Fraud and Fairy Tales: Storytelling and Linguistic Indexicals in Scam E-mails

Most of us receive numerous spam e-mails, texts that in one way or another try to convince us to engage in transactions of enormous sums of money, promising unbelievable benefits. In reality, such scam e-mails are fraudulent attempts to swindle money from unsuspecting internet users. Language, its social contexts, and the composition of texts play a crucial role in the scammers’ strategies in approaching their victims. This article uncovers and discusses some of the linguistic strategies by which scammers try to shape a sense of identity and mutual relationship between themselves and their addressees – in the face of virtual anonymity – and to involve their readers personally. In their attempts to get the recipients involved, scammers combine cultural indexicals, interactional roles, and narrative strategies. Cultural indexicals play a crucial role in scam texts because they can help to underpin the scammers’ identity claims, but they can also uncover obvious mismatches and arouse suspicion. This analysis distinguishes three different narrative strategies in scam e-mails: first-, second-, and third-person narratives. Employing these narrative strategies, scammers establish links with the recipients by combining fictional content with real-world contexts. Some of the narratives display quite elaborate and artful traits and involve prototypical functions of traditional fairy tales. Hereby they implicitly connect the story content with the interactional roles of e-mail communication.

lawyer, asked me to take the role of a distant relative to his deceased client so that I could inherit the client's property. Mr. Afini Mujadeen, a bank officer from Burkina Faso, had discovered an unclaimed fund I could help him claim from the bank. Even UN Secretary General Ban Ki Moon wrote and offered me half a million dollars as compensation for an internet fraud that I may have been victim of. In a hundred e-mails of this kind I was asked to accept the transfer of an overall sum equivalent to 833 million euros to my bank account, in addition to five tons of gold.
I am clearly not the only person to receive such e-mails. This type of spam email is widely known as Nigerian fraud (though they are not all sent from Nigeria), advance-fee scam, or 419 scam. All these e-mails have in common the use of narratives to encourage the recipient to become personally involved with the sender (Blommaert & Omoniyi 2006;Freiermuth 2011;Nikiforova & Gregory 2013). They normally offer the recipient the opportunity to earn substantial sums of money in a short period of time without expending much effort. Instead of becoming rich, scam victims risk losing a lot of money if they do what Samira and the others propose. Victims are normally asked to pay money to the sender in advance to cover bank fees and miscellaneous expenses before the funds can be transferred. While such "fee" payments can add up to considerable sums of money, the victims will never receive the promised funds and will invariably lose their money to the scammers.
The main focus of this article is on the construction of identities and personal relationships by the means of text. All scam e-mails have in common that someone we do not know and with whom we have never been in contact tries to convince us to make a high-involvement decision with an enormous financial and personal impact. Although the opportunity to earn a huge amount of money may represent an attractive offer, acceptance requires an atmosphere of trust and confidence, and a conception of whom we are dealing with. To be able to establish some kind of mutual contact, business relationship, or agreement on a money transfer, writers must establish a mutual understanding of identity, i.e. 'who am I' and 'who are you'. Since there is no other connection between writers and readers of scam e-mails than the e-mail text itself, it provides the only means to establish this sense of mutual understanding.
This article views scam e-mails as multi-voiced, heteroglossic texts that combine different linguistic means of contextualization. It uses a corpus of 100 scam e-mails and presents detailed analyses of characteristic strategies in a set of example texts. Highlighting the dialogic properties of these texts, I analyse the linguistic strategies behind the scammers' attempts to: transmit a sense of identity and authenticity; establish a mutual relationship between sender and receiver; and involve the recipient personally. The linguistic tools analysed encompass cultural indexicals as well as three different narrative strategies. First-, second-, and third-person stories serve to frame the contextual relations derived from the texts, and they construe the relationship between sender and receiver in fundamentally different ways. Some of the messages use a literary pattern that most members of Western societies are very familiar with: Writers design their stories around some prototypical functions of traditional fairy tales and use these to implicitly connect the story content with the interactional roles of the e-mail exchange. While narrative accounts of private and institutional identities normally seem to combine quite well with cultural indexicals such as genre and orthography, my analysis also identifies mismatches between the scammers' identity claims and the indexicals used in the texts.
The article starts with a discussion of identity construction in scam e-mails. The subsequent sections focus on two of the most important means of contextualization and identity construction in scam e-mails: on the use of cultural indexicals and their implementation in the texts, and on narration as an effective means of framing identities and establishing connections between sender and receiver.

