Glossary of Terms

This glossary defines the 528 concepts in this book's learning graph. Each entry links to the chapter where the concept is taught.

Accidental Adversaries Archetype

A systems archetype in which two or more parties with shared goals unintentionally work against each other while trying to solve a common problem, each taking actions that seem sensible but harm the other.

The result is a cycle of growing conflict despite their underlying alignment, a common source of organizational silos.

Example: An IT team locks down laptops for security while a sales team installs its own tools to move faster, and each side's fix makes the other's job harder.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Accidental Adversaries examples

Accidental Competitor

A situation in which two units of the same organization unknowingly work against each other's goals or duplicate each other's efforts, because neither can see what the other is doing.

Looking for these is one of the most reliable ways to find hidden silos.

Example: Two departments each pay outside vendors to build nearly identical customer dashboards in the same year.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Accountability (AI)

The assignment of clear responsibility to specific people or organizations for an AI system's outcomes, including the duty to answer for any harms and to correct them.

See also: AI Governance, Human Oversight

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Accumulation

The process by which a stock builds up or runs down as the difference between its inflows and outflows adds up over time.

See also: Stock, Rate Of Change

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Actuator (Control)

The component of a control system that takes physical action to change a variable, based on the gap between the measured state and the target value.

See also: Sensor (Control), Control System

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Adaptation

The process by which a system changes its structure or behavior in response to changes in its environment in a way that improves its ability to survive or function.

Covered in: Chapter 7: Feedback Resilience and Robustness

Adaptive Capacity

The ability of a system to adjust its structure, behavior, or resources in response to changing or unexpected conditions, and to learn from those adjustments.

See also: Resilience, Adaptation

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Adaptive Management

An approach to running complex systems, such as parks or public programs, that treats each decision as an experiment, monitors the results closely, and adjusts later actions based on what is found.

Example: Park managers try a small controlled burn, count the wildlife that returns, and change the size and timing of the next burn accordingly.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Addiction Cycle

A reinforcing pattern in which reliance on a quick fix weakens the ability to solve the underlying problem, which increases the need for the quick fix.

Example: A company that keeps paying consultants to fix its data problems never trains its own staff, so it needs consultants even more each year.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Adjacency

The property of two vertices being directly joined by an edge, which in a native graph store is recorded as a direct physical reference from each vertex to its neighbors.

See also: Pointer Hopping

Covered in: Chapter 16: Graph Database Architecture

Agent-Based Modeling

A simulation method that represents a system as many individual decision-makers, each following simple rules, and observes the large-scale patterns that result from their interactions.

Example: A simulation of herders sharing a pasture, each deciding whether to add another cow, shows how overgrazing arises without anyone intending it.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Aging Chain

A sequence of connected stocks in which items move from one stage to the next as they age, mature, or advance, with each stage having its own inflows and outflows.

Example: A workforce model moves people from new hires to experienced staff to senior staff to retirees.

See also: Stock, Conveyor Model

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

AI Alignment

The effort to make an artificial intelligence system's goals and behavior match the real intentions and values of the people who build and use it, rather than only the literal target it was given.

Example: A video app told only to maximize watch time may learn to recommend extreme content, which satisfies the literal target but not the designers' real intent.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

AI Dependence

A pattern in which a person or team routes a task entirely through an AI tool without maintaining their own skill at it, so their capability weakens over time.

Example: A student who has an AI write every essay loses the ability to organize an argument and cannot spot when the AI's reasoning is wrong.

Contrast with: Human-AI Collaboration

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

AI Flywheel

A reinforcing loop in which machine learning models make predictions from data, gather feedback on how those predictions turned out, and use that feedback as additional data to make better predictions.

It lets organizations that already hold large amounts of data gain even more, which makes it hard for new competitors to catch up.

Example: A navigation app uses each driver's trip to improve its traffic predictions, which attracts more drivers, whose trips improve the predictions further.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling · AI Flywheel archetype

AI Governance

The policies, processes, roles, and oversight structures an organization or government uses to guide how artificial intelligence systems are developed, deployed, monitored, and held accountable.

Example: A hospital requires every AI diagnostic tool to pass a bias review and assigns a named person to monitor its accuracy each quarter.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

AI Hallucination

An output from a generative AI model that is presented confidently but is false, invented, or unsupported by its sources or training data.

Example: A chatbot lists a research paper, complete with authors and a journal name, that does not exist.

See also: Large Language Model, Human Oversight

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Algorithm

A finite, ordered set of well-defined steps for solving a problem or completing a task, which a computer or a person can follow exactly.

Example: A recipe for baking bread, or the steps a map app follows to find the fastest route.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Algorithmic Bias

A systematic unfairness in the outputs of a computer program, usually learned from historical data, that favors or disadvantages certain groups of people.

Example: A résumé-screening tool trained on past hires, who were mostly men, learns to rank women's résumés lower.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Algorithmic Improvement

The process by which a computer program's outputs become measurably better over time as the program is refined using new data and feedback.

See also: Data Advantage, AI Flywheel

Covered in: Chapter 23: AI Systems Dynamics

Amplification

The strengthening of a signal, change, or effect as it passes through a system, usually by a reinforcing loop, so that the result is larger than the original cause.

Example: A small rumor about a bank's health can grow into a full bank run as worried customers see others withdrawing money.

Contrast with: Dampening

Covered in: Chapter 7: Feedback Resilience and Robustness

Analytical Maturity Level

The third stage of this book's systems-thinking maturity scale, at which an organization maps key relationships and feedback loops before major decisions and routinely separates symptoms from underlying structure.

Example: Before approving a new reorganization, a team draws a causal loop diagram and runs a root cause analysis on the problems it is meant to solve.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Anticipating Side Effects

The deliberate practice of identifying an intervention's likely unintended results before carrying it out, often by walking through each stakeholder's point of view.

Example: A hospital realizes that a smartphone-only booking app would shut out patients without smartphones, so it adds a telephone option.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Antifragility

The property of a system that becomes stronger or more capable as a result of stress, shocks, or variation, rather than merely surviving them.

The term was popularized by writer Nassim Nicholas Taleb.

Example: Muscles grow stronger after exercise, and a software team that practices recovering from planned outages improves its response to real ones.

Contrast with: Fragility

Covered in: Chapter 7: Feedback Resilience and Robustness

API

An Application Programming Interface, a defined set of requests and responses through which one software program can use the data or functions of another without knowing its inner workings.

Example: A weather app asks a weather service's interface for tomorrow's forecast and receives the answer in a standard format.

See also: Interoperability

Covered in: Chapter 18: Data Management and Governance

Architectural Innovation

A change in how an existing product's known components are arranged or connected, without necessarily inventing any single new component.

It can catch established companies off guard, because their expertise is organized around the old arrangement.

Example: The laptop rearranged and shrank the parts of a desktop computer, nearly all of which already existed.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Artificial Intelligence

The field of computer science that builds systems able to perform tasks that normally require human thinking, such as recognizing images, understanding language, making predictions, and making decisions.

See also: Machine Learning, Generative AI

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Attention Economy

A marketplace view in which people's limited focus and time are treated as a scarce resource that companies compete to capture, measure, and sell to advertisers.

Example: Social media apps design endless scrolling and notifications to hold users' attention because advertising revenue depends on it.

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Attractor

A state or pattern of behavior toward which a system tends to move over time and to which it returns after small disturbances.

See also: Equilibrium, Bifurcation

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Attribute

A single named characteristic of an entity that holds a value, such as a customer's name, a product's price, or an event's date.

Example: A student record might store a first name, a grade level, and a date of birth.

Covered in: Chapter 17: Knowledge Representation and Metadata

Authentic Assessment

An evaluation method that asks learners to apply knowledge and skills to realistic, meaningful tasks, rather than to answer isolated test questions.

It reduces the distortions that Campbell's Law predicts for high-stakes test scores.

Example: Instead of a multiple-choice exam, students design a budget for a real school event and present it to the principal.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Automation

The use of machines or software to carry out tasks with little or no human effort, following predefined rules or patterns found in data.

Example: A payroll program calculates and deposits every employee's pay each month without anyone typing in the numbers.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Autonomy

The capacity of a system to choose and adjust its own actions in changing situations without real-time direction from a person.

See also: Automation, Human Oversight

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Aware Maturity Level

The second stage of this book's systems-thinking maturity scale, at which leaders recognize that organizational issues are connected and discuss some secondary effects before deciding, though analysis remains limited and informal.

Example: Leaders ask, "What else might this change affect?" in meetings, but no one draws a diagram or checks data.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Balancing Loop

A closed chain of cause and effect that counters any change with a push in the opposite direction, moving a variable toward a goal or limit and holding it there.

It is also called a negative feedback loop; causal loop diagrams mark it with the letter "B," and older diagrams use the icon of a balance scale.

Example: When a room gets too cold, a thermostat turns on the heat, and when the room warms up, the heat turns off.

Contrast with: Reinforcing Loop

See also: Goal-Seeking Behavior

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Balancing Loop Label

The capital letter "B," often drawn inside a small circular arrow, that is placed on a causal loop diagram to mark a closed circuit that counteracts change and seeks a goal.

See also: Loop Marker, Balancing Loop

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Bifurcation

A point at which a small, gradual change in a system's conditions causes the system's long-term behavior to split into two or more distinctly different possible paths.

See also: Tipping Point, Attractor

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Biological System

A set of interacting living parts, such as cells, organs, or organisms, that together carry out life functions like growth, reproduction, and self-regulation.

See also: Homeostasis

Covered in: Chapter 27: Systems Thinking Across Disciplines

Bottleneck

The single step in a process with the lowest capacity, which caps the throughput of the whole process no matter how fast the other steps run.

Example: If a testing team can review only ten software changes per day, the whole release process moves at ten changes per day.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Bottom-Up Behavior

Structure or activity that arises from the accumulation of many local interactions among lower-level agents rather than being dictated by a central authority.

Example: Students cutting across a campus lawn wear paths along the most convenient routes long before anyone paves a sidewalk there.

Contrast with: Top-Down Constraint

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Bounded Rationality

The idea, developed by economist Herbert Simon, that people make decisions that are reasonable given their limited information, time, and mental capacity, rather than perfectly optimal ones.

Example: A department head picks the software that solves her team's problem without knowing that another team already owns a tool that would work for both.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Brain Drain

The departure of a system's most skilled people toward better opportunities elsewhere, often set off by a cost-cutting quick fix that made staying less attractive.

Example: A company freezes training and promotions to save money, and its most capable engineers, who can most easily find new jobs, leave first.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Brain-Computer Interface

A system that records signals from the nervous system and translates them into commands for an external device, such as a cursor, a robotic arm, or a speech synthesizer.

Example: A person with paralysis moves a cursor on a screen by imagining hand movements.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Buffer

A stock large enough, compared with its flows, to absorb sudden changes and keep a system running smoothly when inputs or outputs vary.

Example: A grocery store's back-room inventory keeps shelves full when a delivery truck is a day late.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Business Glossary

A list of business terms and their agreed definitions, maintained for an organization so that people across departments use the same words to mean the same things.

See also: Common Vocabulary, Data Dictionary

Covered in: Chapter 17: Knowledge Representation and Metadata

Butterfly Effect

The sensitivity of a chaotic system to its starting conditions, in which a tiny difference at the beginning grows into a large difference in outcomes over time.

The name comes from the image of a butterfly flapping its wings in one place and, in theory, influencing a storm far away weeks later.

Example: Two weather simulations that differ only in the fourth decimal place of the starting temperature can predict very different weather two weeks later.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Campbell's Law

The principle, stated by social scientist Donald Campbell, that the more a numeric social indicator is used for important decisions, the more it becomes distorted and the more it corrupts the process it was meant to track.

Example: When school funding depends heavily on test scores, some schools narrow teaching to test preparation or even alter results.

See also: Goodhart's Law, Measurement Trap

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Canonical Schema

A single standardized and approved format for representing a type of data as it moves between computer systems, which each connected system translates to and from.

Converting each system's own format into this shared format is done through schema matching and schema mapping.

Covered in: Chapter 17: Knowledge Representation and Metadata

Capability Erosion

The loss over time of a system's own ability to perform a function, usually because that function has been repeatedly handed to a quick fix instead of being practiced and maintained.

Example: A team that always hires a contractor for database migrations never builds that skill in-house.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Capability Maturity Model

A framework, first developed at Carnegie Mellon University's Software Engineering Institute, that describes an ordered series of stages through which an organization's practices become more consistent, measured, and improved.

Example: The original software model ran from Initial to Repeatable, Defined, Managed, and Optimizing.

See also: Maturity Level, Maturity Assessment

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Carrying Capacity

The largest population or level of activity that an environment can support over the long term without degrading the resources it depends on.

Example: A pasture that can feed 50 cows year after year has a carrying capacity of about 50 cows; adding more slowly destroys the grass.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Cascading Failure

A chain of breakdowns in which the collapse of one part overloads or disables the parts connected to it, causing them to break down in turn.

Example: In a power grid, one overloaded line shuts off and pushes its load onto neighboring lines, which also shut off until a whole region goes dark.

See also: Ripple Effect, Fragility

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Causal Link

A claimed relationship between two variables in which a change in the first actually produces a change in the second, rather than merely happening alongside it.

