0:52 · Chapter 2 · Why the first map looks messy Dual
title: 🧠 🗺️ 🔎 🧭 💡 Mental Model Building - turning complexity into usable understanding cluster: mental-models scores: rank 100% | centrality 100% | emotive 80% | sentiment 89% about: Mental model building is the process of forming a usable internal representation of a complex situation, domain, or system. It supports judgement, planning, explanation, memory, and action by organising scattered facts, relationships, uncertainties, and priorities into a coherent understanding. image: A human head beside a connected map of landmarks, clear calm composition, exploratory mood. Mental Models
🧠 🗺️ 🔎 🧭 💡 Mental Model Building - turning complexity into usable understanding
clustermental-models
scoresrank 100% | centrality 100% | emotive 80% | sentiment 89%
aboutMental model building is the
process of forming a usable
internal representation of a
complex situation, domain, or
imageA human head beside a connected
map of landmarks, clear calm
composition, exploratory mood.
title: 👁️ ⚡ 🧠 🎨 📊 High-Bandwidth Visual Cognition - using vision as a fast channel for thought cluster: mental-models scores: rank 100% | centrality 90% | emotive 70% | sentiment 85% about: High-bandwidth visual cognition is the human capacity to process spatial layout, shape, colour, contrast, scale, motion, and pattern rapidly through sight. It makes diagrams, maps, icons, and visual structure powerful tools for understanding many relationships at once. image: A bright eye scanning a dense but readable map of coloured shapes and paths. Visual Cognition
👁️ ⚡ 🧠 🎨 📊 High-Bandwidth Visual Cognition - using vision as a fast channel for thought
clustermental-models
scoresrank 100% | centrality 90% | emotive 70% | sentiment 85%
aboutHigh-bandwidth visual cognition is
the human capacity to process
spatial layout, shape, colour,
contrast, scale, motion, and
imageA bright eye scanning a dense but
readable map of coloured shapes
and paths.
title: 🧩 🗺️ 🧠 🧬 🔎 Semantic Spaces Made Intelligible - making high-dimensional meaning readable cluster: mental-models scores: rank 100% | centrality 100% | emotive 70% | sentiment 91% about: Semantic spaces made intelligible are hidden meaning structures transformed into visible, interpretable forms. The concept joins machine-readable embeddings with human-readable maps, labels, regions, icons, and explanations so complex conceptual relationships can be inspected and discussed. image: A glowing abstract cloud resolving into a labelled map with paths and regions. Intelligible Spaces
🧩 🗺️ 🧠 🧬 🔎 Semantic Spaces Made Intelligible - making high-dimensional meaning readable
clustermental-models
scoresrank 100% | centrality 100% | emotive 70% | sentiment 91%
aboutSemantic spaces made intelligible
are hidden meaning structures
transformed into visible,
interpretable forms. The concept
imageA glowing abstract cloud resolving
into a labelled map with paths and
regions.
title: 🗺️ 🧭 🧩 📍 🛤️ Topic Map as Medium - a navigable surface for meanings and relations cluster: mental-models scores: rank 100% | centrality 100% | emotive 80% | sentiment 92% about: A topic map as medium is a spatial surface where concepts, entities, events, and relationships become landmarks, paths, neighbourhoods, and regions. It supports navigation, comparison, explanation, memory, collaboration, and expressive arrangement of meaning. image: A city-like topic map with labelled landmarks, paths, icons, and visible regions. Map Medium
🗺️ 🧭 🧩 📍 🛤️ Topic Map as Medium - a navigable surface for meanings and relations
clustermental-models
scoresrank 100% | centrality 100% | emotive 80% | sentiment 92%
aboutA topic map as medium is a spatial
surface where concepts, entities,
events, and relationships become
landmarks, paths, neighbourhoods,
imageA city-like topic map with
labelled landmarks, paths, icons,
and visible regions.
title: 🧬 📍 🧠 🔢 🔎 Semantic Embeddings - numerical positions for related meanings cluster: semantic-map-layout scores: rank 100% | centrality 100% | emotive 40% | sentiment 78% about: Semantic embeddings are numerical vector representations that place texts, topics, images, or concepts near others with related meaning. They enable similarity search, clustering, retrieval, recommendation, projection, topic mapping, and computational treatment of conceptual relationships. image: Glowing points floating in a deep geometric space, connected by faint meaning- lines. Embeddings
🧬 📍 🧠 🔢 🔎 Semantic Embeddings - numerical positions for related meanings
clustersemantic-map-layout
scoresrank 100% | centrality 100% | emotive 40% | sentiment 78%
aboutSemantic embeddings are numerical
vector representations that place
texts, topics, images, or concepts
near others with related meaning.
imageGlowing points floating in a deep
geometric space, connected by
faint meaning-lines.
title: 📉 🗺️ 🧮 🔻 📐 Dimensionality Reduction - compressing high-dimensional meaning into a plane cluster: semantic-map-layout scores: rank 100% | centrality 100% | emotive 40% | sentiment 56% about: Dimensionality reduction converts high-dimensional data into two-dimensional or three-dimensional coordinates for display, analysis, or exploration. It reveals visible patterns but also creates distortion because many hidden dimensions must be compressed into a limited visual space. image: A many-dimensional wireframe squeezed into a flat labelled map, with visible tension. Dimensionality
📉 🗺️ 🧮 🔻 📐 Dimensionality Reduction - compressing high-dimensional meaning into a plane
clustersemantic-map-layout
scoresrank 100% | centrality 100% | emotive 40% | sentiment 56%
aboutDimensionality reduction converts
high-dimensional data into two-
dimensional or three-dimensional
coordinates for display, analysis,
imageA many-dimensional wireframe
squeezed into a flat labelled map,
with visible tension.