Scam e-mails and identity
The relationship between language usage and social and individual identity is a major concern of sociolinguistics and discourse studies (e.g., Bamberg, De Fina, & Schiffrin 2011;Bucholtz & Hall 2010;Johnstone 2010). Identities are normally constructed and negotiated in relation to the diverse immediate and distant contexts that are available to language users in a given situation through a multitude of pragmatic channels. The construction of identities in scam e-mails is an unusual case because there are no given contexts that writers and readers can use and rely on, other than the text itself and the contextual references made relevant in and through the text. In other words, there is no other link between the sender and the receiver than the e-mail text.
The most immediate communicative purpose of scam e-mails is to persuade recipients to engage in immediate interaction with the sender. Therefore, scam e-mails always target individuals (Barron 2006, p. 884). They construe a "oneto-one relationship built on mutual confidence and supported by a realization of mutual benefits" (Blommaert & Omoniyi 2006, p. 592). Other goals of scam emails are to set the groundwork for further communication, to establish solidarity and trust (Freiermuth 2011), to persuade the receiver to engage financially, and finally, to swindle money from the victim and thereby securing financial profit for the sender (e.g., Dobovšek, Lamberger, & Slak 2013;Kich 2005). Establishing a mutual relationship requires a mutual understanding of identities: Who am I, and who are you? Senders of scam e-mails want to signal their victims: I am your reliable and trustworthy partner, you can trust me.
In scam e-mails, the construction of identities and mutual relationships faces two major challenges: Firstly, because they perpetrate criminal activities, it is in the interest of writers to hide their real identities and stay anonymous. Scammers use a number of means to prevent their identities from being traced electronically. The material analysed displays e-mail addresses from providers all over the world that are difficult to locate (cf. Blommaert & Omoniyi 2006). Many writers use different sender and reply addresses, and many e-mails are sent from hidden IP addresses. Secondly, except for an e-mail address, scammers do not have any previous knowledge of the individuals they address in their texts. Unlike some other spam e-mails (e.g., personalized advertisements), none of the e-mails in the corpus addresses the recipient by name. Although Blommaert and Omoniyi (2006) report some cases where recipients are addressed individually; most e-mails are sent to an unknown number of undisclosed recipients.
This means that an anonymous writer seeks to establish a mutual, personal relationship with an individual, but anonymous recipient. A mutual relationship is necessarily based on identity. This dichotomy of anonymity and identity construction is characteristic of scam e-mails. In some cases, attempts to handle this dichotomy in the e-mail texts result in quite artful linguistic outcomes. The only means for spammers to overcome the challenges that arise from this dichotomy is the e-mail text. The text is the only existing link between writer and receiver; and the construction of self and receiver identities are fundamental functions of the text. In their role as e-mail authors, writers are in the position to design a sender identity as well as a sender-receiver relationship by their linguistic choices.
When approaching their victims, writers use the voices of imaginary or real world characters with a name and a history (e.g., Samira Kipkalya Kones, Steve Thompson, Afini Mujadeen, or Ban Ki Moon). These identities are developed in the texts. Receiver identities are always positioned in a dialogic relation to these construed sender identities. (To avoid confusion, I will refer to the writer as writer, while the first-person voice that interacts with the reader will be referred to as sender.) Writers of scam messages use and combine multiple different means of identity construction. These include the linguistic choices that writers make when they compose the texts, as well as the contexts they involve. Identity construction means contextualization. Various contexts are made relevant when writers try to establish mutual relations between sender and receiver: (a) These include large (macro-level) identity categories, such as gender, age, profession, and origin, and the ideologies and expectations connected to these identity categories (cf. Silverstein 2003 on the principle of indexical order). For example, many e-mails construe Africanness as a general identity category, erasing any internal variation of African identities (cf. Gal & Irvine 1995 on the semiotic process of erasure). In doing so, they cater to the (perceived) expectations of Western recipients. Similarly, the material reflects very stereotypical gender identities. (b) Writers choose social roles for their sender characters, i.e., private or institutional. These have a large impact on the linguistic choices in the texts and on the kind of relationship that is construed between sender and receiver. (c) These identity choices affect, and are affected by, interactional positioning, expressed through politeness and emotional stance. In most cases, this means submission to the recipient, but we also find texts where the sender commands the recipient to act.

