Literature Review
Literature Review
Lexical Markers in Fake News
How to cite this article:
Snelgar, Glen (2018). Literature Review. fakernews: http://fakernews.blogspot.com/2018/05/literature-review.html
Research Question(s)
Can fake news articles be
identified using lexical markers?
If so, what lexical
markers are most common in fake news and how reliable an indicator of falsehood
are they?
Abstract
Fake news is a type of tabloid
journalism or propaganda that consists of deliberate misinformation or hoaxes.
It is often created by certain affiliated online communities that are likely to
employ certain types of lexis and focus on certain types of themes. The purpose
of this study is to determine if there are frequent words and lexical markers
which occur in fake news articles that could distinguish them from legitimate
news articles.
Rationale
Previous studies have taken the perspective of a journalism scholar making an informed critical judgement over what constitutes "Fake News." In the seminal 'Stanford Study,' university students were asked to discriminate between genuine news and fake news with varying degrees of success. Therefore, there is a niche available to research how the under-educated lay person can discriminate between genuine and fake news. Also, we can investigate how technology may assist in sorting the real news from the fake.
Definitions
The definition of "Fake News," according to the Cambridge Dictionary is as follows:
noun [ U ] UK /ˌfeɪk ˈnjuːz/ /ˌfeɪk ˈnuːz/
false stories that appear to be news, spread on the internet or using other media, usually created to influence political views or as a joke:
There is concern about the power of fake news to affect election results. https://dictionary.cambridge.org/dictionary/english/fake-news
There are
two different definitions of fake news in the research. The main definition is
a literal definition, and it is favoured by the journalist community. The other
is a discursive definition which is the lexicon used in the genre of fake news.
Fake news
is defined by the false content within the text according to the journalist
community. While this seems simple enough, there are varying types of fake news
which are distinguished by more than just false content. To have a better understanding of fake news,
one must understand what the intention of the creator is. Thus, the definition
of fake news varies depending on the creator.
The definition of
"Fake News," according to the Cambridge Dictionary is as follows
(Cambridge, 2018): false stories
that appear to be news spread on the internet or using other media, usually
created to influence political views or as a joke
The label ‘fake news’ is used frequently within the
journalism discipline to group together all news that is intentionally
misleading and deceptive. Claire Wardle (2017) identifies seven types of fake
news:
1. Satire or parody ("no
intention to cause harm but has potential to fool")
2. False connection ("when
headlines, visuals or captions don't support the content")
3. Misleading content
("misleading use of information to frame an issue or an individual")
4. False context ("when
genuine content is shared with false contextual information")
5. Imposter content ("when
genuine sources are impersonated" with false, made-up sources)
6. Manipulated content
("when genuine information or imagery is manipulated to deceive", as
with a "doctored" photo)
7. Fabricated content
("new content is 100% false, designed to deceive and do harm")
Therefore, these
definitions also include satire and parody as part of fake news. Although
satire certainly contains false content, participants in the political process
may wish to create and consumer satire. The intent of the creator is not to
deceive the consumer, but rather to entertain. If we are to create a tool to
filter fake news from personal news feeds, then we need to ensure that satire
is not filtered out. Therefore, we need a working definition that excludes
satire.
Fake News involves false stories that appear to be news, spread on the internet or using other media, which are intended to deceive to gain political influence
Furthermore, there is also
another definition that we need to use for this study. Since respondents will
be self-selecting what is fake news and what isn't, we need a working
definition that includes all news articles that have been labelled #fakenews by
participants. This definition contradicts what was said above, because it is
based on the perceptions of the consumer, rather than an analysis of the intent
of the creator. However, this is very necessary because texts using this label
will be analysed and compared with a reference corpus. Therefore, all news with
this label is fake news for the purpose of this study. This is based on
analysis of the real life discourse used within the community of consumers of
fake news. Whatever they consider to be fake news is therefore fake news within
this definition.
Recognising fake news
In order to mitigate the
problems associated with fake news, the International Federation of Library
Associations (IFLA, 2018) published a summary in diagram form to assist people
in recognizing fake news. Its main points are:
Consider the source (to
understand its mission and purpose)
Read beyond the headline (to
understand the whole story)
Check the authors (to see if
they are real and credible)
Assess the supporting sources
(to ensure they support the claims)
Check the date of publication
(to see if the story is relevant and up to date)
Ask if it is a joke (to
determine if it is meant to be satire)
Review your own biases (to see
if they are affecting your judgement)
Ask experts (to get confirmation from independent people with
knowledge).
