Chapter 28 Exercises

How to use these exercises: Work through the parts in order. Part A builds recognition skills, Part B develops analysis, Part C applies concepts to your own domain, Part D requires synthesis across multiple ideas, Part E stretches into advanced territory, and Part M provides interleaved practice that mixes skills from all levels.

For self-study, aim to complete at least Parts A and B. For a course, your instructor will assign specific sections. For the Deep Dive path, do everything.

Part A: Pattern Recognition

These exercises develop the fundamental skill of recognizing dark knowledge across domains.

A1. For each of the following scenarios, (i) identify the dark knowledge at play, (ii) explain why it has never been codified, and (iii) assess what would happen if the experienced practitioners suddenly departed.

a) A family bakery that has been making the same sourdough bread for four generations. The recipe is written down. The bread made by the fourth-generation baker is noticeably better than what any skilled baker can produce from the same recipe.

b) A police precinct where experienced officers know which street corners are dangerous at which times, which informants are reliable, and which "routine" calls are actually high-risk.

c) A small law firm where senior partners know which judges prefer which argument styles, which clerks to befriend, and which procedural shortcuts save months of litigation.

d) A rural farming community where elders know the microclimates of every field, the drainage patterns that maps do not show, and the planting schedules that vary by a week from published almanacs.

e) A theatre company where veteran stage managers know the quirks of the lighting system, the timing of the HVAC cycles that affect acoustics, and the placement of props that allows smooth scene transitions.

f) A nuclear power plant where experienced operators know the meaning of sounds and vibrations that no instrument measures, and can diagnose developing problems hours before alarm systems trigger.

g) An experienced kindergarten teacher who can identify which students will struggle with reading eighteen months before any assessment detects the risk.

h) A financial trading floor where experienced traders can sense a shift in market sentiment before it shows up in any data feed.

A2. Classify each of the following as (i) dark knowledge that is hard to articulate (Polanyi's Paradox), (ii) dark knowledge that nobody asks about, (iii) dark knowledge that is politically inconvenient to share, or (iv) dark knowledge that is obvious to insiders. Some may fit more than one category.

a) The collective understanding in a university department about which senior faculty member's research claims are exaggerated.

b) A software team's knowledge of which parts of the codebase are fragile and should not be modified.

c) An airline crew's knowledge of which airports have unreliable de-icing equipment in winter.

d) A chef's collective knowledge of which suppliers deliver inconsistent quality and must be checked.

e) The shared understanding among nurses in a hospital unit about which attending physicians are dangerously overconfident.

f) An experienced pilot's ability to feel when an aircraft is approaching a stall through the controls.

g) A team of firefighters' collective knowledge of how fire behaves differently in buildings with specific construction methods common in their district.

h) The understanding among congressional staffers about which lobbyists' factual claims can be trusted and which cannot.

A3. For each of the following knowledge domains, estimate the ratio of explicit-to-dark knowledge (e.g., 10:90, 30:70, 50:50). Justify your estimate.

a) Commercial aviation (flying a passenger jet)

b) Fine dining (running a Michelin-starred kitchen)

c) Software engineering (building and maintaining a large-scale system)

d) Elementary school teaching (managing and educating a class of 25 seven-year-olds)

e) Diplomacy (negotiating a trade agreement between two nations)

f) Emergency medicine (running an emergency department)

A4. Identify three modern "guild secrets" -- areas where practitioners deliberately maintain dark knowledge as a competitive advantage. For each, assess whether the darkness serves a legitimate purpose or whether it harms the broader community.

A5. The chapter describes the "automation paradox" -- the phenomenon where automating a job reveals (through failures) the dark knowledge that the human worker possessed. Identify three examples of the automation paradox from your own experience or knowledge. For each, describe what dark knowledge was lost and how its absence was discovered.

Part B: Analysis

These exercises require deeper analysis of dark knowledge patterns.

