The suicide note as a genre: Implications for genre theory

Journal of English for Academic Purposes xxx (2015) 1e14

Contents lists available at ScienceDirect

Journal of English for Academic Purposes journal homepage: www.elsevier.com/locate/jeap

The suicide note as a genre: Implications for genre theory Betty Samraj*, Jean Mark Gawron 1 San Diego State University, Department of Linguistics and Asian/Middle Eastern Languages, 5500 Campanile Drive, San Diego, CA 92182, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 July 2014 Received in revised form 16 March 2015 Accepted 17 April 2015 Available online xxx

Although most of the work on genre analysis has attended to academic and professional genres, we attempt to explore and extend key issues in genre theory through a move structure analysis of suicide notes, a non-academic, non-professional genre, which has been the focus of interdisciplinary research. These studies, however, have not considered the suicide note as a genre. We propose a move-based account of this genre despite a variety of challenges these texts pose for the framework, including the lack of a discourse community, great variety in length, and the lack of any identifiable obligatory moves or fixed ordering of moves. The results show that not all genres can be characterized in terms of obligatory and optional moves, and we discuss an alternative way for capturing genre membership. We also show that a genre can exhibit patterns of co-occurrence for moves and steps even when it lacks an identifiable linear order. A computational analysis of a set of lexico-grammatical features of the texts shows that these features found in previous research on suicide notes are actually concentrated in particular moves. This paper concludes with a discussion of the implications of this study for genre theory. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Suicide notes Move structure analysis Linear order Discourse community Genre theory Computational corpus analysis

1. Introduction Swales (1990: 1) states in the first paragraph of his seminal monograph, Genre Analysis, that he is seeking to develop three concepts, “discourse community, genre and language-learning,” as an approach to “the teaching of academic and research English.” Genre Analysis was followed by a large number of studies of academic genres, particularly the research article, with many focusing on the introduction and applying the Create-A-Research-Space (CARS) framework to research article introductions in different disciplines and different languages (see Swales, 2004). Departing from this core set of genre analyses, some relatively recent studies have considered other academic genres such as grant proposals and PhD dissertations (Bunton, 2005; Halleck & Connor, 2006). Genre analysis has also been applied to professional genres, for example, barristers' opinions (Hafner, 2010) and company brochures (Cheng, 2011). The appeal of Swales' framework for describing a genre in terms of functional moves and steps, in addition to the lexico-grammatical features characterizing those moves and steps, has extended to more marginal academic and professional genres, such as letters of appeal to a university (Sadeghi & Samuel, 2013), and recruitment ads (Gillaerts, 2012).

* Corresponding author. Tel.: þ1 619 594 5882; fax: þ1 619 594 4877. E-mail addresses: [email protected] (B. Samraj), [email protected] (J.M. Gawron). 1 Tel.: þ1 619 594 0252; fax: þ1 619 594 4877. http://dx.doi.org/10.1016/j.jeap.2015.04.006 1475-1585/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

2

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

The rich variety of these genre studies has revealed the utility of approaching writing, and also spoken language (see, for example, Orr, 2007), in academic, professional, and a variety of other social contexts in terms of genres, that is, groups of texts fulfilling one or more communicative purposes, characterized by a similarity in discourse organization in terms of functional moves and their constituent steps, produced and consumed by members of a discourse community. These studies in the last 25 years have enhanced our understanding of some of the key aspects of genre theory, such as the complexity of linear and hierarchical ordering of move structure (e.g. Yang & Allison, 2003), the flexibility of some genres (Rowley-Jolivet & CarterThomas, 2005; Sadeghi & Samuel, 2013), the dynamism of genres (Gillaerts, 2012), problematic aspects of privileging communicative purpose in genre categorization (Askehave & Swales, 2001; Swales 2004; Tardy & Swales, 2014), and the occluded nature of some genres (Samraj & Monk, 2008; Swales, 1996). Given that the voluminous research that has followed from Swales (1990) has primarily attended to academic writing, specifically to published academic writing, it might be somewhat surprising that this paper will focus on a definitively nonacademic genre, the suicide note, to explore, question, and, hopefully, extend Swalesian genre theory. It might be worth recalling Miller's (1984: 155) observation that considering “homely discourse” such as the ransom note and the letter of recommendation as potential genres “is not to trivialize the study of genre.” Suicide notes contain many properties that make the application of genre analysis, especially move analysis, problematic. In brief, suicide notes are not associated with a particular discourse community, are semi-occluded, include a wide range in length, and may not have an easily identifiable set of obligatory moves or linear order; as such, they might seem to be an outlier genre. Despite (and because of) these limitations, we aim to show what genre analysis can offer in a study of these important but perplexing texts that have received attention from the disciplines of psychology, forensic linguistics, and computational linguistics. Beginning in the 1950s, suicidologists (Shneidman & Farberow, 1957) have analyzed suicide notes, even focusing on language features, to probe the suicidal mind, and have compared suicide notes produced in different time periods, across cultures, and by different age groups and genders. Many of these studies have analyzed suicide notes using protocol sentences that describe psychological characteristics thought to be associated with suicide, such as unbearable psychological pain (e.g., Leenaars, 1996). Other studies have tried to gain insight into the suicidal mind by gathering statistics on words of particular semantic types using computer programs. One program that has been used in a number of studies (discussed in Section 3.4) is Linguistic Inquiry and Word Count or LIWC (Tausczik & Pennebaker, 2010). Forensic linguists have focused on ways of determining the authenticity of such texts (Olsson, 2009), joined by computational linguists who have applied machine learning methods to the problem, discovering a number of statistically significant differences between fake and authentic notes, including the greater number of words, past tense verbs, nouns, and personal pronouns per authentic note (Pestian, Matykiewicz, & Grupp-Phelan, 2008; Pestian, Nasrallah, Matykiewicz, Bennett, & Leenaars, 2010). Studies focusing on the language of suicide notes, such as those exploring semantic spaces in suicide notes (Matykiewicz, Duch, & Pestian, 2009), or even those labeling themselves as discourse analyses (Edelman & Renshaw, 1982), have not discussed suicide notes as a genre or attempted an analysis of the rhetorical organization of the suicide note. For example, Edelman and Renshaw (1982) use an approach called Syntactic Language Computer Analysis, which uses a set of semantic features classifying nouns, verbs and adjectives, as well as some shallow sentence-level syntactic features, such as part of speech, transitive verb, and modified noun. They were able to discover features that differed significantly between the authentic and fake suicide notes in the Shneidman and Farberow (1957) data set. For example, both the use of negation and use of second person were higher in authentic notes than fake ones. However, the notes are essentially viewed as bags of words, extracted from their function and context, and therefore the interpretation of the features remains quite problematic. In sum, these language studies of suicide notes have revealed the thematic features of suicide notes, for example, suicide as escape and problematic interpersonal relationships, and the psycho-social meanings, such as social detachment, expressed by vocabulary choices, but they have not had much to say about rhetorical organization or communicative purpose. In this study, we perform a genre analysis of a collection of suicide notes, focusing particularly on move structure analysis. We discuss the relationship of a genre to a discourse community using the case of suicide notes, exploring the possibility that some genres may not belong to any specific discourse community. This analysis of suicide notes is one with no opportunity for specialist interviews or observations given the contexts in which the genre is produced and consumed. However, we maintain that an analysis of suicide notes as a genre in terms of moves and steps and accompanied by a linguistic analysis of constituent moves and steps, without the augmentation of ethnographic methods (Tardy & Swales, 2014), can still illuminate and clarify this text type. Askehave and Swales (2001) have pointed out that the communicative purpose of texts may not be immediately apparent and that a linguistic analysis of a group of texts could lead to an understanding of their communicative purpose(s). We attempt a move analysis of a collection of suicide notes in order to identify the communicative purpose(s) for this genre. The results show that not all genres can be characterized in terms of obligatory and optional moves, and we discuss an alternative way for capturing genre membership. We will also show that, although a genre may lack an identifiable linear order of moves, it can still exhibit patterns of co-occurrence for moves and steps. We will further apply the computational tool known as LIWC (Tausczik & Pennebaker, 2010) to perform analyses of a set of lexical and grammatical features of the texts to show that the patterns of language found in previous research on suicide notes are actually characteristic of different functional units. Moreover, our study indicates that studying the move structure can yield a finer-grained analysis of where and why certain expressions occur, underscoring the value of move structure analysis in studies of texts in other social sciences. This paper concludes with a discussion of the implications of this study for genre analysis. Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