title: 🌀 📉 🧩 💥 🏷️ Raw Projection Messiness - why direct 2D maps often become unreadable cluster: semantic-map-layout scores: rank 100% | centrality 90% | emotive 60% | sentiment 21% about: Raw projection messiness is the unreadable display that often results when high- dimensional semantic data is projected directly into two dimensions. It includes crowding, overlap, unstable orientation, misleading distances, tangled labels, and ambiguous visual groupings. image: A cramped map of tangled labels and overlapping icons, messy but analytically clear. Projection Mess
🌀 📉 🧩 💥 🏷️ Raw Projection Messiness - why direct 2D maps often become unreadable
clustersemantic-map-layout
scoresrank 100% | centrality 90% | emotive 60% | sentiment 21%
aboutRaw projection messiness is the
unreadable display that often
results when high-dimensional
semantic data is projected
imageA cramped map of tangled labels
and overlapping icons, messy but
analytically clear.
title: 📍 ⬛ 🖼️ 📏 🏷️ Screen-Space Dimension - giving topics visible size, shape, and area cluster: semantic-map-layout scores: rank 75% | centrality 80% | emotive 40% | sentiment 64% about: Screen-space dimension is the practical requirement that visible interface objects need width, height, shape, padding, label room, and sometimes territorial area. It converts abstract coordinates into readable nodes, icons, labels, cards, clusters, and regions. image: Tiny coordinate points expanding into labelled cards, icons, and soft territories. Screen Space
📍 ⬛ 🖼️ 📏 🏷️ Screen-Space Dimension - giving topics visible size, shape, and area
clustersemantic-map-layout
scoresrank 75% | centrality 80% | emotive 40% | sentiment 64%
aboutScreen-space dimension is the
practical requirement that visible
interface objects need width,
height, shape, padding, label
imageTiny coordinate points expanding
into labelled cards, icons, and
soft territories.
title: 🧩 💥 📍 🏷️ 🚫 Overlap Problem - projected labels and icons colliding on screen cluster: semantic-map-layout scores: rank 100% | centrality 90% | emotive 60% | sentiment 19% about: The overlap problem occurs when projected topics, labels, icons, or cards occupy the same visible area and become hard to read. It arises because mathematical points have no size, while real interface elements need screen space and collision handling. image: Several labelled circles and icons colliding in a cramped map region. Overlap
🧩 💥 📍 🏷️ 🚫 Overlap Problem - projected labels and icons colliding on screen
clustersemantic-map-layout
scoresrank 100% | centrality 90% | emotive 60% | sentiment 19%
aboutThe overlap problem occurs when
projected topics, labels, icons,
or cards occupy the same visible
area and become hard to read. It
imageSeveral labelled circles and icons
colliding in a cramped map region.
title: 🧭 ✅ 🗺️ 📐 🔗 Semantic Integrity - preserving meaningful proximity in a visual map cluster: semantic-map-layout scores: rank 100% | centrality 100% | emotive 60% | sentiment 89% about: Semantic integrity is the quality of a visual layout where distance, grouping, and neighbourhood still reflect meaningful conceptual relationships. It prevents readability improvements, aesthetic tidying, or collision removal from destroying the map’s underlying meaning. image: A compass over related topic nodes that remain close while labels stay readable. Semantic Integrity
🧭 ✅ 🗺️ 📐 🔗 Semantic Integrity - preserving meaningful proximity in a visual map
clustersemantic-map-layout
scoresrank 100% | centrality 100% | emotive 60% | sentiment 89%
aboutSemantic integrity is the quality
of a visual layout where distance,
grouping, and neighbourhood still
reflect meaningful conceptual
imageA compass over related topic nodes
that remain close while labels
stay readable.
title: 🧲 ⚙️ 🧭 📍 ⚠️ Force-Directed Layout Risk - reducing collisions while distorting meaning cluster: semantic-map-layout scores: rank 50% | centrality 70% | emotive 50% | sentiment 34% about: Force-directed layout risk is the danger that simulated node repulsion can improve visual spacing while weakening semantic accuracy. When repelled nodes drift away from meaningful neighbours, the display may become cleaner but less trustworthy as a map of meaning. image: Repelling magnets pushing topic nodes apart while faint semantic lines stretch. Force Risk
🧲 ⚙️ 🧭 📍 ⚠️ Force-Directed Layout Risk - reducing collisions while distorting meaning
clustersemantic-map-layout
scoresrank 50% | centrality 70% | emotive 50% | sentiment 34%
aboutForce-directed layout risk is the
danger that simulated node
repulsion can improve visual
spacing while weakening semantic
imageRepelling magnets pushing topic
nodes apart while faint semantic
lines stretch.
title: 🧭 🚫 🗺️ 📐 ✅ Semantic-Aware Multi-Stage Layout - readable positions constrained by meaning cluster: semantic-map-layout scores: rank 100% | centrality 100% | emotive 70% | sentiment 95% about: Semantic-aware multi-stage layout is a method for producing readable maps by resolving visual collisions while preserving conceptual relationships. It combines projection, overlap removal, semantic constraints, ranking, and stability so the display remains both legible and meaningful. image: A clean constellation of labelled nodes with no collisions, guided by soft semantic lines. Aware Layout
🧭 🚫 🗺️ 📐 ✅ Semantic-Aware Multi-Stage Layout - readable positions constrained by meaning
clustersemantic-map-layout
scoresrank 100% | centrality 100% | emotive 70% | sentiment 95%
aboutSemantic-aware multi-stage layout
is a method for producing readable
maps by resolving visual
collisions while preserving
imageA clean constellation of labelled
nodes with no collisions, guided
by soft semantic lines.