Indexicality and general linguistic properties of scam e-mails
The material analysed consists of one hundred scam e-mails received on an institutional e-mail account in Norway. All e-mails had passed the spam filter. The corpus contains only spam e-mails of the type that is known as Nigerian scam; i.e. it does not contain other types of spam messages, such as advertisements, job offers, or lottery scam. All e-mails in the corpus have in common that they offer some kind of financial benefit to the receiver. The emails were analysed with respect to their generic structures, orthography, language variation, identity categories involved, and the types of stories told.
In general, the linguistic and pragmatic characteristics of the e-mail texts in the corpus vary according to the identities, voices, and strategies chosen by the writers. However, features such as direct, personal addresses (dear friend, my dear, etc.) are recurrent in many of the texts. Grammatical and orthographic errors that are typical of non-native users of English are found in all texts. For a greater overview of these and other linguistic characteristics of scam e-mails, see Schaffer (2012) and Blommaert and Omoniyi (2006). Here, I briefly summarize the generic structures of the e-mail texts and then show how the analysed scam e-mails can be categorized according to the identity categories and types of stories they involve.

Structural variation and role choices
Several authors have described the generic structures of scam e-mails (Barron 2006;Blommaert & Omoniyi 2006;Freiermuth 2011;Kich 2005). Blommaert and Omoniyi (2006) list the following moves that a typical scam e-mail is composed of: (1) a personal, direct address, (2) an apology and introduction, (3) a micro-narrative about the origin of the money that is offered, (4) an invitation to engage in a business transaction, (5) requests for confidentiality, and (6) a closing formula. The e-mail text in Example 1 displays all these elements (numbers in brackets). (1) Dear Friend, (2) I know that this mail will come to you as a surprise as we have never met before, but need not to worry as I am contacting you independently of my investigation and no one is informed of this communication.
(3&4) I need your urgent assistance in transferring the sum of $15,million immediately to your private account.The money has been here in our Bank lying dormant for years now without anybody coming for the claim of it.I want to release the money to you as his business partner, to our deceased customer (the account owner) who died a long with his supposed NEXT OF KIN since 31 January 2007. The Banking laws here does not allow such money to stay more than 13 years,because the money will be recalled to the Bank treasury account as unclaimed fund.
(5) By indicating your interest I will send you the full details on how the business will be executed. Please respond urgently and delete if you are not interested (6) Regards,

Mr Hasib Fatah
Previous research and the corpus used in this study show that the occurrence of different generic moves varies to a considerable degree. Kich (2005) lists an introduction of the sender as a typical move. Freiermuth (2011) and Barron (2006) mention "invite further contacts" or, respectively, "solicit response" as an important move. This function can also be seen in Example 1. Both occur frequently in the material analysed.
The texts analysed range from 22 to almost 2000 words. Especially the narrative parts of the messages vary from extensive and detailed stories to rather brief references to a story that is not told in detail. Example 2 shows a very brief and elliptic version of a scam message. It refers to a story rather than narrating it in the text. Because of the wide range of different structural properties found in the corpus, it is difficult to summarize all texts under one structural pattern. Although a number of typical generic moves recur in many texts in the sample, it is important to stress that writers find a variety of structurally different solutions. Single generic elements can be given a lot of space in the texts, they can also occur in quite compressed forms, or they can be omitted completely. Even length itself can be part of the writer's strategy.
Although all e-mails analysed generally pursue the same communicative goal, differences arise from the roles and voices chosen by the writers, and the stories they construe around them. Generic structure is one means to implement these roles in the texts. A useful categorization of the scam e-mails in the corpus can therefore be made according to the identities and roles the writers chose for themselves, as in Table 1, which also shows the quantitative distribution of identity choices, gender, and types of narrative in the corpus. From the data it is clear that the writer's choice to adopt either a private or an institutional role as part of his or her created sender identity has a major impact on the kind of relationship that is construed between sender and receiver, on the stories told to involve the receiver, and on the generic text structure used to approach the recipient.
In most cases, the gender of the sender is a distinctive factor. Table 1 shows two typical gender-related identities: females acting as private persons and males acting in institutional roles. Together, these account for more than three fourths of all cases in the corpus. This clear-cut allocation of gender to certain social roles reflects a stereotypical role model, according to which women or girls approach the recipients from an inferior position and beg for help. At the same time, men can normally use the authority of their institutional roles as, e.g., lawyers or bank officers. Accordingly, female senders such as Samira Kipkalya Kones frequently express strong emotional stances, while males are more likely to express rational assessments of risks and benefits. Both behaviours are indexical of the stereotypical gender roles writers use in their texts.
The choice of gender roles also projects a recipient identity and thus takes part in the construction of a sender-receiver relationand of the allocation of power. In most cases, then, the receiver is construed as male: Weak women ask (projected) strong men for help, and male lawyers or officers ask male recipients to enter into a business relation with them. Such relationships are construed through narrative strategies. Table 1 shows that the different narrative strategies are connected to different sender identities: Private identities are mainly narrated as first-person narratives, while institutional sender identities are introduced using third-or second-person narratives. The strategic functions of these three different types of narrative are analysed in detail below.