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Causal Loop Diagram

A diagram that shows the variables of a system as labeled nodes joined by arrows showing which variables influence which, used to reveal the closed circuits of influence that drive the system's changes over time.

These diagrams are the main visual language used throughout this book.

Example: A diagram might link "Customers" to "Word of Mouth" to "New Customers" and back to "Customers," showing a reinforcing growth loop.

See also: Node, Causal Link, Loop Polarity

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Cause And Effect

The relationship in which one event or change produces another event or change that would not have happened, or would have been different, without it.

Systems thinking extends this simple idea into loops, where an effect can circle back and change its own cause.

Example: Studying more hours usually leads to higher test scores, and higher scores may then motivate even more studying.

Covered in: Chapter 1: Foundations of Systems Thinking

Cellular Automata

Computational models made of a grid of squares, each in one of a few states, that update step by step according to fixed rules based on the states of neighboring squares, often producing surprisingly complex patterns.

Example: In Conway's Game of Life, cells live or die based only on how many neighbors are alive, yet the grid produces moving shapes and repeating structures.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Central Data Warehouse

A large database, managed by one organization-wide team, that collects cleaned and integrated data from many source systems across an organization for reporting and analysis.

Example: A university loads data nightly from its admissions, registration, and financial aid systems into one repository for campus-wide reports.

See also: Data Lake, ETL Process

Covered in: Chapter 18: Data Management and Governance

Centrality

A family of measures that score how important or influential each vertex is within a graph, based on factors such as its number of connections or how often it lies on paths between others.

Example: An analysis of company email finds the one employee who links several otherwise separate teams.

Covered in: Chapter 16: Graph Database Architecture

Centralization

The gathering of decision-making authority, resources, or data into a single group or location that acts for the whole organization or system, rather than leaving it with local units.

Example: All purchasing decisions for a school district are made by one district office instead of by each school.

Contrast with: Decentralization

Covered in: Chapter 20: Organizational Silos and Silo Busting

Change Management

The structured approach an organization uses to prepare, support, and guide people through a shift in processes, technology, or structure so that the new way is adopted and lasts.

Example: Before switching to a new records system, a clinic trains staff, names champions in each department, and gathers feedback for the first month.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Chaos Theory

The branch of mathematics and science that studies systems whose behavior follows fixed rules yet is so sensitive to starting conditions that long-term prediction becomes practically impossible.

It explains why systems such as the weather can be fully rule-bound and still unpredictable beyond a short time horizon.

Example: Weather forecasts are reliable a few days ahead but not a few months ahead, even with the best computers.

See also: Butterfly Effect, Attractor, Edge Of Chaos

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Chief Data Officer

A senior executive responsible for an organization's data strategy, including data quality, governance, sharing, and the use of data to create business value.

See also: Data Governance

Covered in: Chapter 20: Organizational Silos and Silo Busting

CIO Influence Diagram

A causal map of the loop in which a Chief Information Officer's public support for a knowledge graph pilot raises its funding and visibility, which improves the odds of an early win and further support.

The same loop can spiral downward if an early pilot fails in public.

See also: Influence Diagram

Covered in: Chapter 19: Enterprise Knowledge Graphs

Closed System

A system that exchanges little or no matter, energy, or information with its environment, so that its behavior depends mainly on its own internal parts and relationships.

Contrast with: Open System

Covered in: Chapter 1: Foundations of Systems Thinking

Co-Flow

A secondary accumulation that tracks an attribute traveling along with a primary stock, rising in parallel with it rather than driving it.

Example: A factory tracks units produced as its main stock and the defects carried by those units as a parallel stock.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Code Reuse

The practice of applying an existing software component, model, or solution to a new problem instead of building an equivalent solution from scratch.

See also: Search And Reuse

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Coevolution

The process in which two or more interacting systems or species change over time in response to one another, each shaping the direction of the other's development.

Example: Spam filters improve, spammers invent new tricks to get past them, and the filters improve again in response.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Collective Action Problem

A situation in which everyone in a group would be better off cooperating, but each individual has a personal incentive to act alone, so the group fails to reach the better shared outcome.

Example: Every country benefits from lower global emissions, but each country is tempted to let the others make the costly cuts.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Collective Governance

The capacity of a community to create, enforce, and adjust its own rules for a shared resource, whether natural or digital, without relying on a single central owner.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Collective Impact

A structured framework in which many independent organizations align around one shared agenda, one shared way of measuring progress, and mutually supporting activities to address a large social problem.

Example: Shelters, hospitals, and local employers agree on common goals and shared data to reduce homelessness instead of running separate programs.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Collective Intelligence

The shared problem-solving ability that emerges when many individuals combine their knowledge, judgment, and effort, often exceeding what any single member could achieve.

Example: Thousands of volunteers playing an online puzzle game helped scientists work out the shape of a protein that had puzzled researchers for years.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

College Admissions Algorithm

A prediction model used to help rank or screen applicants to a college or university, often trained on many years of past decisions about which students were accepted.

If past decisions reflect unfair patterns, the model can repeat them, and each new class becomes training data that strengthens the pattern.

See also: Algorithmic Bias, Systemic Inequality

Covered in: Chapter 23: AI Systems Dynamics

Common Pool Resource

A shared supply, such as a fishery, a groundwater basin, or network bandwidth, that is hard to keep people from using and that shrinks with each person's use.

Example: A town's groundwater supply is available to every well owner, and every gallon pumped leaves less for everyone else.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Common Vocabulary

A set of terms with agreed meanings that people across different teams or organizations use consistently, so that they can communicate without misunderstanding.

See also: Business Glossary, Data Standards

Covered in: Chapter 20: Organizational Silos and Silo Busting

Commons Governance Solution

A set of rules, norms, or institutions created and enforced by a shared resource's own users, rather than by private ownership or outside authority, that keeps use within the resource's ability to replenish.

Political economist Elinor Ostrom won a Nobel Prize for showing that real communities manage fisheries, forests, and irrigation systems this way.

Example: Lobster fishers in Maine agree on territories and catch rules among themselves and report violators.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Communication Standards

Agreed-upon formats and protocols that allow independent systems to exchange data so that the sender and the receiver interpret every piece of it in the same way.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling · Communication Standards archetype

Community Detection

A family of methods that divide a network into groups of vertices that are more densely connected to one another than to the rest of the network.

Example: Finding tightly knit friend groups in a school's social network, or clusters of accounts that make up a fraud ring.

See also: Graph Analytics

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Complex Adaptive System

A system made up of many interacting agents that each adjust their behavior based on experience, so that the whole system learns, evolves, and produces patterns that no one designed.

Example: An ant colony, a stock market, the immune system, and an open-source software community all learn and change as their members respond to one another.

See also: Emergence, Adaptive Capacity, Coevolution

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Complexity

The property of a system whose many interacting parts, feedback loops, and nonlinear relationships make its overall behavior hard to predict from knowledge of the parts alone.

Example: City traffic shows this property: a single stalled car can cause a jam that spreads for miles in ways no one planned or can easily forecast.

Covered in: Chapter 1: Foundations of Systems Thinking

Complicated Vs Complex

The distinction between systems with many parts whose behavior can be fully worked out in advance by experts and systems whose interacting, adapting parts produce behavior that cannot be fully predicted.

Complicated problems yield to careful analysis and planning, while complex problems call for small experiments, feedback, and adjustment.

Example: A jet engine is complicated, because engineers can design and predict it; an airline's customer satisfaction is complex, because it shifts with weather, staff morale, competitors, and social media.

See also: Complexity, Wicked Problem

Covered in: Chapter 1: Foundations of Systems Thinking

Compound Interest

Money earned on a loan or deposit that is calculated on both the original amount and all earnings already added, so that a balance grows by a larger amount each period.

Example: $1,000 earning 7% per year grows to about $1,967 in ten years this way, compared with $1,700 if interest were paid only on the original $1,000.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Compounding Advantage

The process by which an early edge in some resource or outcome increases access to further resources, which widens the original edge even more, round after round.

Example: A school with high test scores attracts more funding and stronger teachers, which raises its scores further.

See also: Cumulative Advantage, Matthew Effect

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Compounding Effect

The general process, found well beyond finance, in which a quantity's own growth feeds into the calculation of its next round of growth, so each gain makes the following gain larger.

Example: A viral video's views lead to more recommendations, which lead to more views, and a rumor spreads through a school the same way.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Computational Resource Limit

A ceiling on growth or performance imposed by available processing power, memory, storage, or energy, which no improvement in method can fully overcome.

Example: Training the largest AI models requires so much electricity and specialized hardware that power supply becomes a real barrier.

See also: Scaling Laws, Physical Limits

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Conceptual Data Model

A high-level description of the main things an organization cares about and how they relate, written in plain business terms without technical detail.

Example: A simple diagram showing that Customers place Orders and that Orders contain Products.

See also: Logical Data Model, Physical Data Model

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Condition

A node in a causal loop diagram that represents a target or state set from outside the system, such as a desired temperature, rather than one the system computes for itself.

See also: Set Point, Outcome

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Constants And Parameters

The numeric settings a system operates under, such as tax rates, prices, quotas, speed limits, or a thermostat setting, which together make up the weakest category of leverage point.

Example: Raising the minimum wage by a dollar is a real and immediate change, but employers can absorb it by adjusting hours or prices.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Constitution-Level Rule

A provision that governs how the other policies and laws of a system may be created or changed, such as a nation's procedure for amending its founding charter or a company's procedure for amending its bylaws.

Example: A school board policy stating that any change to the grading policy requires a public hearing and a two-thirds vote.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Constraint

Any factor that restricts what a system can do or how far it can grow, such as a limit on time, money, space, rules, or physical resources.

See also: Bottleneck, Limiting Factor

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Content Moderation Policy

The set of rules a digital platform enforces about which posts, videos, or other material its systems may recommend, demote, or remove.

These rules decide which signals the platform's learning loops may use; without them, a loop that seeks engagement may discover that outrage and falsehoods hold attention best.

Example: A video site stops recommending videos that make false health claims, even when those videos get many views.

Covered in: Chapter 23: AI Systems Dynamics

Context Graph

A connected data model structured to represent the surrounding circumstances of a particular situation or request, such as a user's recent activity, location, and related entities, so an application returns facts framed by their setting.

Example: An AI assistant answering "When is my next meeting?" draws on a connected model of the user's calendar, time zone, and current project.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Control System

An arrangement of a sensor, a comparison against a set point, and an actuator that work together in a balancing loop to keep a variable near its target.

Example: A car's cruise control senses speed, compares it with the chosen speed, and adjusts the throttle to close the gap.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Conveyor Model

A stock-and-flow structure in which everything entering a stock leaves after exactly the same fixed transit time, like items riding a moving belt from one end to the other.

Example: A two-year associate degree program can be modeled so that each entering class leaves as graduates exactly two years later.

See also: System Delay, Aging Chain

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Corrective Action

A response that closes the gap between a goal and actual performance by improving performance, rather than by changing the goal.

Contrast with: Goal Erosion

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Correlation Vs Causation

The distinction between two quantities that merely rise and fall together and one quantity actually producing a change in another.

Mistaking a shared pattern for a causal link leads to fixes that target the wrong thing.

Example: Ice cream sales and drowning rates rise together in summer, but ice cream does not cause drowning; hot weather drives both.

Covered in: Chapter 1: Foundations of Systems Thinking

Creative Destruction

The process, named by economist Joseph Schumpeter, in which new innovations sweep away established companies, products, and jobs, freeing resources for new and more productive uses.

Example: Online maps and GPS put many paper map publishers out of business while creating new jobs in software and delivery services.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Credit Scoring System

A prediction model that estimates how likely a borrower is to repay a loan, based on the borrower's financial history and related information.

People who were unfairly denied loans in the past have thinner financial histories, so such a model may keep rating them as risky regardless of their current situation.

See also: Algorithmic Bias

Covered in: Chapter 23: AI Systems Dynamics

Crisis Point

The moment at which the accumulated gap between an eroded goal and a system's real needs becomes too large to ignore, forcing an abrupt and often costly reckoning.

See also: Goal Erosion

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Cross-Functional Team

A group of people from different departments or areas of expertise who work together toward a shared goal, bringing their different perspectives and knowledge to a single problem.

Example: A product launch team that includes an engineer, a designer, a marketer, a lawyer, and a customer support lead.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Cross-Generational Knowledge

Know-how deliberately passed from older or departing workers, or from the founders of a family business, to those who follow them, so that an organization's capability survives retirements and departures.

Example: A retiring master electrician spends her final year training two apprentices on the building's oldest wiring.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Cross-Team Collaboration

The practice of separate groups within an organization sharing information, resources, and decisions and working jointly on common goals, rather than working only within their own group.

Example: The security team and the developers review new features together early, instead of security checking finished work at the end.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Cumulative Advantage

The compounding process by which a small early lead in resources, status, or opportunity grows into a large and lasting gap, because each gain makes the next gain easier to win.

Example: A researcher whose first paper is widely cited gets more invitations and grants, which lead to more papers and still more citations.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Customer 360

A unified profile of a buyer that combines identity with the full history of purchases, support requests, and marketing interactions, so every team works from one complete picture.

Example: A support agent can see that a caller made a large purchase last week and has opened two complaints this month.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Cypher

A declarative query language, first created for the Neo4j database, that describes graph patterns with text symbols, such as parentheses for vertices and arrows for edges.