3

We also acknowledge that suicide notes are emotionally intense texts, which express sorrow and suffering, and feel that a move structure analysis that sheds light on the communicative purpose of these texts can be another means of giving voice to the meanings being expressed by the writers in their final pieces of communication. It is hoped that the results of this analysis, besides extending genre theory, can contribute in a small way to therapeutic work with suicidal individuals. 2. Methodological issues 2.1. Data In this section we first discuss what is meant by a suicide note in order to distinguish it from texts that might share related social purposes, such as entries in a diary expressing suicidal thoughts, but which are not themselves considered suicide notes. Antoon Leenaars is one of the primary proponents of approaching the study of suicide through content analysis of suicide notes; he has argued forcefully for the use of personal documents in the understanding of the suicidal mind. In the following passage, Leenaars (1988) defines the suicide note and identifies the primary goal of his study: They are the unsolicited productions of the suicidal person, usually written minutes before the suicidal death. They are an invaluable starting point for comprehending the suicidal act and for understanding the special features of the people who actually commit suicide and what they share in common with the rest of us who have only been drawn to imagine it. (p. 34) Leenaars (1988: 37) also discusses the suicide diary, which he describes as “a suicide note with a history” and goes on to state that it “is a lengthy, literate document kept over a fairly long period of time” where “the diarist writes explicitly about suicide, including his (sic) suicidal thoughts, impulses, reflections, and resistances; and, the diarist commits suicide.” A recent article in psychology (Lester, 2010) compares the linguistic features found in suicide notes, a journal entry, and taped messages left by a suicide, treating them as different kinds of texts. Other related genres would include suicide posts, which are entries made on web sites devoted to the topic of suicide (Prokofyeva, 2013), and suicide poems (Lester & Mc Swain, 2010). The suicide notes that form the data set for this study are part of a collection of 222 suicide note text files given to the second author by the forensic linguist, John Olsson of the Forensic Linguistics Institute in the United Kingdom (http://www. thetext.co.uk/) from his own personal collection of notes from around the world. All of these text files had been verified to be authentic and include both public texts available on the world-wide-web and texts that had been given to Olsson by coroner's offices and agencies such as the British Transport Police (Olsson, 2009). Included in his collection are also texts published by Art Kleiner (1981) collected from coroner's offices. Some of the texts had been given numerical codes (such as JCB1) with no other information. Others have basic demographic information such as “Greek male, 15, 2000.” We reduced the original set by removing texts written by famous suicides, such as Virginia Wolf, texts that were diary entries (some texts from Art Kleiner's collection), suicide notes from politically-motivated suicides such as suicide bombers, long treatises such as those written by suicides from Jonestown, translations of notes from languages other than English, and notes that accompanied obvious murder suicides. The goal of this culling was to arrive at a set of texts that might be considered more prototypical suicide notes in order to increase the ease of conducting a move analysis. Several of the 222 text files were parts of single suicide notes as indicated by their numbering and these parts were then collated to form complete texts. The deletions and collations resulted in a total of 114 suicide notes. Our final set included notes from different parts of the world, such as Botswana, the United Kingdom, and the United States, and were written by both males and females encompassing a large age range. 2.2. Analyses The main analysis conducted was an analysis of rhetorical moves of the suicide notes, since move structure analysis is a crucial component of genre analysis. The rhetorical purposes of individual moves contribute to the communicative purpose of the genre as a whole. (See Samraj (2014) for a discussion of move structure analysis in genre theory.) Following regular practice in move structure analysis, the texts were read a number of times to arrive at a preliminary list of constituent rhetorical functional units, which were then modified inductively through recursive analyses of the texts (Tardy & Swales, 2014). Once the final set of moves and constituent steps had been identified, the moves and steps in each text were marked and computationally quantified for the frequencies of moves and steps across the whole data set in order to arrive at conclusions about the salience of certain moves and steps for this genre. We also computed the amount of textual space taken by a given move or step as another indicator of its importance in this genre. The order of moves and steps in each note was also examined to determine any recurring pattern in linear order of rhetorical moves in this genre. Both co-occurrence and sequencing matrices were constructed for all the moves. A co-occurrence matrix simply contains the number of times two moves (or steps) appear in the same note. A sequencing matrix is a square table with M rows and M columns, where M is the number of moves. The number in row 1, column 2 represents the number of times Move 2 follows Move 1. Linguistic analysis was conducted for moves and steps in this genre, and the whole suicide note using the computer program Linguistic Inquiry and Word Count or LIWC (Pennebaker, Chung, Ireland, Gonzales, & Booth, 2007), which computes the percentages of words associated with different categories out of the total number of words in a set of texts. These categories may be themes such as death, positive and negative emotions, and social processes, or they may be function words Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

4

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

such as the first person and second person pronouns (Tausczik & Pennebaker, 2010). LIWC analysis has figured prominently in work done by psychologists and computational linguists on various kinds of texts produced by suicidal people, such as diaries (Pennebaker & Stone, 2004), poems (Stirman & Pennebaker, 2001), and suicide notes by attempters and completers (Handelman & Lester, 2007). Researchers who have used LIWC on suicide texts have analyzed whole texts to reach conclusions about the focus of the authors of these texts. In our analysis, we wanted to investigate whether different themes such as social processes or a focus on self (by use of the first person singular pronoun) might be associated with particular rhetorical moves. By conducting a move analysis of the suicide note and associating the different emotions and foci that psychologists have located in suicide texts with particular rhetorical functions, we hoped to gain a better understanding of what function these emotions and foci were attached to. Therefore, separate text files were created for each move and each step, each containing every instance of that move or step in our data set, and these files were analyzed using LIWC. 3. Results and discussion We begin this section by making some preliminary observations about the suicide notes in the data and in the following section discuss the notion of discourse community in relation to suicide notes. Sections 3.3 and 3.4 discuss the results from the move structure and linguistic analysis. 3.1. Preliminaries We begin our discussion of results with a few broad observations about the suicide notes that were analyzed. Spelling, grammar and punctuation errors were corrected in order to increase the accuracy of the LIWC analysis but were not corrected for the move structure analysis. The 114 notes include significant variation in length, ranging from four to 820 words with a median of 65 words. More than 68% of the texts are 100 words or shorter. A recent study of suicide notes collected in Birmingham found that suicide notes had a median size of around 77 words and also exhibited a rather large variation in length (Shapero, 2011: 100e102). The smaller median for our data set probably stems from our exclusion of suicide notes that were also political treatises, which tended to be longer. A histogram depicting note length is given as Fig. 1. 3.2. Discourse community and suicide notes The discourse community is a key concept in Swales' (1990: 24e27) genre theory, as genres are communicative events that arise within social contexts. The community of users of a particular genre, known as the discourse community associated with that genre, is said to be characterized by a number of defining characteristics, such as having a number of public goals, lines of communication between members, and some ratio of novice and expert members. What is interesting about a genre such as the suicide note is the absence of an identifiable discourse community. Although suicide notes are produced in specific social contexts when a person is on the verge of ending his or her life and is left behind for the consumption of various audiences, such as family members or no one in particular, or the authorities, it is not the case that the writers of these suicide notes have read other notes or produced multiple iterations of these texts over time. (It is true that those who attempt suicide more than