Linguistic indexicals: style, orthography, and language choice
In most e-mail texts, sender identities are enhanced by linguistic indexicals. These signal the sender's belonging to large sociocultural categories such as region, profession, gender, place, as well as situational identities on the level of interpersonal interaction. In sociolinguistics, indexicality is understood as a sign relationship (going back to the work of C.S. Peirce) that creates links between linguistic forms and social meanings, normally based on sociocultural experiences and ideologies about certain language users and ways of language use (Bucholtz & Hall 2010;Silverstein 2003). It is the fundamental semiotic means to construct social identities through linguistic forms (Bucholtz & Hall 2010). As Blommaert and Omoniyi (2006) argue, the use of such cultural indexicals shows that writers of scam messages are culturally literate in a globalized world: Writers know how to employ cultural indexicals effectively as a resource although their recipients live in culturally different surroundings. However, some cases also show that the indexicality of their linguistic choices can work against the scammers' intentions.
Style. In sociolinguistics, style has been frequently investigated as an indexical marker of social identity (e.g., Eckert & Rickford 2001). In the scam e-mails analysed, it seems that the writers know how to apply stylistic variation to underpin the sender-identity they construe in their texts. Writers who claim for themselves an institutional role (as lawyer, bank officer or UN official) are more likely to use long and complex clause constructions, such as in Example 3.