B1. Dark Knowledge Audit. Choose an organization you know well (your workplace, your school, a community organization). Conduct a dark knowledge audit:

a) Identify at least five pieces of dark knowledge that the organization relies on but has never documented.

b) For each piece, classify it using the four reasons dark knowledge stays dark (hard to articulate, nobody asks, politically inconvenient, obvious to insiders).

c) For each piece, assess the risk of loss. How many people currently hold this knowledge? What would trigger its loss (retirements, reorganization, turnover)?

d) For each piece, assess the consequences of loss. What would go wrong if this knowledge disappeared? How long would it take to rediscover it?

e) Propose a preservation strategy for the three most critical pieces of dark knowledge you identified. Be realistic about what can and cannot be made explicit.

B2. Organizational Amnesia Case Analysis. Research a specific case of organizational amnesia -- an instance where an organization lost critical dark knowledge and suffered consequences. Good candidates include: NASA's loss of Apollo-era manufacturing knowledge, the loss of institutional memory during corporate mergers, the degradation of military capabilities following rapid drawdowns, or the loss of manufacturing expertise when plants are closed and reopened.

a) Describe the dark knowledge that was lost.

b) Identify the mechanism of loss (retirement, layoff, reorganization, automation, other).

c) Document the consequences of the loss.

d) Assess whether the loss could have been prevented, and at what cost.

e) Evaluate whether the organization has since recovered the lost knowledge, and if so, how.

B3. Extraction Method Evaluation. The chapter describes five methods for extracting dark knowledge: ethnography, apprenticeship, debriefing, storytelling, and knowledge engineering.

a) For each method, identify a specific type of dark knowledge it is best suited to capture. Give a concrete example.

b) For each method, identify a specific type of dark knowledge it cannot capture. Explain why.

c) Propose a sixth method not discussed in the chapter. Describe how it works, what types of dark knowledge it can capture, and what its limitations are.

d) Design a "dark knowledge preservation program" for a specific organization, combining multiple extraction methods. Justify your choice of methods and explain how they complement each other.

B4. The Automation Paradox -- Extended Analysis. The chapter introduces the automation paradox: automating the explicit portion of a job reveals the importance of the dark portion.

a) Identify a job that has been partially automated. Describe the explicit tasks that were automated and the dark knowledge that was not.

b) Assess the consequences. Has the automation produced the expected improvements? Has it created unexpected failures? Do the failures match the automation paradox prediction?

c) The automation paradox suggests that dark-knowledge-heavy jobs are more difficult to automate well. Rank the following jobs from most to least dark-knowledge-dependent, and justify your ranking: airline pilot, tax preparer, emergency room nurse, factory assembly worker, restaurant server, software architect.

d) Propose a framework for assessing a job's "dark knowledge quotient" -- the proportion of the job that relies on dark knowledge. What factors would you measure? How would you measure them?

B5. The Oral Tradition Paradox. The chapter argues that literacy destroyed dark knowledge by replacing embodied, contextual, integrated oral traditions with abstracted, decontextualized written texts.

a) Is this argument too strong? Identify specific types of knowledge that literacy preserves better than oral tradition.

b) Is there a modern equivalent of the oral-to-literate transition? Is the shift from human expertise to AI-based knowledge systems creating a similar dynamic? What dark knowledge might be lost?

c) Identify a modern community that maintains important knowledge primarily through oral tradition (examples: specialized trades, ethnic communities, religious orders, military units). Assess the vulnerability of this oral tradition to disruption.

Part C: Application to Your Own Domain

These exercises connect dark knowledge to your area of expertise.

C1. Identify the three most important pieces of dark knowledge in your professional or academic field -- knowledge that practitioners widely share and rely on but that has never been formally documented.

a) For each piece, describe the knowledge and explain how you know it exists (if it has never been documented, how did you learn about it?).

b) Classify each using the four reasons dark knowledge stays dark.

c) Assess the vulnerability of each piece to loss. What would cause it to be lost? How would you know it was gone?

d) For each piece, propose a method for partial extraction. What could be made explicit? What would inevitably remain dark?