Fig. 1. Histogram of note lengths.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

5

Table 1 Moves and steps in suicide notes with frequencies of occurrence. Moves

Steps

1. Addressing recipient 2. Providing explanation 2.1 Providing larger context or background 2.2 Giving reason(s) or justifying 2.3 Taking responsibility or ascribing responsibility to others 2.4 Expressing feelings about act 3. Apologizing 3.1 Apologizing with or without explanation 3.2 Asking for forgiveness 4. Describing one's state at death and after 5. Expressing directions and wishes

4.1 Describing state at death 4.2 Describing state after death 5.1 Expressing short-term wishes 5.2 Expressing long-term wishes

6. Describing or announcing death 7. Bidding farewell 7.1 Expressing love or positive feelings towards others 7.2 Expressing hate or negative feelings towards others 7.3 Expressing thanks 7.4 Expressing (imagined) others' feelings 7.5 Expressing good-bye and good wishes 8. Commentary on life 9. Meta-commentary on letter 10. Problem

Texts with move or step Total: 114

Move or step size Total: 11, 799 words

No.

%

No.

%

56 93 7 89 15

49.1 81.6 6.1 78.1 13.2

173 5760 168 5140 283

1.47 48.8 1.4 43.6 2.4

11 40 32 12 23 9 17 48 26 31 10

9.7 35.1 28.1 10.5 20.2 7.9 14.9 42.1 22.8 27.8 8.8

169 692 506 186 402 132 270 2105 1285 820 522

1.4 5.9 4.3 1.6 3.4 1.1 2.3 17.8 10.9 6.9 4.4

67 47

58.8 41.2

1630 853

13.8 7.2

7

6.1

135

1.1

15 4 52 7 8

13.2 3.5 45.6 6.1 7.0

280 102 260 126 185 205

2.4 0.9 2.2 1.1 1.6 2.6

once might produce notes multiple times.2) The absence of an identifiable discourse community with expert members also means that drawing on specialist informant views about this genre is not an option in conducting genre analysis. Understanding the reasons for various rhetorical and linguistic choices cannot be informed by the views of writers of these texts unless we have access to attempters. Even if suicide note writers have not been acculturated into a specific discourse community to attain skills in writing suicide notes, psychologists have noted differences between authentic and simulated notes such as genuine note writers giving more information regarding final affairs and writing longer texts. Black (1993: 702) contends that these differences “highlight the discrepancy between cultural ideas about suicide and the reality of suicide.” This shows two things. First, with this genre, it seems that texts produced in a role-play differ from authentic texts. Second, the fact that systematic differences between authentic and simulated suicide notes have been found indicates that, even in the absence of a discourse community, features of the context of situation such as the writer's relation with the audience and author intention (Bhatia, 1993) exert a systematic influence on the shape of the text. 3.3. Rhetorical moves 3.3.1. Description of moves Table 1 gives the nine rhetorical moves that were postulated to account for the functions expressed in the set of suicide notes analyzed in addition to a category called “problem.” This last category includes instances of text that could not be ascribed a rhetorical function either because the language was difficult to understand or because the language seemed to be expressing a function not found in any other suicide note analyzed. Some moves contain constituent steps and are more complex than others. This list of moves and constituent steps reveal the multiple communicative purposes of this genre, as will be discussed below. First, it is clear from Table 1 that no move is an obligatory move, if presence in 100% of the texts analyzed serves as the criterion (Samraj, 2014). What is not clear from the table is that there is no easily discernible preferred order of moves or steps, and we will demonstrate this in detail below. Thus the order of moves in the following discussion as well as the move numbering is at least partly arbitrary.

2 Psychologists distinguish between attempters and completers of suicides and have compared the notes produced by these two groups of people and have found some important differences across the texts (Handelman & Lester, 2007).

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

6

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

Move 1, Addressing Recipient, is similar to what others have analyzed as salutation and has been referred to as a structural move in contrast to a rhetorical move (Sadeghi & Samuel, 2013). Addressing the recipient, as in “Mom and Dad,” (SN 66)3 or “My darling,” (SN 92), has been analyzed as an independent move although it does not have the same sort of rhetorical weight as the other moves in the schema provided in Table 1. As seen from Table 1, Move 1 is found in nearly half the texts. Most structural features seen in business or professional letters are absent in suicide notes and only two structural elements, addressing recipient and expressing good-bye (Step 7.5), were found; both have been incorporated in the move-step analysis. The signature, as in “Yours sincerely, [Name]” (SN 24) or simply a statement of the name of the note author, has been analyzed as a possible instantiation of Step 7.5 in Move 7, the Bidding Farewell move, discussed below. The move that is most frequently present is Move 2, Providing Explanation, appearing in nearly 82% of the texts. Its importance is also manifested by the move taking up almost 50% of textual space in suicide notes. It seems that, for the 18e38% of suicides who leave a note behind (Leenaars, 1988), providing an explanation for the act is a key purpose of the text. It is also noteworthy that over 73% of the very short notes (30 notes with fewer than 20 words) contain this move. This move is realized through a number of steps, none of which is obligatory. A key step in Move 2 is that of providing one or more reasons for taking one's life or justifying the act, Step 2.2. The importance of this step is seen in a number of ways. First, it is present in 78% of all notes. Second, it can be fairly extended, as shown by the amount of textual space it occupies. Third, it is the most frequent choice in notes with just one step. Two examples of Step 2.2 are provided below: (1) After almost a year I don't seem to be getting any better on the medication e I am still afraid to go out & can't drive the car. Therefore I have decided to take the matters into my own hands. (SN 23) (2) I am not worthy of you, you are the perfect wife you deserve so much better … It changed my life forever I haven't been the same since it affected me in a way I never felt possible. I am filled with so much hate, hate toward myself hate towards God and unimaginable emptyness it seems like everytime we do something fun I think about how [Name] wasn't here to share it with us and I go right back to anger. (SN 32) Step 2.3, taking responsibility or ascribing responsibility to others for the act, is closely related to the one of providing reasons. Whether one takes responsibility for the act or blames it on another usually follows from the explanation provided, although a statement of responsibility can be found without a statement of reason, as in example 3 below. Example 4 illustrates Step 2.2 followed by Step 2.3: (3) PS Dont blame yourself it's all my fault (SN 77) (4) I have been depressed for almost a year now. Don't blame anyone but me. (SN 113) In some cases, a description of the larger context or some background event different from the precipitating event might also be discussed as a part of this move and is labeled Step 2.1: (5) For 23 years we lived happy together. Our married life was ideal … (SN 95) The final step in this move, Step 2.4, where the author discusses his or her feelings about committing the act, appears infrequently: (6) I would have preferred to die with more dignity. (SN 3) The second most frequent move in the suicide note is Move 7, Bidding Farewell, which can be realized through one or more of five different steps: expressions of positive or negative feelings towards others (Steps 7.1 and 7.2), a statement of thanks to recipient or others (Step 7.3), a prediction of others' feelings (Step 7.4) or an expression of good-bye and good wishes (Step 7.5). This is a fairly broad move with the general function of a farewell, but it is rhetorically quite distinct from the other major moves in this genre such as providing an explanation for the suicide or giving directions to family and friends about arrangements after the author's death. A number of constituent functions have been identified as ways of bidding farewell. Expressing one's love or other positive feelings (Step 7.1) for those being left behind can appear alone as a farewell expression as in example 7, where it is the last sentence in the note, or it can appear with a goodbye (Step 7.5), which could be a simple instance of good-bye and a signature (example 8), or a longer expression of good wishes (not shown, for reasons of length): (7) I do love you (SN 14) (8) I love you sweetie, love always mommy (SN 10) The author can also bid farewell in two other ways: by expressing hate or other negative feelings towards others (Step 7.2), for example, “damn those coyotes” (SN 53) or by expressing appreciation to others (Step 7.3), as in the following example:

3

All suicide notes (SN) have been anonymized and will only be referred to by the number we assigned them.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

7

(9) Thank you for all your attempts to get me back to normal living. (SN 26) In the majority of cases, we see Step 7.1 and positive emotions spike (see results of linguistic analysis below), but in the smaller set in which step 7.2 occurs, we see a burst of negative emotions. In a few cases, the leave-taking includes a prediction of what others will feel (Step 7.4): (10) I know that you, [Name] will be sorry but it seems to be the best way out of the situation at present. (SN 23) This move is striking in its function to maintain and perhaps even strengthen social relationships and bears resemblances to moves that could appear in any other letter expressing leave-taking (except for the permanence in leave-taking in this case). The two most frequent constituent steps are ones where the author expresses his or her love for those being left behind and the last good-bye step (7.5). Step 7.5 includes simple instances of good-byes and signatures, which might explain its high frequency. The rhetorical function of the third most frequent move (excluding Move 1) in the suicide note, Move 5, Expressing Directions and Wishes, is somewhat more pragmatic than those of the first two more frequent moves. Directions for practical, business-like actions such as locating passwords to bank accounts or making arrangements for one's funeral can be given in a suicide note and have been analyzed as the first step realizing this move (see example 11), and long-term wishes such as those concerning care for a child being left behind or less tangible wishes about how they want to be remembered (example 12) are analyzed as belonging to the second step in this move: (11) Oh yeah I want [Priest's name] to speak @ the funeral his words are beautiful a body viewing & open casket is up to family. (SN 11) (12) All of you, my dear ones, I ask to keep my memory alive in your hearts e To live on in the hearts of our dear ones is all that I can conceive of immortality. Please think of me kindly. Remember that which was good and lovely in our relationship … (SN 88) Apologizing, Move 3, is found in a little more than a third of the texts and is executed in one or two steps: apologizing with or without an explanation (Step 3.1) or asking for forgiveness (Step 3.2). The apology could be directed towards the recipient(s) or someone else, in which case the recipient might be asked to relay the apology or the request for forgiveness to a third party. The apology move does not take up much textual space, only about 6%, even though it appears in about 35% of the texts. In comparison, Move 5, which also only contains two steps, takes up about three times as much textual space. Move 3, Apologizing, is another move that fulfills a social rhetorical function. Excerpts 13 and 14 are examples of Steps 3.1 and 3.2 respectively: (13) To all of my friends I'm so sorry for what I've done to you and put you through (SN 84) (14) Please forgive me. (SN 95) The remaining moves listed in Table 1 are not very frequent and also do not occupy much textual space in the texts. In Move 4, authors describe their mental state at the time of death and also speak (lovingly) about their future after death. This move never appears alone. The key step in this move is Step 4.2, which expresses beliefs or wishes for the time after death (example 15), and is found in 17 out of 23 instances of this move (74%). Another step in which the writer speaks of their attitude towards death (Step 4.1), is exemplified in 16 below. Together these two steps convey the author's feelings about death and what they believe or hope will follow death. As with the other moves, each step can appear without the other step: (15) … when your time comes to go to Heaven you will find your ole Pappy waiting for you. (SN 90) (16) I am now free. I am no longer in pain and I no longer hate myself. (SN 73) It might seem that Step 4.1 overlaps with the last step of Move 2 (Providing Explanation), but Step 2.4 concerns expressions of feelings regarding the decision to take one's life, whereas Step 4.1 expresses feelings toward death in general. Step 2.4 instead seems to be a coda to the reasons given for the decision to end one's life (see example 6 above). The boundary between Step 2.4 and Step 4.1 is no doubt a fuzzy one. The difficulty of drawing boundaries between certain moves has been noted in earlier research. As Bloor (1998: 60) points out, identifying moves can be a “difficult and contentious” activity. The final three moves identified in suicide notes each occupy less than 5% of textual space and appear in fewer than 10% of the texts. Some authors either announce their suicide or describe the methods used, occasionally in great detail (Move 6), as seen in the following examples: (17) I could wish that I had, for my goodby kiss, a 0.38 police special with which I have made some good scores e not records but at least made my mark. Instead, I have this black bitch e bitch, if the word is not familiar to you e but at least an honest one who will mean what she says. Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

8

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

The neighbors may think it's a motor backfire, but to me she will whisper e “Rest e Sleep.” (SN 93) (18) I have decided to end my life. (SN 68) In Move 8, the author makes a commentary about life in general, taken somewhat broadly to encompass generalizations about death. This move is a mixture of commentaries on the author's life and commentaries on the author's view of life in general: (19) I think it is wonderful the way God has looked after me and as you know I don't think death is the end of the story (SN 28) (20) As you sowed, so shall you reap. (SN 85) In some cases, what might seem to be a commentary on life serves as an explanation for the termination of life and is considered Step 2.2 of Move 2, as in the example below. Although the first sentence is phrased as a commentary on life in general, the second makes clear that this is a reason for the suicide: (21) lifes a bitch. I cant fuckin' take it no more, (SN 58) In Move 9, the final move in our schema, the author comments on the note itself, for example, about its brevity, as in the examples 22 and 23 below: (22) I'm sorry this is short. (SN 66) (23) This note should be pretty easy to understand. (SN 60) Move 2, Providing Explanation, occurs in 93 of the total 114 texts in our data set. No single move is obligatory in the remaining 21, but all of the notes contain either Move 5, Expressing directions, Move 3, Apologizing, or Move 7, Bidding Farewell. It seems that the suicide note genre cannot be characterized by the presence of one or more obligatory moves. Instead of concluding that this genre does not contain any obligatory move, as others have concluded about genres such as the grant proposal (Halleck & Connor, 2006), we would like to suggest that some genres can best be characterized by the obligatory choice of one or more moves out of a set of core moves. The results of our analysis suggest that it is obligatory for suicide notes to contain at least one of these four core moves. These moves indicate the range of communicative purposes for this genre: it could be a text explaining the reason for the suicide, apologizing for the act, expressing wishes for the future, or taking leave. The four core moves are quite entangled. 58 of the 114 notes in our data have two or three core moves, 15 have all four core moves. Thus, this is a collection of texts in which as many as four communicative purposes may be in play, with all logically possible pairings found. Beyond these core moves, the suicide note genre also contains a number of other less frequent moves, which also take less textual space. These moves, which appear with at least one core move, we propose to call minor moves. As mentioned earlier, the notes vary in length. We analyzed the correlation of note length with number of individual steps in a note and ascertained that there was a strong positive correlation 0.769 (p < .001). The correlation between length and number of moves was significantly weaker at 0.614 (p < .001). We conclude that the longer texts differ from the shorter ones not so much in containing a greater variety of moves but in being realized by a greater variety of constituent steps. Although some short notes (with fewer than 20 words) contain only one step of one move, other short notes fulfill a number of communicative purposes as seen in the following note, which only contains 13 words but includes four moves: (24) Darling [Name] Please forgive me I can't go on. I do love you (SN 14) In this short note, the author addresses the recipient (Move 1), asks for forgiveness (Move 3), provides an explanation (Move 2) and finally bids farewell (Move 7), showing that suicide notes, even when they are not very long, can be rhetorically complex. To summarize, the rhetorical structure of the suicide note can be described through a set of nine moves, some with constituent steps. Some of these moves can be considered core moves, with Move 2, Providing Explanation, being the most frequent and developed move. In addition, suicide notes also fulfill the interpersonal purpose of apologizing and bidding farewell and the more pragmatic purpose of giving directions and stating wishes for the future. These core moves embody the multiple communicative purposes of this genre. In addition, some other minor moves can also be present in suicide notes. 3.3.2. Linear order and co-occurrence of moves Speech genres have been found to be fluid in their structure and determining the order of rhetorical moves has been a challenge for analysts of such genres (Rowley-Jolivet & Carter-Thomas, 2005). The same can probably be said about texts that