Example 3
However, I wish to personally bring to your urgent notice, so that you will be aware that your unclaimed payment data's is still reflecting in our central computer as unpaid and un-claimed owed debt in my department where I newly work as the new appointed (UN) lawyer in charge of approving & releasing of all unpaid debts recently recovered back by the (UN) office in London been our branch after discovering that your payment has been delayed unnecessary due to one reason or another while auditing was going on this week.
This example shows one single sentence consisting of 91 words. The sentence contains several repetitions (bring to your urgent noticethat you will be aware; your unclaimed paymentunpaid and un-claimed owed debt; in our central computerin my department) so that some propositional content is presented several times in different ways. These repetitions and the combination of multiple minor clauses, in addition to a number of grammatical errors, result in a construction that readers hardly can keep track of. Though "garbled syntax" (Schaffer 2012, p. 169) is a typical phenomenon in many spam e-mails, this effect might be intended: The (questionable) content is hidden in grammatical complexity while the chosen style indexes this as an official, and therefore serious, piece of writing. In addition, authority and credibility is enhanced by the double mentioning of the UN (in brackets and capital letters) that sticks out from the dense text and easily catches the reader's attention. As an outcome, the indexical meaning of the chosen style might have a stronger impact on many readers than the propositional content.
Language errors. Besides long sentences, we find numerous orthographic, grammatical, and lexical errors in the texts. In her e-mail, Samira Kipkalya Kones uses sentences like this one: After the burial of my father, my stepmother and uncle conspired and sold my father's property to an Italian Expert rate which the shared the money among themselves and live nothing for me. The attribute "garbled syntax" definitely applies to this sentence, too. Lexical, grammatical, and orthographic errors are typical of scam e-mails. Of a hundred texts examined, not a single one is completely free of such errors. The case of scam e-mails demonstrates that language errors are a type of sociolinguistic variation that transports indexical meaning. Various scholars ascribe special importance to this deviation from grammatical and orthographic norms (Blommaert & Omoniyi 2006;Nikiforova & Gregory 2013;Schaffer 2012). It is easy to recognize that the texts have been written by non-native speakers with varying fluency in English (Schaffer 2012).
There are two fundamentally different explanations of how the indexicality of language errors affects the credibility of the senders' voices and identity claims. Blommaert and Omoniyi (2006) consider such errors a major obstacle for scammers. Scammers seem to be fully competent in the technology of electronic communication, and they appear to be culturally literate; i.e., they master the specific generic and stylistic characteristics that are needed to convince their addressees. But their texts reveal an incomplete control over the linguistic details of written English, including accent, writing style, and politeness formulae. Rather than enhancing a construed sender voice, the text would then expose the voice of the writer. Blommaert and Omoniyi (2006, pp. 602-603) conclude that "the nonnativeness of their English, in short, exposes their messages as cases of fraud; it destroys the whole indexical edifice they attempted to build in their message." It is, however, possible that receivers with restricted knowledge of English do not recognize these signals of fraud because of the non-nativeness of their own English.
Together with Ottenheimer and Ottenheimer (2012), Nansen (2008) and Nikiforova and Gregory (2013) suggest another explanation of the consequences of such indexicals. They interpret the use of bad grammar as an intentional, strategic choice. Grammatical errors are used to reinforce stereotypical images of Africans, i.e. writers use language ideology to index Africanness (Ottenheimer & Ottenheimer 2012). Grammar is used to construe a stereotypical African identity inferior to Western culture. Together with the narratives told by scammers, grammatical errors are used "to maintain the postcolonial perceptions of power in which the receiver in a First-World country is always recognized as superior to the sender of the scam" (Nikiforova & Gregory 2013, p. 398). Nansen (2008) considers this strategy fundamental to the scammers' success.
The effect of this indexicality depends on the identities writers construe in their narratives. In Samira Kipkalya Kones's narrative, language errors help to foster her identity as a weak, young African woman in a difficult situation. They amplify Samira's inferior standing vis-à-vis the Western addressee she asks for help (Example 8 below). However, the claim that language errors help to maintain Western post-colonial perceptions, which, in turn, enhance the success of the scam, can only be true when the scammer construes an African identity (or one that implies post-colonial perceptions of power). When scammers in their stories claim to be from, e.g., England or the United States, this identity should be reflected in an adequate orthography and grammar. In Steve Thompson's email, which presents the sender as an engineer from London, grammatical errors immediately arouse suspicion. Native speakers and experienced ESLusers can easily detect the mismatches between indexicality and claimed identity.
Language choice. Mismatches between indexicality and claimed identity expose the scammers' insufficient control of grammar and orthography, and in this way, narrows down the success of the scam. This effect is even more evident when the scammer tries to accommodate linguistically to his recipients in non-English speaking countries (according to the country code top-level domain of the e-mail address). I received these messages on a Norwegian e-mail account, and a few messages were in Norwegian. Example 4 shows extracts from an e-mail from Dr. Anthony Davis from a bank in London, who claims to be British. The Norwegian text displays numerous errors, including those relating to verb forms, pronouns, lexis, and the mixing of languages (English words in the text).
'Be on the *confirmation page of this message and *show your interest I will give you more information.' Messages like this one are obviously translated from English using a machine translation service on the internet. In (a), the machine obviously did not translate the words URGENT PLEASE because they were written in capital letters. The result is a sentence that makes little sense in Norwegian, though Norwegian readers (with knowledge of English) would understand what is meant. Some expressions only make sense when translated back into English, as in (b). The Norwegian word sikkerhetskopiere (literally 'make a safety copy') does not make sense in this context. It has most likely been (mis-)translated from English back up. 'To back up the claim' makes sense. This is, however, not visible in the Norwegian text. Also (c) would be identified as not correct by any Norwegian reader, both lexically and grammatically. Bekreftelsessiden ('the confirmation page') does not make sense in the given context. The verb form viser (inflected present of vise 'show') fails to replace the English gerund phrase showing your interest since there is no equivalent construction in Norwegian. Here, the writer would have needed to choose a completely different grammatical construction.
In cases like this one, it is clear that the senders have little control of the grammar and lexicon they use. It is very unlikely that such errors which obscure the meaning of the text are intended; or at least, they do not fulfil their intended function. Therefore, we can say that attempts to accommodate linguistically to recipients in non-English speaking countries are very likely to fail.

