Friday, 13 January 2017

1.      A Historical Overview & The Emergence of Online Publications
According to Chovanec’s (2014) publication dedicated to the emerging genre of the football match reports in the Times in the 1860s, football match reports, and sports reporting in general, were a major development of 19th century journalism. This paper, the most comprehensive of its kind on the topic of football match reports in Britain, suggests the years between 1862-1877 represent a formative age for the genre, with people hired specifically for the purpose of reporting on sports matches during this time.
     In the past, football reports were brief and dedicated to a sport still in its infancy in the 1860s. They contained uncertain, technical and circumstantial information, sometimes detailing the length of the pitch and the rules of the game to audiences unfamiliar with the sport (Chovanec, 2014). Today, with football now an institutionalised and incredibly popular mass spectator sport, football match reports are a permanent and recognisable fixture in national newspapers and their respective websites.
     National newspapers began publishing online in the early to middle 1990s (Thiel, 1998). The Guardian, for example, launched its website in September 1995 and just five months later, a dedicated site for the Euro’96 competition was launched, In 1999, the Guardian Unlimited network of websites was launched comprising ‘Football Unlimited’ among three other networks and in September of that year the site registered one million users (“History of the Guardian Website, 2010). As Thiel (1998) rightly suggests, online media is so popular because of its ubiquitousness. Thiel (1998) comments that online media is ‘never put to bed’, meaning news is available at any time of the day for any reader with access to the website. However, although most likely true at the time of writing, Thiel’s (1998) assertion that writers ‘disappear’ online is one that has now been proven wrong. With the emergence of social media platforms such as Twitter, the writer’s name at the bottom of an online article often contains a link to their personal Twitter profile where readers can access more content written by the writer and find out more about the writer’s personality. Twitter gives a voice to journalists outside of the article where they are more at liberty to disclose their personal views and to interact with their audience.
     In 2003, 35% of Americans were reported to go online for news once a week (Kawamoto, 2003). That number is bound to have increased. Nic Newman, a Research Associate at Reuters Institute, wrote in May, 2016 that 51% of the Digital News Report 2016 sample said they use social media as a source of news each week. When one looks at readership totals for the UK’s top circulating newspapers published by the National Readership Survey in 2014, it is clear to see how many people go online for their news, with the Daily Mail recording over 20 million reads on electronic platforms.
Figure 1: National Readership Survey figures, 2014. (Ponsford, 2015)
     With the rise in popularity of the journalist has come a higher demand for more opinionated pieces of writing. Football match reports are a sub-genre of news reporting and were traditionally objective pieces of writing that reported facts (Politis, 2008). Today, match reports ‘fall in between the grey area between news and more observational writing’ (Prior, 2016). According to the Football Editor of the Sun, Charlie Wyatt, writers are now ‘encouraged to give opinion rather than just ball-by-ball coverage’ and this development marks a significant change in the way football match reports have developed over time. One would argue that the emergence of social media, especially Twitter, has motivated this change. Readers are now more inclined to read a particular piece of writing if it has been written by a particular author. Social media allows readers to interact with and keep up to date with a particular journalist’s material, and many readers now have a significant interest in the opinions of the writers they follow. Accounts dedicated to the sport and football sections of newspapers are widely followed online. For example, the ‘Guardian Sport’ account (@guardian_sport) has 712 thousand followers while the ‘Mirror Football’ account is followed by 424 thousand people. The Twitter profile of the Times’ Chief Football Writer, Henry Winter, which is followed by 1.2 million people, pays testament to the rise in popularity of the journalist; they are no longer just a name at the bottom of the page, but an online personality.
2.      Stereotypes: Tabloids vs Broadsheets
Football match reports, like other news pieces, can carry the overall style employed by a particular newspaper. While ‘tabloid’ and ‘broadsheet’ technically refer to the two main physical formats of British newspapers, they also connote a certain style of reporting (Rogers, 2016). Traditionally, the British tabloid press has a reputation for ‘emotionally charged journalism’, notorious for a focus on scandal and sensationalised human interest stories. Its style is reader friendly and content is ‘malignly concerned with titillating readers with the worst in human behaviour’ (Sterling, 2009). Sterling’s (2009) description of British tabloid newspapers is overwhelmingly negative, but does uncover some characteristics that still exist in today’s newspapers. If football match reports do exhibit the same characteristics seen in the regular news stories of tabloid newspapers, it’s possible that they will have a human focus and consist of reporting high in emotion.
