**Author:** [[3. Resources/Authors/Colin Ware|Colin Ware]] # Visual Thinking for Information Design ![[Visual Thinking for Information Design Visual.excalidraw]] ## Book Overview ### About the Author Colin Ware is Director of the Data Visualization Research Lab at the University of New Hampshire. He has a background in both computer science and cognitive psychology, making him uniquely positioned to bridge the gap between perception science and practical design. He is also author of the more comprehensive textbook *Information Visualization: Perception for Design*, now in its fourth edition. This book, *Visual Thinking for Information Design*, is a more accessible introduction to the same core ideas. ### Central Thesis The brain is not a passive receiver of visual information but an active pattern-seeking engine. Effective visualization design requires understanding how this neural machinery works. Ware introduces **predictive cognition theory**—the idea that memories are not passive repositories of our past but *predictive mental models* that guide future actions. Graphics should be thought of as *cognitive tools* that enhance and extend our brains. Just as a hammer extends our physical capabilities, well-designed visualizations extend our cognitive capabilities by transforming data into forms our pattern-finding visual system can efficiently process. The key insight is that visual thinking consists of a series of **visual queries**—acts of attention that drive eye movements and tune our pattern-finding circuits. We only perceive a tiny amount of information at any instant (through the fovea), but can sample rapidly through eye movements. Effective design supports these visual queries by making important patterns easy to find and interpret. ### Why This Book Matters This book provides a science-based foundation for information design decisions. Rather than relying on aesthetic intuition alone, designers can understand *why* certain visual choices work better than others based on how the brain actually processes visual information. The book is practical—filled with concrete design guidelines derived from perceptual research—while remaining grounded in cognitive neuroscience. **Readwise Highlights**: No Readwise highlights available for this book. --- ## Chapter Summaries ### Preface Ware introduces **predictive cognition theory**: memories serve as predictive mental models guiding future actions rather than being passive repositories of the past. He argues we should think about graphic designs as *cognitive tools* that enhance and extend our brains. Graphics allow us to offload cognitive work—like using a subway map to plan a route rather than memorizing the entire transit system. ### Chapter 1: Visual Queries The central concept is that visual thinking consists of a series of **visual queries**—acts of attention driving eye movements and tuning our pattern-finding circuits. The brain doesn't capture everything at once; the fovea (high-resolution center of vision) is only about thumbnail-sized at arm's length. We sample the visual world through rapid eye movements. Key concepts: - **The illusion of seeing**: We feel we see a rich visual world, but actually perceive very little at any instant. Change blindness experiments show people fail to notice major changes (50% didn't notice when a person they were talking to was swapped for someone else mid-conversation) - **Three-stage processing model**: Features → Patterns → Objects, processed through V1/V2 cortex up to inferotemporal cortex - **Nested loops model**: Outer loop (problem-solving steps), middle loop (visual search via eye movements), inner loop (pattern testing at each fixation) - **Distributed cognition**: Cognition extends outside the skull—the world itself serves as external memory ### Chapter 2: What We Can Easily See This chapter explores which visual features "pop out" and can be found instantly versus those requiring slow serial search. Key concepts: - **Feature channels**: Different visual properties (color, orientation, size, motion) are processed by separate neural channels in V1 cortex - **Pop-out effects**: Features processed preattentively can be found instantly regardless of how many distractors are present - **Biased competition**: The brain can tune attention to enhance specific feature channels - **Conjunctive searches are hard**: Finding a green square among green circles and red squares requires serial search because the target isn't unique in any single feature channel - **Motion is special**: Moving objects trigger an orienting response (evolutionary adaptation for predator detection) Design principle: Use different feature channels to code different types of searchable information. ### Chapter 3: Structuring Two-Dimensional Space Pattern processing in the brain is hierarchical—simple features combine into patterns, which combine into objects. Understanding this hierarchy is key to effective spatial layout. Key concepts: - **2.5D space**: Visual space has essentially 2.05 dimensions—the image plane (up-down, sideways) contains far more information than depth (towards-away) - **Binding**: How the brain combines fragmented features into continuous contours and coherent patterns - **Generalized contours**: The brain can extract boundaries from many types of discontinuity (luminance, color, texture, motion) - **Texture discrimination**: Easy when textures differ in basic V1 properties; hard when differences are at higher levels - **Spatial metaphors**: George Lakoff's insight that spatial relationships are fundamental to reasoning—terms like "connected to," "inside," and "part of" are foundational to both language and visual design Design principle: Relationships between entities can be established through connecting contours, proximity, alignment, enclosure, color, texture, and common movement. ### Chapter 4: Color Color perception begins with