title: ⚓ 🧭 🔒 🔄 📌 Orientation Stability - keeping regenerated maps aligned over time cluster: semantic-map-layout scores: rank 75% | centrality 80% | emotive 50% | sentiment 79% about: Orientation stability is the property that a regenerated map keeps familiar direction, landmarks, and spatial memory across updates. It helps users compare versions, recognise regions, trust continuity, and add or remove topics without losing orientation. image: A compass pinning a regenerated topic map so its landmarks remain aligned. Stable Orientation
⚓ 🧭 🔒 🔄 📌 Orientation Stability - keeping regenerated maps aligned over time
clustersemantic-map-layout
scoresrank 75% | centrality 80% | emotive 50% | sentiment 79%
aboutOrientation stability is the
property that a regenerated map
keeps familiar direction,
landmarks, and spatial memory
imageA compass pinning a regenerated
topic map so its landmarks remain
aligned.
title: 📌 🔄 🧭 ⚓ 📐 Two-Point Rotation Anchor - aligning layout orientation with stable landmarks cluster: semantic-map-layout scores: rank 25% | centrality 60% | emotive 30% | sentiment 78% about: A two-point rotation anchor is a layout alignment technique that uses two stable reference topics to determine map rotation after regeneration. It reduces random orientation changes and preserves visual continuity across stochastic projection or layout runs. image: Two pinned landmarks rotating a map into alignment, precise technical diagram. Rotation Anchor
📌 🔄 🧭 ⚓ 📐 Two-Point Rotation Anchor - aligning layout orientation with stable landmarks
clustersemantic-map-layout
scoresrank 25% | centrality 60% | emotive 30% | sentiment 78%
aboutA two-point rotation anchor is a
layout alignment technique that
uses two stable reference topics
to determine map rotation after
imageTwo pinned landmarks rotating a
map into alignment, precise
technical diagram.
title: ⬛ 📍 🗺️ 🧵 📦 Regions Not Points - representing topics as areas with shape and extent cluster: semantic-map-layout scores: rank 75% | centrality 80% | emotive 50% | sentiment 86% about: Regions not points is the idea that concepts and categories can occupy visible areas rather than zero-size coordinates. It supports boundaries, territory, membership, overlap, enclosure, influence, and map-like interpretation of conceptual space. image: Map dots expanding into labelled territories with soft boundaries and clear regions. Regions
⬛ 📍 🗺️ 🧵 📦 Regions Not Points - representing topics as areas with shape and extent
clustersemantic-map-layout
scoresrank 75% | centrality 80% | emotive 50% | sentiment 86%
aboutRegions not points is the idea
that concepts and categories can
occupy visible areas rather than
zero-size coordinates. It supports
imageMap dots expanding into labelled
territories with soft boundaries
and clear regions.
title: 📦 🧵 🗺️ ⬛ 🔲 Category Boundaries - visible outlines for conceptual grouping cluster: semantic-map-layout scores: rank 50% | centrality 70% | emotive 50% | sentiment 79% about: Category boundaries are visible enclosures such as boxes, contours, hulls, borders, or shaded areas that show which items belong together. They make cluster membership legible, separate neighbouring territories, and help viewers read conceptual structure spatially. image: Soft boxes and outlines surrounding related map nodes, tidy visual grouping. Boundaries
📦 🧵 🗺️ ⬛ 🔲 Category Boundaries - visible outlines for conceptual grouping
clustersemantic-map-layout
scoresrank 50% | centrality 70% | emotive 50% | sentiment 79%
aboutCategory boundaries are visible
enclosures such as boxes,
contours, hulls, borders, or
shaded areas that show which items
imageSoft boxes and outlines
surrounding related map nodes,
tidy visual grouping.
title: ⬡ 📍 🧊 🗺️ 📐 Voronoi Topic Regions - partitioning space around nearest topic centres cluster: semantic-map-layout scores: rank 0% | centrality 50% | emotive 30% | sentiment 72% about: Voronoi topic regions divide a map into cells around topic centres, where each position belongs to the nearest centre. They can visualise local territory, neighbourhood influence, screen-space ownership, and approximate conceptual regions around individual topics. image: An irregular polygon map divided around labelled topic points, crisp geometric mood. Voronoi Regions
⬡ 📍 🧊 🗺️ 📐 Voronoi Topic Regions - partitioning space around nearest topic centres
clustersemantic-map-layout
scoresrank 0% | centrality 50% | emotive 30% | sentiment 72%
aboutVoronoi topic regions divide a map
into cells around topic centres,
where each position belongs to the
nearest centre. They can visualise
imageAn irregular polygon map divided
around labelled topic points,
crisp geometric mood.
title: 🎨 🖼️ 📍 ✨ 🏷️ Generated Topic Icons - visual symbols for individual map concepts cluster: visual-language-zui scores: rank 75% | centrality 80% | emotive 70% | sentiment 86% about: Generated topic icons are small topic-specific images created or selected to make abstract or textual concepts recognisable. They improve scanning, memory, emotional tone, visual differentiation, landmark formation, and zoomable navigation when paired with labels and placement. image: A grid of small topic icons hovering over labelled map nodes, playful and systematic. Topic Icons
🎨 🖼️ 📍 ✨ 🏷️ Generated Topic Icons - visual symbols for individual map concepts
clustervisual-language-zui
scoresrank 75% | centrality 80% | emotive 70% | sentiment 86%
aboutGenerated topic icons are small
topic-specific images created or
selected to make abstract or
textual concepts recognisable.
imageA grid of small topic icons
hovering over labelled map nodes,
playful and systematic.
title: 👤 🖼️ 🗺️ 📍 ✨ Topics as Recognisable Entities - making abstract nodes visually memorable cluster: visual-language-zui scores: rank 75% | centrality 80% | emotive 70% | sentiment 89% about: Topics as recognisable entities are abstract concepts rendered as memorable visual objects through icons, labels, scale, position, and style. This makes a semantic map feel like a navigable environment with landmarks rather than a plain scatterplot. image: Abstract dots transforming into memorable small characters, objects, and landmarks. Topic Entities
👤 🖼️ 🗺️ 📍 ✨ Topics as Recognisable Entities - making abstract nodes visually memorable
clustervisual-language-zui
scoresrank 75% | centrality 80% | emotive 70% | sentiment 89%
aboutTopics as recognisable entities
are abstract concepts rendered as
memorable visual objects through
icons, labels, scale, position,
imageAbstract dots transforming into
memorable small characters,
objects, and landmarks.