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

9

are not codified, texts for which exemplars are not easily available, and those produced without much planning, such as the suicide note. Therefore, arriving at generalizations about the order of moves for this genre posed some challenges. It seems that this genre is not characterized by fixed positions for moves. Even so, a few broad observations can be made about the linear order of the moves, mainly with regard to the beginning and end of the suicide note. In addition, we will also discuss patterns of co-occurrence of certain steps and moves in this genre. The Addressing Recipient Move mostly appears as the first move, if it is present in the text, and never as the last. Interestingly, notes without the Addressing Recipient Move tend to be shorter and contain fewer moves and steps. Seventy percent of notes above the median length of 65 words contained Move 1 while only 28% of notes below the median length contained it. When Move 1 is not present, most of the openings in notes with at least three individual steps are instances of Move 2, Providing Explanation, but there are a few interesting cases of suicide notes opening with instances of Move 5, Expressing Directions and Wishes. Move 7, Taking Leave, appears in around 60% of the suicide notes and does not always close the suicide note. Although Step 7.5 is the most frequent closing step when this move is present, there are several notes in which Step 7.1, Expressing Love, is the final step. Consider the following example: (25) I could've never asked for a better mom. I know I probably never told you enough but I really love you (SN 52) If we consider the concept of a closing move to be meaningful only in texts with at least three moves, the two most common closing moves are Move 7 (Farewell) and Move 5 (Expressing Directions). Move 5, however, occurs much more frequently in the middle of a note. Some generalizations can also be made about which moves appear early in the suicide note and which appear in the second half. Move 2, Providing Explanation, tends to appear early in a suicide note. Often Step 2.2 begins this move and is followed by Step 2.3, which either expresses responsibility for the act or ascribes it to another. What is interesting is that it is not uncommon to find a cycling of the Apology Move (M3) and the Explanation Move (M2), where pairs of these two moves are repeated in a text, as in the example below: (26) I am so sorry for this. (Move 3, Step 3.1) I feel I just can't go on. I have always tried to do the right thing but where there was once great pride now its gone … (Move 2, Step 2.2) I just can't be any good to you or myself. The pain is overwhelming. (Move 2, Step 2.1) Please try to forgive me. (Move 3, Step 3.2) (SN 51) The kind of repetition of move sequence seen in the example above could be a result of a text written without much planning or revision. It could also signal the close relationship between the functions of two moves in this example. An author may provide an apology before presenting the explanation and another part of the explanation might warrant further apology, as in example 26. Another set of repetitions is that of Move 3 and Move 7, where authors apologize and express their love or thanks to those left behind. Minor moves do not have any fixed position in the genre. In addition, the Apology move exhibits a large degree of flexibility in the company it keeps. On the whole, although we can make some qualified generalizations about the order of moves in this genre, most moves in the suicide note are characterized by a lack of a specific position in the linear order of the genre. This is illustrated in Fig. 2, a graph showing the distribution of steps following the two steps of Move 5. The lines for Steps 5.1 and 5.2 both have multiple spikes, indicating multiple possible following steps, and demonstrating that neither of these steps has any strong relationship with any following step (or move). In a loosely structured genre like the suicide note, there might still be patterns of co-occurrence of moves rather than patterns of contiguous moves. To probe this possibility, we produced co-occurrence tables for the suicide notes at both the move and step levels, searching for significant clustering of rhetorical functions. The statistical significance of these cooccurrences was computed, because there are unequal numbers of certain moves and steps in the data as discussed above and seen in Table 1. The measure of statistical significance used was Pointwise Mutual Information or PMI (Church & Hanks, 1990), a measure of strength of association commonly used in computational linguistics. Positive (or negative) values indicate greater (or lesser) than expected numbers of co-occurrences, given the overall frequencies of the two moves. For example, the high number of co-occurrence of Step 5.1 with Step 2.2 is not very significant because of the high frequency of Step 2.2; this is indicated through the low statistical significance of 0.12 for this co-occurrence.4 Some interesting patterns of co-occurrences are nevertheless seen in the use of certain steps, particularly with Step 7.1, Expressing Love. It is interesting that 80% of the instances of Step 7.3, Expressing Thanks, co-occur with Step 7.1. But more interesting is the finding that out of 7 instances of Step 7.2, 5 instances also co-occur with Step 7.1 indicating that expressions of hate are not mutually exclusive with expressions of love. The two steps that realize the Apology Move also co-occur

4 In the interest of space, we only provide a table with the co-occurrence numbers and statistical significance for Moves 2e5 and 7 in the paper. Cooccurrence numbers and statistical significance for all the steps can be found under Supplementary data.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

10

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

Fig. 2. The following step distribution for the two steps of Move 5. (‘#’ indicates the end of the note.)