Language choice and religious indexicals.
In addition to the few cases where messages are written in languages other than English, we find some texts that use single phrases in other languages. Such choices must be considered intentional. In Example 5, this choice supports the construction of the writer's regional and religious identity.

Example 5
Assalam O Alaikum wa Rahmatullah wa Barakaatuh, Can you help to complete the transfer of us$28,000,000.00 to your account?
The Arabic phrase clearly functions as a cultural indexical. It is a frequently used standard phrase that states, 'may the peace, mercy, and blessings of Allah be with you'. The writer uses this Arabic opening in a text written in English and addressed to a member of a Western culture. A recipient in a Western country will automatically draw conclusions about the cultural background of the sender. Besides such shifting of linguistic codes, religious background is signalled in some texts through closing formulae such as Allah bless you or God bless you or addresses such as dearest in Christ. Such religious references (either explicit or implicit) introduce a moral voice in the texts that serves to enhance credibility and to appeal to the recipients' readiness and moral commitment to help.

Engaging with the recipient: narratives
The indexical meanings and structural features discussed above interact with and support the stories that form the core of the scam. Narratives are the most important tools to construe a mutual relation between the sender (constructed fake identity) and the receiver. They are elaborate linguistic tools for contextualization that can fuse close and distant contexts, as well as the contents and contexts of telling. In the case of e-mail scam, narratives are used to establish links between persons that have never been in contact with each other before. The narratives also establish the contextual frames within which cultural indexicals make sense in the texts.
To achieve this, writers use the narrative contents as well as the contextual frames in which the narratives are told and those that are aroused through the linguistic indexicals in the texts. Narratives in scam e-mails are fundamentally dialogic. They are addressed to a recipient; their primary goal is to involve and engage the recipient; and they are used to justify the sender's reason for contacting the recipient. Furthermore, narratives do not even need an elaborate internal structure. Many narratives used in the scam e-mails analysed are rather short references to more complex stories that are not told in detail (cf. Example 2, above), but they are effective as such (cf. Bamberg et al. 2011;Georgakopoulou 2007).
In the following sections I present three different e-mail texts and show how writers use narratives strategically to establish links between the sender (firstperson voice), the story content, and the recipient. These represent three general strategies of approaching the recipient: first-person, second-person, and third-person narratives; i.e., strategies differ according to whom the story is about. Depending on the type of narrative, the sender identity is construed as institutional or private. The three types of narrative construe three fundamentally different relationships between sender and recipient.
In a typical third-person narrative, the connecting link between sender and receiver is a third person that has something in common with the recipient (e.g., being non-African, being from the same country, or having the same last name).
In most cases, that person has died and left behind a considerable amount of money. The sender (in the role of a lawyer or bank officer) and the recipient have to collaborate in some way to inherit the money.
Second-person narratives often draw on the authority of an international institution, such as the United Nations or the World Bank. The sender takes the role of a representative of that institution and offers the recipient financial support or a compensation for a problem, damage, or loss. The recipient is in the focus of what has happened. To avoid obvious mismatches, second-person stories are never told in detail but rather referred to.
First-person narratives are spun around the sender as the main character of the story. These are normally the most elaborate narratives. The sender (often a female private person) recounts a serious problem (mistreatment by others, disease, etc.) and asks the recipient to solve her problem. This narrative strategy often uses literary elements that are well-known from classical fairy tales.

Interactional roles and identities
The writers' narrative strategies are not only limited to the use of story contents to engage with the recipients. They also take advantage of their position as writers and narrators. Following Goffman's (1981) discussion of the speaker role, one can identify three main functions in the role of the writer, animator, author, and principal. The writer acts as an animator when he physically produces utterances (i.e., written text) and addresses the interlocutor (recipient).
In the function of an author, he choses what is narrated, and how. Through the linguistic interaction (i.e., the text) the writer establishes the position and authority of a principal. This role corresponds with the sender identity construed in the texts.
These functions reveal some of the possibilities that a scammer has when approaching a potential victim by means of an e-mail text. Writers do not have any previous knowledge of their recipients, but they try to encourage them to engage in personal interaction. When doing this, they take advantage of the interactional roles and functions of the e-mail communication. In the role of an author, the writer can design the text and the story in a way that projects a recipient identity (adaptable to a large variety of potential recipients). The writer can make textual choices that add authority to the role of principal; and he positions himself not only by claiming an identity, but also by relating it to the addressee. All three types of stories discussed here reflect the strategic combination of such elements of interactional context with elements of narrative content.