C2. Describe a situation in your field where dark knowledge loss had concrete consequences -- where an experienced cohort departed and the resulting gap in dark knowledge produced measurable degradation in performance, quality, or outcomes.

a) What dark knowledge was lost?

b) How was the loss discovered?

c) What was the cost of the loss (in time, money, quality, safety, or lives)?

d) Could the loss have been prevented? If so, at what cost?

C3. Design a "dark knowledge preservation program" for your organization or field. Your program should:

a) Identify the most critical dark knowledge at risk of loss.

b) Assess the feasibility of extraction (what can be made explicit and what cannot).

c) Specify preservation methods (apprenticeship, debriefing, storytelling, documentation, or others).

d) Estimate the cost of the program and compare it to the estimated cost of knowledge loss.

e) Address the political obstacles -- the reasons some dark knowledge is kept dark deliberately and how your program would navigate those dynamics.

C4. Assess the "dark knowledge quotient" of your own job. What proportion of what you do relies on dark knowledge -- knowledge that you possess but could not fully document in a manual? Give three specific examples of dark knowledge you use regularly and explain why each resists codification.

Part D: Synthesis

These exercises require integrating ideas across multiple chapters.

D1. Dark Knowledge and Boundary Objects. Chapter 27 argued that boundary objects enable cooperation by creating shared surfaces for coordination. Chapter 28 argues that dark knowledge fills the gaps that boundary objects leave open.

a) Identify a boundary object in your field. What dark knowledge do practitioners bring to their use of this boundary object? What would happen if the boundary object were used by someone who lacked this dark knowledge?

b) Can a boundary object transmit dark knowledge? Or do boundary objects, by their nature, work only with explicit knowledge? Defend your answer with examples.

c) When dark knowledge is lost, do boundary objects become less effective? Explain the mechanism by which dark knowledge loss degrades the functioning of boundary objects.

d) Propose a boundary object specifically designed to preserve dark knowledge across generations of practitioners. What would it look like? How would it work? Is this a contradiction in terms?

D2. Dark Knowledge and Paradigm Shifts. Chapter 24 described how paradigm shifts transform what a field can see and know. Chapter 28 describes dark knowledge as what a field knows but never writes down.

a) When a paradigm shift occurs, what happens to the dark knowledge accumulated under the old paradigm? Is it lost, transformed, or carried forward?

b) Is dark knowledge paradigm-dependent? Does the dark knowledge of pre-germ-theory physicians (their intuitions about which treatments worked and which did not) have value within the germ theory paradigm?

c) Can dark knowledge cause a paradigm shift? If a community's dark knowledge consistently contradicts the explicit paradigm -- if practitioners know, through dark channels, that the official theory is wrong -- can this dark knowledge eventually force a paradigm revolution?

d) Kuhn argued that practitioners in different paradigms "live in different worlds." Does this mean they possess different dark knowledge? If two communities with different paradigms both treat the same patient population, would their dark knowledge converge (because they are responding to the same clinical realities) or diverge (because they are interpreting those realities through different frameworks)?

D3. Dark Knowledge and the Map-Territory Distinction. Chapter 22 argued that maps are useful simplifications of reality, and that confusing the map with the territory is a universal failure mode.

a) The chapter argues that explicit knowledge is the map and dark knowledge is the territory. Push this metaphor: in what ways does dark knowledge resemble "the territory" more closely than explicit knowledge? In what ways does the metaphor break down?

b) Knowledge management systems (databases, wikis, documented procedures) are maps of organizational knowledge. Assess a knowledge management system you have used: what territory does it miss? What dark knowledge does it fail to capture?

c) Can the map-territory confusion cause dark knowledge loss? If an organization believes that its documented procedures capture everything important (confusing the map with the territory), how does this belief contribute to the loss of dark knowledge?

d) Is there a "dark map" -- a representation of dark knowledge that is itself dark? Can communities develop shared, unwritten understandings of their own dark knowledge? Give examples.

