Advanced English II (BFE19013)

Media Discourse Analysis I: What It Is and How to Do It

Instructor: Miao Zhang 張淼

What is “media discourse”?

Plain definition: the language of the media — news, interviews, broadcasts — made for an audience not in the room.

Public — made for everyone, not private
Manufactured — someone wrote, edited, and packaged it on purpose, with a purpose
On record — it leaves traces we can collect and study

Why should we care about it?

Media don’t just report the world — they construct versions of it, choice by choice.
Every word is a decision: this word, this quote, this number — not the alternatives.
Those choices add up to how the public sees the world and how they would do about it.

Media discourse analysis = looking closely at those choices and asking why.

How is media discourse analysis done?

Lens 1 — the text: the words themselves. Naming, describing, metaphors, numbers, grammar.
Lens 2 — the producer: who made this, for whom, and why. Outlet, genre, audience, purpose.
Lens 3 — what’s missing: who never gets to speak, which questions are never asked.

Critical Discourse Analysis (CDA)

Studying how power and ideology hide in everyday language — how texts represent people and events in ways that serve some interests more than others.

Lens 1 in detail (1/2): nomination & predication

Referential / nomination — what do they call things? A “system”? A “machine”? Students as “products”? (Refugees as flood?)
Predication — adjectives and metaphors that carry evaluation: rigorous, overwhelming, race, engine.

Spot the pattern: which words belong together?

“…thousands of illegal immigrants to sneak into Britain…” — Daily Star

“bogus asylum seekers have been allowed to pour into Britain to live off our taxes.” — The Sun

Spot the pattern: which words belong together?

Which words seem to belong together — same topic, same picture?
What picture of refugees and asylum seekers do they paint?

Semantic fields: words travel in packs

A semantic field = a set of words related in meaning, clustering around one topic or image.

The writer chooses fields — and the choice quietly builds the message.

The fields they found

Semantic field	Examples from the articles	Effect
Illegal entry	sneak into, break into, pour into Britain	newcomers = intruders
Number	thousands, massive, influx	an overwhelming quantity
Burden	live off our taxes, benefits bill	newcomers = a cost

(Baker et al. 2008)

How to find semantic fields

Underline the loaded words — repeated or vivid nouns and verbs.
Cluster them — which share a topic? Label each cluster.
Count the clusters — which field dominates? Which never appears?
Ask why — why this word from the field, not a neutral one? (sneak into vs. arrive in)

How to find semantic fields

Corpus shortcut (coming weeks): keywords + concordance lines, grouped into categories — Baker et al. (2008) sorted collocates of refugees / asylum seekers into eight fields, then read the lines to interpret them.

Metaphor: one thing said as another

Metaphor = an abstract thing talked about in concrete terms.

Metaphor: one thing said as another

In the quotes above, pour makes people into water.

Metaphor: one thing said as another

Water has no face, no voice, no rights — dehumanisation (Baker et al. 2008).
Water is out of control — the topos of number/threat in a single image (Van der Valk 2000, cited there).
The writer never said anything “bad” — the metaphor did the work.
One metaphor is a flourish; a repeated metaphor (pour, flood, stream) is a strategy.

Quick check: spot the field, spot the metaphor

In pairs, 4 minutes. Two (constructed) headlines about universities:

A. “Universities race to climb the rankings as Asian rivals muscle in.”

B. “A generation of brilliant minds is being locked out of the lab.”

Which semantic field? Which metaphorical domain?

Quick check — what you should have found

A — competition/sport field: race, climb, rivals, muscle in → education = a contest, rankings = the scoreboard. Learning never appears.
B — enclosure field: locked out, lab → talent = people shut outside; the metaphor hides who holds the key and why the door is shut.
Both are “just metaphors” — and both do evaluative work: DHA strategy 2, predication.

Lens 1 in detail (2/2): numbers & grammar

Numbers — rankings, statistics, percentages. They feel objective — but someone chose which numbers to print.
Grammar — who acts and who suffers: “China is pouring products” vs. “grants were frozen”.

Lens 2 in detail: who made this, for whom?

Every media text is oriented to a target audience — and shaped for them.

Genre signals it: a literary interview (book readers) vs. a news feature (general public, policymakers).
Register & tone adjust to the audience: personal storytelling vs. data and official voices.
O’Keeffe: with the audience absent, media design a relationship — friendly, alarming, authoritative.

Lens 2 in detail: who made this, for whom?

Ask of any article: who is this written for, and what is it trying to make them feel or do?

Lens 3 in detail: absences are evidence too

A text tells you not only what the world looks like, but what is not worth mentioning.

Who is quoted — and who never gets a voice?
Which inconvenient facts or viewpoints are simply absent?
If two “balanced” articles leave out opposite things, what does each audience never learn?