     However, it is important not to tinge all UK tabloids with the same brush. Sparks (2000) suggests that three different types of tabloid newspaper exist, with ‘serious popular’ tabloids straying away from the sensationalistic stereotype and reporting ‘some serious news’ with extensive treatment of health and personal finance. Examples include the Daily Mail and the Daily Express.
     ‘Classic tabloids’, on the other hand, are aimed at a poorer working class readership and are dominated by scandal, celebrity news and TV (Sparks, 2000). Examples include British ‘red tops’ such as the Sun, the Daily Mirror and the Daily Star. It is possible that ‘classic tabloids’ and ‘supermarket tabloids’, which place an even greater focus on scandal and sexual content, are responsible for the negative reputation of British tabloid newspapers portrayed in Sterling’s (2009) publication.
     Broadsheets receive a more positive review in Sterling’s (2009) book. They have a reputation of seriousness and have a serious and educated readership. Rogers (2016) describes the tone employed by broadsheet newspapers as being ‘sober ‘and their coverage as being ‘in-depth’. The differences presented by Sparks (2000), Sterling (2009), and Rogers (2016) between tabloid and broadsheet newspapers advises that one should expect to find differences in their respective football match reports.
3.      The Football Match Report: Structure & Content
     The football match report can be considered as its own text type, part of a larger genre of football writing. While the football match report is defined by its own internal, linguistic characteristics, football writing in general is based on external, non-linguistic criteria such as audience, purpose and activity type (Lee, 2001). Supporting this claim is the fact that football match reports have a discernible structure. According to “How To Write a Sports Report” (n.d.),football match reports contain the result of the match, which teams played and which players scored the goals. Incidents that occur in the game are described, and quotes from managers and players after the game may also be included. However, while the article does provide a very basic overview of how match reports are produced, it fails to disclose the effects of the ever-growing accessibility of online media on the development of football match reports over time. While match reports in newspapers are published and released to the public the next day, online media is immediate and often free to read. As a result, many football journalists reporting on matches are now required to submit ‘on the whistle’ match reports which, for obvious reasons, may not include post-match quotes from post-match press conferences which usually commence at least 10 minutes after the final whistle. Nevertheless, the basic, recognisable structure of the football match report, which cannot effectively meet the requirements of its audience without referencing the score, which teams played and events that occur in the game,  has remained intact and this is what makes it analysable from a linguistic perspective.
     Editors are generally flexible in what they permit from their writers, so long as facts are reported accurately. Although there are no written rules (more unwritten rules that journalists abide by) to enforce how a football match report should be structured, newspapers often have house style guides to ensure material is clear. According to former Sports Editor and current Digital Editor of the Guardian newspaper, Ian Prior, the Guardian has a house style guide that ‘makes several recommendations in terms of clarity and brevity but places few restrictions on vocabulary other than that which is offensive, outdated or plain incorrect’. Football Editor at the Sun, Charlie Wyett, explains that writers are ‘urged to express ourselves and use a bit of humour’, but insists ‘we are never told what to write’. It is clear then that factuality is of paramount importance to the football match report.
     In terms of word length, and although the issue of space does not apply to online editions in the way that it does to printed newspapers, editors do impose word count restrictions on football match reports written in both contexts for purposes of readability. When asked about the restrictions placed on him when writing match reports, journalist Sam Lee, who writes for football news website, said that ‘they tell us to keep it between 400-500 words, but if there’s a good talking point it can reach up to 800 words’. In the present study, the Daily Mail sub-corpus had the highest average word count with 818 while the Daily Express had the lowest with 408.2 words. A football match report shorter than 400 words is likely to lack depth, while a report longer than 800 words may lose the interest of the reader. The average word count of the 45 match reports compiled in the present study is 638.
4.      The Football Match Report: Language
     Bergh & Ohlander (2016) argue the language of football is the world’s biggest ‘special language’ in their article on football verbs and transivity. As one has discussed above, football’s following and its status as a mass spectator sport means it can also be considered as a general or public language (Bergh & Ohlander, 2016). The article posits that special languages owe much to technical vocabulary. A game of football and the events which occur in it, of course, cannot be described accurately without reference to technical terms such as ‘defenders’, ‘midfielders’ and ‘attackers’, to name but a few. Berg & Ohlander (2016) also make an accurate point on object omission, explaining that the ‘ball’ in a football match is so crucial that it need not be mentioned by the writer/commentator reporting on the game.