Example: The query MATCH (p:Person)-[:WORKS_AT]->(c:Company) RETURN p.name, c.name lists every person and the company where that person works.

See also: GQL, Graph Query Language

Covered in: Chapter 15: Graph Theory Fundamentals

Dampening

The weakening of a signal, change, or effect as it passes through a system, usually by a balancing loop or buffer, so that swings grow smaller over time.

Contrast with: Amplification

Covered in: Chapter 7: Feedback Resilience and Robustness

Data Advantage

The competitive edge a platform gains from having more, or more relevant, information about user behavior than its rivals, information often produced only by already having many users.

Example: The most-used search engine sees the widest range of searches and clicks, something a smaller rival cannot buy at any price.

Covered in: Chapter 23: AI Systems Dynamics

Data Denormalization

The process of combining data from several related tables into fewer, wider tables, copying some values into more than one place so that reading the data becomes faster.

Covered in: Chapter 18: Data Management and Governance

Data Dictionary

A catalog that lists every data element in a database or system along with its name, meaning, data type, allowed values, and source.

Example: An entry might say that "enroll_dt" is a date field holding the day a student first registered, taken from the admissions system.

Covered in: Chapter 17: Knowledge Representation and Metadata

Data Domain

A logical grouping of related information entities that share a common business meaning or ownership, such as Customer, Product, or Finance.

It gives governance a natural unit that is smaller than the whole company but larger than a single table.

See also: Data Governance, Data Steward

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Data Governance

The framework of roles, responsibilities, policies, and processes an organization uses to manage the availability, quality, security, and proper use of its data.

Example: A company decides who may approve changes to customer data, how long records are kept, and who may see salary information.

Covered in: Chapter 18: Data Management and Governance

Data Integration

The process of combining data from different sources into a unified, consistent view so that it can be used together for reporting, analysis, or applications.

Example: Merging customer records from a store's website, its phone orders, and its loyalty program into one view of each shopper.

Covered in: Chapter 18: Data Management and Governance

Data Integration Cost

The total time, money, and effort needed to connect data across systems, which rises sharply as the number of separately designed systems grows.

With direct connections between every pair of systems, this cost grows roughly with the square of the number of systems.

Covered in: Chapter 18: Data Management and Governance

Data Lake

A large storage repository that holds raw data from many sources in its original format, whether structured or unstructured, until someone needs to use it.

Example: A company stores website clicks, sensor readings, emails, and spreadsheets together without first reorganizing them.

Covered in: Chapter 18: Data Management and Governance

Data Lineage

A record of where a piece of data came from and every system, step, and transformation it passed through on the way to its current form and location.

See also: Data Provenance

Covered in: Chapter 18: Data Management and Governance

Data Literacy

The ability to read, understand, question, work with, and communicate using data, including recognizing when numbers or charts are misleading.

Example: A manager who notices that a chart's vertical axis starts at 90 instead of 0 and exaggerates a small change.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Data Mart

A smaller, subject-focused store of data built to meet the reporting needs of one department or business area, such as sales, finance, or human resources.

Example: The marketing team's own reporting database holds only campaign and customer response data.

Covered in: Chapter 18: Data Management and Governance

Data Provenance

Documentation of the origin of a piece of information and the history of who created, owned, or changed it, used to judge whether it can be trusted.

It differs from data lineage, which traces the technical path through systems and transformations rather than origin and custody.

Example: A newsroom records that a photo came from a named freelance photographer and has not been edited.

See also: Data Lineage

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Data Quality

The degree to which a set of data is accurate, complete, consistent, up to date, and suitable for the purpose for which people intend to use it.

Example: A mailing list full of outdated addresses and duplicate names scores poorly, because letters go to the wrong places.

Covered in: Chapter 18: Data Management and Governance

Data Set

A structured collection of related records or measurements, usually organized as rows and columns or as a group of files, gathered for a purpose such as analysis or model training.

Example: A table of daily weather readings for one city over twenty years.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Data Silo

A store of information held by one department or application that is not easily reachable by, or compatible with, other parts of the organization.

Example: The sales team keeps its customer notes in a spreadsheet that the support team cannot see.

See also: Organizational Silo, Data Integration

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Data Standardization Cycle

A reinforcing loop in which each additional system that adopts a shared set of data conventions lowers the integration cost for the next system to join, which makes adoption even more attractive.

See also: Data Standards, Data Integration Cost

Covered in: Chapter 19: Enterprise Knowledge Graphs

Data Standards

Agreed conventions for how data is named, defined, formatted, and coded, so that data from different sources can be combined and compared reliably.

Example: Every department agrees to record dates as year-month-day and to use two-letter codes for U.S. states.

Covered in: Chapter 17: Knowledge Representation and Metadata

Data Steward

A person assigned responsibility for the definition, quality, and proper use of a particular set of data within an organization.

Example: A product manager who decides what "product category" means and fixes miscategorized products in the catalog.

Covered in: Chapter 18: Data Management and Governance

Data Stewardship

The day-to-day work of overseeing a set of data, including defining it, monitoring its quality, resolving problems, and making sure it is used appropriately.

See also: Data Steward, Data Governance

Covered in: Chapter 18: Data Management and Governance

Decentralization

The spreading of decision-making authority, resources, or data across many local groups or locations within an organization or system, so that each can act on its own.

Example: Each store manager chooses local products to stock, rather than following a list from headquarters.

Contrast with: Centralization

Covered in: Chapter 20: Organizational Silos and Silo Busting

Deep Leverage Point

A place to intervene that changes the rules governing a system, or who has the power to create and change those rules, reaching beyond any single feedback loop.

Example: Changing which department has authority to approve new vendor contracts, rather than changing the dollar amount that needs approval.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Defined Maturity Level

A stage in a capability maturity model at which a process is documented, standardized, and understood well enough to be taught to a new team member.

See also: Repeatable Maturity Level

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Delay

The interval of time between a cause occurring and its effect appearing somewhere else in a system, during which the cause has happened but its results cannot yet be seen.

Gaps like this are a leading cause of overshoot and oscillation, because people keep acting before they can see the results of earlier actions.

Example: A new hire takes months to become fully productive, so adding staff does not speed up a late project right away.

See also: Time Delay, Feedback Delay, System Delay

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Delayed Consequence

A result of an action that appears only after a significant time has passed, often too late to be clearly linked back to the decision that caused it.

See also: Delay, Unintended Consequence

Covered in: Chapter 7: Feedback Resilience and Robustness

Denormalized Data Model

A relational data design that deliberately combines data from several tables into fewer, wider tables, accepting duplicated values in exchange for simpler loading and faster reads.

Such designs are easier to load into a database and fast for fixed reports, but harder to keep consistent and harder to query in new ways.

Example: An order table that repeats the customer's full name and address on every order row.

Contrast with: Normalized Data Model

Covered in: Chapter 18: Data Management and Governance

Diffusion Process

The spread of something, such as an idea, technology, behavior, or disease, through a population over time as it passes from members who have it to the members connected to them.

See also: Epidemic Model, Technology Adoption

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Digital Commons

A shared pool of information, software, or knowledge that many people can use and contribute to, such as open-source code, open data, or online encyclopedias.

Example: Wikipedia and the Linux operating system are maintained by volunteers and freely used by millions.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Digital Divide

The gap between people or communities who have reliable access to computers, the internet, and the skills to use them well and those who do not.

It tends to widen over time because those with access gain skills and opportunities that bring still more access.

Example: During remote schooling, students without home internet fell behind classmates who could attend online classes easily.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Digital Transformation

The broad process of changing an organization's processes, products, culture, and business model by adopting computing, internet, and data-driven ways of working in place of manual or paper-based ones.

Example: A bank moves from paper loan applications and branch visits to online applications with instant decisions.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Digital Twin

A virtual representation of a real-world object, process, or organization that is kept current with data from its real counterpart so it can be monitored, analyzed, or simulated.

Enterprise knowledge graphs can serve this role by modeling how real-world components interact with each other.

Example: An airline keeps a virtual copy of each jet engine, fed by sensor data, to predict when parts will need replacing.

Covered in: Chapter 19: Enterprise Knowledge Graphs

Diminishing Returns

The pattern in which each additional unit of effort or investment produces a smaller gain than the unit before it.

Example: The first hour of studying for a test helps a lot, the fifth hour helps less, and the tenth hour may barely help at all.

See also: S-Curve, Economic Limits

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Directed Graph

A graph in which every edge points one way, from a starting vertex to an ending vertex, so that the relationship it records runs from one to the other and not back.

Example: On a social app where following is one-way, you can follow a celebrity without that celebrity following you back.

Contrast with: Undirected Graph

Covered in: Chapter 15: Graph Theory Fundamentals

Distributed Control

An arrangement in which decision-making authority is spread among many parts of a system, each acting on local information, rather than held by a single central authority.

Contrast with: Centralization

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Distributed Graph Database

A graph database whose vertices and edges are stored and processed across many connected servers, so that it can hold more data and answer more queries than one machine could.

See also: Scale-Out Graph Architecture, Scale Out

Covered in: Chapter 16: Graph Database Architecture

Drifting Goals Archetype

A systems archetype in which a persistent gap between a target and actual performance is closed by gradually lowering the target rather than by improving performance.

See also: Goal Erosion, Eroding Goals

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Drifting Goals examples

Dunbar's Number

The estimated limit, proposed by anthropologist Robin Dunbar, of about 150 stable personal relationships that one person can maintain, set by the brain's capacity to track social ties.

Example: Some companies split teams or offices once they grow past about 150 people so that everyone can still know one another.

See also: Physical Limits

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Dynamic Equilibrium

A state in which a stock stays at a constant level because its inflows and outflows are equal, even though material or information keeps moving through the system.

Example: A lake whose water level holds steady while rivers flow in and out at equal rates is in this state.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Dynamics

The patterns of change over time in the stocks, flows, and connections of a system, such as growth, decline, oscillation, or leveling off.

Example: A new app's user base might show slow early growth, then a rapid rise, and finally a plateau.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Echo Chamber

A social environment in which people mainly encounter opinions that match their own, because they choose like-minded sources and communities, so that those opinions are repeated and strengthened.

It differs from a filter bubble in that people build it through their own choices rather than having software build it for them.

Example: An online group where members share only posts that agree with the group's views and criticize anyone who disagrees.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Ecological System

Any system studied in terms of the relationships between living things and their physical surroundings, at scales from a single pond to the whole planet, including the human activities that affect them.

Example: The global carbon cycle links forests, oceans, the atmosphere, and the burning of fossil fuels.

See also: Ecosystem

Covered in: Chapter 27: Systems Thinking Across Disciplines

Economic Complexity Index

A measure, developed by economist Ricardo Hausmann and physicist César Hidalgo, of how much productive know-how a country holds, based on the diversity and sophistication of the products it successfully exports.

Example: Countries that export a wide range of advanced machinery and electronics rank near the top, while countries that export mainly a few raw materials rank low.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Economic Limits

Constraints on growth that arise when the cost of obtaining more of a resource rises faster than the value gained, so growth stops being worthwhile before it becomes physically impossible.

See also: Physical Limits, Diminishing Returns

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Economic System

The set of institutions, rules, and relationships through which a society produces, distributes, and uses goods and services, including its markets, businesses, banks, workers, and government agencies.

See also: Market System

Covered in: Chapter 27: Systems Thinking Across Disciplines

Ecosystem

A community of living organisms in a particular area together with the soil, water, air, and climate they interact with, functioning as a single unit through flows of energy and nutrients.

Example: On a coral reef, fish, coral, algae, and the chemistry of the seawater all depend on one another.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Ecosystem Balance

A condition in which a shared natural resource is used no faster than it naturally replenishes, allowing it to remain healthy indefinitely.

See also: Carrying Capacity, Sustainable Growth

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Edge (CLD)

An arrow in a causal loop diagram drawn from one node to another, pointing from a cause toward the effect that it produces.

See also: Causal Link, Node, Edge (Graph)

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Edge (Graph)

A connection between two vertices in a graph data structure that represents a relationship between the things they stand for and may carry a direction, a type, and properties.

See also: Vertex, Edge (CLD)

Covered in: Chapter 15: Graph Theory Fundamentals

Edge Of Chaos

A transition zone between rigid order and complete disorder, thought to exist in many kinds of systems, where bounded instability allows a constant, creative interplay between stability and change.

Systems in this zone are often the most adaptable, and within enterprise knowledge graphs it is the region that tends to give the highest return when modeled.

Example: A company with some firm rules but freedom for teams to experiment can adapt faster than one that is either tightly controlled or completely unstructured.

See also: Chaos Theory, Complex Adaptive System

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Edge Traversal Performance

The speed of moving across a single connection between two vertices, which in a native graph database stays roughly constant no matter how large the database grows.

See also: Pointer Hopping, RDBMS JOIN

Covered in: Chapter 16: Graph Database Architecture

EdTech Funding Gap

The disparity between wealthy school districts that can afford AI-powered learning tools and the teacher training to use them and poorer districts that cannot, a difference that tends to widen over time.

Example: A suburban district gives every teacher AI tutoring tools and training, while a rural district cuts its training budget to save money and falls further behind.

Covered in: Chapter 23: AI Systems Dynamics

Education System

The network of schools, teachers, students, families, curricula, funding, and policies that together shape how people learn and are taught.