three cone types but is transformed in V1 into three opponent channels: red-green, yellow-blue, and black-white (luminance). Key concepts: - **Opponent process theory**: Neural mechanisms compute differences between cone signals, creating three perceptual channels - **Luminance channel superiority**: The black-white channel conveys far more spatial detail than chromatic channels; motion, stereoscopic depth, and shape-from-shading all rely on luminance - **Simultaneous contrast**: Colors are distorted by surrounding colors - **Unique hues**: Red, green, yellow, blue, black, and white are special—each produces an extreme signal on one channel while being neutral on others Design principles: - Use luminance contrast for detailed information (text, fine lines) - Use high-saturation colors for small areas, low-saturation for large areas - Limit color codes to 6-12 reliably distinguishable colors - Red-green color blindness affects 8% of males ### Chapter 5: Getting the Information: Visual Space and Time This chapter explores depth cues and the costs of navigating information spaces. Key concepts: - **Depth cues**: Occlusion (strongest), linear perspective, texture gradients, shading, cast shadows, depth of focus, atmospheric contrast - **Stereoscopic depth**: Best for guiding hand movements to nearby objects; 20% of people lack stereoscopic vision - **Structure from motion**: More powerful than stereopsis—rotation reveals 3D structure dramatically - **2.5D design principles**: Layout in the image plane with judicious use of depth; minimize occlusion; show joints clearly; keep text in the image plane Key insight: The cost of accessing information varies enormously—eye movements (0.1 second), mouse clicks (1.5 seconds), walking, driving. Design should minimize cognitive costs for common queries. ### Chapter 6: Visual Objects, Words, and Meaning Visual working memory is severely limited—we can hold only about three meaningful "chunks" at once. Key concepts: - **Nexuses of meaning**: Temporary bindings of visual, verbal, and motor information that form during each eye fixation - **The three-object limit**: A fundamental bottleneck in visual thinking - **Scene gist**: We can extract the general meaning of a scene in a single glance (100ms), though not details - **Concept proxies**: Visual symbols can serve as external memory, allowing faster concept activation than retrieval from internal memory ### Chapter 7: Visual and Verbal Narrative This chapter distinguishes visual and language-based thinking, using sign language as a key example. Key concepts: - **Principle of arbitrariness**: Language symbols (spoken, written, or signed) are arbitrary—nothing about "dog" is dog-like - **Visual logic vs. verbal logic**: Language excels at conditional logic ("if," "and," "but"); visual media excel at structural relationships ("connected to," "inside," "part of") - **Deixis**: Pointing gestures link visual and verbal information—fundamental to effective presentations - **Mirror neurons**: Support learning by imitation and reading others' emotions - **The cognitive thread**: The sequence of concepts held in working memory; good design captures and guides this thread Design principles for presentations: Slides should be primarily images and diagrams; narrative should come from the speaker; deictic gestures link the two modalities. ### Chapter 8: Building Mental Models: Why We Present with Visualizations The purpose of presentations is to build or enhance **predictive mental models** in the audience's minds. Key concepts: - **Predictive cognition**: The brain's fundamental purpose is predicting the future; we predict results of eye movements, hand movements, driving behaviors - **Episodic memory as simulation**: Remembering the past and imagining the future are the same process—running simulations using stored mental models - **Why we remember**: According to Mahr and Csibra, episodic memory evolved specifically to construct stories and communicate our mental models to others - **Agent-based thinking**: People understand stories better when framed in terms of animate agents causing effects Design implications: - Know your audience's existing mental models and how they can be extended - Use conventional visualizations (bar charts, node-link diagrams) unless innovation is central to the message - Photographs enrich mental models incomparably - Simple cartoons can convey mechanical models effectively - Assembly diagrams support both narrative and self-paced model construction ### Chapter 9: Creative Meta-seeing This chapter covers the design process itself (read in a previous session). Key concepts: - **Constructive seeing**: Designers mentally add patterns to sketches to test possible changes - **Critical seeing**: Analyzing which combinations of patterns best support cognitive tasks - **The sketch-critique cycle**: Externalize ideas, critique them, refine ### Chapter 10: The Dance of Meaning The concluding chapter synthesizes all concepts into a unified view of visual thinking as a dance between internal cognition and external visual displays. **Twelve-point summary**: 1. The fovea provides high-resolution vision; we sample through rapid eye movements 2. Visual queries drive attention and eye movements 3. Early feature analysis (orientation, color, motion, size) happens simultaneously across the visual field 4. Two pathways: "what" (object identification) and "where" (action guidance) 5. Temporary nexuses of meaning are constructed 1-3 times per second 6. Visual displays support cognition by providing concept proxies 7. Pattern finding—seeing connections—often solves cognitive problems 8. Language processing uses specialized brain areas with different logic (ifs, ands, buts) 9. Attention is multifaceted: early feature biasing, eye movement control, working memory binding 10. Constructive seeing allows designers to mentally test additions to