title: 🌫️ ⭐ 📏 🔍 👁️ Salience, Rank, Scale, and Opacity - encoding attention in the display cluster: visual-language-zui scores: rank 100% | centrality 90% | emotive 60% | sentiment 88% about: Salience, rank, scale, and opacity are visual priority signals that control which topics appear large, bright, faint, foregrounded, or hidden. They encode attention, importance, centrality, context, and zoom level without forcing every priority into one score. image: Large bright landmark nodes fading into smaller translucent peripheral nodes. Visual Priority
🌫️ ⭐ 📏 🔍 👁️ Salience, Rank, Scale, and Opacity - encoding attention in the display
clustervisual-language-zui
scoresrank 100% | centrality 90% | emotive 60% | sentiment 88%
aboutSalience, rank, scale, and opacity
are visual priority signals that
control which topics appear large,
bright, faint, foregrounded, or
imageLarge bright landmark nodes fading
into smaller translucent
peripheral nodes.
title: 🔍 🗺️ 🪜 📍 🧭 Zoomable User Interface - multi-scale navigation through information space cluster: visual-language-zui scores: rank 100% | centrality 100% | emotive 80% | sentiment 93% about: A zoomable user interface is an information space where users move between overview and detail by zooming. It supports landmarks, progressive disclosure, local inspection, nested structure, scale-dependent labels, and navigation through large visual knowledge maps. image: A magnifying lens revealing nested topic layers inside a large visual map. ZUI
🔍 🗺️ 🪜 📍 🧭 Zoomable User Interface - multi-scale navigation through information space
clustervisual-language-zui
scoresrank 100% | centrality 100% | emotive 80% | sentiment 93%
aboutA zoomable user interface is an
information space where users move
between overview and detail by
zooming. It supports landmarks,
imageA magnifying lens revealing nested
topic layers inside a large visual
map.
title: 🔍 🪜 🧭 🏷️ ✨ Progressive Zoom Disclosure - revealing more meaning as scale increases cluster: visual-language-zui scores: rank 75% | centrality 80% | emotive 60% | sentiment 88% about: Progressive zoom disclosure reveals information gradually as scale increases. A distant view shows major landmarks and broad structure, while closer views reveal additional topics, labels, icons, edges, examples, regions, and fine- grained semantic detail. image: A stepped zoom path from a few large landmarks to many smaller labelled nodes. Zoom Disclosure
🔍 🪜 🧭 🏷️ ✨ Progressive Zoom Disclosure - revealing more meaning as scale increases
clustervisual-language-zui
scoresrank 75% | centrality 80% | emotive 60% | sentiment 88%
aboutProgressive zoom disclosure
reveals information gradually as
scale increases. A distant view
shows major landmarks and broad
imageA stepped zoom path from a few
large landmarks to many smaller
labelled nodes.
title: 🖼️ 🗣️ 🎨 🏷️ 🧩 Visual-Linguistic Themes - recurring meaning patterns in images and words cluster: visual-language-zui scores: rank 50% | centrality 70% | emotive 60% | sentiment 82% about: Visual-linguistic themes are recurring meaning patterns expressed through both words and images, such as food, fire, trade, danger, status, work, or domestic life. They align labels, icons, colour, metaphor, and spatial placement into a coherent visual vocabulary. image: Small themed icons and labels forming a coherent visual vocabulary on a map. Visual Themes
🖼️ 🗣️ 🎨 🏷️ 🧩 Visual-Linguistic Themes - recurring meaning patterns in images and words
clustervisual-language-zui
scoresrank 50% | centrality 70% | emotive 60% | sentiment 82%
aboutVisual-linguistic themes are
recurring meaning patterns
expressed through both words and
images, such as food, fire, trade,
imageSmall themed icons and labels
forming a coherent visual
vocabulary on a map.
title: 🪜 🔤 🖼️ 📐 🔎 Abstraction Levels - choosing how concrete or symbolic each visual element should be cluster: visual-language-zui scores: rank 75% | centrality 80% | emotive 50% | sentiment 83% about: Abstraction levels describe how concrete or symbolic a representation is, from realistic image to simplified icon, diagrammatic mark, category label, or abstract symbol. The right level affects recognition speed, nuance, ambiguity, compactness, and zoom suitability. image: A staircase from concrete object pictures to simple symbols and labels. Abstraction
🪜 🔤 🖼️ 📐 🔎 Abstraction Levels - choosing how concrete or symbolic each visual element should be
clustervisual-language-zui
scoresrank 75% | centrality 80% | emotive 50% | sentiment 83%
aboutAbstraction levels describe how
concrete or symbolic a
representation is, from realistic
image to simplified icon,
imageA staircase from concrete object
pictures to simple symbols and
labels.
title: 🗣️ 🗺️ ✍️ 🧩 ✨ Map as Expressive Interface - using the visual system to compose meaning cluster: expressive-interface scores: rank 100% | centrality 100% | emotive 80% | sentiment 93% about: A map as expressive interface is a visual system used not only to view information but to compose meaning. Selection, grouping, placement, proximity, annotation, and rearrangement become acts of explanation, argument, storytelling, and design. image: Hands arranging icons and labels on a map as if composing a visual sentence. Expressive Map
🗣️ 🗺️ ✍️ 🧩 ✨ Map as Expressive Interface - using the visual system to compose meaning
clusterexpressive-interface
scoresrank 100% | centrality 100% | emotive 80% | sentiment 93%
aboutA map as expressive interface is a
visual system used not only to
view information but to compose
meaning. Selection, grouping,
imageHands arranging icons and labels
on a map as if composing a visual
sentence.