This model will be useful for my research as a model for analysing the
intent of the creator.
Existing Scholarship
The information below is a short summary of existing articles that may be relevant to this study:
The Stanford Study
What is more relevant to this study is that the researchers also found that some readers are susceptible to fake news and even believed in placebos planted by the researchers.
• 8% of readers of fake news recalled and believed in the content they were reading
• the same share of readers (8%) also recalled and believed in "placebos" – stories they did not actually read, but that were produced by the authors of the study
Therefore, this suggests that some people are susceptible to fake news stories, even if they don't read them entirely.
2016 United States Presidential Election
There has been a concern
recently about the impact of fake news on politics, including the power of fake
news to affect election results.
Perhaps the most significant
focus of study on fake news has been the 2016 United States Presidential
election, featuring current US president, Donald Trump. The Trump team created
a narrative accusing mainstream media of generating fake news, and they filled
the vacuum with their own fake news stories (Kirby, 2016). There was a
concerted, organised pattern of deliberately misleading stories disseminated to
consumers, many of these originated from outside of the U.S.
According to Kirby (2016), many
online pro-Trump fake news stories are being sourced out of a small city in
Macedonia, where approximately seven different fake news organizations are
employing hundreds of teenagers to rapidly produce and plagiarize
sensationalist stories for different U.S. based companies and parties.
Therefore, news stories originating from abroad could be flagged or even
filtered out by computer software to prevent such stories reaching the
electorate.
Furthermore, according to Watson (2017), in
the 2016 American election, Russia paid over 1,000 internet trolls to circulate
fake news and information about Hillary Clinton, and they also had the power to
create social media accounts that resembled voters in important swing states
that spread influential political standpoints.
One fake news writer, Paul
Horner said that he had an "enormous impact" on the 2016 U.S.
presidential election (Gunaratna, 2016). His stories consistently appeared in
Google's top news search results, were shared widely on Facebook and were taken
seriously. Horner later claimed that his work during this period was intended
"to make Trump's supporters look like idiots for sharing my
stories"(Gunaratna, 2016). Instead, they had the opposite effect of
reinforcing their world view. In this case, computer software could target the
creators who could be flagged or even blocked if they were found to have
created a certain amount of fake news.
The main issue with all of this
research is that it concerns the U.S. environment only. There was very little
past research on the New Zealand environment, or even any other place for that
matter. This may make it difficult to calibrate computer software to work as a
global tool to filter out content if we only have information from the U.S.
context. I may need to conduct some original research to make it relevant to
the New Zealand context.
Social Media and Fake News
Social media sites and search
engines, such as Facebook and Google, are at the forefront of the global
phenomenon of fake news. It is only fitting that they should be taking measures
to combat fake news. Facebook plays a huge part in spreading fake news. Fake
news was shared more frequently on Facebook than legitimate news stories
(Silverman, 2017).
Analysts explained that this is
because fake news often panders to expectations or is otherwise more exciting
than legitimate news (Agrawal, 2017). Users play a major role in feeding into
fake news stories by making sensationalised stories "trend",
according to BuzzFeed media editor Craig Silverman (Davis, 2016).
The most salient issue in social media news is the usage of the ‘filter bubble’ that has been created that gives the viewer, on social media platforms, a specific piece of the information knowing they will like it. Filter bubbles develop when algorithms select stories to appear in a consumer's news feed based on stories that have been viewed and 'liked' before. Therefore, participants who view a lot of fake news may eventually become inundated by fake news as the filter bubble reaches a critical point. Those participants may find themselves cut off by the flow of fake news, unable to return to the mainstream.
Negroponte (1995) anticipated a world where news through technology becomes progressively personalized and news media constantly adapted content to reader preferences. This prediction has since been reflected in news and social media feeds of modern day.
These social media corporations have received criticism for facilitating the spread of fake news and need to take measures to explicitly prevent the spread of fake news. Algorithms that create filter bubbles could instead create filter blockers that filter out and block and fake news content. My research aims to create a typology that will inform these algorithms of what is and isn’t fake news.