Example: A change in college entrance tests ripples back into what high schools teach and what tutoring companies sell.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Embedding (AI)

A list of numbers that represents a word, sentence, image, or other item as a point in a mathematical space, placed so that items with similar meanings end up close together.

Example: The number lists for "puppy" and "dog" lie close together, while those for "puppy" and "spreadsheet" lie far apart.

See also: Vector Database

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Emergence

The process by which new patterns, structures, or behaviors arise at the level of the whole system from the interactions of simpler parts following local rules.

Example: A traffic jam forms from many drivers each braking a little; no single driver decides to create the jam.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Emergent Property

A characteristic of a whole system that none of its individual parts has on its own and that arises only from the way the parts interact.

See also: Emergence

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Emergent Risk Management

An approach to handling danger that accepts that some threats appear only in whole-system patterns, and therefore focuses on watching for those patterns as they form and building spare capacity to absorb surprises.

Example: A bank watches for unusual behavior spreading across many accounts at once, rather than only checking each transaction on its own.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Emerging Technology System

A system built around a new tool or scientific capability that is still early in its development and adoption, whose real capabilities, risks, and workable business models remain uncertain.

See also: Technology Adoption, S-Curve

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Employee Career Path Loop

A reinforcing cycle in which workers who use a knowledge graph of skills, roles, and projects to plan their next job move keep their own information current, which makes the graph more accurate for the next person who uses it.

See also: Enterprise Knowledge Graph

Covered in: Chapter 19: Enterprise Knowledge Graphs

Enterprise Knowledge Graph

A connected layer of data, stored as entities and the relationships among them, that links information about an organization's people, products, customers, processes, and systems across departmental boundaries with shared meaning, so it can be queried as one whole.

It is this book's central technical tool for breaking down organizational silos.

Example: A manufacturer links its parts, suppliers, factories, orders, and customers so that it can instantly see which customers a delayed shipment will affect.

See also: Graph Database, Silo Busting, Digital Twin

Covered in: Chapter 19: Enterprise Knowledge Graphs

Entity

A distinct real-world or conceptual thing, such as a person, organization, product, place, or event, about which data is recorded.

Example: In a school database, each student, teacher, course, and classroom is a separate thing with its own record.

Covered in: Chapter 17: Knowledge Representation and Metadata

Entity Resolution

The process of determining which records from one or more data sources refer to the same real-world thing, and then linking or merging those records.

Example: Recognizing that "J. Smith, 12 Oak St." and "John Smith, 12 Oak Street" are the same customer.

Covered in: Chapter 18: Data Management and Governance

Environment (System)

Everything outside a system's boundary that affects the system or is affected by it, including the sources of its inputs and the destinations of its outputs.

Example: For a small bakery, the environment includes customers, flour suppliers, the local economy, the weather, and health inspectors.

Covered in: Chapter 1: Foundations of Systems Thinking

Epidemic Model

A mathematical model that divides a population into groups, such as susceptible, infected, and recovered, and uses the rates of movement between the groups to simulate how something spreads.

Example: Public-health teams use such models to predict flu cases, and marketers use the same math to predict how a video will spread online.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Equilibrium

A state in which the influences acting on a system are in balance, so that the system has no internal tendency to change unless something disturbs it.

Example: A ball resting at the bottom of a bowl stays put; nudge it and it rolls back to the bottom.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Equitable Intervention Point

A specific place within a loop of compounding advantage where a deliberate change can interrupt the compounding and produce a fairer distribution of resources going forward.

Example: Blind review of grant applications hides applicants' names and schools so reviewers judge only the proposal.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Eroding Goals

The observable pattern of targets or standards slipping lower over time through a series of small, reasonable-seeming adjustments, each made in response to disappointing performance.

It is commonly used as another name for goal erosion.

See also: Goal Erosion

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Escalation Archetype

A systems archetype in which two parties each see the other's actions as a threat and respond by increasing their own efforts, so both keep raising the stakes in a spiral.

Example: Two stores cut prices to beat each other, again and again, until neither makes a profit.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Escalation examples

ETL Process

The three-step process of extracting data from source systems, transforming it into a consistent format, and loading it into a target system such as a data warehouse.

Example: Each night, sales records are copied from store registers, converted to a common currency and date format, and added to the company warehouse.

Covered in: Chapter 18: Data Management and Governance

Events Layer

The visible top level of the iceberg model, made up of individual happenings that can be observed directly, such as a crash, a missed deadline, or a sudden jump in sales.

Example: "The payroll system failed on Friday afternoon."

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Experience Curve

The pattern in which an organization's total cost to produce each unit of a product falls by a fairly steady percentage every time its total cumulative output doubles.

Example: The price of solar panels has dropped sharply as the total number manufactured worldwide has doubled again and again.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Explainability

The degree to which the reasons behind an AI system's specific output or decision can be described in terms that a person can understand and check.

Example: A loan-approval model reports that an application was declined mainly because of a high debt-to-income ratio and a short credit history.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Explicit Knowledge

Know-how that can be written down, recorded, and shared in documents, databases, manuals, or diagrams, so others can pick it up without direct contact with its source.

Example: A step-by-step guide for resetting a customer's password.

Contrast with: Tacit Knowledge

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

The common human tendency to underestimate how large a quantity growing by a fixed percentage will become, because early growth of this kind looks almost the same as steady, straight-line growth.

Example: Folding a sheet of paper in half 42 times would, in theory, make it thick enough to reach the Moon, a result that surprises almost everyone.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Exponential Growth

An increase in which a quantity rises by a fixed percentage of its current size in each period, so the amount added becomes larger every period.

Example: A bacteria colony that doubles every hour grows from 1 cell to more than 1,000 cells in ten hours.

Contrast with: Linear Growth

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Faceted Search

An interface for finding information that lets users narrow a large set of results step by step by choosing values for several independent attributes, such as type, date, owner, or topic.

Example: An online shoe store lets shoppers filter by size, color, brand, and price range at the same time.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

FAIR Data Principles

A set of guidelines stating that research and other shared information are most useful when they are Findable, Accessible, Interoperable, and Reusable by both people and machines.

Example: A university publishes a climate study's measurements with a permanent identifier, a clear license, a standard file format, and a full description.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Feedback

The process in which information about the result of an action is routed back to influence the next action, so that a system's outputs return to become part of its inputs.

Example: A student reads a teacher's comments on an essay and uses them to improve the next draft.

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Feedback Delay

The lag between an effect actually happening and that effect being noticed and returned along a loop to the point where decisions are made, so that choices rest on outdated information.

Example: Monthly sales reports mean a store manager learns about a drop in sales weeks after it began.

See also: Delay, System Delay, Oscillation

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Feedback Loop

A closed circuit of cause-and-effect links in which a change in one variable eventually returns, through other variables, to affect that same variable.

Circular chains like these, rather than straight lines of cause and effect, explain most of the persistent patterns found in systems.

Example: Hunger leads to eating, eating reduces hunger, and reduced hunger leads to stopping eating.

See also: Reinforcing Loop, Balancing Loop

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Feedback Mechanism Design

The deliberate creation of a way to observe how a system actually responds to a change, so that wrong assumptions about its effects can be caught and corrected early.

Example: Pairing a short post-visit patient survey with usage data from a new booking app.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Filter Bubble

A narrowing of the news and ideas a person sees, caused by personalization software that shows more of what the person already clicks on and less of everything else.

Example: A reader who clicks mostly on sports stories gradually stops seeing any news about science or local government.

See also: Echo Chamber, Social Feed Ranking Loop

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Fishbone Diagram

A cause-and-effect chart, also called an Ishikawa diagram, laid out like the skeleton of a fish, with the problem written at the head and possible causes sorted into categories along branching lines.

Example: A restaurant with slow service lists possible causes under bones labeled People, Equipment, Methods, and Supplies.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Five Whys

A root-cause technique in which a person repeatedly asks what caused a problem, usually about five times in a row, with each answer becoming the subject of the next question.

Example: The server went down because the disk was full; the disk was full because logs were never deleted; logs were never deleted because no one owned that task, which points to a missing responsibility rather than a hardware fault.

See also: Root Cause Analysis, Fishbone Diagram

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Fixes That Fail Archetype

A systems archetype in which a rapid remedy relieves a problem in the short term but produces delayed side effects that make the original problem worse over time.

Example: A company cuts maintenance spending to boost profits, but more equipment breaks down later, costing more than the savings.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Fixes That Fail examples

Flow

The rate at which material, money, people, or information moves into, out of, or between accumulations in a system, measured per unit of time.

Example: Gallons per minute from a faucet, dollars per month of salary, and new students per year are all rates of this kind.

See also: Stock, Inflow, Outflow

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Foreign Key

A column in one relational table whose values point to the unique row identifiers of another table, creating a link between related records.

Example: The customer ID stored in each row of an Orders table points to the matching row in the Customers table.

Covered in: Chapter 18: Data Management and Governance

Fragility

The property of a system that is easily damaged or broken by shocks, stress, or variation, often because it has few buffers and tightly coupled parts.

Example: A just-in-time supply chain with a single supplier for one key part can halt a whole factory when that supplier has a fire.

Contrast with: Robustness

Covered in: Chapter 7: Feedback Resilience and Robustness

Fraud Detection Graph

A connected data model of accounts, people, devices, addresses, and transactions used to uncover suspicious rings by following shared links that separate records would hide.

Example: Twenty new credit card accounts that share three phone numbers and one laptop reveal a coordinated scheme.

See also: Community Detection, Graph Analytics

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Free Rider Problem

The situation in which people benefit from a shared resource or effort without contributing their fair share, which can cause the resource to be underfunded or overused.

Example: In a group project, one student does little work but receives the same grade as the students who did most of it.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Fundamental Solution

An intervention on a problem's underlying cause that is substantial enough to resolve the problem permanently rather than suppressing it for a while, so it does not return once the intervention ends.

Example: Drinking water cures a dehydration headache for good, whereas a painkiller only hides it until the medicine wears off.

Contrast with: Symptomatic Solution

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Generative AI

Artificial intelligence systems that create new content, such as text, images, audio, video, or computer code, based on patterns found in large collections of existing examples.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Geographic Knowledge Stickiness

The tendency of certain kinds of know-how, especially skills learned through hands-on experience, to stay concentrated in particular places because they spread mainly through face-to-face contact.

Example: Much of the world's most advanced chip-making skill remains clustered in a few regions, even though the underlying science is published worldwide.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Global AI Governance

The emerging set of international norms, agreements, and coordinating institutions aimed at managing the risks and benefits of artificial intelligence across national borders.

It remains far less developed than national rules, because no global body yet holds the authority that a national regulator holds at home.

See also: AI Governance

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Global Pattern

A large-scale structure or behavior visible across an entire system that results from many local interactions, even though no individual agent can see it or intends to create it.

See also: Emergence, Local Interaction

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Goal Erosion

The process of closing a gap between a target and actual performance by lowering the target to match performance, rather than improving performance to meet the target.

Example: A team that keeps missing its target of 20 completed tasks per sprint quietly lowers the target to 15.

Contrast with: Corrective Action

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Goal-Performance Gap

The measured difference between the level a system is aiming for and the level it is actually achieving, which prompts either an improvement in performance or a lowering of the target.

See also: Corrective Action, Goal Erosion

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Goal-Seeking Behavior

A pattern of change in which a system moves toward a target value, quickly at first when the gap is large and more slowly as the gap closes.

Example: A cup of hot coffee cools quickly at first, then more and more slowly as it nears room temperature.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Goals Of The System

The purpose that a system's behavior actually serves, as shown by what the system consistently produces, which may differ from its officially stated mission.

Changing this purpose redirects every feedback loop in the system, which is why it ranks as a very powerful leverage point.

Example: A hospital's stated mission is patient health, but if its budget rewards the number of procedures performed, its behavior will follow the procedure count.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Golden Record

The single most accurate and complete version of the information about one entity, created by merging and cleaning duplicate or conflicting entries from several sources.

Example: Three slightly different entries for one supplier are merged into one entry with the correct legal name, address, and tax number.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Goodhart's Law

The principle, named for economist Charles Goodhart, that when a measure becomes a target, it ceases to be a good measure, because people change their behavior to hit the number.

Example: A call center that rewards short call times finds agents hanging up on customers with difficult problems.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Governance Model

The defined set of roles, decision rights, processes, and rules that determines who makes which decisions about a shared resource and how those decisions are made.

See also: Data Governance, Collective Governance

Covered in: Chapter 20: Organizational Silos and Silo Busting

GQL

The international standard language for retrieving and updating data in property graphs, published in 2024 as ISO/IEC 39075, which draws heavily on the pattern-matching style of Cypher.

Its name is short for "Graph Query Language," and it plays the same role for graph databases that SQL plays for relational databases.

See also: Cypher, Graph Query Language

Covered in: Chapter 15: Graph Theory Fundamentals

Graph (Data Structure)

A data structure made up of a set of points, called vertices, and a set of connections between pairs of those points, called edges, used to represent things and the relationships among them.

Example: A social network can be stored with each person as a vertex and each friendship as an edge.