sketches 11. Long-term memories are executable predictive models, not static storage 12. Effective presentations transmit predictive mental models from presenter to audience **Four design implications**: 1. Support the brain's pattern-finding capability—transform information into identifiable patterns 2. Optimize the cognitive process as nested loops 3. Consider cognitive economics—costs of learning new tools vs. benefits 4. Design for the audience's existing mental models --- ## Key Themes and Highlights ### The Brain as Prediction Engine > The fundamental purpose of the brain is predicting the future. Our brains constantly make predictions—about eye movement results, hand movements, driving situations. When predictions fail, exceptions trigger higher-level attention. This is extremely efficient: most of the time things go as planned and lower-level actions run automatically. Memory itself serves prediction, not reminiscence. We remember poorly because recall is secondary to the primary function of acting on predictive models. ### Visual Queries and the Nested Loops Model Visual thinking operates through nested loops: - **Outer loop**: High-level problem-solving steps - **Middle loop**: Visual search via eye movements - **Inner loop**: Pattern testing at each fixation point Design should support efficient visual queries at every level. ### The Power and Limits of Working Memory We can hold only ~3 meaningful chunks in visual working memory simultaneously. This severe bottleneck shapes all visual cognition. External displays serve as concept proxies, dramatically expanding our effective working capacity—but only for information we've learned to associate with those proxies. ### 2.5D Design Perception is fundamentally 2.5-dimensional. The image plane (up-down and sideways) contains vastly more information than depth. Design principles: - Lay out information primarily in the image plane - Use depth cues selectively for specific goals - Minimize occlusion - Keep text in the image plane (perspective-distorted text is hard to read) ### Color as Three Opponent Channels Color vision creates three channels: red-green, yellow-blue, and black-white (luminance). The luminance channel is far more capable—it alone conveys fine detail, motion, depth from shading, and stereoscopic information. **Critical principle**: Detailed information requires luminance contrast. Text must have strong light-dark contrast regardless of hue. ### Distributed Cognition Cognition extends beyond the skull. The world itself is its own memory—we sample what we need through visual queries rather than storing everything internally. Graphics serve as cognitive tools extending our capabilities. --- ## Analytical Reading ### Main Arguments 1. Visual thinking is an active process of querying the environment, not passive reception 2. Understanding neural processing explains *why* design principles work 3. Predictive cognition provides a unified theory of memory, action, and communication 4. Effective design supports the brain's pattern-finding machinery 5. The goal of presentations is transmitting executable mental models 6. Visual and verbal processing are complementary with distinct strengths ### Methodology Ware bridges cognitive neuroscience and practical design by: - Explaining neural mechanisms (V1 feature detection, opponent color channels, what/where pathways) - Deriving concrete design principles from those mechanisms - Providing visual examples demonstrating each principle - Connecting low-level perception to high-level cognition (mental models, narrative) ### Strengths - Grounded in solid perceptual science while remaining accessible - Practical design guidelines backed by theory - Excellent visual illustrations - Useful framework (visual queries, nested loops) for analyzing any visualization - Integration of cognitive science with design practice ### Limitations - Some neuroscience is necessarily simplified - Design examples skew toward scientific visualization - Less coverage of aesthetic/emotional aspects of design - Written primarily for designers who want scientific grounding; may be too technical for some audiences --- ## Personal Reflection ### Key Takeaways The "visual queries" framework is immediately useful—when designing or evaluating any visualization, ask: What queries will users make? What patterns will they search for? How easily can those patterns be found? The predictive cognition theory reframes presentations entirely. The goal isn't to present information but to build or modify executable mental models that the audience can use to take effective action. The 2.5D design concept clarifies when 3D is appropriate (rarely) and why flat design often works better (the brain processes image-plane patterns far more efficiently than depth). ### Questions Raised - How do these principles apply to interactive and animated visualizations specifically? - What's the relationship between aesthetic pleasure and cognitive efficiency in design? - How do individual differences in visual working memory capacity affect design choices? ### Connections to Other Works - Edward Tufte's *The Visual Display of Quantitative Information* (complementary focus on data integrity) - Don Norman's *The Design of Everyday Things* (affordances and cognitive engineering) - Daniel Kahneman's *Thinking, Fast and Slow* (System 1/2 maps loosely to feature vs. pattern processing) - Andy Clark's work on predictive processing and extended mind - George Lakoff's *Metaphors We Live By* (spatial metaphors underlying cognition) --- ## Reading Notes Second edition of a well-regarded book on visualization design. Colin Ware writes from the intersection of computer science and perceptual psychology. The book is more accessible than his comprehensive *Information Visualization: Perception for Design* while covering the core ideas. 212 pages, published by Morgan Kaufmann (Elsevier), 2022.