D4. Dark Knowledge and the Adjacent Possible. Chapter 25 argued that innovation is constrained exploration of an expanding space of possibilities.

a) Does dark knowledge expand the adjacent possible? If a community possesses dark knowledge that is unavailable to others, does this give it access to innovations that others cannot reach?

b) Can dark knowledge loss shrink the adjacent possible? If a community loses dark knowledge (through organizational amnesia, cultural disruption, or generational turnover), does this close doors that were previously open?

c) The adjacent possible explains why some innovations appear "ahead of their time" -- they require preconditions that do not yet exist. Can dark knowledge serve as a precondition? Can a discovery be impossible because the required dark knowledge does not yet exist in the community?

d) Multiple discovery (Ch. 26) occurs because the adjacent possible creates the same opportunities for independent discoverers. But if a critical precondition is dark knowledge rather than explicit knowledge, would multiple discovery be less likely? Explain.

Part E: Advanced Challenges

These exercises push beyond the chapter's material into deeper or more speculative territory.

E1. Research Julian Orr's Talking About Machines: An Ethnography of a Modern Job (1996), which studied Xerox photocopier repair technicians. Assess how Orr's findings about the role of storytelling in maintaining technical knowledge relate to the chapter's concept of dark knowledge. What does Orr's work reveal about the limitations of documented procedures as a knowledge storage mechanism?

E2. Research the concept of "organizational memory" as developed by James P. Walsh and Gerardo Rivera Ungson ("Organizational Memory," Academy of Management Review, 16(1), 1991). Compare and contrast their framework for organizational memory with this chapter's concept of dark knowledge. Are they describing the same phenomenon? What does dark knowledge capture that organizational memory does not, and vice versa?

E3. Research the history of the Stradivarius violin -- the instruments made by Antonio Stradivari in Cremona, Italy, in the seventeenth and eighteenth centuries. Modern luthiers, despite having access to advanced materials science, acoustic analysis, and CT scanning of original instruments, have been unable to consistently replicate the tonal qualities of Stradivarius violins. Analyze this case using the dark knowledge framework. What dark knowledge did Stradivari possess? Why has it resisted reconstruction? Is this a case of genuinely inarticulate knowledge, deliberately hidden knowledge, or something else?

E4. The chapter describes the "automation paradox" -- automating the explicit portion of a job reveals the importance of the dark portion. Research the implementation of a specific AI/automation system (e.g., automated radiology reading, automated trading systems, autonomous vehicles, or AI-assisted medical diagnosis). Assess: did the automation paradox manifest? What dark knowledge did the automated system fail to capture? What consequences resulted?

E5. Design a research study to test the hypothesis that organizations with strong dark knowledge preservation practices (mentorship programs, storytelling cultures, low turnover, gradual succession planning) outperform organizations that rely primarily on documented procedures.

a) Define "dark knowledge preservation practices" and "organizational performance" in measurable terms.

b) Identify a population of organizations to study and a method for assessing the quality of their dark knowledge preservation practices.

c) Specify your dependent variable (performance) and your method for measuring it.

d) Identify confounding variables and explain how you would control for them.

e) Predict what you would expect to find if the hypothesis is correct, and what evidence would falsify it.

Part M: Mixed Practice (Interleaved Review)

These exercises mix concepts from Chapters 22-28 to build integrated understanding across all of Part IV.

M1. A manufacturing company acquires a smaller competitor and shuts down the competitor's factory, planning to produce the competitor's products at its own facilities using the competitor's documented processes. Analyze this situation using concepts from Chapters 22-28:

a) Dark knowledge (Ch. 28): What dark knowledge is at risk of being lost in the factory closure? How would you identify it before the closure occurs?

b) Tacit knowledge (Ch. 23): How does the individual tacit knowledge of the acquired company's workers differ from the collective dark knowledge of the acquired company as an organization? Which is easier to preserve?

c) Map-territory (Ch. 22): The acquiring company believes the documented processes (the map) are sufficient to replicate the competitor's production quality. Is this a map-territory confusion? What territory might the map be missing?

d) Boundary objects (Ch. 27): The documented processes served as boundary objects between different teams in the original factory. Will these boundary objects function the same way in the new factory, with different teams? What dark knowledge did each original team bring to their use of the boundary objects?

e) Paradigm shifts (Ch. 24): If the two companies used different manufacturing paradigms (different approaches to quality control, different production philosophies), will the acquiring company's paradigm blind them to aspects of the competitor's dark knowledge that only make sense within the competitor's paradigm?