     The significance of the ball is also noted in Leitner & Hesslemann (1996), Politis (2008) and Muller (2008). In their respective articles, Leitner & Hesslemann (1996) regard the ball as the central object, Politis (2008) as ‘the focus of the game’, while Muller (2008) accurately claims that the pronoun ‘it’ in a piece of football commentary usually refers to the ball. Skill or failure in controlling the ball successfully can constitute the focus of attention in a football match report (Leitner & Hesselmann, 1996). However, as Bergh & Ohlander (2016) state, the ball is not always explicitly referenced. For example, standard body points often have derived action meanings such as ‘header’ and ‘left-footer’ which reference particular types of contact with the ball (Leitner & Hesslemann, 1996). It is common in a football match report for a writer to say ‘he headed into the back of the net’ rather than ‘he headed the ball into the back of the net’ because the ball is central to all actions completed in a football match. It must also be noted that derived action meanings from particular body parts such as those listed above can also be described by the writer. For example, ‘he directed a powerful header into the net’. In this sense, Bergh & Ohlander’s (2016) article was useful in putting forward a selection of possible node words to be investigated in the present study.
     From a methodological point of view, Politis’s (2008) article on lexical variation in written sports reports was particularly useful. The article discusses the core vocabulary of written sports reports and posits that shooting is the prototypical action in football. Additionally, Politis (2008) offers a list of statistically dominant modifying evaluative adjectives from sports reports in two Greek newspapers, including ‘short’, ‘long’, ‘deep’, ‘straight’, ‘volleyed’, ‘impressive’, ‘powerful’ and ‘unstoppable’. As football match reports contain a large number of adjectives and actions described by particular adjectives that corpus software cannot detect without manual input, Politis’s (2008) article was extremely useful in directing me towards a subset of node words appropriate for investigation.
5.      Adjectives & Evaluation
Central to the football match report are adjectives. As “How To Write a Sports Report”( n.d.) convey, the Six Ws (Who, What, Why, Where, When and How) are incredibly important in the football match report. Match reports are written for an audience who are presumed to have not seen the game and therefore accurate description is of paramount importance to the genre.  It is therefore taken as given that adjectives have a prominent role in the core vocabulary of football match reports (Caldas-Coulthard & Moon, 2010).
     The present study is largely based on the findings of Caldas-Coulthard & Moon’s (2010) paper which suggests adjectives are a distinguishing characteristic between tabloid and broadsheet newspapers. The study uses corpus methodology to investigate how social actors are classified in the public discourse of the media and findings suggest that tabloid and broadsheet newspapers use adjectives differently. For example, when using the adjective ‘curvy’, The Sun used it predominantly to describe women as objects of desire. Comparatively, broadsheet newspapers were found to use the same adjective mainly to comment on shape and design. This suggests tabloid newspapers may have more a human focus and that adjectives used in these newspapers are often sexually oriented.
     The investigation is also relevant to the present research in that it uses corpus methodology as a research tool. In comparison to my own study, however, Caldas-Coulthard & Moon (2010) retrieve their data from the ‘Bank of English’ corpus, a 450 million word corpus implemented in 2002-2003. Although the corpus used in this study is much larger than the bank of 28,725 words my own study uses, I was keen to provide an as up-to-date as possible analysis of adjectives being used in football match reports being written today. Caldas-Coulthard & Moon’s (2010) investigation requires a much larger corpus for the researchers to acquire an as representative as possible sample of the nature of the labels which provide categorisation of gender relations. The Bank of English corpus serves its purpose effectively in their study, whereas a more specific study such as my own can make more use of a custom-made corpus such as the Antconc corpus.
     In terms of limitations, Caldas-Coulthard & Moon (2010) only compare adjectives used in the Sun and those in what they term ‘quality newspapers’. The study therefore does not deal with a truly representative sample of tabloid newspapers. The term ‘quality newspapers’ is often used to describe broadsheet newspapers, but it also suggests that the authors may have conducted analysis with a preconceived bias towards broadsheet newspapers. Additionally, only the use of three adjectives chosen by the authors is analysed, meaning the study is very limited in what it can tell us about the general use of adjectives in tabloid and broadsheet newspapers. Nevertheless, the idea that tabloid newspapers may have more of a human focus is an interesting one and a finding that my own study supports.