See also: Vertex, Edge (Graph), Graph Database

Covered in: Chapter 15: Graph Theory Fundamentals

Graph Algorithms

Step-by-step computational procedures designed to analyze networks of vertices and edges, such as finding shortest routes, ranking important vertices, or detecting clusters.

Example: PageRank scores web pages by counting and weighting the links that point to them.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Graph Analytics

The use of computations on connected data, stored as vertices and edges, to uncover patterns, influential members, clusters, and paths that are hard to see in tables.

Example: A phone company finds customers whose friends recently switched carriers, because they are more likely to switch next.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Graph Database

A database that stores items as vertices and the relationships among them as edges, each of which can carry attributes, so that connections can be followed directly at query time.

Because it does not compare primary and foreign keys in JOIN operations at query time, it can follow chains of relationships far faster than a relational database, especially on highly connected data.

Example: A bank stores customers, accounts, devices, and addresses as a network so it can quickly spot accounts that share the same phone and address.

Covered in: Chapter 16: Graph Database Architecture

Graph Neural Networks

Machine learning models that work on information structured as vertices and edges by repeatedly combining details from each vertex's neighbors to make predictions about vertices, edges, or whole structures.

Example: Predicting whether a new chemical compound is toxic by treating its atoms as vertices and its chemical bonds as edges.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Graph Query Language

Any computer language designed specifically for retrieving, searching, and updating data stored as vertices and edges, expressing requests as patterns of connections to match rather than as table combinations.

See also: Cypher, GQL

Covered in: Chapter 16: Graph Database Architecture

Graph Systems Thinking

A multidisciplinary field of study that applies feedback loops, archetypes, and leverage points to understand and explain the business value of connected data, especially the value of linking information across organizational silos.

It is often abbreviated GST.

Covered in: Chapter 19: Enterprise Knowledge Graphs

Graph Traversal

The process of visiting vertices in a graph by moving from one vertex to its neighbors along edges, following a set of rules, in order to find paths, patterns, or connected items.

Example: Finding a user's "friends of friends" means stepping from the user to each friend and then to each of their friends.

Covered in: Chapter 15: Graph Theory Fundamentals

Graph-Optimized Hardware

Computer processors, memory, and storage whose layout and design are tuned to speed up the rapid, irregular memory lookups needed to follow connections among vertices, rather than general-purpose workloads.

Covered in: Chapter 16: Graph Database Architecture

Growth By Acquisition

An expansion strategy in which an organization gets bigger by purchasing other companies, rather than only by expanding its own operations, bringing in their separate systems, cultures, and vocabularies.

Example: A company that buys five smaller firms over a decade ends up with five billing systems and five different meanings of "customer."

Covered in: Chapter 20: Organizational Silos and Silo Busting

Habit Formation Loop

A reinforcing cycle in which a cue triggers a routine, the routine delivers a reward, and the reward strengthens the craving for that cue, making each repetition more likely.

Example: A phone notification leads to checking social media, which delivers a pleasant message, which makes the next check more likely.

See also: Reinforcing Loop, Addiction Cycle

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Health Disparities

Preventable differences in rates of illness, injury, disability, or death between groups of people defined by factors such as income, race, geography, or disability status.

Example: In many regions, rural residents die of heart disease at higher rates than city residents, partly because hospitals are farther away.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Healthcare AI Access Disparity

The gap between well-resourced hospital systems that can afford AI diagnostic and administrative tools and under-resourced clinics and rural hospitals that cannot, which deepens differences in the quality of care.

See also: EdTech Funding Gap, Health Disparities

Covered in: Chapter 23: AI Systems Dynamics

Hierarchy (Systems)

The arrangement of a system into nested levels, in which each level is made up of the subsystems below it and is itself a part of the level above, with material or information moving up and down this tree-like structure.

Example: Individual workers are nested in teams, teams in departments, and departments in the whole company.

See also: Subsystem, Organizational Structure

Covered in: Chapter 20: Organizational Silos and Silo Busting

High-Leverage Intervention

A proposed change that works on a system's rules, its ability to reorganize itself, its goals, or its underlying paradigm, usually hard to achieve but capable of transforming behavior.

Contrast with: Low-Leverage Intervention

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Highly Connected Data

Information in which the relationships among records matter as much as the records themselves, and in which typical questions require following several relationships in a row.

Example: Detecting a fraud ring means following links from accounts to shared phone numbers, addresses, and devices, several steps deep.

Covered in: Chapter 16: Graph Database Architecture

Holism

An approach to understanding that studies a whole as a single unit, with attention to the relationships and interactions among its parts rather than to each part alone.

Example: A coach who watches how players move together as a team, and not just each player's statistics, is taking this approach.

Contrast with: Reductionism

Covered in: Chapter 1: Foundations of Systems Thinking

Homeostasis

The ability of a living organism or other system to keep an internal condition, such as temperature or chemical balance, within a narrow range despite outside changes.

Example: The human body holds its temperature near 37 degrees Celsius by sweating when hot and shivering when cold.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Hub-And-Spoke Architecture

An integration design in which every system connects to a single central translator and router of data, rather than connecting directly to every other system.

With this design, adding a new system requires only one new connection instead of one connection to each existing system.

Example: A hospital links its lab, pharmacy, billing, and records systems through a single integration engine instead of wiring each pair together.

Contrast with: Point-To-Point Integration

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Human Flourishing Feedback

A loop of information deliberately built into a system's design to track whether the system is truly improving people's well-being, rather than only a narrow efficiency or engagement number.

Example: A learning app measures whether students' understanding and confidence grow, not just how many minutes they spend in the app.

See also: AI Alignment, Feedback Mechanism Design

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Human Oversight

The ongoing monitoring of an automated or AI system by responsible people who can review its behavior, step in when something goes wrong, and change or stop it.

It differs from a human-in-the-loop design, where a person approves individual outputs inside the process itself.

See also: Human-In-The-Loop, Accountability (AI)

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Human-AI Collaboration

A way of working in which a person uses an AI tool to extend a skill that the person continues to practice and understand, rather than handing that skill over entirely.

Example: A developer uses an AI assistant to draft routine code but still reads, understands, and could rewrite every line.

Contrast with: AI Dependence

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Human-Centered Design

An approach to creating products, services, and systems that begins with the needs, abilities, and experiences of the people who will use them and involves those people throughout development and testing.

Example: A hospital watches elderly patients try to book appointments before building a new scheduling system.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Human-In-The-Loop

A system design in which a person reviews, approves, corrects, or overrides an automated system's outputs at key steps, with those decisions often fed back to improve the system.

Example: A fraud detection tool flags suspicious payments, and a human analyst decides which ones to block.

See also: Human Oversight, Automation

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Iceberg Model

A thinking tool that pictures a situation as a mostly submerged mass of ice with four levels: visible events above the waterline, and patterns, underlying structures, and mental models hidden progressively deeper below it.

Example: A crashed server is the visible event; monthly crashes are the pattern; no testing environment is the structure; and the belief that speed matters more than quality is the mental model.

See also: Events Layer, Mental Models Layer

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Implementation Strategy

A deliberate plan for how, when, and in what order an intervention will be rolled out, often starting small before expanding.

See also: Microstrategy

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Incentive Structure

The set of rewards and penalties, such as pay, bonuses, promotions, recognition, and budgets, that shapes what behavior people in an organization find worthwhile.

See also: Misaligned Incentive, Shared Metrics

Covered in: Chapter 20: Organizational Silos and Silo Busting

Induced Demand

The increase in use of a resource or service that occurs when its supply is expanded, often erasing the relief the expansion was meant to provide.

Example: New highway lanes make driving faster for a while, which encourages more people to drive until the road is as crowded as before.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Inference Engine

The software component that applies logical rules to a set of stored facts in order to derive new conclusions or answer questions.

See also: Knowledge Base Reasoning

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Infinite Alphabet Metaphor

A way of picturing an economy in which each distinct productive capability is a letter and each product is a word spelled from a combination of letters, so economies with more letters can spell more and harder words.

Example: A country holding "letters" for precision machining, chemistry, and logistics can spell "medical devices," while one holding only farming and mining cannot.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Inflow

A rate of movement that adds to a stock, increasing the amount stored for as long as it runs, such as deposits into an account or births into a population.

Contrast with: Outflow

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Influence Concentration

The outcome in which a small number of highly connected members of a network gather a disproportionate share of its total sway over opinion, information flow, or the division of resources.

Example: A handful of social media accounts with millions of followers shape which news stories most people see on a given day.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Influence Diagram

A type of causal graph that shows how individuals, groups, or decisions affect one another's thinking and choices, making the paths of persuasion and support easy to trace.

Example: A diagram might show how a CIO's public support shapes a project's funding, which shapes team morale, which shapes the project's results.

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Information Flow Structure

The pattern of who can see which facts, reports, and signals, when, and through which channels, a pattern that can change behavior even when rules and incentives stay the same.

Example: When a utility printed each household's electricity use next to the neighborhood average, many households cut their use without any change in price.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Information Pollution

The decline in the overall reliability of a shared environment of news, facts, and online content, caused by the buildup of low-quality, misleading, or false material.

See also: Misinformation, Digital Commons

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Initial Advantage

A starting edge in some resource or attribute, often small and often a matter of luck rather than merit, that a competitive process amplifies rather than corrects.

See also: Cumulative Advantage

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Initial Success

The genuine short-term improvement that a quick fix produces immediately after it is applied, which makes the fix look effective and encourages its repeated use.

See also: Quick Fix, Symptom Relief

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Input

Any matter, energy, information, money, or effort that enters a system from its environment and can be used or transformed by the system's internal processes.

Contrast with: Output

Covered in: Chapter 1: Foundations of Systems Thinking

Integrated Maturity Level

The fourth stage of this book's systems-thinking maturity scale, at which feedback loops, archetypes, and leverage points are built into strategic decisions and cross-functional work runs on a shared systems-thinking vocabulary.

Example: Every major proposal includes a short section naming the archetypes it could trigger and the leverage points it targets.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Interconnection

A relationship or link through which one part of a system affects another part, such as a flow of material, money, or information between them.

Example: The link between a store's sales data and its purchasing team matters: when that data arrives late, the store orders the wrong products.

Covered in: Chapter 1: Foundations of Systems Thinking

Interdependence

A condition in which two or more parts of a system rely on one another, so that a change in one part affects the ability of the others to function.

See also: Interconnection

Covered in: Chapter 1: Foundations of Systems Thinking

Interoperability

The ability of different systems, organizations, or products to exchange data and use it correctly, with both sides understanding its meaning in the same way.

Example: A patient's records move from one hospital's system to another's, and the second hospital reads every allergy and medication correctly.

Covered in: Chapter 18: Data Management and Governance

Intervention Point

The specific place within a system where a change is actually introduced, such as a particular screen, policy, or process step.

It applies the idea of a leverage point at the practical level of deciding what to touch first.

Example: A school chooses to change its homework policy first rather than its grading scale.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

ISO Definition

A statement of a term's meaning written to follow the ISO/IEC 11179 metadata registry guidelines, making it precise, concise, distinct from other entries, non-circular, and free of context-specific business rules.

These guidelines, which come from the international committee for metadata registry standards, are the ones used for every entry in this glossary.

Example: "A closed chain of cause and effect that counters change" avoids the word "balancing" when defining a balancing loop, so it is non-circular.

See also: Metadata Registry, Business Glossary

Covered in: Chapter 17: Knowledge Representation and Metadata

Job Recommendation Platform

An online service that matches people looking for work with job postings, or suggests candidates to employers, often by studying which past candidates were hired.

If it copies past hiring patterns, it can keep showing the same kinds of people the same kinds of jobs, narrowing opportunity instead of widening it.

See also: Algorithmic Bias, Recommendation Engine

Covered in: Chapter 23: AI Systems Dynamics

JOIN Fear Modeling

The defensive habit among database designers of avoiding designs and queries that require combining many tables at once, often by merging tables or simplifying the model, because such queries are known to run slowly.

Example: A team squeezes its product data into one wide table and loses the ability to record how parts relate to one another.

Covered in: Chapter 16: Graph Database Architecture

Knowledge Base

An organized store of facts, rules, and relationships about a subject area, kept in a form that people or computer programs can search and reason over.

Example: A help desk's collection of known problems, their causes, and their fixes, linked so that a support agent can find the right answer quickly.

Covered in: Chapter 17: Knowledge Representation and Metadata

Knowledge Base Reasoning

The process of deriving new facts or answering questions by applying logical rules to the facts and relationships held in an organized store of information.

Example: Because the store records that Ana manages Ben and Ben manages Cal, the system concludes that Cal is in Ana's reporting chain.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Knowledge Decay

The loss of a skill's accuracy or usefulness over time, either because the person holding it stops practicing or because the facts it was based on change.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Knowledge Embodiment

The process by which abstract know-how becomes built into a physical product, a piece of software, or a documented process, so that people who do not personally hold it can still use it.

Example: A jet engine carries decades of engineering know-how that the pilots who fly it never need to learn.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Knowledge Graph Adaptability

The capacity of an enterprise knowledge graph to take on new kinds of entities, attributes, and relationships as needs change, without the costly restructuring a rigid table design would require.

Example: Adding a new "Supplier Risk" category next quarter means adding new vertices and edges, not rebuilding existing tables.

Covered in: Chapter 19: Enterprise Knowledge Graphs

Knowledge Path Dependence

The tendency of an organization to keep choosing the tools and methods it already knows well, because its accumulated expertise and documentation make them the easiest option, even after better choices exist.