Third-person narratives
Mr. Afini Mujadeen (Example 6) presents himself as a bank officer from Burkina Faso. He claims for himself an African identity and an institutional position. His use of genre and vocabulary indexes this institutional role. The plot he uses is described as a "dormant account" narrative (Blommaert & Omoniyi 2006;Freiermuth 2011). It occurs frequently in the corpus. In most cases, the sender is male and has an institutional position (bank officer, state official, lawyer). He uses his institutional identity to approach the recipient; pursuing private interests that demand a high degree of confidentiality from both partners (sender and receiver). Base on my security report, these funds can be claimed without any hitches as no one is aware of the funds and it's closing balance except me and the customer who is (Now Deceased) therefore, I can present you as the Next of Kin and we will work out the modalities for the claiming of the funds in accordance with the law. Now,if you are interested and really sure of your trustworthy,accountability and confidentiality on this transaction without dissapointment, please call me +22-74-52-19-04 and contact me through email for further details.i expect your letter.
our sharing ratio will be 50% for me and 40% for you.while 10% will be for the neccesary expenses that might occur along the line.
i expect your call and reply sincerely Mr Afini Mujadeen.
To establish a relation between the sender and the receiver, the writer uses a third-person story that can seem credible. In e-mails of this type, writers exploit the potential of narratives to contextualise and establish relations between originally independent and distant persons, events, and contexts. The narrative combines elements that can be linked to both the sender and the recipient. The third-person character is Western and thus somewhat closer to the supposed recipient. He or she was involved in a tragic accident. To confirm the authenticity of such events, many writers even provide links to news websites.
This tragic background story is only mentioned through a brief reference in the text. Assuming that the recipient has previous knowledge of the accident, the sender treats it as a shared story (cf. Georgakopoulou 2005). Shared attitudes about the tragic event serve to strengthen the link between sender and receiver.
The advantage of such third-person stories is that they offer the creative freedom to develop characters that function as links between the sender and the receiver. Figure 1 summarises how third-person stories establish such links.

THIRD PERSON STORY
· shared experience/knowledge of event · construction of sender identity · shared attitudes about tragic event · ...

Second-person narratives
Example 7 represents a typical second-person scam narrative. These second person narratives situate the sender at an official institution. Normally the institution, and not the sender as an individual person, addresses the recipient.
In some cases, e-mails of this type are composed in a completely impersonal style. The narrative is usually centred on the addressee, elliptic, and told without details so that it can fit most receivers. To address the recipient as "Dear scam victim", as in Example 7, seems rather impudent. The message is signed with the name of UN Secretary General Ban Ki Moon; hence, this e-mail construes the sender in an institutional role. As the direct address indicates, this scam type focuses on the recipient rather than on the identity of the sender or of a third person. It uses the authority and the reputation of an official institution (in this case, the United Nations), but it does not develop an individual sender identity. The passive, impersonal style with numerous nominalizations maintains a personal distance. Together with the use of imperatives, it enhances the superior role and the authority of the institution.
The story is not elaborate either: The recipient is the victim of a scam, and the UN will pay a compensation. Second-person narratives in scam e-mails are necessarily short and elliptic. Were the narrative to contain too many details, the receiver would easily detect obvious mismatches. Instead, the writer tries to make use of the relation that is construed between the institution and the recipient.