     While adjective use in tabloid and broadsheet newspapers has received some attention, research into the use of adjectives in football writing has been comparatively sparse. Wiredu & Anderson’s (2008) investigation into adjectives in football reporting was therefore of particular interest to myself when conducting the present study. The report that analysed 121 news reports in the Daily Graphic, Ghana, found that adjectives were used to qualify four main categories: matches, teams, players and fans.
     Wiredu & Anderson (2008) found that most adjectives were used to describe matches and that 62.23% of these adjectives were ‘personal assessments’. Contrastingly, ‘performance’ and ‘assessment’ adjectives (32.39% and 40.85%) were used the  most to describe players, while ‘behaviour’, ‘size’ and ‘emotion’ adjectives (30.44%, 21.74% and 26.8%) described fans. Other classifications included ‘physical appearance’, ‘age’, and ‘basic elements’. 
     In relation to the development of football match reports over the years, Wiredu & Anderson (2008) make an interesting point, stating that the audience want football reports to reflect the tension in the game and that in correspondence with the rise in popularity of sports, the media have a duty to enhance the enjoyment of sports in their reports. A way in which they can do this, surely, is to enhance their description of events.
     In general, I found that Wiredu & Anderson’s (2008) results mirrored my own in many ways. Wiredu & Anderson (2008) recognise that there are separate areas of the football that require description in a report and are also aware of the different types of adjectives used to describe these areas. I found their classifications to be highly representative of the various areas (teams, players, fans, etc.) in need of attention. However, while the report is extremely useful in terms of building a model for adjective categorisation, it analyses football reports from only one newspaper and therefore cannot offer any detailed insight into the differences between tabloid and broadsheet newspapers.
     As Wiredu & Anderson’s (2008) publication evidences, adjectives carry different functions. They can be descriptive e.g. ‘small’ or emotive e.g. ‘disgusting’, however one would argue that both of these functions fall into a larger function of evaluation. As Bednarek (2006) explains in her highly enriching book dedicated to evaluation in media discourse, evaluation is a device for interpreting the world and can simultaneously be used to express the writer’s opinion, to construct relations between the writer and reader, and to organise the text. Evaluation is therefore of paramount importance to the football match report writer who is expected to impart his or her opinion on events that happened in the game through the use of adjectives.
     In her book, Bednarek (2006) questions how news writers express opinions about events, people and situations they are reporting on and whether tabloid and broadsheet newspapers differ in terms of how they express opinion. Significantly, the study analyses ten hard news stories reporting on the same issues in the same year from the Guardian, the Independent, the Daily Telegraph, the Times, the Financial Times, the Daily Express, the Daily Mail, the Daily Mirror, the Daily Star and the Sun. Issues reported on include ‘Israel’, ‘Rio’, ‘HIV’ and ‘Barclays’. Methodologically, the present study was highly influenced by Bednarek’s (2006) work in that match reports analysed were all dedicated to the same five games to ensure similar events were being described. I also chose to analyse nine of the ten newspapers Bednarek analyses in her own study as they represent nine of the United Kingdom’s top circulating newspapers. The Financial Times was not included in the present study as it does not publish football match reports.
     Bednarek (2006) references the work of Francis (1995), who devised six evaluative parameters assuming speakers can evaluate the world as:
·         Good or bad – the parameter of EMOTIVITY
·         Important or unimportant – the parameter of IMPORTANCE
·         Expected or unexpected – the parameter of EXPECTEDNESS
·         Comprehensible or incomprehensible – the parameter of COMPREHENSIBILITY
·         (not) possible or (not) necessary – the parameter of POSSIBILITY/NECESSITY
·         Genuine or fake – the parameter of RELIABILITY
Francis’s (1995) model shows that different kinds of description exist and that evaluative writing/speech can be categorised. Essentially, as Bednarek (2006) suggests, the categorisation of different evaluative methods can give an indication as to whether newspapers differ in terms of how they express opinion.