M2. An indigenous community partners with a university to document its traditional ecological knowledge before the last elders who possess it pass away. Analyze using Chapters 22-28:

a) What forms of dark knowledge does the community possess? Classify them using the chapter's typology (hard to articulate, nobody asks, politically inconvenient, obvious to insiders).

b) Which documentation methods (ethnography, apprenticeship, debriefing, storytelling, knowledge engineering) are most appropriate for different types of the community's dark knowledge?

c) The written documentation produced by the university researchers will be a map of the community's knowledge territory. What will it miss? How can the map-territory gap be minimized?

d) The community's knowledge was maintained through oral traditions that integrated knowledge with practice, context, and embodiment. What is lost in the transition to written documentation? Is there a way to preserve the integrated, embodied quality of the knowledge?

e) How might the university researchers' paradigm (Western scientific frameworks) distort their understanding of the community's knowledge? What dark knowledge might they fail to recognize because it does not fit their paradigm?

M3. A veteran software architect announces her retirement in six months. She has worked on the company's core platform for fifteen years. Analyze using Chapters 22-28:

a) What dark knowledge does she likely possess about the system? Categorize it (debugging instincts, system folklore, deployment lore, architectural intuition, team dynamics).

b) Design a six-month knowledge transfer program. Which extraction methods would you use for each category of knowledge? What trade-offs are involved?

c) The company's documentation (the map) captures the system's explicit architecture. What territory does she know that the documentation misses?

d) Her knowledge functions as a boundary object between different teams that interact with the platform. When she leaves, what gaps in cross-team coordination will emerge?

e) Is it possible that some of her knowledge is paradigm-dependent -- that it makes sense within the current architectural paradigm but might not apply if the system is redesigned? How should this affect the knowledge transfer strategy?

M4. A hospital system implements an AI-assisted diagnostic tool that draws on the medical literature to suggest diagnoses and treatments. Analyze using Chapters 22-28:

a) The AI was trained on the medical literature -- the explicit knowledge of medicine. What dark knowledge is not in its training data? How might this gap affect its performance?

b) The AI's suggestions are a map of the clinical territory. When is confusing the AI's suggestion with the clinical reality (map-territory confusion) most dangerous?

c) If clinicians begin to rely on the AI for diagnoses they previously made using clinical intuition, will the clinical community's dark knowledge atrophy? What would be the consequences?

d) The AI functions as a boundary object between different clinical teams. How does each team's dark knowledge affect how they interpret and use the AI's suggestions?

e) The AI captures the current medical paradigm's explicit knowledge. What happens when a paradigm shift occurs in medicine -- when the guidelines change? Will the AI's "knowledge" become an obstacle to change, in the same way that the old guard's commitment to the old paradigm slows paradigm shifts?

M5. Design a "Part IV Knowledge Assessment" for an organization. The assessment should evaluate the organization's knowledge health across all seven epistemological dimensions:

a) Map-territory (Ch. 22): Does the organization confuse its models, metrics, and documentation with the reality they represent?

b) Tacit knowledge (Ch. 23): Does the organization recognize and support the tacit knowledge of its individual experts?

c) Paradigm awareness (Ch. 24): Is the organization aware of the paradigm within which it operates, including the paradigm's blind spots?

d) Adjacent possible (Ch. 25): Is the organization positioned to exploit emerging possibilities, and does it have the dark knowledge needed to do so?

e) Multiple discovery readiness (Ch. 26): Is the organization monitoring its adjacent possible for convergent developments that might make its current approaches obsolete?

f) Boundary objects (Ch. 27): Are the organization's boundary objects well-designed, and do practitioners bring sufficient dark knowledge to their use?

g) Dark knowledge (Ch. 28): Does the organization have mechanisms for identifying, preserving, and transmitting its dark knowledge? What is its organizational amnesia risk?