Affiliation and Searchable Talk
Social semiotic theory looks at the discourse of the text and text participants to create a genre of text. Fake news is a genre of text with its own discourse and lexicon which can be grouped around the hashtag #fakenews.
Social media allows consumers to form groups with others who consume the same content and produce the same discourse. Zappavigna (2015) argues that we are currently witnessing a cultural movement from online conversation to what can be termed 'searchable talk'-online talk where people affiliate by making their discourse findable (for example, via metadata such as Twitter hashtags) by others holding similar interests. Therefore, consumers of fake news can create imagined communities across borders and share fake news with each other.
In this way, whole communities can be trapped inside the ‘filter bubble’ with other like-minded users based on what they have previously searched for. Zappavigna argues that an important dimension of social media discourse is its searchability. A key semiotic resource supporting this function is the hashtag, a form of social tagging that allows authors to embed metadata in social media posts. “While popularly thought of as topic-markers, hashtags are able to construe a range of complex meanings in social media texts (Zappavigna, 2012).”
In my research, the hashtag #fakenews is used first as a label (informative) and as normative meta-discourse which creates a social group of similar users. This on-line imagined community consumes, shares and perhaps even produces fake news texts affiliated with this discourse. As the consumers are creating these communities and affiliating with each other, the intent of the creator of the fake news is no longer relevant. Therefore, what has been created is a new demographic that determines truth according to their own norms. This is important for my research, because a lot of what is labeled #fakenews is in reality factual content. It has been labeled as fake news simply because members of that community chose to do so.
Demographics
Who are the people that consume and believe in fake news?
Research in America has looked
at who tends to consume the most fake news in order to create a typology of
this demographic.
A 2018 study at Oxford
University (Narayanan et al., 2018) found that a network of Trump's supporters
consumed the "largest volume of 'junk news' on Facebook and Twitter".
On Facebook, the skew was even greater. There, "extreme hard right pages –
distinct from Republican pages – share more junk news than all the other audiences
put together. A study by Guess, Nyhan, & Reifler (2018) from Princeton
University, Dartmouth College, and the University of Exeter has also examined
the consumption of fake news during the 2016 U.S. presidential campaign. The
findings showed that Trump supporters and older Americans (over 60) were far
more likely to consume fake news than Clinton supporters. Those most likely to
visit fake news websites were the 10% of Americans who consumed the most
conservative information. According to Guess et al. (2018) there was “an 800%
difference in the consumption of fake news stories as related to total news
consumption between Trump supporters 6.2% and Clinton supporters 0.8%”. Brendan
Nyhan, one of the researchers, emphatically stated in an interview on NBC News:
"People got vastly more misinformation from Donald Trump than they did
from fake news websites -- full stop." (Guess, Nyhan, & Reifler,
2018).
This research would appear to
correlate with the Stanford Study described above. That study found the
majority of people discriminated against fake news, but that 8% of participants
believed in fake news, and even believed in stories planted by the researchers.
Implications for my research
There appears to be a community, at least in
the U.S., that regularly consumes fake news and believes in the content. Their
discourse may be distinguished from the discourse of the mainstream to reveal typical
lexical markers of fake news consumers.
Furthermore, there appears to be a typology
used to create these stories (proven by the use of placebos) that can be
applied to alternative facts while still being consumed by this community. This
also forms part of fake news discourse which can gives a picture of what fake
news looks like.
Therefore, there is a niche available to
research how the under-educated lay person can discriminate between genuine and
fake news.
Also, we can investigate how technology may
assist in sorting the real news from the fake. In an effort to reduce the effects of fake
news, fact-checking websites, including Snopes.com and FactCheck.org, have
posted guides to spotting and avoiding fake news websites. In addition,
artificial intelligence is one of the more recent technologies being developed
in the United States and Europe to recognize and eliminate fake news through
algorithms (Marr, 2017).
Perhaps
in the future artificial intelligence will replace journalists and critically
discriminate between the genuine news and the fake news.
Conclusion
Previous studies have taken the perspective
of a journalism scholar making an informed critical judgement over what
constitutes "Fake News." However, we can’t always rely on critical
analysis of the intent of the fake news creator. In the seminal 'Stanford
Study,' university students were asked to discriminate between genuine news and
fake news with varying degrees of success. A number of the participants recalled
reading and believing in fake news stories, and even believed in placebos
planted by the researchers.