Example: A company with years of experience on one database product keeps choosing it for every new project, even for data it handles poorly.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Knowledge Representation

The field of artificial intelligence devoted to expressing information about the world in a form that a computer system can use to reason and solve complex tasks.

Example: Encoding facts such as "every employee works in exactly one department" so a program can answer questions and detect errors on its own.

Covered in: Chapter 17: Knowledge Representation and Metadata

Knowledge Spillover

The process by which know-how created by one organization becomes available to others without full payment to the original creator.

Example: A competitor hires away a trained engineer and gains skills that the first company spent years developing.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Knowledge Vs Information

The distinction between facts that describe what is the case and the experience-based understanding that tells a person what to do about those facts in a new situation.

Example: A forecast of heavy rain is a fact; realizing that you should leave early and avoid the low road that floods is the understanding built from experience.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Labeled Property Graph

A property graph in which each vertex carries one or more type names, such as Person or Company, and each edge carries a relationship type, making it easy to find all items of a given category.

Example: Vertices labeled Person and Company are joined by an edge of type WORKS_AT, which itself stores a start date.

See also: Cypher, GQL

Covered in: Chapter 15: Graph Theory Fundamentals

Large Language Model

A neural network with billions of internal settings, trained on huge amounts of written text to predict the next word or word-piece in a sequence, which lets it generate, summarize, translate, and answer questions in ordinary human writing.

Example: Chat assistants that draft emails, explain homework problems, or write computer code are built on these models.

See also: Generative AI, AI Hallucination

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Law Of Space

The principle that economic worth built on know-how tends to concentrate geographically, because such know-how stays rooted in particular places and communities.

Example: Fine watchmaking has stayed concentrated in a few regions of Switzerland for centuries.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Law Of Time

The principle that the economic worth of a piece of know-how tends to decline as it becomes more widely known and built into competing products.

Example: The first company to sell a touchscreen smartphone enjoyed a large technical lead that faded as competitors learned to build similar phones.

See also: Law Of Space, Law Of Value, Market Renewal

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Law Of Value

The principle that an economic actor captures worth not simply by possessing know-how but by building it into a product or service that a market will pay a premium for.

Example: Raw, unbranded coffee beans sell for a small fraction of the price of the same beans sold as a branded specialty roast.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Learned Helplessness

A state in which people who have repeatedly failed to change their situation stop trying, even when conditions change and success has become possible.

See also: Goal Erosion

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Learning Curve

The pattern in which the time and effort an individual worker or team needs for a specific repeated task fall with practice, quickly at first and then more slowly.

Example: The tenth time an assembly worker installs a part, the job takes noticeably less time than the first.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Learning Organization

A company or institution, as described by Peter Senge, that continually expands its ability to create the results it wants by building shared vision, open reflection on mental models, group reflection and skill-building, and systems thinking.

Example: A software company holds blame-free reviews after every outage and shares the lessons with all teams so the same mistake is not repeated.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Legacy System

An older computer system or application that remains in use because important work depends on it, even though it is outdated, costly to maintain, or hard to connect with newer systems.

See also: Technical Debt, Path Dependence

Covered in: Chapter 20: Organizational Silos and Silo Busting

Level Confusion Mistake

The error of treating an intervention aimed at one depth of the leverage points hierarchy as though it worked at another depth, such as expecting a new mission statement to change daily decisions the way a new rule would.

Example: A company announces a "customer first" value but changes nothing about how its sales staff are paid or promoted.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Leverage Point

A place within a system where a relatively small, well-chosen change can produce a large and lasting shift in the system's overall behavior.

Systems scientist Donella Meadows ranked twelve kinds of these, from adjusting numbers, the weakest, to changing paradigms, the strongest.

Example: Showing each household its energy use next to its neighbors' average use reduced consumption far more than a small price increase.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Leverage Points Hierarchy

This book's grouping of Donella Meadows' twelve ranked places to intervene in a system into four bands, shallow, structural, deep, and transformative, ordered by increasing power and increasing difficulty.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Leverage Points Iceberg

A diagram that arranges Donella Meadows' ranked list of places to intervene by depth below a waterline, with the easiest, weakest options near the surface and the deepest, most powerful ones far below.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Limiting Factor

The one constraint that is in shortest supply relative to need and therefore sets the upper bound on a system's performance or growth at that moment.

Once it is relieved, a different constraint usually takes its place as the new ceiling.

Example: For decades, slow JOIN performance in relational databases held back efforts to build detailed, highly connected models of the real world inside a database.

See also: Constraint, Limits To Growth Archetype

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Limits To Growth Archetype

A systems archetype in which a reinforcing loop of expansion eventually runs into a balancing loop created by some constraint, so that expansion slows, stops, or reverses.

Example: A popular new restaurant grows quickly until long waits and a crowded kitchen drive customers away.

See also: Limiting Factor, S-Curve, Carrying Capacity

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Limits to Growth examples

Linear Growth

An increase in which a quantity rises by the same fixed amount in each time period, so that its graph against time has a constant slope.

Example: Saving $20 every week grows a jar of cash to $20, $40, $60, and $80 over four weeks.

Contrast with: Exponential Growth

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Linear Maturity Level

The first stage of this book's systems-thinking maturity scale, at which an organization makes decisions using direct, short-term cause-and-effect reasoning with little awareness of feedback loops or side effects.

Example: "Sales are down, so cut prices" is decided without asking how competitors or profit margins will respond.

See also: Linear Thinking, Aware Maturity Level

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Linear Relationship

A connection between two variables in which a change in one always produces a proportional change in the other, so that plotting them together gives a constant slope.

Contrast with: Nonlinear

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Linear Thinking

A way of reasoning that assumes each cause leads directly to one proportional effect in a straight chain, with no feedback, delays, or side effects.

Example: "If we hire more programmers, the project will finish sooner" ignores the time needed to train newcomers and the extra coordination a larger team requires.

Contrast with: Systems Thinking

Covered in: Chapter 1: Foundations of Systems Thinking

Link Prediction

A method for estimating which missing or future connections between vertices are most likely, based on the existing pattern of connections.

Example: A professional networking site suggests "people you may know" because you share many mutual contacts.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Linked Data

A method of publishing structured information on the web so that each item has its own unique web address and points to related items at other addresses, allowing computers to follow the connections.

Example: A library's online record for a book points to the web addresses that identify its author and its publisher in other collections.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Local Interaction

An exchange or influence between neighboring agents in a system that is based only on the information each one can directly sense, without knowledge of the whole system.

See also: Global Pattern, Bottom-Up Behavior

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Logical Data Model

A detailed description of the entities, attributes, keys, and relationships needed to meet business requirements, independent of any particular database product.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Loop Dominance

The condition in which one feedback loop has the strongest influence on a system at a given time, so that the system's overall behavior follows that loop's pattern.

See also: Shifting Dominance

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Loop Identification

The process of finding each closed path of causal links in a diagram and determining whether that path amplifies change or counteracts it.

Real diagrams often contain several overlapping loops, and the system's behavior depends on which one is strongest at a given time.

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Loop Marker

A small symbol placed at the center of a closed circuit of links on a causal loop diagram to record its polarity, so that readers do not need to recheck its links.

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Loop Polarity

The classification of a closed chain of causal links as either reinforcing or balancing, based on whether a change traveling around the chain is amplified or counteracted.

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Low-Leverage Intervention

A proposed change that works on numeric settings or on physical stocks and flows, usually easy to carry out but limited in how much lasting effect it has.

Contrast with: High-Leverage Intervention

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Low-Resolution Data Model

A data model that stores only summarized values, such as monthly totals by region, rather than individual records, so it answers broad reporting questions quickly but cannot answer detailed ones.

Example: A sales table with one row per region per month can show total spring sales but cannot say which customers bought what.

Covered in: Chapter 18: Data Management and Governance

Machine Learning

A branch of artificial intelligence in which computer programs improve their performance at a task by finding patterns in example data, rather than by following rules written step by step by a programmer.

Example: An email filter improves at spotting spam by studying millions of messages that people have already marked as spam or not spam.

See also: Training Data, Neural Network

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Market Equilibrium

The state in which the quantity of a good that buyers want at the current price equals the quantity that sellers offer, so the price has no tendency to change.

Example: When too many strawberries arrive at a farmers' market, sellers cut prices until buyers purchase the rest.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Market Renewal

The ongoing process by which new products, technologies, and business models replace aging ones, driven by the spread of know-how that wears away the edge of existing leaders.

See also: Creative Destruction, Law Of Time

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Market System

A way of organizing trade in which prices set by the voluntary exchanges of many buyers and sellers guide what is produced, how much, and for whom.

Example: Rising coffee prices lead farmers to plant more coffee trees, which a few years later pushes the price back down.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Master Data

The core, shared information about an organization's key business entities, such as customers, products, suppliers, employees, and locations, that many processes and systems depend on.

See also: Reference Data, Master Data Management

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Master Data Management

The set of processes, tools, and governance an organization uses to create and maintain one consistent, accurate version of its core shared data, such as customers, products, and suppliers.

See also: Master Data, Golden Record, Entity Resolution

Covered in: Chapter 18: Data Management and Governance

Material Stocks And Flows

The physical capacities and structures of a system, such as warehouse space, road networks, or a reservoir's volume, considered as places where an intervention can be made.

Example: Building a bigger warehouse adds buffer space but does not change the ordering rules that keep filling it with the wrong products.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Matrix Organization

An organizational structure in which employees report to two managers at once, typically one for their functional specialty and one for a project or product line.

Example: A software tester reports to the head of quality assurance and also to the manager of the mobile app project.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Matthew Effect

The pattern, named by sociologist Robert Merton after a verse in the Gospel of Matthew, in which people or groups who already have status or resources receive still more, while those with little fall further behind.

Example: Well-known scientists often receive more credit for a shared discovery than lesser-known coauthors who did equal work.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Maturity Assessment

An evaluation that compares an organization's current practices against the ordered stages of a capability framework to determine which stage it has reached and what it needs to advance.

Example: A survey and interviews show that a company maps feedback loops for large projects but not for everyday decisions.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Maturity Level

One stage in an ordered scale of an organization's capability, defined by the set of practices the organization performs routinely and reliably at that stage.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Measurement Obsession Mistake

The error of fixating on whichever single number is easiest to track as proof that a deep change is working, while the underlying mindset or structure stays the same.

Example: A company meets every target attached to its new "stakeholder value" goal while managers still make decisions exactly as before.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Measurement Trap

The situation in which a system starts managing to a stand-in number instead of the real goal the number was meant to approximate, so the number and the goal drift apart.

Example: A hospital cuts reported waiting times by moving patients into hallway beds sooner, while actual care does not improve.

See also: Proxy Metric, Campbell's Law, Goodhart's Law

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Medium-Leverage Intervention

A proposed change that works on the strength of feedback loops or on the structure of information flows, requiring more effort than adjusting numbers but producing more durable effects.

See also: Structural Leverage Point

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Mental Model

A person's internal picture of how something works, built from experience and beliefs, that shapes what the person notices, expects, and decides.

These pictures are powerful because they are usually invisible to the people who hold them.

Example: A manager who believes "people only work hard when watched" designs very different policies than one who believes "people want to do good work."

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Mental Models Layer

The deepest level of the iceberg model, made up of the beliefs, assumptions, and values that lead people to create and keep the structures above them.

Example: "Leaders see spending on servers as a cost to cut rather than an investment in reliability."

See also: Mental Model, Mindset Or Paradigm

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Metadata

Descriptive information about a set of data, such as its meaning, source, format, owner, creation date, or permitted uses, that helps people and systems find and understand it.

Example: A photo file's metadata records the date it was taken, the camera model, and the location.

Covered in: Chapter 17: Knowledge Representation and Metadata

Metadata Registry

A system that stores, manages, and provides access to agreed definitions and standards for data elements, supporting a shared understanding of data across different systems and organizations.

See also: ISO Definition, Metadata

Covered in: Chapter 17: Knowledge Representation and Metadata

Metcalfe's Law

The principle, attributed to Ethernet co-inventor Robert Metcalfe, that the value of a communications network grows roughly with the square of the number of its connected users, because possible connections grow that fast.

Example: A network of 10 phones allows 45 possible pairs of callers, while a network of 100 phones allows 4,950.

See also: Network Effect, Quadratic Scaling

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Microstrategy

A small-scale, low-risk move or pilot, such as one team's experiment, undertaken to test whether a new practice works before committing the whole organization to it.

Example: One department tries drawing causal loop diagrams before its planning meetings for three months before anyone proposes a company-wide rollout.

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Mindset Or Paradigm

The shared, often unspoken set of beliefs and assumptions about how the world works from which a system's goals, rules, and structures arise.

Example: The belief that "each department owns its own data" shapes budgets, software purchases, and job descriptions across an entire company.

See also: Paradigm Shift, Mental Models Layer

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Misaligned Incentive

A reward or penalty that encourages people in one part of a system to act in ways that work against the goals of the larger system.

Example: Paying a help desk by the number of tickets closed encourages staff to close tickets quickly rather than solve problems fully.

See also: Incentive Structure, Suboptimization

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Misinformation

False or inaccurate claims or content that are spread from person to person regardless of whether the person sharing them intends to deceive, often because they are surprising or emotionally appealing.