title: ✋ 🧩 🖼️ 📍 🔄 Drag-and-Drop Arrangement - manually placing images and ideas on a canvas cluster: expressive-interface scores: rank 50% | centrality 70% | emotive 80% | sentiment 89% about: Drag-and-drop arrangement is direct manipulation of images, labels, notes, concepts, or fragments on a spatial canvas. It preserves the expressive freedom of paper collage while adding digital search, linking, scaling, recomputation, and editable structure. image: Cut-out images, notes, and labels being moved by hand on a digital canvas. Arrangement
✋ 🧩 🖼️ 📍 🔄 Drag-and-Drop Arrangement - manually placing images and ideas on a canvas
clusterexpressive-interface
scoresrank 50% | centrality 70% | emotive 80% | sentiment 89%
aboutDrag-and-drop arrangement is
direct manipulation of images,
labels, notes, concepts, or
fragments on a spatial canvas. It
imageCut-out images, notes, and labels
being moved by hand on a digital
canvas.
title: 🔁 📍 🧬 🧭 🔎 Reverse Projection - inferring semantic meaning from a chosen map position cluster: expressive-interface scores: rank 75% | centrality 90% | emotive 60% | sentiment 89% about: Reverse projection estimates a high-dimensional representation from a chosen low-dimensional map position. It allows manual placement to become a semantic query about what an object would mean relative to existing concepts, examples, and embedding geometry. image: An arrow running from a placed map node back into a glowing high-dimensional cloud. Reverse Projection
🔁 📍 🧬 🧭 🔎 Reverse Projection - inferring semantic meaning from a chosen map position
clusterexpressive-interface
scoresrank 75% | centrality 90% | emotive 60% | sentiment 89%
aboutReverse projection estimates a
high-dimensional representation
from a chosen low-dimensional map
position. It allows manual
imageAn arrow running from a placed map
node back into a glowing high-
dimensional cloud.
title: 🧠 ✋ 🔁 🧩 💡 Moving Things as Thinking - spatial arrangement as cognitive work cluster: expressive-interface scores: rank 100% | centrality 90% | emotive 80% | sentiment 93% about: Moving things as thinking is the cognitive practice of using spatial rearrangement to compare, test, refine, and discover relationships. Physically or digitally moving concepts can make implicit associations visible and support reasoning through external structure. image: A hand moving symbolic objects while thought lines connect and rearrange. Move to Think
🧠 ✋ 🔁 🧩 💡 Moving Things as Thinking - spatial arrangement as cognitive work
clusterexpressive-interface
scoresrank 100% | centrality 90% | emotive 80% | sentiment 93%
aboutMoving things as thinking is the
cognitive practice of using
spatial rearrangement to compare,
test, refine, and discover
imageA hand moving symbolic objects
while thought lines connect and
rearrange.
title: 📄 ✋ 🖼️ 🧩 🖊️ Paper-Like Digital Canvas - keeping free spatial arrangement in software cluster: expressive-interface scores: rank 50% | centrality 60% | emotive 70% | sentiment 85% about: A paper-like digital canvas preserves the freedom of arranging notes, images, sketches, and labels in open space while adding computational affordances. It supports provisional thinking, loose grouping, recomposition, annotation, search, zoom, and structured reuse. image: Paper notes and digital icons sharing a single open canvas, tactile mood. Digital Canvas
📄 ✋ 🖼️ 🧩 🖊️ Paper-Like Digital Canvas - keeping free spatial arrangement in software
clusterexpressive-interface
scoresrank 50% | centrality 60% | emotive 70% | sentiment 85%
aboutA paper-like digital canvas
preserves the freedom of arranging
notes, images, sketches, and
labels in open space while adding
imagePaper notes and digital icons
sharing a single open canvas,
tactile mood.
title: 🕸️ 📉 🧭 🔎 ⚖️ Local-Global Layout Tension - good neighbourhoods can still make a distorted whole cluster: shared-semantic-geometry scores: rank 75% | centrality 80% | emotive 50% | sentiment 33% about: Local-global layout tension is the trade-off where a projection preserves nearby relationships but distorts the overall shape of a dataset. It matters when a map must support both neighbourhood inspection and reliable interpretation of global conceptual structure. image: A tidy local web embedded inside a warped larger net, analytical mood. Local Global
🕸️ 📉 🧭 🔎 ⚖️ Local-Global Layout Tension - good neighbourhoods can still make a distorted whole
clustershared-semantic-geometry
scoresrank 75% | centrality 80% | emotive 50% | sentiment 33%
aboutLocal-global layout tension is the
trade-off where a projection
preserves nearby relationships but
distorts the overall shape of a
imageA tidy local web embedded inside a
warped larger net, analytical
mood.
title: 📐 🧬 🤖 🔁 🧠 Shared Representational Geometry - meaningful shapes inside learned spaces cluster: shared-semantic-geometry scores: rank 100% | centrality 100% | emotive 70% | sentiment 91% about: Shared representational geometry is the idea that learned systems may contain structured shapes for concepts, sequences, categories, and behaviours. It links semantic maps, embeddings, neural representations, interpretability, transfer, and cross-model comparison. image: A luminous geometric shape appearing inside several layered representation spaces. Shared Geometry
📐 🧬 🤖 🔁 🧠 Shared Representational Geometry - meaningful shapes inside learned spaces
clustershared-semantic-geometry
scoresrank 100% | centrality 100% | emotive 70% | sentiment 91%
aboutShared representational geometry
is the idea that learned systems
may contain structured shapes for
concepts, sequences, categories,
imageA luminous geometric shape
appearing inside several layered
representation spaces.