significantly with Step 7.1, Expressing Love, which is not surprising, as pairings of Moves 3 and 7 occur repeatedly in the suicide notes. Expressions of long term wishes (Step 5.2), also have a high co-occurrence rate with Step 7.1, unlike Directions for short-term plans (Step 5.1), which tend to be much more business-like requests, such as those pertaining to bank accounts. Step 7.3, Expressing Thanks, also co-occurs with Step 5.2 revealing that those who express long-term wishes in their notes are also more likely to express thanks and love to the recipients of the notes. Thus, there are frequent associations of particular steps from different moves or even the same moves. A few noteworthy co-occurrences can be seen at the level of whole moves (see Table 2). The co-occurrence of moves that were significant according to the PMI numbers were those between Moves 3 and 7, Apologizing and Bidding Farewell, between Moves 5 and 7, Expressing Directions and Wishes and Bidding Farewell, and between Move 4, describing one's state at and after death, with Move 5, directions to loved ones about long- and short-term concerns. The findings from our analysis of co-occurrence of moves or steps in the texts being analyzed show that, for genres where strong patterns in linear order are not present, an analysis of co-occurrence of particular moves or steps might provide another means to explore the nature of a genre. For the suicide note, the analysis of co-occurrence revealed a clustering in notes of expression of thanks, love, and apologies as well as a clustering of long-term wishes with expressions of love and thanks. In addition, note writers who spoke about their state at and after death also tended to inform readers of their short- and long-term wishes for them. The cooccurrence analysis also reveals the degree to which the four core moves are entangled in the texts. None of the four moves has any strong positive or negative associations with any other. Rather, they combine freely. 3.3.3. Conclusion on move structure analysis The move structure analysis has shown that rhetorical moves can be identified even in a genre with no associated discourse community. The analysis revealed that suicide notes can be characterized by a set of nine moves, some of which contain constituent steps. Even short suicide notes were shown to contain rhetorical complexity. Although the Explanation move is the key move in this genre both in terms of frequency and textual space taken, it is not an obligatory move. Instead, the suicide note is best described as comprising four core rhetorical moves, the Explanation Move (M2), the Apology Move (M3), the Farewell Move (M7) and the Directions and Wishes Move (M5). Our results suggest that the rhetorical structure of some genres can be accounted for by a choice of at least one move out of a set of a core moves, instead of a list of obligatory moves. This analysis revealed the robustness of move structure analysis as a tool for understanding the communicative purposes of a text that is at the nexus of interdisciplinary research. The move structure analysis also revealed that the suicide genre does not have a clearly identifiable linear order although some generalizations can be made about the position of some core moves. Moreover, we have shown that even in a genre with a loose linear structure, it might be possible to identify some move and step co-occurrence tendencies. Table 2 Number of co-occurrences of two moves with significance (given after /). Move

2

3

4

5

7

Total texts

2 3 4 5 7

0 30/0.084 19/0.013 36/0.084 56/0.024

30/0.084 0 8/0.009 19/0.121 33/0.339

19/0.013 8/0.009 0 14/0.364 16/0.169

36/-0.084 19/0.121 14/0.364 0 39/0.349

56/0.024 33/0.339 16/0.169 40/0.349 0

93 40 23 48 67

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

11

3.4. Text analysis All 114 texts were analyzed for a subset of the lexical categories used by LIWC, a lexical categorization tool that has made appearances in several recent studies of suicide notes (Lester, 2010; Pennebaker & Stone, 2004), as well as in the clinical psychological literature generally (Pennebaker & Stone, 2003). LIWC has a lexicon of over 4000 common words classified into over 80 categories, but only a few have been shown to be useful in suicide studies; for example, Handelman and Lester (2007) compared suicide notes of attempters and completers and showed that attempters had fewer positive emotions than completers. Following up on the recent work on suicide-related texts, we used six of the LIWC categories: “Negative emotion,” “Positive emotion,” “I”, “You”, “Pronoun,” and “Social”. The first two categories are related to the degree of positive and negative emotion being expressed, the last four to the degree of self-absorption and disconnectedness of the author. The innovation in our work is to couple LIWC with move by move and step analyses, which shows that the degrees of positive and negative emotion and social connectedness are not invariant features of the texts, but are focused in particular moves. This is shown in Table 3 below. Overall, the suicide notes in our data set exhibit a rate of positive emotion words of 5.52% (for example, “confidence,” “better,” “easy,” “handsome” and “romantic”). This is comparable to the rate observed in Handelman and Lester (2007: 103), 4% for males and 7% for females. More interesting, from our point of view, is the significant variability in positive language across the moves: From Table 3, it is clear that positive emotion words are particularly salient in Moves 4 and 7. This is not unexpected, given that 4 predicts the suicide's state in death and thereafter, and that 7 is the farewell move. In Move 4, especially in Step 4.2, death is often depicted as something to look forward to (see examples 15 and 16), and Move 7, Bidding Farewell, is an occasion for expressing positive emotions such as love and thanks towards those left behind. On the other hand, positive emotion is lowest in the move where the author is providing the explanation for the suicide (3.25% for Move 2), which mostly includes discussion of the many unfortunate factors that led to the decision to end one's life. We turn, then, to negative emotions. Previous research on suicide (Lester, 2010) has not found the negative emotion rate useful in analyzing these text types. Our results show that negative emotions do play a role, but they are concentrated in a few steps or moves; consequently, the counts are diluted when the texts are considered only as whole notes. The move with the highest percentage of words conveying negative emotion is Move 3, where the author apologizes and asks for forgiveness, but the percentages for the two steps in Move 3 differ sharply. Apologizing for unfortunate features of one's life and for the act of suicide itself (Step 3.1) entails using words that LIWC counts as negative, for example, “loser” and “trouble,” while requesting forgiveness (Step 3.2) entails the use of much more positive words such as “hope” and “kindly.” Interestingly, Move 7, mentioned above in connection with a concentration of positive emotion, also contains a concentration of negative emotion in Step 7.2, where the author expresses hate or negative feelings towards others in leave-taking. In fact, the negative emotion rate in 7.2 is the second highest for all steps (10.53%). What we see, then, is that treating the notes as indivisible units conceals the conflicting signals in many of these texts. Previous analyses of suicide notes and diaries have pointed to the use of first person singular pronouns as indicators of selfabsorption and social disconnectedness (Pennebaker & Stone, 2004). Another indicator of the lack of social connectedness is a

Table 3 Results of LIWC analysis in percentage of words.

Move 2 Step 2.1 Step 2.2 Step 2.3 Step 2.4 Move 3 Step 3.1 Step 3.2 Move 4 Step 4.1 Step 4.2 Move 5 Step 5.1 Step 5.2 Move 6 Move 7 Step 7.1 Step 7.2 Step 7.3 Step 7.4 Step 7.5 Move 8 Move 9 Complete texts

I

Neg emot

Pos emot

Pronoun

Social

You

11.05 8.28 11.18 10.07 11.56 16.19 15.13 19.25 12.29 14.18 11.36 8.46 8.72 8.07 10.90 10.63 14.02 9.02 9.25 9.00 2.88 6.25 11.11 10.69

3.69 2.96 3.37 9.03 5.20 8.78 10.70 3.21 1.97 4.48 0.73 1.14 0.63 1.93 1.72 1.27 0.58 10.53 0 1.00 0.36 2.34 1.59 2.98

3.25 7.10 3.10 2.78 4.62 5.62 1.66 17.11 9.09 8.21 9.52 5.18 3.30 8.07 3.06 13.23 15.06 4.51 13.88 10.00 12.23 3.91 1.59 5.52

20.14 24.85 19.43 28.82 22.54 25.93 25.09 28.34 20.15 18.66 20.88 19.54 18.93 20.48 17.59 25.92 31.05 27.07 22.78 27.00 12.23 14.84 24.87 20.98

7.07 14.20 6.61 13.19 4.05 10.70 9.23 14.97 8.60 3.73 10.99 12.08 10.21 14.94 3.06 23.93 30.01 15.04 14.59 11.00 23.38 11.72 10.58 10.75

1.81 5.92 1.46 5.21 2.89 4.80 4.80 4.81 4.42 1.49 5.86 3.80 2.51 5.78 0.57 9.97 11.24 7.52 10.32 10.00 6.83 3.91 2.65 3.61