The majority of those who read
these fake news stories are able to discriminate between genuine and
false news stories. However, there is an important minority who are unable or
unwilling to think critically about news. This significant group is consuming,
sharing and creating a great volume of fake news stories.
Fake news can have an effect on
the democratic process by spreading false news stories that get shared and
talked about. This demographic has had a significant affect on the political
process in the U.S., and there is the potential that this phenomenon could
spread to other jurisdictions.
Social media is taking over
from mainstream media as the primary medium for consuming news. These social
media companies do not employ journalists to filter out false news stories, and
they may be conversely pushing fake news stories to their consumers through use
of what's called the filter bubble. These companies have an obligation to do
more in combating fake news.
There needs to be a new way to
identify fake news so that computer software can flag or filter out fake news
content according to set algorithms. These algorithms need to be programmed
with the relevant discourse of fake news that is distinguished fro mainstream
news and political satire.
My research seeks to discover what
lexical markers are most common in fake news and how reliable an indicator of
falsehood they are.
References
Agrawal, N. (2017). Where fake news came from – and why some readers
believe it. Retrieved from Los Angeles Times:
http://www.latimes.com/nation/la-na-fake-news-guide-2016-story.html
Allcott, H., & Gentzkow, M. (2017). Social Media and Fake News in the
2016 Election. Journal of Economic Perspectives, 31(2), 211-236.
Retrieved from https://web.stanford.edu/~gentzkow/research/fakenews.pdf
Cambridge. (2018). Cambridge Dictionary. Retrieved from Cambridge:
https://dictionary.cambridge.org/dictionary/english/fake-news
Davies, D. (2016). Fake News Expert on How False Stories Spread And Why
People Believe Them. Retrieved from NPR:
https://www.npr.org/2016/12/14/505547295/fake-news-expert-on-how-false-stories-spread-and-why-people-believe-them
Guess, A., Nyhan, B., & Reifler, J. (2018, June 23). Selective
Exposure to Misinformation: Evidence from the consumption of fake news during
the 2016 U.S. presidential campaign. Retrieved from Dartmouth:
https://www.dartmouth.edu/~nyhan/fake-news-2016.pdf
Gunaratna, S. (2016). Facebook fake news creator claims he put Trump in
White House. Retrieved from CBS News:
https://www.cbsnews.com/news/donald-trump-election-facebook-fake-news-creator-paul-horner-claims-responsibility/
IFLA. (2018). How To Spot Fake News. Retrieved from The
International Federation of Library Associations and Institutions:
https://www.ifla.org/publications/node/11174
Kirby, E. J. (2016). The city getting rich from fake news.
Retrieved from BBC News: http://www.bbc.com/news/magazine-38168281
Marr, B. (2017). Fake News: How Big Data And AI Can Help. Retrieved
from Forbes:
https://www.forbes.com/sites/bernardmarr/2017/03/01/fake-news-how-big-data-and-ai-can-help/#34a4a58270d5
Narayanan, V., Barash, V., Kelly, J., Kollanyi, B., Neudert, L.-M., &
Howard, P. N. (2018, June 23). Polarization, Partisanship and Junk News
Consumption over Social Media in the US. Retrieved from Project on
Computational Propaganda:
http://comprop.oii.ox.ac.uk/research/polarization-partisanship-and-junk-news/
Negroponte, N. (1995). Being Digital. Michigan: Vintage Books.
Silverman, C. (2017). This Analysis Shows How Fake Election News
Stories Outperformed Real News on Facebook. Retrieved from Buzzfeed:
https://www.buzzfeed.com/craigsilverman/viral-fake-election-news-outperformed-real-news-on-facebook?utm_term=.mcWPJ58bN#.shQVqardX
Wardle, C. (2017). Fake news. It's complicated. Retrieved from
First Draft News: https://firstdraftnews.org/fake-news-complicated/
Watson, K. (2017). Russian bots still interfering in U.S. politics
after election, says expert witness. Retrieved from CBS News:
https://www.cbsnews.com/news/russian-bots-still-interfering-in-u-s-politics-after-election-expert/
Zappavigna, M. (2012). Discourse of Twitter and Social Media: How We
Use Language to Create Affiliation on the Web. UK: Bloomsbury Academic.
Zappavigna, M. (2015). Searchable talk: the linguistic functions of
hashtags. Social Semiotics, 25, 1-18.
Comments
Post a Comment