A sentence and a photograph can describe the same event without preserving the same reality.
A map and a video can represent the same city while serving completely different cognitive tasks.
A spreadsheet and a story can both describe a school, a company or a country, yet one preserves quantities and the other preserves sequence, voice and lived meaning.
Media formats are different projection systems. Each preserves some dimensions of reality by sacrificing others.
This article belongs to the How Media Works series. It builds on How Media Recording Works, which explains how events become capturable tokens. Here the question is different: once reality is recorded, which format should carry which part of it?
1. There Is No Universal Media Format
If one format could preserve everything perfectly, humans would not need so many media forms.
We use prose, photographs, equations, diagrams, tables, audio, video, maps and databases because each is good at carrying different structures.
A format is therefore not a cosmetic wrapper. It is part of the representation itself.
Change the format and you change what becomes easy to see.
2. Text Preserves Propositions
Text is exceptionally good at preserving explicit symbolic relationships.
It can describe cause, sequence, qualification, intention, argument and abstraction. It can preserve laws, instructions, stories, contracts and theories with relatively little storage.
But text often loses tone, facial expression, exact spatial arrangement, smell, texture and much of the sensory environment.
Text is powerful because it compresses experience into symbolic structure.
3. Images Preserve Spatial Arrangement
Images excel at showing what was beside what, how something looked, what colour relationships existed and what visual detail occupied a scene.
A photograph can communicate in one second what paragraphs may take minutes to describe.
Yet images are weak at explaining invisible causes. A photograph of a collapsed bridge shows consequence, not necessarily engineering failure mode.
Images are dense in visible information but sparse in explicit causal explanation.
4. Audio Preserves Voice and Time
Audio carries rhythm, timing, accent, tone, silence, emphasis and emotional texture.
A transcript can preserve the words of a speech. The recording preserves how those words sounded.
Audio therefore carries dimensions of human communication that text often compresses away.
But audio is weak at spatial layout and visual reference unless the speaker explicitly describes them.
5. Video Preserves Change Across Time
Video combines moving images with time, and usually sound.
It is powerful for processes: a machine operating, a person speaking, a storm developing, an experiment unfolding.
Yet video is expensive in attention. A receiver may need ten minutes to experience ten minutes of footage.
Text can often summarise those ten minutes in seconds, though at the cost of detail.
Richness and efficiency often trade off against each other.
6. Maps Preserve Spatial Relationships Selectively
A map does not try to look like the world from eye level.
It deliberately transforms the world so relationships become easier to reason about.
A transit map may distort geographic distance while preserving station order and connections. A topographic map preserves elevation. A political map foregrounds borders. A weather map foregrounds atmospheric conditions.
The map succeeds by discarding most physical detail.
A good representation does not preserve everything. It preserves what the task requires.
7. Diagrams Preserve Structure
Diagrams are designed representations.
They can show how parts connect, how flow moves, where forces act or how a hierarchy is organised.
A circuit diagram does not show the physical appearance of the circuit. It shows logical and electrical relationships.
A biological diagram may exaggerate size or colour to make structure legible.
The distortion is useful because it serves explanation rather than documentary resemblance.
8. Charts Preserve Quantitative Pattern
Charts transform numerical relationships into visual form.
A table may contain exact values while a chart makes trend visible.
Line charts preserve change over time. Bar charts emphasise comparison. Scatter plots expose relationships. Histograms show distributions.
Charts are powerful because the human visual system detects pattern rapidly.
But axes, scales and omitted variables can frame interpretation, which is why chart literacy belongs inside media framing literacy.
9. Tables Preserve Exact Comparison
Tables are less dramatic than charts but often better when exact values matter.
A table allows the receiver to inspect rows and columns systematically. It supports lookup, comparison and auditing.
The weakness is cognitive load. Large tables can hide pattern that a chart reveals instantly.
That is why strong reports often use both: chart for pattern, table for detail.
10. Databases Preserve Queryable Structure
A database is not designed primarily for a human to read from top to bottom.
It is designed so many records can be stored in structured fields and retrieved by query.
This makes databases powerful civilisational memory systems.
But the schema determines what can be stored cleanly. What does not fit the fields may be simplified or lost.
11. Equations Preserve Relationships With Extreme Compression
Mathematical notation can compress a large family of relationships into a small symbolic form.
An equation may represent infinitely many cases while using only a line of symbols.
This is one of the highest-compression media forms humans have developed.
Its strength is precision. Its weakness is accessibility: without the decoding knowledge, the symbols are opaque.
High compression often requires high receiver competence.
12. Stories Preserve Human Causality
Narrative is a format for organising events through sequence, agents, goals, conflicts and consequences.
Stories are memorable because human cognition is highly responsive to intention and sequence.
But narrative can overcompress complexity into heroes, villains and clean causal arcs.
Reality often contains systems rather than single protagonists.
13. Lists Preserve Category and Sequence
Lists are simple but powerful. They turn a set into inspectable units.
A checklist supports execution. A ranked list creates hierarchy. A timeline list supports chronology. A glossary stabilises vocabulary.
But lists can imply that categories are equally important or independent when they are not.
14. Interactive Media Preserves Choice
Interactive media adds user action to the representation.
A map can be zoomed. A dataset can be filtered. A simulation can be changed. A webpage can expose layers progressively.
This gives the receiver some control over framing.
Instead of accepting one static view, the user can ask new questions of the representation.
15. Multimodal Media Combines Strengths
Many powerful media objects combine formats.