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

12

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

reduction in the use of language referring to others, such as second and third person pronouns and expressions such as “anybody” and “somewhere,” and a reduction in the use of words LIWC labels “social.” The “social” category includes lexical items describing social relationships and social activities, for example, nouns such as “children” and “friend” and verbs such as “e-mail” and “talk.” The results in Table 3 again indicate that the use of the first person pronoun, second person pronoun, and pronouns in general can also vary across moves. The high frequency of first person pronouns, and the low frequency of second person pronouns in Move 2, and especially in the step which provides reasons for the suicide (Step 2.2), shows the suicides' focus on themselves, feelings, and agentiveness when describing the reasons for taking their lives. In contrast, Move 7, where the author bids farewell, contains almost equal proportions of use of the first and second person pronouns, supporting the social nature of this move and the expression of meaning relating to both self, the recipient, and others. Interestingly, Move 7, especially Step 7.1, where love is being expressed to the reader and others, shows the largest percentage of words from the category, “social” (30.01% in contrast to 6.61% for Step 2.2) and general “pronouns” as well (31.05%). In this section, we have tried to show that pairing analysis of the rhetorical structure of a genre with lexical analysis can offer a more nuanced understanding of texts such as suicide notes and ultimately of the mind crafting the text. Previous studies have not found that negative emotions or evidence of social connectedness are strong features, but we argue that analyzing the texts as a whole may actually underestimate the degree to which negative emotions and strong social connections are still felt, and also expressed, when the communicative purpose calls for it. 4. Conclusions and implications for genre analysis This genre analysis of suicide notes, a non-academic, semi-occluded genre has some implications for genre theory. This study indicates that it is perhaps not the case that all genres with specifiable communicative purposes and rhetorical organizations belong to a discourse community of users. The semi-occluded nature of suicide notes does not explain the absence of a discourse community since members of some discourse communities, such as a disciplinary community, can produce and consume both occluded genres, such as manuscript reviews, and other non-occluded genres, such as book reviews and research articles. The case of suicide notes seems to indicate the need for constructing higher-level categories of genres in genre theory. It seems that one category of genres is produced by members of identifiable discourse communities while another is not associated with a specific discourse community, one that has a set of public goals, a mix of novice and expert members, genres, and specific lexis, as characterized in Swales (1990). Ransom letters, also of interest to forensic linguists, is another genre without an identifiable discourse community. In sum, texts with no identifiable discourse community may also constitute genres, which share a rhetorical structure and communicative purposes. We postulate then that some genres, such as suicide notes, belong to the broader speech community as such and not to a specific discourse community where members become experts through acculturation. It might be argued that authors of suicide notes have learned the structure of this genre by reading exemplars available in the media (due to its semi-occluded status). However, the research on simulated versus authentic texts provides evidence against this argument. Fake suicide notes, presumably modeled on what is publicly available of this genre, can be distinguished from authentic suicide notes by trained suicidologists and forensic linguists (Leenaars, 2004; Olsson, 2009), suggesting that it is the authentic communicative purposes that drive the genre's rhetorical features and not authors' emulation of available exemplars. With the absence of an identifiable discourse community, analysis of the suicide note as a genre calls for a greater emphasis on the role of social context. Suicide notes are inscriptions that accompany an act, as a caption accompanies a figure in a text. Here the Systemic Functional Linguistics notion of genre as a type of text in our larger culture enacting a specific type of situation (Martin & Rose, 2003: 7) seems relevant. In the tradition of Systemic Functional Linguistics, it might be useful in further studies to consider the type of social context shaping the genre of suicide note. In terms of methodology, this study has revealed that move structure analysis is a robust tool for elucidating the communicative purposes of a semi-occluded genre, even when there is little opportunity for employing an ethnographic methodology. Move structure analysis as a methodology provides a viable means for understanding a genre that might embody within it great variation in structure and length. The results provide further evidence that short texts can be rhetorically complex. More importantly, the results indicate that it may not always be useful to label moves as obligatory or optional. Instead, we offer a different way of categorizing moves as either core moves or minor moves. This analysis of suicide notes shows that some genres may not be defined by a set of obligatory moves. Instead, all texts belonging to a genre may require at least one move from a set of core moves. As such, “what holds membership together is not a shared list of defining features, but inter-relationships of a somewhat looser kind” (Swales, 1990: 49). The texts that belong to this genre do so not because of one privileged communicative purpose but because they share a family resemblance to each other (Swales, 1990: 52) through these core moves. This study has also shown that a written genre that is not codified might bear some similarity to spoken genres in terms of a lack of linear order, recycling of moves, and fluidity of structure. In such a case, it may be useful to explore co-occurrence of moves in a text when the contiguous presence of moves is difficult to identify. Although this study has concerned an avowedly non-academic genre, we would like to speculate briefly on implications of our findings for EAP. In particular, student academic writing both at the undergraduate and graduate levels has been discussed in previous research as a means of acculturation of novices into disciplinary communities (e.g., Prior, 1998; Walvoord & McCarthy, 1990). In fact, some writing-in-the-disciplines research has pointed to variation in discoursal expectations even within the same discipline (e.g., Herrington, 1985; Samraj, 2004). The notion of discourse community has been employed in Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

