from individuals steeped in some of the cultures discussed in the literature. From the Korean perspective, Moon (Reference Moon2002) argues that while the situation may be changing, ‘Korean students copy . . . and it has been allowed. More precisely, it has been ignored because plagiarism is not a concern of teachers in academic settings in Korea’ (p. 1351). This view was taken up in a reflective piece on cultural differences as an explanation for plagiarism (Sowden Reference Sowden2005) which stimulated disagreement in print from academics from Vietnam and China: ‘the Vietnamese terms for plagiarism . . . have the same or even more negative connotations [than the English word]’ (Ha Reference Ha2006: 76); ‘all my teachers warned us not to copy others’ work. In fact, the concept of “plagiarism” as an immoral practice has existed in China for a very long time’ (Liu Reference Liu2005: 235). As one of the participants in Tang's (Reference Tang and Tang2012) study of junior EFL researchers explains, although the practices of Chinese writers’ language reuse that have been described in the literature reminded him of his own experiences of learning English at a Chinese college (‘i.e., noting down useful expressions in a notebook and using them in my speaking and writing’, p. 214), he strongly disagreed with claims in the literature about plagiarism being acceptable in Chinese culture, which he considered essentialist and ‘largely based on assumption of the cultural acceptability of plagiarism rather than solid evidence’ (p. 215).

The role of culture as an explanatory factor in plagiarism in the work of L2 writers is thus far from clear. If an answer is discoverable, it will require more research based on direct comparisons. However, it may not be possible, or worthwhile, to unpick cultural values from the web of language skills, discoursal adjustments and individual perceptions and practices in which they are embedded. Taking plagiarism as an example of a problem which has controversially been pursued from a contrastive rhetorical perspective, Bloch (Reference Bloch2008) acknowledges that temptations toward moralising, essentialism and reductionism are present but argues that it is possible to resist them and that by making the effort, rich understandings can be achieved. Perhaps the real take-home message is that while understanding our students’ backgrounds may inform us about their plagiarism, a more fruitful enterprise may be studying their plagiarism for what it can teach us about their experiences as writers.

6.3 Plagiarism and pedagogy

As Wette notes, the literature on plagiarism tends to be ‘more problem-oriented than solution or practice-oriented’ (Reference Wette2010: 159). There is, however, a growing awareness among scholars and practitioners alike that the solution to problems relating to plagiarism lies in education rather than punitive measures. Recently there have been numerous descriptive accounts of, and suggestions for, plagiarism pedagogy, which can be grouped into two broad approaches: (i) educating students explicitly about plagiarism and (ii) teaching source use and referencing in greater depth. The two approaches are not seen as mutually exclusive; in fact in many accounts they are combined (e.g. Barks & Watts Reference Barks and Watts2001; Bloch Reference Bloch2012). In this section we first summarise these two foci of pedagogical activity and advice before reviewing empirical research on the effectiveness of pedagogical interventions aimed at preventing plagiarism.

7.2 Research ethics

One concern shared by studies in this area, whatever their methodology, is research ethics, of particular importance when dealing with such a sensitive issue as plagiarism. Not only is it essential to protect the identity of participants, but a series of additional measures need to be taken. Researchers tend to avoid using the term ‘plagiarism’ in research instruments not only because it may be understood in different ways by different individuals but also because of the stigma surrounding it and the potential offence it may cause. When interviewing students and teachers about potentially unethical behaviour, researchers ensure that questions are asked in a non-judgmental and non-threatening way. However, researchers may themselves face ethical dilemmas, such as when they discover instances of unacceptable source use in student writing that was not discovered by markers, and where marks have already been awarded, particularly in the case of high-stakes genres, such as dissertations and theses.