See also: Information Pollution, Echo Chamber

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Mob Behavior

A rapid, self-amplifying pile-on in which many individuals, each reacting to the visible reactions of others, together produce an outcome far more extreme than any one of them intended.

Example: A single critical post about a small business leads thousands of strangers to flood it with angry reviews within hours.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Model Assumptions

The beliefs, simplifications, and starting values that a model builder accepts as true in order to build a model, whether they are stated openly or left unstated.

See also: Model Limitations, Sensitivity Analysis

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Model Drift

The gradual decline in a deployed machine learning model's accuracy, caused by changes in the real world that make current data differ from the data the model was trained on.

Example: A shopping-prediction model trained before a pandemic performs poorly once people suddenly start buying most things online.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Model Limitations

The conditions, questions, or time spans for which a model's results are not reliable because of what the model leaves out or simplifies.

Knowing these limits prevents people from trusting a model's answers in situations it was never built to handle.

Example: A traffic model built for weekday commutes cannot predict the crowds on the night of a large stadium concert.

See also: Model Assumptions, Model Validation

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Model Training

The process of repeatedly showing examples to a machine learning model and adjusting its internal settings to shrink the difference between its outputs and the correct answers.

See also: Training Data, Overfitting

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Model Validation

The process of checking whether a model's structure and results match real-world data and observations closely enough to be trusted for its intended purpose.

See also: Model Limitations, Scenario Testing

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Moore's Law

The observation, made by Intel co-founder Gordon Moore in 1965, that the number of transistors on a computer chip doubles about every two years, making computing power steadily cheaper.

It is a famous case of sustained exponential growth that is now slowing as it reaches physical limits.

Example: A smartphone today has more computing power than the computers that guided the Apollo missions to the Moon.

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Multi-Agent System

A computer system made up of several independent software programs, each able to sense its surroundings, make decisions, and act on its own, that interact to reach individual or shared goals.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Multi-Stakeholder System Design

An approach to creating a system that deliberately serves several groups with different, sometimes conflicting interests at the same time, rather than optimizing for a single type of user.

Example: A scheduling system that delights patients but buries billing staff in manual work has solved only half of the problem.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Native Graph Database

A graph database whose storage engine records each vertex's connections as direct references to neighboring vertices, instead of rebuilding those connections through index lookups each time a query runs.

Example: Asking for a person's friends of friends of friends stays fast even as the database grows to millions of people.

See also: Pointer Hopping, Adjacency

Covered in: Chapter 16: Graph Database Architecture

Negative Causal Link

A link, marked with a minus sign, in which a change in one variable produces a change in the variable it influences in the reverse direction, all else being equal.

Example: More exercise tends to lower resting blood pressure, and less exercise tends to raise it.

Contrast with: Positive Causal Link

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Negative Externality

A cost or harm imposed on third parties who are not directly involved in an economic transaction or decision and who receive no compensation for it.

Because the decision-maker does not pay the full social cost, these hidden costs drive the Tragedy of the Commons by splitting individual incentives from collective outcomes.

Example: A fishing company that catches more fish in a shared lake keeps all the extra revenue, while the cost of a shrinking fish population is spread among every other fisher and future generations.

Contrast with: Positive Externality

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Negative Feedback

Returned information or influence that pushes a system in the opposite direction from a change, shrinking the change and keeping the system near a target.

Contrast with: Positive Feedback

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Negative Feedback Strength

The degree of force with which a balancing loop corrects a departure from its target, treated as a setting that can be deliberately raised or lowered.

Example: A thermostat that reacts to a half-degree change keeps a room far more comfortable than one that waits for a three-degree change.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Negative-Link Counting Rule

A shortcut for finding a loop's polarity from the number of minus-sign links met in one full trip around it: an even number, including zero, marks a reinforcing loop, and an odd number marks a balancing loop.

Example: A thermostat loop contains exactly one minus-sign link, where a warmer room lowers heater output, so the loop is balancing.

See also: Loop Polarity, Loop Identification

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Network Effect

The phenomenon in which a product or service becomes more valuable to each user as more people use it, which creates a reinforcing loop of growth.

Example: A messaging app is useless if none of your friends have it and very useful once all of them do, so each new user attracts more.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Network Externality

The economic term for the benefit, or occasionally the cost, that one person's joining or using a shared service creates for its other users, which the new member neither pays for nor receives.

See also: Network Effect, Positive Externality

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Network Topology

The overall pattern of connections among a set of linked members, such as a chain, a star, a mesh, or a structure dominated by a few hubs, which strongly shapes how things spread among them.

Example: A rumor spreads much faster through a school where a few popular students know everyone than through one made of small, separate friend groups.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Network Trust

The tendency for belief in a claim to spread more readily along existing social or professional connections than across unconnected groups, so that a well-connected person's claims reach and persuade far more people.

See also: Influence Concentration, Misinformation

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Neural Network

A computing model made of many layers of simple connected units, loosely inspired by brain cells, whose connection strengths are adjusted during training so that the model can recognize patterns in data.

Example: A network trained on thousands of labeled photos can tell whether a new photo shows a cat or a dog.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Node

A labeled element in a causal loop diagram that stands for a single variable, meaning a quantity that can meaningfully increase or decrease over time.

See also: Edge (CLD), Variable (CLD), Vertex

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Nonlinear

Describing a relationship in which a change in one component does not produce a proportional change in another, so that a small change may cause a very large effect, or a large change almost none.

Human intuition expects proportional results, so people need practice to anticipate how these relationships play out over the long term.

Example: In ecology, a small rise in atmospheric carbon dioxide can shift global temperature far more than expected; in business, a small drop in product quality can trigger bad reviews, fewer recommendations, and a steep fall in sales.

Contrast with: Linear Relationship

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Nonlinear Behavior

The actual patterns of change a system displays when its underlying relationships are not proportional, including sudden acceleration, slowing, reversal, or a jump to an entirely new pattern.

See also: Nonlinear, S-Curve, Tipping Point

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Nonlinear Feedback

A loop of influence whose strength is not fixed but changes with the current value of the variables involved, so that the loop can grow stronger, weaker, or even reverse direction as conditions change.

Example: On an empty highway one more car barely slows traffic, but on a crowded highway one more car can trigger a sudden jam.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Normalized Data Model

A relational data design that divides data into many related tables so that each fact is stored in only one place, faithfully capturing complex real-world relationships without duplication.

When such a model captures the real world accurately, many business units can share it, but highly normalized designs need many joins and did not scale well on older relational technology.

Example: A customer's address is stored once in a Customers table rather than repeated on every order.

Contrast with: Denormalized Data Model

Covered in: Chapter 18: Data Management and Governance

Ontology

A formal, machine-readable specification of the kinds of things in a subject area, their properties, and the relationships and rules that connect them, used to give data a shared meaning.

Example: A healthcare model might state that a Physician is a kind of Person who can treat a Patient and prescribe a Medication.

See also: Taxonomy, Semantic Web

Covered in: Chapter 15: Graph Theory Fundamentals

Open Standard

A communication standard whose full specification is published for anyone to read and implement without special permission or fees, allowing products from many different makers to work together.

Example: Because the rules for email are openly published, a message sent from one company's mail service arrives correctly at any other provider.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Open System

A system that exchanges matter, energy, or information with its environment across its boundary, so that outside conditions can reshape its behavior.

Nearly every real-world system, from a living cell to a company, is of this kind.

Example: A city takes in food, water, workers, and data every day and sends out goods, trash, and commuters.

Contrast with: Closed System

Covered in: Chapter 1: Foundations of Systems Thinking

Opposite-Direction Link

A plain-language name for a causal relationship in which the effect moves the reverse way from the cause, so an increase produces a decrease and a decrease produces an increase.

See also: Negative Causal Link

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Organizational Culture

The shared values, beliefs, habits, and unwritten rules that shape how people in an organization behave, communicate, and make decisions.

Example: In some workplaces admitting a mistake is praised as learning; in others it is quietly punished.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Organizational Maturity

The degree to which an organization's practices are consistent, documented, measured, and continuously improved across the whole organization, rather than depending on a few talented individuals.

See also: Maturity Level

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Organizational Silo

A department, team, or business unit that operates in isolation from the rest of its organization, keeping its own data, goals, tools, and vocabulary and sharing little with others.

Example: Marketing and customer support each keep separate customer lists, so neither knows a customer who just complained is about to receive a sales promotion.

See also: Data Silo, Silo Busting

Covered in: Chapter 20: Organizational Silos and Silo Busting

Organizational Structure

The formal arrangement of roles, reporting lines, departments, and decision-making authority within an organization, which determines who works with whom and who can decide what.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Oscillation

A repeated rising and falling of a variable around a target or average level, typically caused by a balancing loop that contains a significant delay.

Example: A shower that responds slowly to the faucet makes people overcorrect, swinging between too hot and too cold.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Outcome

The variable in a diagram or model whose behavior over time the analyst most wants to observe, explain, or predict.

See also: Variable (CLD), Condition

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Outflow

A rate of movement that drains a stock, decreasing the amount stored for as long as it runs, such as withdrawals from an account or deaths in a population.

Contrast with: Inflow

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Output

Any product, service, waste, energy, or information that a system releases into its environment as a result of its internal processes.

Contrast with: Input

Covered in: Chapter 1: Foundations of Systems Thinking

Overfitting

A modeling error in which a model matches its training examples so closely, including their random noise, that it performs poorly on new data it has not seen before.

Example: A student who memorizes the answers to one practice test does well on it but poorly on a real test with different questions.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Overshoot

The condition in which a variable moves past its target or limit because the corrective response arrives too late or too strongly.

See also: Oscillation, Overshoot And Collapse

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Overshoot And Collapse

A pattern of change in which a growing system exceeds the carrying capacity of its environment, damages the resources it depends on, and then declines sharply.

Example: Deer on an island with no predators multiply, overgraze the vegetation, and then die off in large numbers when the food runs out.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Paradigm Shift

A fundamental change in the basic assumptions and worldview a community uses to understand a problem, after which old questions, methods, and solutions look very different.

Example: Astronomy moved from an Earth-centered view of the universe to a Sun-centered one; organizations make a similar change when they stop treating data as each department's property and start treating it as a shared asset.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Parameter Change

The concrete act of adjusting a numeric setting within a system, such as a price, budget, or threshold, which is usually the first and shallowest intervention people attempt.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Path Dependence

The tendency for early choices to limit later options, so that a system's present state is shaped heavily by the particular sequence of past decisions rather than only by current conditions.

Example: The QWERTY keyboard layout was designed for mechanical typewriters that jammed, yet it survives on touchscreens because switching now costs too much.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Patterns Layer

The second level of the iceberg model, made up of trends and repeated behaviors that become visible only when individual events are tracked over time.

Example: "The payroll system fails at the end of every quarter."

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Personal Resilience

An individual's capacity to adapt to and recover from setbacks, stress, or disruption, built from buffers, support networks, and flexible habits in much the same way an organization builds robustness.

Example: A worker with some savings, a network of friends, and several skills recovers from a layoff faster than one without them.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Personal Thinking Toolkit

A practical, portable set of diagnostic questions a person carries into everyday decisions to apply systems ideas without needing a whiteboard or formal diagram.

Example: "What loop am I in?" "What happens after the first effect?" "What small change could I test first?"

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Phase Transition

An abrupt change in a system from one overall state or organization to a very different one when an underlying condition crosses a critical value.

Example: Water turns to ice at 0 degrees Celsius; in the same way, a quiet online forum can shift almost at once into a heated, hostile space once enough angry posters join.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Physical Data Model

The specific implementation of a design in a particular database product, including actual table names, column types, indexes, and storage settings.

Example: A table named ORD_HDR with a date column named ORD_DT and an index on the customer ID column.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Physical Limits

Constraints on growth imposed directly by finite material resources or by the laws of nature, which cannot be overcome with more money or effort.

See also: Economic Limits, Constraint

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Platform Dynamics

The broad set of feedback effects that shape how a service connecting many users, buyers, or sellers grows, competes, and sometimes comes to dominate a market.

Example: A ride-sharing app attracts drivers because it has riders and riders because it has drivers, which makes it hard for a newcomer to compete.

Covered in: Chapter 23: AI Systems Dynamics

Point-To-Point Integration

A way of connecting systems in which each pair that needs to share data is joined directly by its own custom link, so the number of links grows rapidly as systems are added.

Example: Ten systems that each need to talk to all the others require up to 45 separate custom links.

Contrast with: Hub-And-Spoke Architecture

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Pointer Hopping

The technique of moving from one vertex to a neighboring vertex by following a stored direct memory reference, rather than searching an index, so each step takes about the same small amount of time.

Example: Moving from a customer to that customer's orders takes one quick step whether the database holds a thousand customers or a billion.

Covered in: Chapter 16: Graph Database Architecture

Policy Resistance

The tendency of a system to counteract, weaken, or neutralize a well-intentioned intervention, so that the problem the intervention targeted returns despite the effort spent.

It usually happens because the intervention addressed one loop while leaving the loops that created the problem in place.

Example: A city adds highway lanes to reduce congestion, but new drivers fill them within a few years.