title: 🧬 📐 🔁 🤖 🪞 Cross-Model Geometric Patterns - recurring structures across different models cluster: shared-semantic-geometry scores: rank 75% | centrality 80% | emotive 60% | sentiment 88% about: Cross-model geometric patterns are recurring internal structures found across different trained models. They suggest that independently learned systems may converge on similar representational organisation for meaning, order, category, relation, or behaviour. image: Three translucent model spaces with matching shapes highlighted in each one. Cross-Model Shapes
🧬 📐 🔁 🤖 🪞 Cross-Model Geometric Patterns - recurring structures across different models
clustershared-semantic-geometry
scoresrank 75% | centrality 80% | emotive 60% | sentiment 88%
aboutCross-model geometric patterns are
recurring internal structures
found across different trained
models. They suggest that
imageThree translucent model spaces
with matching shapes highlighted
in each one.
title: 🌐 📐 🗣️ 🔤 🌉 Cross-Lingual Semantic Shapes - similar meaning structures across languages cluster: shared-semantic-geometry scores: rank 75% | centrality 90% | emotive 80% | sentiment 92% about: Cross-lingual semantic shapes are comparable representational structures for meanings expressed in different languages. They concern shared conceptual geometry beneath different vocabularies, scripts, grammars, translations, cultural contexts, and multilingual communication. image: Words from several languages curving into the same glowing geometric shape. Language Shapes
🌐 📐 🗣️ 🔤 🌉 Cross-Lingual Semantic Shapes - similar meaning structures across languages
clustershared-semantic-geometry
scoresrank 75% | centrality 90% | emotive 80% | sentiment 92%
aboutCross-lingual semantic shapes are
comparable representational
structures for meanings expressed
in different languages. They
imageWords from several languages
curving into the same glowing
geometric shape.
title: 👤 🍽️ 🧠 ⭐ 📜 Person-Specific Meaning - the same concept carrying different life associations cluster: shared-semantic-geometry scores: rank 75% | centrality 80% | emotive 80% | sentiment 86% about: Person-specific meaning is the way a concept gains different salience, emotion, memory, context, and association for different individuals. It matters for biographical maps, personal knowledge systems, recommendation, education, and humane comparison of lived experience. image: Two people seeing the same dining icon surrounded by different associated nodes. Personal Meaning
👤 🍽️ 🧠 ⭐ 📜 Person-Specific Meaning - the same concept carrying different life associations
clustershared-semantic-geometry
scoresrank 75% | centrality 80% | emotive 80% | sentiment 86%
aboutPerson-specific meaning is the way
a concept gains different
salience, emotion, memory,
context, and association for
imageTwo people seeing the same dining
icon surrounded by different
associated nodes.
title: 🌀 📐 🤖 🧬 🎛️ Manifold Steering - analysing behaviour through representational geometry cluster: shared-semantic-geometry scores: rank 75% | centrality 80% | emotive 60% | sentiment 86% about: Manifold steering studies structured geometric surfaces inside neural network representations and how movement through those surfaces relates to model behaviour. It treats behaviour as shaped by representation geometry rather than by isolated features alone. image: A curved manifold surface gently shaped by arrows inside a model space. Manifold Steering
🌀 📐 🤖 🧬 🎛️ Manifold Steering - analysing behaviour through representational geometry
clustershared-semantic-geometry
scoresrank 75% | centrality 80% | emotive 60% | sentiment 86%
aboutManifold steering studies
structured geometric surfaces
inside neural network
representations and how movement
imageA curved manifold surface gently
shaped by arrows inside a model
space.
title: 📅 📐 🌀 🔢 🤖 Calendar Concept Shapes - ordered dates and weekdays inside model geometry cluster: shared-semantic-geometry scores: rank 25% | centrality 50% | emotive 40% | sentiment 78% about: Calendar concept shapes are geometric structures representing ordered temporal ideas such as dates, months, or weekdays inside a model’s internal space. They provide a concrete example of familiar conceptual sequences appearing as learned spatial patterns. image: Calendar days arranged around a smooth loop inside a glowing model space. Calendar Shapes
📅 📐 🌀 🔢 🤖 Calendar Concept Shapes - ordered dates and weekdays inside model geometry
clustershared-semantic-geometry
scoresrank 25% | centrality 50% | emotive 40% | sentiment 78%
aboutCalendar concept shapes are
geometric structures representing
ordered temporal ideas such as
dates, months, or weekdays inside
imageCalendar days arranged around a
smooth loop inside a glowing model
space.
title: 🛡️ 🔎 🧠 🤖 ⚠️ Interpretability and AI Safety - understanding representations to reduce risk cluster: shared-semantic-geometry scores: rank 75% | centrality 80% | emotive 80% | sentiment 69% about: Interpretability and AI safety concern understanding what AI systems internally represent, how those representations influence behaviour, and where risks may arise. The goal is to improve reliability, debugging, alignment, oversight, misuse prevention, and responsible deployment. image: A shield and magnifying glass over a neural geometry diagram, serious careful mood. AI Interpretability
🛡️ 🔎 🧠 🤖 ⚠️ Interpretability and AI Safety - understanding representations to reduce risk
clustershared-semantic-geometry
scoresrank 75% | centrality 80% | emotive 80% | sentiment 69%
aboutInterpretability and AI safety
concern understanding what AI
systems internally represent, how
those representations influence
imageA shield and magnifying glass over
a neural geometry diagram, serious
careful mood.
title: 🧱 🔤 🖼️ 🧭 📐 Visual Language Grammar - structured meaning from text, image, layout, and symbols cluster: collective-sensemaking scores: rank 100% | centrality 100% | emotive 70% | sentiment 93% about: Visual language grammar is the structured use of text, images, icons, arrows, regions, scale, colour, layout, and abstraction to communicate meaning. It turns visual artefacts into interpretable explanations rather than decorative collections of marks. image: Letters, icons, arrows, panels, and regions arranged as a clear grammar chart. Visual Grammar
🧱 🔤 🖼️ 🧭 📐 Visual Language Grammar - structured meaning from text, image, layout, and symbols
clustercollective-sensemaking
scoresrank 100% | centrality 100% | emotive 70% | sentiment 93%
aboutVisual language grammar is the
structured use of text, images,
icons, arrows, regions, scale,
colour, layout, and abstraction to
imageLetters, icons, arrows, panels,
and regions arranged as a clear
grammar chart.