     It is also true that categorisation can help to explain evaluation in many other genres of writing. As an example, Swales & Burke (2003) investigated the differences in evaluative adjectives across academic registers and devised a coherent seven category model to classify adjectives used in academic articles taken from the Michigan Corpus of Academic English (MICASE) corpus:
·         Acuity – smart, stupid
·         Aesthetic Appeal – beautiful, elegant
·         Assessment – good, bad
·         Deviance – likely, unlikely (how closely related something is to what you would expect)
·         Relevance – how close something is to the topic or field being discussed
·         Size – small, large
·         Strength – weak, strong
When compared to Francis’s (1995) model, the above list of categories is much clearer in that it provides examples for each category/parameter. Further, Swales & Burke’s (2006) model contains categories that can account for a much larger spectrum of adjectives. For example, ‘Assessment’ adjectives in this model could accommodate both the ‘EMOTIVITY’ and ‘IMPORTANCE’ parameters in Francis’s (1995) model, as both parameters include general, opinionated evaluations of situations. It is also difficult to envisage where ‘Strength’ and ‘Size’ evaluations would fit into the Francis model, meaning it would have to be adapted to account for a wider range of evaluations or adjectives. In conclusion, Swales & Burke’s (2003) model is clearer, more accommodating and, as the present study illustrates, much easier to apply to different genres than Francis’s (1995) model.
6.      Corpus Studies
     Corpus methodology allows the researcher to analyse a generalizable set of data, such as adjectives in the football match reports of tabloid and broadsheet newspapers, and to analyse language its broadest sense (Paltridge, 2006). With the aid of corpus analysis, one is able to make comparisons between textual objects at a large scale and test intuitions about texts from other digital methods (Froelich, 2015).
     In her article on corpora and discourse, Bednarek (2009) identifies three kinds of corpus study. This three-pronged approach involves a) large-scale computerised corpus analysis, b) semi-automated small-scale corpus analysis, and c) manual analysis of individual texts. When dealing with large samples of text of over 100,000 words, large-scale computerised corpus analysis is most suitable. Material in the corpora is usually representative of the variety of language for which it has been designed and computer assistance can uncover features of language that are inaccessible to intuition/can’t be discovered through analysis of one or a few texts (Bednarek, 2009). Nevertheless, as Bednarek (2009) explains, the sheer amount of data present in a large-scale computerised corpus means that researchers can only execute searches for formally defined items.
     In a study of adjectives in football match reports where context is very much crucial, semi-automated small-scale corpus analysis is effective. For example, the adjective ‘huge’ may appear in a corpus but may be used to describe the importance of something, rather than the size. In a large-scale study, analysing and contextualising the appearance of every single adjective could prove to be a time-consuming task. Semi-automated small-scale is optimal for sample sizes large enough to require some computer assistance, but also small enough to be manually analysed. This type of analysis makes use of qualitative and quantitative data and is large enough to show a certain degree of representativeness. Alternatively, a researcher may wish to complete manual analysis of texts when dealing with one or a few individual texts (Bednarek, 2009).
     If one is to use Bednarek’s (2009) definitions, Wyatt & Hadikin (2015) conducted a large-scale computerised analysis of football expression in three corpora. Using Sketch Engine, a corpus manager and analysis software developed to enable linguists, lexicographers and translators to search language corpora in more than 80 languages, Wyatt & Hadikin (2015) analysed three sets of data from the British National Corpus (BNC), the SiBol/Port corpus, and the enTenTen corpus. While the BNC is a long-standing corpus consisting of 100 million words from written texts largely produced between 1960-1993, the SiBol/Port corpus of 787,000 UK broadsheet newspaper articles and the enTenTen corpus of 19 billion words solely from internet articles are relatively modern in comparison, having been created in 1993 and 2013, respectively (Wyatt & Hadikin, 2015). While all three corpora are useful for large scale studies such as Wyatt & Hadikin’s (2015), they consist of too many words for semi-automated small scale studies and manual analysis.

     A corpus software toolkit more appropriate for smaller analyses is the Antconc software package designed by Laurence Anthony. In her article on the software, Froelich (2015) describes Antconc as the ‘standalone software package for linguistic analysis of texts, freely available for Windows, Mac OS and Linux’. Instead of rooting through piles of data in corpora such as enTenTen, Antconc allows the researcher to search for concordances, collocations, N-grams and clusters in their own pre-selected data which must be converted into a .txt file before inserting into the software. Antconc is therefore highly convenient for researchers conducting very specific investigations.