Lens 3 in detail: absences are evidence too

CDA’s core habit: account for absences as well as presences — why these choices, among all the alternatives?

A CDA checklist: the five DHA strategies (1/2)

Referential / nomination — how are people and things called and categorised?
Predication — what qualities and attributes are attached to them?
Argumentation — which topoi justify the view? (threat, success, fairness…)

A CDA checklist: the five DHA strategies (2/2)

Perspectivation — whose voice and whose side does the text take?
Intensification / mitigation — what is stressed, what is played down?

(Discourse–Historical Approach, Reisigl & Wodak 2001; Baker et al. 2008, Table 1)

An example

The buzzwords reflecting the frustration of China’s young generation

Session 2

Survey + group CDA workshop

Our lab rats for today

Reading A — Peter Hessler (2024): Peter Hessler on China’s Evolving Education System.

Reading B — NPR (2026): How the U.S. is losing ground to China in university research.

Reading C - Wang Aihua, People’s Daily (2025): Beyond exams, China envisions versatile future for students.

Same broad topic: Chinese education and scientific research.

Pick A and either B or C and analyze using the approach we just learned. Then compare your findings.

Step 1: The survey

In your group, carefully read the sourcees (the first one is an online interview):

What is each article’s topic and tone?
Who does it seem written for?
Which one interests your group most — and why?

Then claim your article (first come, first served — one group per article if possible).

Step 2: Group roles & ground rules

Facilitator — keeps the group moving through the worksheet
Recorder — writes down every finding and every disagreement
Presenter — delivers the report in Session 3
Everyone — a reader and a coder first

Golden rule: complete your own reading and notes before you discuss. Separate eyes first — shared conclusions second.

Step 3: The CDA worksheet

Referential / nomination: how are students, teachers, education, society called and categorised? Which semantic fields dominate the vocabulary?
Predication: what adjectives, metaphors, evaluative words attach to them? Which metaphorical domains keep repeating?
Argumentation: which topoi justify the article’s view? (threat, success, tradition, fairness, freedom…)

Step 3: The CDA worksheet

Perspectivation: who speaks (quoted voices), whose side is the text on?
Intensification / mitigation: what is (not) stressed, what is hedged or played down?

Why CDA first, corpus later?

CDA teaches you to see the choices — naming, framing, absences. That skill comes first.
The risk of close reading alone: maybe you picked a weird example. Maybe it’s just this one article.
Corpus methods (weeks 4+) answer that: do these patterns repeat across hundreds of articles?

Today’s job: produce one solid, evidence-backed close reading. Scale comes next.

Session 3

Group presentations

Presentation format

Your report should cover:

The article — what it is, who it’s for
The construction — 1–2 DHA strategies, each with a quoted example
The absence — one thing the article makes hard to see
Your finding(s) (30 sec)

Listener’s task

While other groups present, note on your worksheet:

Which strategies keep recurring across articles? (nomination? threat topos? numbers?)
Do the articles construct Chinese education/research differently — and can you link that to their audiences?
Which article would you want to fact-check or compare against a corpus?

After the talks: the pattern question

Discussion:

Same broad topic, different constructions. What explains the differences — facts, or audiences & purposes?
If you had 100 articles instead of 4, how would you check whether your pattern is real?

That question — “is it just this text, or a pattern?” — is exactly where corpus linguistics comes in next.

Wrap-up

Keywords in media discourse analysis (1/4)

Term	Plain meaning
Media discourse	Language of media, made for an absent audience (O’Keeffe)
Public / manufactured / on record	Its three properties — traceable, designed, open to study
Audience design	Shaping a text for its intended audience
Framing	The angle a text takes on an event (education as tradition vs. as race)
Semantic field	Words from one meaning-area clustering in a text (power, dawn, grasp)

Keywords in media discourse analysis (2/4)

Term	Plain meaning
Stance	The attitude a text signals toward its topic
Representation	How a text builds a picture of people/events
Ideology	Values & interests hidden in those representations
Absence	What a text leaves out — evidence too
Metaphor / metaphorical domain	Abstract said as concrete; repeated domains = strategy (Baker et al. 2008)

Keywords in media discourse analysis (3/4)

Term	Plain meaning
CDA	Critical Discourse Analysis — power in language
DHA	Discourse–Historical Approach (Reisigl & Wodak) — our five strategies
Nomination/referential & predication	Strategies 1–2: naming, describing
Topos / argumentation	Strategy 3: topoi (threat, success, fairness…)

Keywords in media discourse analysis (4/4)

Term	Plain meaning
Perspectivation	Strategy 4: viewpoint & voice
Intensification / mitigation	Strategy 5: stressing vs. softening
Corpus / keyness / collocation / concordance	Tools for pattern-checking at scale (Baker et al. 2008)