Analyzing Language 2018
Core Curriculum Uni Bern
Core Curriculum Uni Bern
-
- 1 / 150
-
Flashcards
Differences between Spoken and Written language
What are some of the different approaches to written language out there?
Language and the media (media discourse analysis)
• Looks at mass media, and news media:
how does the news communicate people, ideas and events, and what ”counts” as news?
Digital discourse
• Looks at so-called ‘new’ media (compared to the ‘old’ media that existed before digital technology like smartphones and the internet) and social media (digital platforms for social interaction, networking and content-sharing).
How do digital media affect the nature of written communication, and how do people use the technology strategically for social/cultural purposes?
Linguistic landscapes (LL)
• Looks at how written language appears in the environment, and the words (and other visual modes) displayed in public spaces.
What languages are used in a given space? What functions and symbolic purposes does language have when it is displayed in space?
Stylistics / poetics
• Looks at literary use of language, or verbal (written) art – bridging literary and linguistic research.
Orthography
• Looks at sets of conventions for writing a language, and how they change/develop over time.
Critical discourse analysis (CDA), aka critical discourse studies (CDS)
Approaches to written language: sociolinguistic vs. discourse analytic
sociolinguistic: identify linguistic features, and/or structural patterns that typify certain different kinds of written language: the reader is meant to feel about the question being asked
• e.g. describing how a women’s magazine’s advice column is structured, and the linguistic features which stand out
discourse analytic:
not just set out to describe the characteristic linguistic features of media texts, but also ask questions about how the story is framed, how the reader is positioned in relation to the text, and how linguistic choices may point a reader to a particular interpretation.
• e.g. asking critical questions of how the women’s magazine frames the person asking advice, and how the reader is meant to feel about the question being asked
Digital discourse / computer mediated communication
Zappavigna (2011) presents an analysis of discourse on Twitter, in which she describes hashtags as markers of ‘ambient affiliation’. The ‘ambient’ element of this affiliation comes about through:
the sense that any of a hashtag’s users may not have interacted directly and likely do not know each other, and may not interact again, and;
the fact that it could not occur without adequate search functionality (aka ‘searchability’).
Linguistic landscapes (LL)
Landry and Bourhis (1997:25) define linguistic landscapes as the combination of “the language of public road signs, advertising billboards, street names, place names, commercial shop signs, and public signs on government buildings”.
LL studies the way language marks public space, and the ”symbolic construction of public space” (Ben-Rafael).
As a simple example of how we can ‘read’ street signs, Backhaus (2007) distinguishes between ‘top-down’ and ‘bottom-up’ linguistic landscaping:
’Top-down’ landscaping is done by state or local authorities, and have kinds of ‘official’ status
‘Bottom-up’ landscaping is done by ordinary citizens in banal and often informal ways
Although signs, street and place names, and advertising have been and to a certain extent still are the objects of LL investigation, the built environment, food, tattoos, bodies and T- shirts have also been analysed through LL research.
Critical discourse analysis (CDA)
Critical Discourse Analysis (CDA) is an approach that highlights the social, ideological and political dimension/s of discourse.
Being critical means revealing connections between language, power and ideology which are usually hidden or taken-for-granted.
CDA regards discourse not just as describing existing reality, but augmenting and actively shaping our understanding of reality – considering the ways discourse relates to how power and control are exercised.
In the 1970ies, linguists set up a framework for describing what they argued were ideologically-significantpatterns in the wording and grammar of headlines and news reports.
CDA explicitly shows how discourse is not only produced by societies, but shapes and structures them, and maintain an explicit orientation to social problems.
It should identitfy a pattern in a text, or a set of related texts. You have to be able to show that a piece of discourse is not isolated or unique!
It should focus on what is covert in a text. It is less interesting to find racism on a fascist Facebook page than it is to find the subtle ways that speakers express racist sentiments without appearing overtly racist (see Teo 2000)
Ideology
Representations of the world, worldviews, and views on how society should be organised:
‘A systematic body of ideas, organised from a particular point of view'
Everyone has ideologies. CDA’s treatment of ideology is generally associated with views of the world linked to power and exploitation.
CDA provides us with tools for exploring what options communicators use, and why, and what the consequences are of these choices. Crucuially, we think of these choices in terms of power relations and ideology.
Headlines
Headlines “express, in a highly concise form, the crux of the news event and ... orient the reader to process the text in a pre-determined direction “ (Teo 2000:13)
They are the 'story in microcosm' (Bell 1991:174).
Covert patterns for racism in the news
NEWSPAPER HEADLINES AND LEADS
GENERALIZA TION
QUOTATION PATTERNS
OVER-LEXICALIZA TION
‘results when a surfeit of repetitious, quasi-synonymous terms is woven into the fabric of news discourse, giving rise to a sense of overcompleteness’ (Teo 2000:20)TRANSITIVITY
‘the study of what people are depicted as doing and refers, broadly, to who does what to whom, and how’ (Machin and Mayr 2012:104)THEMA TIZA TION
‘the positioning of a piece of information in a clause is indicative of the kind of prominence or foregrounding the writer wishes to attribute to it’ (Teo 2000:30)LEXICAL COHESION
e.g. in Teo (2000), the ‘juvenilization’ and ‘Asianization’ of the ST and militarization of the police.
Corpus
collection of “texts” compiled for a particular reason (comparison of differentvarieties or types of text; dialect; types of discourse, different representations etc).
Token
an example of a word in a corpus
Type
unique ‘stems’ in a corpus (eg cut, cuts, cutting)
Frequency
the sheer number of examples of a form/word in a corpus (most frequentwords in English?: the, be, of, and, a, in...)