A textbook may use prose, diagrams, photographs and equations. A documentary combines video, narration, maps, archival records and music. A scientific paper combines text, tables, charts and formulae.
Multimodal representation works because one format can supply what another lacks.
Redundant formats can increase understanding when each contributes a distinct dimension.
16. But More Formats Can Also Create Noise
Adding media does not automatically add understanding.
An unnecessary animation can distract from a chart. Decorative images can occupy attention without adding evidence. Repeating the same claim in many formats can create the appearance of depth without new information.
The right question is not “How much media can we add?”
It is “Which format reduces uncertainty about this question?”
17. Every Format Has a Native Question
- Text: What is being argued, explained or specified?
- Image: What did this look like?
- Audio: What did this sound like?
- Video: How did this change over time?
- Map: Where is it in relation to other things?
- Diagram: How are the parts connected?
- Chart: What quantitative pattern exists?
- Table: What are the exact values?
- Database: What structured records can be retrieved?
- Equation: What relationship holds across cases?
This does not mean formats are locked to one use. It means each has representational strengths.
18. Format Conversion Is Never Neutral
When media moves from one format to another, some structure changes.
Speech becomes transcript. Video becomes summary. Data becomes chart. Article becomes headline. Map becomes verbal directions.
Every conversion chooses what to preserve.
Format conversion is another compression step.
19. The Transcript Problem
A transcript preserves words but may lose timing, sarcasm, accent, laughter and overlapping speech.
It is therefore easier to search and quote than audio but less complete as a record of vocal performance.
Good systems preserve both when stakes justify it: transcript for access, audio for source fidelity.
20. The Screenshot Problem
A screenshot converts an interactive, contextual digital object into a static image.
The conversion can lose links, metadata, edit history, surrounding thread and machine-readable text.
This is why a screenshot is often weaker evidence than the source page itself.
It preserves appearance while discarding structure.
21. The Summary Problem
A summary converts a long token into a shorter token.
The benefit is attention efficiency. The cost is lost detail and qualification.
A good summary preserves the argument’s load-bearing structure. A bad summary preserves only the conclusion.
This becomes especially important when AI systems summarise large corpora.
22. AI Is a Format Transformer
AI can transform text into image, audio into transcript, documents into summaries, datasets into explanations and prompts into synthetic media.
This makes format conversion faster than ever.
It also makes category discipline more important. A generated visual explanation is not the same as a captured photograph. A summary is not the same as the source. A synthetic voice is not the same as a recording of the person.
This connects directly to How Media Authenticity Works.
23. Format Determines Searchability
Text is easy to index. Images require captions, metadata or machine vision. Audio and video become easier to search when transcribed. Databases are highly searchable when fields are structured well.
Thus format affects whether future users can retrieve a record at all.
This connects format to media memory.
24. Format Determines Distribution Cost
A line of text is cheap to transmit. High-resolution video is heavier. Interactive environments may require software, bandwidth and compatible devices.
Historically, format has therefore been constrained by infrastructure.
Cheap printing favoured text. Broadcast technology favoured audio and later video. Smartphones favour short multimodal content. High bandwidth enables live video.
Distribution architecture shapes which formats become culturally dominant.
25. Format Shapes Memory
People remember differently depending on how information was represented.
A photograph can become a visual anchor. A slogan becomes verbal memory. A melody becomes auditory memory. A map becomes spatial memory. A formula becomes procedural or symbolic memory.
The format influences the form in which the past later returns to the mind.
26. Format Shapes Authority
People often infer credibility from format.
A polished chart feels analytical. A photograph feels evidential. A table feels precise. A documentary feels immersive.
These feelings can be useful heuristics, but they can also mislead.
A fabricated chart can still look mathematical. A synthetic image can look photographic. A large table can contain poor data.
The rhetorical authority of a format is not the same as the evidential quality of its content.
27. Format Literacy
A strong receiver asks different questions of different media.
- For text: what definitions and assumptions organise the claim?
- For images: what lies outside the frame?
- For audio: what context is unavailable visually?
- For video: what happened before and after the clip?
- For charts: what do the axes and denominator do?
- For maps: what projection and purpose shape the view?
- For databases: what fields were never recorded?
- For summaries: what qualification was removed?
28. Choose Format by Information Loss
When designing media, ask what would be lost if this information were expressed in another form.
If exact wording matters, preserve text. If tone matters, preserve audio. If motion matters, preserve video. If spatial relation matters, use a map or diagram. If quantitative pattern matters, use a chart. If exact quantities matter, provide a table.
The best format is therefore a fit between question and preserved structure.
29. The Format Matrix
We can compress the system into a simple matrix of strengths:
- Text: semantics and argument.
- Image: visible spatial detail.
- Audio: acoustic timing and tone.
- Video: audiovisual change through time.
- Map: spatial relation and location.
- Diagram: abstract structure and mechanism.
- Chart: quantitative pattern.
- Table: exact comparison.
- Database: structured retrieval.
- Equation: general relation with extreme compression.
- Interactive media: user-directed exploration.
30. Final Thesis
Media does not merely carry information.
The format determines which dimensions of reality survive the journey.
Text preserves symbolic relationships. Images preserve visual arrangements. Audio preserves sound through time. Video preserves audiovisual change. Maps preserve spatial structure. Charts preserve quantitative pattern. Databases preserve queryable records.
Each format is powerful because it is selective.
Media literacy is partly the ability to know what a format can show, what it cannot show, and what may have been lost when reality was translated into it.
Continue with the canonical series at How Media Works | Reality, Representation, Memory and the Human Interface.