MDIA5001 Chap.8 Writing for multimedia platforms
Writing for multimedia platforms
Week 7 starts from the claim that every message runs on several modes at once, and that words on their own are the rare case rather than the normal one. Meaning is made through several simultaneous modes, language, image, sound, gesture and space, each with its own affordances, and the writer's job is to orchestrate them rather than to layer them. Language is good at argument, sequence, abstraction and precise attribution.
Image is good at spatial relationships, emotional immediacy and simultaneity. Sound carries intimacy, pace and atmosphere. Video does movement over time and demonstrates process. Which mode dominates shifts with platform, genre and audience expectation.
Affordance is then defined precisely as the range of functions and constraints an object provides for and places upon its users, and the lecture separates three levels that students routinely collapse: a feature is what exists, an affordance is what the feature makes possible, and the outcome is what somebody actually does with it.
The same feature affords different things to different people, which is why a reporter and a reader can look at the same button and see reach and social currency respectively.
Four platform interaction models follow, each rewarding and penalising different writing: infinite scroll rewards hooks and visual hierarchy and punishes preamble; search intent rewards directness and keywords and punishes narrative delay; a subscription feed rewards depth and craft and punishes thin content; an algorithmic feed rewards emotional resonance and shareability and punishes nuance.
The relationship between words and images gets its own grammar, borrowed from semiotics: anchorage, where text fixes the meaning of an ambiguous image, which is what most news captions do; relay, where text and image together make a meaning neither produces alone; redundancy, where text describes what the image already shows and wastes a mode; and contradiction, used deliberately for irony.
Writing for video adds a counterpoint principle, script and picture should not say the same thing at once, and audio adds the constraint that comprehension is linear. Six named mistakes close the week, and they are the marking traps for the multiplatform assignment.
What this chapter covers
- 01
Every message runs on several modes at once, and the writer's job is orchestration rather than layering
- 02
Modal affordances: what language, image, sound and video are each good at
- 03
Affordance defined, and the three levels of feature, affordance and outcome
- 04
Why the same feature affords different things to a journalist and to a reader
- 05
Four platform interaction models, and what each rewards and penalises
- 06
Anchorage, relay, redundancy and contradiction as four caption relationships
- 07
Why describing the image is the most common caption failure
- 08
Writing for video: the counterpoint principle and sentence length as edit rhythm
- 09
Writing for audio: linear comprehension, conversational syntax and the spoken sentence test
- 10
Data visualisation as writing: chart title, annotation and standfirst as three distinct jobs
- 11
Six common mistakes in multiplatform content, and how a marker sees each one
Fixing four captions using the four text and image relationships
- +1Caption one is pure redundancy. The image already shows a collapsed seawall, so the words add nothing and one of your two modes has been spent saying what the other said. Replace it with what the picture cannot carry: when it failed, which section of coast it protected, and how many properties sit behind it. That is anchorage, and it is what a news caption is for.
- +1Caption two is a label rather than a claim, which is the chart equivalent of redundancy: it describes the axes a reader can already see. A chart title should make the argumentative point in one active sentence, so rewrite it as the finding, something like 'The shoreline has retreated further in the last three years than in the previous seven.' The axis labels can keep the descriptive job.
- +1Caption three fails on a different axis. It is anchorage in form but empty in substance, because 'local resident' identifies nobody and lets the reader weigh nothing. Name the person, give their relation to the story, and the portrait becomes evidence rather than decoration. This is also where an assignment can quietly breach consent rules, so the caption and the permission are one decision, not two.
- +1Caption four is the only one already working. The image of an empty chamber and the line about a debate on Tuesday combine into something neither carries alone, a decision made in a room the public was not in. That is relay. Leave it, and note that relay is the relationship worth aiming for once the basic anchorage job is done.
Key terms
- Multimodality
- The principle that meaning is made through several simultaneous modes, language, image, sound, gesture and space, each carrying what the others cannot.
- Affordance
- The range of functions and constraints an object provides for, and places upon, the people using it, which is distinct from the feature that creates it and from what anyone actually does.
- Anchorage
- The caption relationship in which the words settle which reading of an open image is intended, which is the ordinary job of a news caption.
- Relay
- The caption relationship in which text and image together produce a meaning neither carries alone, common in photo essays and the strongest use of two modes.
- Redundancy
- The caption relationship in which the words restate what is already visible, spending one of the two modes on nothing.
- Counterpoint principle
- The rule for scripted video that words and pictures should not say the same thing at the same time, letting the visuals carry action while the script carries cause, context and consequence.
- Scrollytelling
- Interactive long form storytelling in which content is triggered by scroll position, so the reader's physical movement paces the narrative.
Writing for multimedia platforms FAQ
What does platform adaptation mean, in a way a marker can see?
It means each asset was built for the way its platform is used, and the visible evidence is that the assets differ in more than shape. A carousel can build an argument slide by slide because it is swipeable, so slide one carries a single strong statistic, the middle slides carry one fact each, and the last slide carries the call to action.
A short video has to land its hook within the first seconds and be legible with the sound off, so it needs on screen text as structure rather than decoration. A thread is headline driven and rewards a fresh angle rather than the headline again. A professional network post leads on the policy or industry consequence.
What a marker sees when adaptation is fake is uniformity: the same sentence in five boxes, one register everywhere, and a caption that narrates the image. Those three are the named mistakes and they are visible in seconds.
How many pieces do I need, and does a video count as one of them?
The requirement is either three to five distinct media assets drawn from a single coherent news story, or one short video of 60 to 90 seconds as an alternative.
The University's own course record phrases this in a way that reads as though a video were one of the three to five pieces, but the assignment brief and the Week 1 slides both make it an alternative to them, so treat it as an either or and confirm on the course site if you are unsure.
If you choose assets, they can all live on one platform, for example a single carousel of several slides, or they can be spread across three platforms with one piece each. If you choose the video, the structure the brief suggests runs hook, key facts, context, an attributed expert line and a call to action.
Do I have to actually publish this content?
No. The task is explicitly a simulation of a professional workflow: you create the content as though you were working in a newsroom, and you submit the files rather than posting them. If you do choose to post your own work, do not affiliate it with the University.
The submission itself is a single archive containing the assets in the permitted formats, plus a document carrying the link to the news story you adapted and all the textual material, captions and hashtags included. If your video is hosted rather than uploaded, the link has to work when the marker clicks it, because a broken or inaccessible link is treated as a failure to submit the work at all.
Assessment move
Start from one story and hold it fixed. The most common way this assignment goes wrong is that the assets drift onto different angles, which breaks the coherence requirement and makes each piece weaker. Write the single sentence your whole package is making before you open any design tool, and pin it somewhere visible.
Second, decide the modal division of labour explicitly: write down what the image will carry, what the text will carry, and what neither can, then check no two are doing the same job. That is the anchorage and relay test applied in advance rather than as a repair. Third, test on a phone, every time.
The course notes that most digital content is consumed on mobile, and a carousel that reads beautifully on a laptop routinely fails at phone size on type that is too small and captions that are truncated. Fourth, write your captions as standalone units.
Research the course cites finds that many readers treat the caption as the whole story and never follow the link, so a caption that only makes sense alongside the article has already lost most of its audience. Fifth, run the six mistakes as a checklist over the finished package before you zip it.
Working through Writing for multimedia platforms in MDIA5001? Sia is AskSia’s AI Arts and Humanities tutor — ask any MDIA5001 Writing for multimedia platforms question and get a clear, step-by-step explanation grounded in how MDIA5001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