First-person narratives
Let us, finally, return to the narrative of Samira Kipkalya Kones, the young woman from Kenya. In her e-mail (Example 8) Samira tells the story of how her father died, how she was mistreated by her stepmother and her uncle, how she discovered her father's secret bank account, and how she hopes to solve her problem. This is not only a long, elaborate, relatively detailed, and emotionally loaded narrative about the fate of a young Kenyan woman. Samira's story bears many of the classical components of a traditional folktale. We can easily recognize two typical fairy tale characters: a weak, young woman and an evil stepmother. More than the characters themselves, their functions are important for the narrative and its ability to establish links between the sender and the receiver. The writer uses these narrative functions to get the reader involved in the story.
In his analysis of Russian folk tales, the folklorist V. Propp (1968) outlines a number of limited, functional elements that constitute the fundamental components of a fairy tale. The functions of the characters serve as stable, constant elements in the tale, independent of how and by whom they are fulfilled (Propp 1968, p. 21). This includes that the functions of the hero can be fulfilled by a prince on a horse as well as by an internet user with a computer and a bank account.
Of the prototypical narrative characters described by Propp, we find a victim, a villain, and a hero in Samira's story. Samira presents herself as the victim who is in a precarious situation and in danger of getting killed by her stepmother. Her vulnerable position as an African woman is reflected in her weak grammar and orthography and in her submissive tone (e.g., PLEASE I NEED YOUR HELP; I beg you not to disclose it till i come over). In addition, she exposes her emotional state. Samira's stepmother and uncle fulfil the functions of the villain. According to Propp (1968, p. 30), one of the most important functions of the villain is that he "causes harm or injury to a member of a family." This function is narrated in the second paragraph of Samira's e-mail. As Propp states, "this function is exceptionally important, since by means of it the actual movement of the tale is created" (Propp 1968, p. 30).
The reader, originally in the passive role of a recipient, is included in the narrative. The writer, as author, ascribes the functions of the hero to the recipient. Here, the e-mail fulfils a function that Propp (1968, p. 36) defines as connective incident: "Misfortune or lack is made known; the hero is approached with request or command; he is allowed to go or he is dispatched". This function is important because it brings the hero into the tale. Here, it helps to fuse a story character with a participant of the interaction (the recipient, in this case), and it connects thepresumably fictionalcontent of the story with a real world context. Figure 3 illustrates how such first-person stories involve narrative characters to establish links between the sender and the recipient.

Conclusion
When approaching their victims, scammers face at least one major challenge: There is no previously existing link between the originator of a scam e-mail and the recipient. To convince their victims to engage personally in a mutual interaction and even to transfer money, scammers have to present reasonable and credible explanations for why they are addressing a particular addressee, and they have to provide evidence that their offer is authentic and trustworthy. This task centrally involves the construction and positioning of identities.
Combining fictional story contents and characters with the interactional roles of e-mail communication is a central part of this. Narratives by first-, second-, and third-person characters offer different ways of establishing links between the sender and the receiver. Such narratives can either take an elaborate form or simply refer to another story. They always construe a sender identity, position the recipient, and link the offer to both participants, although they do so in different ways.
The manner in which cultural indexicals are used in the scam e-mails investigated plays an important role in underpinning the identities that scammers develop in their narratives. In most cases scammers seem to have good control of the genres that are indexical of the institutional or private positions to which they claim to adhere. Frequently occurring language errors can, in some cases, index the socially weak and subordinate position of a private person begging for help, in line with many first person narratives. In many other texts, however, insufficient orthographic and grammatical skills expose a lack of coherence with the identities developed in the narratives.
Having analysed the linguistic choices of scammers, we cannot draw fully reliable conclusions about which linguistic choices were made on purpose and which not. It is difficult to determine if and why writers of scam e-mails make certain linguistic choices deliberately, first and foremost, because we cannot ask them. This article, however, shows how such linguistic choices achieve functional effects, and how these get connected strategically to persuade the recipient. While it is not guaranteed that scammers think the same way as analysts, we can clearly conclude that scammers dispose linguistic tools that enable them to convince unsuspicious recipientsin some cases successfully.
In some of the texts analysed, these linguistic strategies even feature quite elaborate and artful traits.
The analyses show also that the cultural indexicals in the texts can support the constructed identities and enhance the impression of authenticity. At the same time, indexical signs in the texts can make visible the obvious gaps between the constructed sender identity and the writer's real intentions, separating the writer and sender voices from each other and exposing who is the real 'villain' and who the victim in scam interactions. Most recipients are able to recognize these signs as indexes of scam. Thus, after analysing the communicative strategies used in the data, it seems likely that the best way to succeed as a scammer is by sending out as many e-mails as possible. Sadly, this quantity still allows for scammers to achieve their goals with their communicative strategies and to cheat quite a number of victims.