13

studies of student writing, with the instructor as gatekeeper and students being acculturated into the discoursal values of a specific disciplinary community or sub-community. (See Wilder (2012) for further discussion of the use of discourse community in studies of student writing.) However, these specific classroom discourse communities are not often strongly tied to easily identifiable genres characterized by a list of rhetorical moves. Our finding that not all genres can be characterized by a list of obligatory moves might be relevant to analysis of student writing. Student writing could include greater variation in rhetorical moves and their linear order due to complexly layered communicative purposes. Just as not all identifiable genres (such as suicide notes and ransom notes) can be related to specific discourse communities of participants, perhaps not all discourse communities with expert gatekeepers give rise to the production and consumption of groups of texts with easily identifiable communicative purposes and move structures. This speculation needs further consideration in work on disciplinary writing, especially in undergraduate courses. Most importantly, we believe this study of suicide notes shows how move structure analysis can be profitably applied to a non-academic, non-professional genre and have real-world application. In this regard, by performing the same sorts of lexicogrammatical analysis used by psychologists on a set of texts whose rhetorical constituents had already been identified, we were able to show the utility of the move structure analysis in further elucidating the nature of suicide notes: Specifically, features believed to characterize suicide notes as a whole in fact shift in value across different rhetorical moves. In so doing, we would like to state that 25 years after the publication of Genre Analysis, Swalesian move analysis remains a robust method of analysis, which holds much promise for the analysis of genres far beyond academic and professional genres in multiple disciplines. Acknowledgment The authors would like to thank John Olsson, head of the Forensic Linguistics Institute, for generously making his data available. Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jeap.2015.04.006 References Askehave, I., & Swales, J. M. (2001). Genre identification and communicative purpose: a problem and a possible solution. Applied Linguistics, 22(2), 195e212. Bhatia, V. K. (1993). Analysing genre: Language use in professional settings. London: Longman. Black, S. T. (1993). Comparing genuine and simulated suicide notes: a new perspective. Journal of Consulting and Clinical Psychology, 61(4), 699. Bloor, M. (1998). English for specific purposes: the preservation of the species. English for Specific Purposes, 17(1), 47e66. Bunton, D. (2005). The structure of PhD conclusion chapters. Journal of English for Academic Purposes, 4(3), 207e224. Cheng, W. (2011). ‘Excellence, always’: A genre analysis of engineering company brochures. Intercultural Interactions in Business and Management (pp. 45e72). Bern: Peter Lang. Church, K. W., & Hanks, P. (1990). Word association norms, mutual information, and lexicography. Computational Linguistics, 16(1), 22e29. Edelman, A. M., & Renshaw, S. L. (1982). Genuine versus simulated suicide notes: an issue revisited through discourse analysis. Suicide and Life-threatening Behavior, 12(2), 103e113. Gillaerts, P. (2012). From job announcements to recruitment advertising: the evolution of recruitment ads in a Flemish newspaper (1946e2010). Linguistic Insights e Studies in Language and Communication, 159, 263e276. Hafner, C. A. (2010). A multi-perspective genre analysis of the barrister's opinion: writing context, generic structure, and textualization. Written Communication, 27(4), 410e441. Halleck, G. B., & Connor, U. M. (2006). Rhetorical moves in TESOL conference proposals. Journal of English for Academic Purposes, 5(1), 70e86. Handelman, L. D., & Lester, D. (2007). The content of suicide notes from attempters and completers. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 28(2), 102e104. Herrington, A. J. (1985). Writing in academic settings: a study of the contexts for writing in two college chemical engineering courses. Research in the Teaching of English, 19(4), 331e361. Kleiner, A. (1981). How not to commit suicide. The CoEvolution Quarterly, 30, 89e108. Leenaars, A. A. (1988). Suicide notes: Predictive clues and patterns. New York: Human Sciences Press. Leenaars, A. A. (1996). Suicide: a multidimensional malaise. Suicide and Life- Threatening Behavior, 26(3), 221e236. Leenaars, A. A. (2004). Psychotherapy with suicidal people: A person-centred approach. John Wiley & Sons. Lester, D. (2010). The final hours: a linguistic analysis of the final words of a suicide. Psychological Reports, 106(3), 791e797. Lester, D., & McSwain, S. (2010). Poems by a suicide: Sara Teasdale. Psychological Reports, 106(3), 811e812. Martin, J. R., & Rose, D. (2003). Working with discourse: Meaning beyond the clause. Bloomsbury Publishing. Matykiewicz, P., Duch, W., & Pestian, J. (2009). Clustering semantic spaces of suicide notes and newsgroups articles. In Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing (pp. 179e184). Association for Computational Linguistics. Miller, C. R. (1984). Genre as social action. Quarterly Journal of Speech, 70(2), 151e167. Olsson, J. (2009). Wordcrime: Solving crime through forensic linguistics. Bloomsbury Publishing. Orr, W. W. (2007). The bargaining genre: a study of retail encounters in traditional Chinese local markets. Language in Society, 36(1), 73e103. Pennebaker, J. W., Chung, C. K., Ireland, M., Gonzales, A., & Booth, R. J. (2007). The development and psychometric properties of LIWC2007. Austin, TX: LIWC. Net. Pennebaker, J. W., & Stone, L. D. (2003). Words of wisdom: language use over the life span. Journal of Personality and Social Psychology, 85(2), 291. Pennebaker, J. W., & Stone, L. D. (2004). What was she trying to say? A linguistic analysis of Katie's diaries. In D. Lester (Ed.), Katie's diary: Unlocking the mystery of suicide (pp. 53e79). Brunner-Routledge. Pestian, J. P., Matykiewicz, P., & Grupp-Phelan, J. (2008). Using natural language processing to classify suicide notes. In Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing (pp. 96e97). Association for Computational Linguistics. Pestian, J., Nasrallah, H., Matykiewicz, P., Bennett, A., & Leenaars, A. (2010). Suicide note classification using natural language processing: a content analysis. Biomedical Informatics Insights, 2010(3), 19e28.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006

14

B. Samraj, J.M. Gawron / Journal of English for Academic Purposes xxx (2015) 1e14

Prior, P. (1998). Writing/disciplinarity. A sociohistoric account of literate activity in the academy. Mahwah, NJ: Lawrence Erlbaum. Prokofyeva, T. (2013). Language use in two types of suicide texts. Unpublished MA thesis. Linkoping University. Rowley-Jolivet, E., & Carter-Thomas, S. (2005). The rhetoric of conference presentation introductions: context, argument and interaction. International Journal of Applied Linguistics, 15(1), 45e70. Sadeghi, V., & Samuel, M. (2013). Genre analysis of the letters of appeal. Discourse Studies, 15(2), 229e245. Samraj, B. (2004). Discourse features of the student-produced academic research paper: Variations across disciplinary courses. Journal of English for Academic Purposes, 3(1), 5e22. Samraj, B. (2014). Move structure. In K. P. Schneider, & A. Barron (Eds.), Pragmatics of discourse (pp. 385e405). Berlin: Mouton de Gruyter. Samraj, B., & Monk, L. (2008). The statement of purpose in graduate program applications: genre structure and disciplinary variation. English for Specific Purposes, 27(2), 193e211. Shapero, J. J. (2011). The language of suicide notes. Doctoral dissertation. University of Birmingham. Shneidman, E. S., & Farberow, N. L. (1957). Clues to suicide. New York: McGraw- Hill. Stirman, S. W., & Pennebaker, J. W. (2001). Word use in the poetry of suicidal and nonsuicidal poets. Psychosomatic Medicine, 63(4), 517e522. Swales, J. (1990). Genre analysis: English in academic and research settings. Cambridge University Press. Swales, J. (1996). Occluded genres in the academy: the case of the submission letter. In E. Ventola, & A. Mauranen (Eds.), Academic writing: Intercultural and textual issues (pp. 45e58). John Benjamins. Swales, J. (2004). Research genres: Exploration and applications. Cambridge: Cambridge University Press. Tardy, C., & Swales, J. M. (2014). Genre analysis. In K. P. Schneider, & A. Barron (Eds.), Pragmatics of discourse (pp. 165e187). Berlin: De Gruyter Mouton. Tausczik, Y. R., & Pennebaker, J. W. (2010). The psychological meaning of words: LIWC and computerized text analysis methods. Journal of Language and Social Psychology, 29(1), 24e54. Walvoord, B. E., & McCarthy, L. P. (1990). Thinking and writing in college: A naturalistic study of students in four disciplines. Urbana, IL: National Council of Teachers. Wilder, L. (2012). Rhetorical strategies and genre conventions in literary studies: Teaching and writing in the disciplines. SIU Press. Yang, R., & Allison, D. (2003). Research articles in applied linguistics: moving from results to conclusions. English for Specific Purposes, 22(4), 365e385. Betty Samraj is Associate Professor of Linguistics at San Diego State University. Her main research interests are in academic writing in different disciplines and genre analysis. She has published her research in journals such as English for Specific Purposes, Journal of English for Academic Purposes, and TEXT and TALK. Her dissertation, The Nature of Academic Writing in an Interdisciplinary Field, was completed in 1995 at the University of Michigan. John M. Swales was her dissertation advisor. Jean Mark Gawron is Professor of Linguistics at San Diego State University. His research interests include Statistical Parsing, Distributional Semantics, and Formal Semantics. He has used distributional semantics to analyze the language of white militant, climate change, and anti-vaccine websites, focusing on the discovery of linguistic group membership markers.

Please cite this article in press as: Samraj, B., & Gawron, J. M., The suicide note as a genre: Implications for genre theory, Journal of English for Academic Purposes (2015), http://dx.doi.org/10.1016/j.jeap.2015.04.006