Purpose
~5 minLLMs tend to fall back on bland cartographic defaults unless a designer names the intended visual form. This lab teaches you to describe map aesthetics as a set of promptable dimensions: balance, fragmentation, specificity, motion, hierarchy, contrast, complexity, symbol type, generalization, and basemap detail.
Core ideaForm is rhetoric. A map can whisper, warn, explain, provoke, stabilize, or dramatize before anyone reads the legend.
Setup: Choose One Map Mission
~10 minChoose a topic that can plausibly support multiple visual tones. Use the same mission for every part so changes in response come from form language, not topic changes.
- Heat risk and cooling access in a city
- Earthquake activity in California
- Noise complaints in a neighborhood
- Wildfire evacuation readiness
- Transit access to clinics or grocery stores
- Invasive species spread across a region
Map mission:
Topic:
Place:
Audience:
Purpose:
Desired emotional tone:
What the map should make visually dominant:
What the map should downplay:
What the map must not imply:
Part 1 — Build a Dondis Vocabulary Deck
~20 minCreate a personal vocabulary deck from Figure 8.3. Write each dimension as a promptable slider rather than a label to memorize.
| Dimension | Extreme A | Extreme B | Map effect | Useful when... |
|---|
| Balance | Uneven layout | Balanced layout | Example: Steers attention | Example: Audience needs quick comparison |
| Structure | Single map | Map series | Example: Steers attention | Example: Audience needs quick comparison |
Starter examples only; expand this in your own notes or submission document.
Mini-deliverableComplete at least eight rows. You will use this deck as your style vocabulary for the rest of the lab.
Part 2 — Prompt Four Contrasting Map Forms
~35 minAsk an LLM to produce four concept descriptions for the same map mission. If your free LLM can generate images, request rough image drafts. If not, request detailed visual briefs.
Using my map mission, create four distinct map-form concepts. Do not write code. For each concept, specify layout, visual hierarchy, symbol strategy, basemap treatment, color/contrast, typography tone, and user impression.
Concept 1: balanced, top-down, minimalist basemap, simple hierarchy, stable/static.
Concept 2: uneven, fragmented, high contrast, accented hierarchy, dynamic/active.
Concept 3: map series, specific, multivariate, detailed basemap, complex hierarchy.
Concept 4: abstract, generalized, icon-symbol based, fluid layout, expressive style.
For each, explain what the form makes users feel or believe before they read the legend.
Map mission:
[PASTE]
Choose the two most rhetorically different outputs and annotate what changed.
Part 3 — Conjoin Dimensions into a Purposeful Visual Dialect
~30 minNow combine dimensions deliberately. Avoid asking for isolated gimmicks. Choose a coherent visual dialect that matches the map's purpose.
Design one coherent visual dialect for my map mission by combining Dondis-style dimensions. Use this structure:
- primary style dialect: minimalist modern / data visualization / historical-retro / realistic-photographic / experiential-artistic / hybrid
- layout: balanced or uneven, with rationale
- continuity: fragmented or fluid, with rationale
- hierarchy: accented, flat, simple, or complex
- abstraction level: abstract, specific, generalized, or detailed
- symbol logic: icon, generic, multivariate, or univariate
- contrast: high or minimized
- basemap: detailed or minimalist
- motion: dynamic or stable
- typography tone:
- title tone:
- what this style helps users understand:
- what this style risks exaggerating or hiding:
Return the result as a reusable style schema.
ReminderStrong style choices exclude things. A style schema should say what the map refuses to do.
Part 4 — Translate the Schema into a Production Prompt
~25 minTurn the schema into a prompt that could guide a map-making agent or image-generation model.
Convert my style schema into a production-ready map design prompt. The prompt should:
1. name the map purpose and audience
2. declare the visual dialect early
3. specify Dondis-style form dimensions explicitly
4. describe layout, hierarchy, symbols, basemap, contrast, typography, and title behavior
5. include constraints that prevent default generic styling
6. include a self-check asking the assistant to explain how the output follows the style schema
Style schema:
[PASTE]
Run the production prompt in a new chat. Save the result or image.
Part 5 — Critique Aesthetic Fit
~25 minUse critique to judge whether the aesthetic serves the map instead of merely looking interesting.
Critique this map design output against my style schema and map mission. Evaluate:
- whether the output preserves the intended visual dialect
- which Dondis dimensions are strongest
- which dimensions drifted or became generic
- whether the visual hierarchy matches the map's message
- whether the style supports credibility, emotion, and usability for the audience
- one style choice I should keep
- one style choice I should revise or remove
Do not praise the design generally. Be specific.
Mini-deliverableCreate a style-fit scorecard with five criteria and evidence from the output.
Part 6 — Style Drift Repair
~20 minIteration often pulls maps back toward defaults. Practice repairing drift without rewriting the whole design.
Revise the production prompt to prevent the style drift identified in the critique. Keep the map's purpose and data unchanged. Add only the minimum new language needed to preserve the intended visual dialect. Then list the exact phrases you added and the drift each phrase is meant to prevent.
Submission
~15 min- Map mission brief
- Dondis vocabulary deck with at least eight dimensions
- Four contrasting map-form concepts
- Reusable style schema
- Production-ready map design prompt
- LLM output or image description
- Style-fit scorecard
- Drift-repair prompt
- A 300-500 word reflection explaining how visual language changed your ability to direct the map
How to Do Well A friendly self-check: aim to use precise visual vocabulary, combine dimensions coherently, connect style to audience and purpose, identify drift with evidence, and revise prompts without falling back on vague words like clean, nice, modern, or professional. Think of this as the success path while you work, not a gotcha at the end.
Psst! Before You Turn This In...
~3 minCan you tell Ian thinks this is really, really important yet? Good. Keep going.
Do the tiny-but-mighty judgment check. If these answers are fuzzy, revise the map idea before polishing the prompt.
- Purpose: Can you say what this map helps someone understand or decide?
- Evidence: Can you point to the data, source, or context behind the claim?
- Omission: Did you leave out, downplay, or defer anything that would distract or mislead?
- Risk: What could be overstated, exposed, or misunderstood?
- Human check: What must be verified outside the LLM?
- Stop rule: What would make you redesign, withhold, or simplify this map?
Ian's recurring refrain Prompt frameworks are scaffolds. The LLM can suggest, critique, and surprise you; you still own the cartographic judgment.
Reference
For use with Prompt Cartography: Interactive Web Map Design with LLMs, CRC Press. www.promptcartography.com