title: 📚 🎨 🗣️ 🧱 🖼️ Robert Horn Visual Language - historical grounding for text-image communication cluster: collective-sensemaking scores: rank 75% | centrality 80% | emotive 70% | sentiment 89% about: Robert Horn visual language refers to a body of work on combining words, images, diagrams, symbols, layout, and abstraction into systematic visual communication. It provides historical grounding for information murals, knowledge maps, and text-image explanation. image: An open visual-language book with diagrams, arrows, labels, and image panels. Horn Visual Language
📚 🎨 🗣️ 🧱 🖼️ Robert Horn Visual Language - historical grounding for text-image communication
clustercollective-sensemaking
scoresrank 75% | centrality 80% | emotive 70% | sentiment 89%
aboutRobert Horn visual language refers
to a body of work on combining
words, images, diagrams, symbols,
layout, and abstraction into
imageAn open visual-language book with
diagrams, arrows, labels, and
image panels.
title: 🧱 🗺️ 👥 📝 🔗 Information Murals and Mess Maps - large shared artefacts for complex understanding cluster: collective-sensemaking scores: rank 75% | centrality 80% | emotive 70% | sentiment 89% about: Information murals and mess maps are large visual artefacts that combine text, diagrams, categories, timelines, images, evidence, and relationships. They help groups examine complexity, preserve context, surface disagreements, and build shared understanding. image: A wall-sized visual map covered with diagrams, notes, images, and connecting lines. Info Murals
🧱 🗺️ 👥 📝 🔗 Information Murals and Mess Maps - large shared artefacts for complex understanding
clustercollective-sensemaking
scoresrank 75% | centrality 80% | emotive 70% | sentiment 89%
aboutInformation murals and mess maps
are large visual artefacts that
combine text, diagrams,
categories, timelines, images,
imageA wall-sized visual map covered
with diagrams, notes, images, and
connecting lines.
title: 🧑‍🔬 👥 🗺️ 🔗 💬 Cross-Specialist Collaboration - helping different experts share partial knowledge cluster: collective-sensemaking scores: rank 75% | centrality 80% | emotive 80% | sentiment 88% about: Cross-specialist collaboration is the coordination challenge faced by people with different expertise, evidence, vocabulary, incentives, and responsibilities. Visual maps can expose dependencies, gaps, conflicts, shared reference points, and opportunities for synthesis. image: Experts from different fields gathered around a shared visual map with linked notes. Specialist Collab
🧑‍🔬 👥 🗺️ 🔗 💬 Cross-Specialist Collaboration - helping different experts share partial knowledge
clustercollective-sensemaking
scoresrank 75% | centrality 80% | emotive 80% | sentiment 88%
aboutCross-specialist collaboration is
the coordination challenge faced
by people with different
expertise, evidence, vocabulary,
imageExperts from different fields
gathered around a shared visual
map with linked notes.
title: 🦠 👥 🗺️ ⚠️ 🧑‍🔬 Pandemic and Biorisk Collaboration - coordinating specialists around shared threat maps cluster: collective-sensemaking scores: rank 50% | centrality 60% | emotive 80% | sentiment 61% about: Pandemic and biorisk collaboration requires clinicians, modellers, biosecurity specialists, policy experts, communicators, and institutions to share fast- changing knowledge. Visual threat maps can reveal risks, responsibilities, dependencies, uncertainty, and urgent coordination needs. image: Specialists around a bio-risk map with virus icons, documents, and connecting lines. Biorisk Collab
🦠 👥 🗺️ ⚠️ 🧑‍🔬 Pandemic and Biorisk Collaboration - coordinating specialists around shared threat maps
clustercollective-sensemaking
scoresrank 50% | centrality 60% | emotive 80% | sentiment 61%
aboutPandemic and biorisk collaboration
requires clinicians, modellers,
biosecurity specialists, policy
experts, communicators, and
imageSpecialists around a bio-risk map
with virus icons, documents, and
connecting lines.
title: 🧑‍🤝‍🧑 🧠 🗺️ 🔗 💡 Collective Sensemaking Tools - shared media for coordinating understanding cluster: collective-sensemaking scores: rank 100% | centrality 100% | emotive 90% | sentiment 94% about: Collective sensemaking tools help groups build, compare, revise, and act on shared understanding of complex domains. They connect individual mental models with deliberation, education, coordination, disagreement, memory, accountability, and cooperative action. image: A group building a shared map together, with paths linking many viewpoints. Collective Tools
🧑‍🤝‍🧑 🧠 🗺️ 🔗 💡 Collective Sensemaking Tools - shared media for coordinating understanding
clustercollective-sensemaking
scoresrank 100% | centrality 100% | emotive 90% | sentiment 94%
aboutCollective sensemaking tools help
groups build, compare, revise, and
act on shared understanding of
complex domains. They connect
imageA group building a shared map
together, with paths linking many
viewpoints.
title: 🧪 🎨 🗺️ 🤖 ✨ New Media Grammar - emerging conventions for generated, spatial, and interactive media cluster: collective-sensemaking scores: rank 75% | centrality 80% | emotive 80% | sentiment 92% about: New media grammar is the emerging design language of generated images, embeddings, zoomable canvases, dynamic ranking, interactive maps, multimodal symbols, and adaptive interfaces. It extends older diagram traditions into computational and personalised media. image: A laboratory of icons, maps, generated images, and interface layers forming a new alphabet. New Media Grammar
🧪 🎨 🗺️ 🤖 ✨ New Media Grammar - emerging conventions for generated, spatial, and interactive media
clustercollective-sensemaking
scoresrank 75% | centrality 80% | emotive 80% | sentiment 92%
aboutNew media grammar is the emerging
design language of generated
images, embeddings, zoomable
canvases, dynamic ranking,
imageA laboratory of icons, maps,
generated images, and interface
layers forming a new alphabet.