Lemma
(basically) a word stem, the base form of a word.
Concordance
a display of all the search items in a corpus
Collocation
a sequence of words that co-occur more frequently than you’d expect bychance
Collocations: example
Collocations with make:
plans and decisions, negativity, meal/food/drinks, activity/creation, business, produce a reaction
Twitter as linguistic data
Many millions of users, short texts, relatively easily gatherable...more than 500million sent each day in 2015, data are informal (often)
Aims of the App
to tap into people’s fascination with dialect and languagechange;
to collect dialect data from users across England, and in so doing:
Assess the validity of collecting non-standard dialect data using such approaches, complementary to other methods;
Evaluate the extent to which crowdsourced dialect data can be collated, analysed and presented in informative and meaningful ways.
- To collect a nationwide survey of the dialects of England for the first time since the 1950s/1960s
EDA: User background information
43,000 people completed the App quiz and completed the socialbackground information questions (110,000 in all did the quiz...).
Social background information asked about:
Age;
Gender;
Ethnicity;
Age of completion of full-time education;
Highest qualification;
How far they travelled to school or work;
How many times they have moved in past 10 years.
An update in May 2016 used existing quiz answers to improve the prediction accuracy of the app.
Reflections on collecting dialect data with apps
We were not with the users when using the app, so we had no control over data collection:
■ They could have been messing around, choosing random answers, pretending to be from Newcastle when they are from Norwich, or London when they are from Liverpool;
■ They could have answered more than once;
■ Some people reported that they had such mobilebiographies, they ‘didn’t know’ where they were ‘from’;
Big data, bad data?
The app does not provide a sociolinguistically typical dataset:
Small samples of carefully chosen speakers from one or few localities;
Tendency to favour ‘native’-speaking, often working class speakers, with geographically and socially stable biographies;
Long stretches of informal conversation;
Consequently:
No sensitive ethnographic information about speakers
No focus on speakers likely to produce the most non-standard, vernacular speech
No ability to consider informal speech, or, in any depth, connected speech, many tokens of the same variable, or linguistic constraints on variation.
BUT:
Regional and national overviews, of the kind carried out in the Survey of English Dialects, have traditionally been painfully slow, and expensive, and yet equally methodologically problematic.
The App was relatively quick to produce, relatively cheap, and has crowdsourced masses of data in a short space of time. It is unlikely there will ever be national surveys using traditionaltechniques....
Collecting intuitions
If we ask people, ‘how do you pronounce this word?’, we are collecting people’s intuitions about their own languagebehaviour, not people’s own language behaviour
Our intuitions are most often wrong when contemplating socially salient linguistic features (Labov 1996). Given that sociolinguists and dialectologists almost always deal with such features, this is a problem for the app.
Multimodality
- communication works across different modes
- language only works in conjunction with other modes
- a need to understand other modes
Linguistic Landscapes
-> way of thinking about signs & words, language & space
- words are meaningful through their placement
- places are always organized and structured by language/s
- cultural / political struggles over language / space are examined
visual discourse methods
- linguistic landscapes
- cultural studies
- visual content analysis
- iconography
- (visual) semiotics
- social semiotics
(visual) semiotics
Barthes: "Italianicity"
First order of meaning: Denotation -> Common sense
Second order of meaning: Connotation -> associated feeling
Third order of meaning: Myths -> underlying cultural meaning
Social semiotics
- meaning-making as a social practice: as (inter)action
"What is the image doing to us?"
- Analyzing "semiotic resources" and "meaning potentials"
Representational meanings: looks at the representation: what is in the image?
Compositional meanings: how the image is arranged
Interactional meanings: how the image engages us
Narrative vs. conceptional images
Narrative (a little girl as a representation for immigrants) vs. conceptional images (the masses as a representation for immigrants)
Information value
how is key information distributed or arranged?
on the bottom of the page, or on top? large or small?
Framing
how information is connected or disconnected in the image / text
connection through: framelines, colour / shape, empty spaces, vetors
What is in the background, what is highlighted?
Salience
What elements of the image stand out and how?
- Size, forground, background, colour, contrast, texture, placement / symmetry
Contact
How represented participants engage viewing participants
Model smiling at you whilst holding product, etc.
Social distance
How intimate (or "close") do represented participants feel to us?
Power / Involvement
relative status of represented and viewing participants
-> viewing angles, dominance
"This could be you if you buy our products" is evoked
Content analysis
Looking for patterns in everyday talk, texts and practices.
When you look at many different examples, you start to see the patterns in communication.
We are all globalizing the world and doing it in the same way —> banal globalization
Definition Berelson
„a research technique for the objective, systematic and quantitative description of the manifest content of communication.“ —> but there is no absolute objective, but CA is fairly objective and quantitative.
Definition Krippendorf
„a research technique for making replicable and valid inferences from texts (or the meaningful matter) to the context of their use“
—> replicability and transparency is very important; some else should be able to do the same.
Inferences = interpretation;
CA sits more or less between quantity and quality. To explain what we see in the given context of the data is important
Key features of Content analysis
- Addresses and/or follows the analysis of newspapers and media texts
- Involves the enumeration of recurrent patterned themes or trends
- Focuses on the identification of specific features of texts
- Should be organizes as a replicable systematic process
- Uses relatively large samples of data (e.g. news articles)
- Considers both “manifest” and “latent” content
—> not only what is seen, but also what is not seen.
- Often focused on dominant representations
Key steps of Content analysis
- Research questions --> priority / salience, bias, history
- Sampling
- Units of analysis --> identify variables and values
- Coding
- Analysis