title: 📜 🗺️ 🏛️ 👤 🕯️ Samuel Pepys Life Map - biographical topics arranged as historical terrain cluster: case-study scores: rank 25% | centrality 60% | emotive 60% | sentiment 72% about: The Samuel Pepys life map is a biographical semantic map of people, places, events, institutions, possessions, illnesses, anxieties, and historical episodes connected with Samuel Pepys. It demonstrates how a life can be arranged as navigable conceptual terrain. image: A seventeenth-century diary page unfolding into a labelled city-like topic map. Pepys Map
📜 🗺️ 🏛️ 👤 🕯️ Samuel Pepys Life Map - biographical topics arranged as historical terrain
clustercase-study
scoresrank 25% | centrality 60% | emotive 60% | sentiment 72%
aboutThe Samuel Pepys life map is a
biographical semantic map of
people, places, events,
institutions, possessions,
imageA seventeenth-century diary page
unfolding into a labelled city-
like topic map.
title: 👤 ⭐ 📜 🧠 🗺️ Biographical Salience - deciding what matters inside a life map cluster: case-study scores: rank 50% | centrality 70% | emotive 80% | sentiment 79% about: Biographical salience is the judgement of what matters inside a life story or personal history. It weighs events, relationships, duties, illnesses, places, emotions, possessions, and historical forces to decide what should appear prominent in a life map. image: A life-map timeline with some personal landmarks glowing brighter than others. Bio Salience
👤 ⭐ 📜 🧠 🗺️ Biographical Salience - deciding what matters inside a life map
clustercase-study
scoresrank 50% | centrality 70% | emotive 80% | sentiment 79%
aboutBiographical salience is the
judgement of what matters inside a
life story or personal history. It
weighs events, relationships,
imageA life-map timeline with some
personal landmarks glowing
brighter than others.
title: 🪨 ⚕️ 😟 📜 🕯️ Pepys Medical Anxiety - bodily pain as a salient life-map concern cluster: case-study scores: rank 0% | centrality 40% | emotive 90% | sentiment 11% about: Pepys medical anxiety refers to Samuel Pepys’s bladder stone, surgery, cold- triggered pain, bodily fear, and recurring health preoccupation. It is an example of intimate suffering becoming a highly salient concern inside a historical biography. image: A dark stone beside old medical instruments and a diary page, anxious mood. Medical Anxiety
🪨 ⚕️ 😟 📜 🕯️ Pepys Medical Anxiety - bodily pain as a salient life-map concern
clustercase-study
scoresrank 0% | centrality 40% | emotive 90% | sentiment 11%
aboutPepys medical anxiety refers to
Samuel Pepys’s bladder stone,
surgery, cold-triggered pain,
bodily fear, and recurring health
imageA dark stone beside old medical
instruments and a diary page,
anxious mood.
title: ⚓ 📒 🏢 📜 👤 Pepys Work and Institutions - Navy duties as a semantic neighbourhood cluster: case-study scores: rank 0% | centrality 50% | emotive 40% | sentiment 68% about: Pepys work and institutions refers to Samuel Pepys’s naval administration, official duties, workplace geography, documents, rank, and institutional responsibility. It represents public role and bureaucratic systems within a broader biographical map. image: An anchor, ledgers, and dockside buildings arranged as a tidy topic cluster. Pepys Work
⚓ 📒 🏢 📜 👤 Pepys Work and Institutions - Navy duties as a semantic neighbourhood
clustercase-study
scoresrank 0% | centrality 50% | emotive 40% | sentiment 68%
aboutPepys work and institutions refers
to Samuel Pepys’s naval
administration, official duties,
workplace geography, documents,
imageAn anchor, ledgers, and dockside
buildings arranged as a tidy topic
cluster.
title: 🔥 🏙️ 📜 ⚠️ 🗺️ Historical Catastrophe Landmark - disaster as an emotionally charged map region cluster: case-study scores: rank 0% | centrality 50% | emotive 90% | sentiment 9% about: A historical catastrophe landmark is a disaster event that becomes a major emotional and spatial anchor in a historical or biographical map. It captures destruction, fear, displacement, civic memory, rebuilding, and the way catastrophe reorganises attention. image: A burning old city skyline with a cleared space around it on a historical map. Catastrophe
🔥 🏙️ 📜 ⚠️ 🗺️ Historical Catastrophe Landmark - disaster as an emotionally charged map region
clustercase-study
scoresrank 0% | centrality 50% | emotive 90% | sentiment 9%
aboutA historical catastrophe landmark
is a disaster event that becomes a
major emotional and spatial anchor
in a historical or biographical
imageA burning old city skyline with a
cleared space around it on a
historical map.
title: 🌱 🗺️ 🔜 🧪 🛠️ Prototype First Steps - early system work with a clear future direction cluster: case-study scores: rank 25% | centrality 60% | emotive 80% | sentiment 91% about: Prototype first steps describe an early but functional system that demonstrates a promising direction while leaving major design, evaluation, interaction, and product questions open. It signals credible progress, incomplete polish, and a path for iterative development. image: A small green shoot growing beside an early topic-map interface. First Steps
🌱 🗺️ 🔜 🧪 🛠️ Prototype First Steps - early system work with a clear future direction
clustercase-study
scoresrank 25% | centrality 60% | emotive 80% | sentiment 91%
aboutPrototype first steps describe an
early but functional system that
demonstrates a promising direction
while leaving major design,
imageA small green shoot growing beside
an early topic-map interface.

Transcript

Raw Projection Messiness

0:52 · chapter 2 · high-D → 2D

Video and live map are synced to the current topic.