Covered in: Chapter 7: Feedback Resilience and Robustness

Political System

The set of institutions, rules, and processes through which a society makes collective decisions, distributes power, and resolves conflicts, such as elections, legislatures, courts, and parties.

See also: Public Policy Feedback

Covered in: Chapter 27: Systems Thinking Across Disciplines

Population Dynamics

The patterns by which the size and makeup of a group of organisms change over time because of births, deaths, and movement in and out of the group.

Example: A deer herd grows quickly after its predators disappear, then crashes when its food runs out.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Portfolio Assessment

An evaluation method that judges learning from a collection of a student's work gathered over time, showing growth and depth rather than a single score.

See also: Authentic Assessment

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Positive Causal Link

A link, marked with a plus sign, in which a change in one variable produces a change in the variable it influences in the same direction, all else being equal.

The plus sign describes direction only; it does not mean the result is good.

Example: More rainfall leads to more crop growth, and less rainfall leads to less growth.

Contrast with: Negative Causal Link

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Positive Externality

A benefit received by people who are not part of a transaction or decision and who did not pay for or ask for it.

Example: An engineer who carefully documents a tricky bug fix helps every future colleague who searches the codebase for the same problem.

Contrast with: Negative Externality

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Positive Feedback

Returned information or influence that pushes a system further in the direction it is already changing, amplifying the original change.

"Positive" here means amplifying, not good; this kind of feedback can drive runaway collapse as easily as healthy growth.

Example: A microphone held near a speaker picks up its own sound, which grows louder with each pass until it squeals.

Contrast with: Negative Feedback

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Positive Feedback Strength

The degree to which each cycle of a reinforcing loop amplifies the one before it, treated as a setting that can be deliberately raised or lowered.

Example: A referral program whose bonus grows with each friend a person has already referred speeds up far more than one with a flat reward.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Power To Transcend Paradigms

The ability to recognize that every worldview, including one's own favorite, is a constructed lens rather than literal truth, and to switch lenses deliberately when a situation calls for a different one.

Donella Meadows ranked this as the most powerful leverage point of all.

See also: Mindset Or Paradigm, Paradigm Shift

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Predator-Prey Dynamics

The linked rises and falls in the numbers of a hunting species and the species it feeds on, in which each group's size drives the other's with a delay.

Example: In Canada, lynx numbers rise after snowshoe hare numbers rise and fall after the hares decline, in cycles of about ten years.

See also: Oscillation, Population Dynamics

Covered in: Chapter 27: Systems Thinking Across Disciplines

Prediction Model

A mathematical or machine learning model that uses patterns found in past data to estimate an unknown or future value, such as a price, a risk, or a likely choice.

Example: A model estimates how many students will enroll next fall based on past enrollment, local birth rates, and application numbers.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Predictive Feedback Cycle

A reinforcing loop in which an enterprise knowledge graph's forecasts or recommendations produce outcomes that are recorded back into the graph, improving the data available for the next forecast.

Example: A model flags customers likely to cancel, and whether each one actually stays or leaves is added back to the graph to sharpen future predictions.

See also: AI Flywheel

Covered in: Chapter 19: Enterprise Knowledge Graphs

Primary Key

A column, or set of columns, whose value uniquely identifies each row in a relational database table, so that no two rows in that table can share the same value.

See also: Foreign Key

Covered in: Chapter 18: Data Management and Governance

Principle Of Relatedness

The observation that a country or region is much more likely to succeed at making a new product when that product requires capabilities close to ones its economy already has.

Example: A country that makes refrigerators can move into air conditioners far more easily than into medicines.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Product Space

A network map in which each exported good is a node and two goods are linked when the same countries tend to export both successfully, showing which goods draw on similar capabilities.

Dense regions of the map, such as machinery and electronics, offer many nearby next steps, while sparse regions, such as raw materials, offer few.

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Prompt Engineering

The practice of designing and refining the instructions, context, and examples given to a generative AI model to get more accurate, useful, or reliable outputs.

Example: Asking a chatbot to "explain feedback loops to a ninth grader in three short paragraphs with one school example" instead of just "explain feedback loops."

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Property Graph

A graph data model in which both vertices and edges can store key-value pairs describing them, such as a person's name or the date a relationship began.

Example: An edge linking an employee to a company can record that the job began in 2020 and is full-time.

Covered in: Chapter 15: Graph Theory Fundamentals

Protocol Governance

The process and institutions that decide how a shared set of technical communication rules changes over time and who has a voice in those decisions, balancing the stability users depend on against needed improvements.

Example: Web standards bodies gather browser makers, researchers, and companies to agree on new features before they are added to the rules every browser follows.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Proxy Metric

A measurable quantity used to stand in for a goal that is harder to measure directly, on the assumption that the two move together.

Example: A company uses "hours logged in" as a stand-in for employee productivity, even though time online does not equal useful work.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Public Policy Feedback

The loop connecting a government decision's actual implementation to its real-world outcomes, which then shape the next round of decisions, often after a delay of several years.

Example: A tax credit for electric cars raises sales, and the resulting sales data leads lawmakers to extend or change the credit years later.

See also: Delay, Political System

Covered in: Chapter 27: Systems Thinking Across Disciplines

Quadratic Scaling

A growth pattern in which a quantity increases in proportion to the square of some input, so that doubling the input roughly quadruples the result.

Example: Connecting every pair of 10 systems needs 45 links, but connecting every pair of 20 systems needs 190 links.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Quantum Computing Claims

Public statements about the abilities of computers built on the physics of subatomic particles that often run ahead of the machines' independently demonstrated performance on real-world problems.

A systems thinker's task is to separate reproducible benchmark results from promotional forecasts, not to dismiss the technology.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Quick Fix

An intervention that relieves the visible sign of a problem rapidly, usually with little cost or effort, but without addressing whatever is actually generating the problem.

Example: Restarting a crashed server every morning keeps an app running without anyone finding the memory leak behind the crashes.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Rate Of Change

The speed at which a quantity increases or decreases per unit of time, found for a stock by subtracting its total outflow from its total inflow.

Example: A town that gains 300 residents a year and loses 200 is growing by 100 residents per year.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Rational Self-Interest

The assumption or practice of each individual making choices that bring the most benefit to themselves, which can produce poor results for a group when a resource is shared.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

RDBMS JOIN

An operation in a relational database that combines rows from two or more tables by matching values in related columns, usually a foreign key in one table to a primary key in another.

Each added join makes a query more expensive, so questions that follow many relationships can become very slow.

Example: Listing each order with its customer's name requires joining the Orders table to the Customers table on the customer ID.

See also: Primary Key, Foreign Key, JOIN Fear Modeling

Covered in: Chapter 16: Graph Database Architecture

RDF

The Resource Description Framework, a World Wide Web Consortium standard that represents information as subject-predicate-object statements, called triples, with each item identified by a web address.

Example: The statement "Ada works at Acme" is stored as the triple (Ada, worksAt, Acme).

Covered in: Chapter 15: Graph Theory Fundamentals

Recommendation Engine

Software that suggests products, content, or connections to a user based on that user's past behavior, the behavior of similar users, and the relationships among items.

Example: An online bookstore suggests a new mystery novel because readers who bought your last three books also bought it.

See also: Recommendation System Refinement

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

A reinforcing loop in which more viewing, buying, or listening activity produces more data about what users want, which trains a better suggestion model, which generates still more activity to learn from.

Example: A streaming service that suggests shows based on what you finished watching keeps you watching, which gives it even more to learn from.

Covered in: Chapter 23: AI Systems Dynamics

Reductionism

An approach to understanding that breaks a whole into its smallest parts and studies each part separately, on the assumption that the whole can be explained entirely from those parts.

It works well for machines with fixed designs but can miss behavior that appears only when parts interact.

Example: Studying each organ separately tells a doctor a great deal, but not how stress, sleep, and diet combine to affect a patient's overall health.

Contrast with: Holism

Covered in: Chapter 1: Foundations of Systems Thinking

Reference Data

A fixed or slowly changing set of permitted values used to categorize or classify other information, such as country codes, currency codes, units of measure, or product categories.

See also: Master Data, Data Standards

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Reification Challenges

The practical difficulties that arise in triple-based graph models when describing a property of a relationship itself, which requires turning that relationship into an extra node with its own statements.

Example: Recording that "Ada has worked at Acme since 2020" takes several extra triples in RDF, but only one property on an edge in a property graph.

See also: RDF, Property Graph

Covered in: Chapter 15: Graph Theory Fundamentals

Reinforcement Learning

A training approach in which a software agent improves its decisions through trial and error, receiving rewards or penalties from its environment for the actions it takes.

Example: A program becomes skilled at a video game by playing millions of rounds and keeping the moves that raise its score.

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Reinforcing Loop

A closed chain of cause and effect in which a change in any variable travels around and returns to push that same variable further in the same direction, producing growth or decline that speeds up.

It is also called a positive feedback loop and is the engine behind both virtuous and vicious cycles.

Example: More users make a social app more valuable, which attracts still more users.

Contrast with: Balancing Loop

See also: Vicious Cycle, Virtuous Cycle

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Reinforcing Loop Label

The capital letter "R," often drawn inside a small circular arrow, that is placed on a causal loop diagram to mark a closed circuit whose changes amplify themselves.

See also: Loop Marker, Reinforcing Loop

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Relational Database

A database that stores data in tables of rows and columns and connects records in different tables through matching identifier values, usually searched and updated with the SQL language.

Example: A store keeps one table of customers and another of orders, linked by a customer ID number.

Covered in: Chapter 18: Data Management and Governance

Relationship (Data Modeling)

A meaningful association between two entities that is recorded in a data model, such as a customer placing an order or an employee belonging to a department.

See also: Entity, Edge (Graph)

Covered in: Chapter 17: Knowledge Representation and Metadata

Repeatable Maturity Level

A stage in a capability maturity model at which a process has been carried out successfully more than once and can reasonably be expected to succeed again, though it may not yet be written down.

See also: Defined Maturity Level

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Resilience

The ability of a system to keep functioning in a stable way, or to recover its function, when conditions change or disruptions occur.

Example: A power grid with many small local generators keeps most lights on when one plant fails.

Covered in: Chapter 7: Feedback Resilience and Robustness

Resistance To Leverage Change

The tendency of a system, and especially of people whose position depends on its current arrangement, to push back against an intervention in proportion to how deeply that intervention reaches into the system.

Example: A proposal to change a company's goal from shareholder value to stakeholder value meets far more opposition than a proposal to change a price.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Resource Allocation

The general process of deciding how a limited supply of money, time, people, attention, or materials is divided among competing uses or claims.

See also: Resource Concentration

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Resource Concentration

The outcome that results when a process of dividing up money, time, attention, or materials keeps routing more of them toward whichever claim already holds the most, so they gather in fewer and fewer hands.

Example: A video platform recommends the videos that already have the most views, so a few videos collect most of the audience.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Resource Depletion

The using up of a natural or shared supply faster than it can be replaced, until what remains is too small to support the activities that depend on it.

Example: Overfishing of Atlantic cod in the late twentieth century shrank the population so much that the fishery off Newfoundland was closed.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Reverse Tragedy Of The Commons

A situation in which a shared resource becomes more valuable, rather than less, as more people use it and voluntarily contribute to it.

Example: Wikipedia's quality has generally improved as its number of volunteer editors has grown.

Contrast with: Tragedy Of The Commons Archetype

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Rich Picture

An informal, hand-drawn diagram that uses sketches, symbols, and short notes to capture the people, concerns, conflicts, and relationships in a messy real-world situation.

Example: Before redesigning how patients leave a hospital, staff sketch patients, nurses, pharmacists, and insurers, adding speech bubbles that show each group's frustrations.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Ripple Effect

A spreading series of consequences in which one change affects nearby parts of a system, which then affect parts that are farther away.

Example: When a large employer closes a factory, local restaurants, schools, and home prices all feel the impact over the following years.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Robustness

The ability of a system to keep performing normally across a known range of stresses and variations without needing to change its structure.

It differs from resilience, which also covers recovering from and adapting to shocks that were not planned for.

See also: Resilience, Fragility

Covered in: Chapter 7: Feedback Resilience and Robustness

Root Cause

The deepest underlying condition that, if changed, would stop a problem from coming back, as opposed to the surface events through which the problem shows up.

Example: A website crashes on busy days; the underlying condition is not heavy traffic itself but a database design that was never built to handle more than a few hundred users at once.

Contrast with: Symptom

Covered in: Chapter 1: Foundations of Systems Thinking

Root Cause Analysis

A structured method for tracing a problem backward through its chain of contributing factors in order to find the underlying conditions that keep producing it.

Example: After a data breach, a team traces events backward and discovers that the real problem was an unclear rule about who reviews access permissions.

See also: Five Whys, Fishbone Diagram, Root Cause

Covered in: Chapter 1: Foundations of Systems Thinking

Root Cause Solution

An intervention that targets the actual underlying condition generating a problem, rather than the visible sign of the problem, so that the conditions producing the trouble are changed.

See also: Root Cause, Fundamental Solution

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Rules Of The System

The formal and informal policies, laws, incentives, and punishments that define which actions are allowed, rewarded, or forbidden within a system.

Example: A company that bases bonuses on team results rather than individual results changes how employees share information.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

S-Curve

A pattern of growth that starts slowly, speeds up rapidly, and then levels off as it approaches a limit, producing a line shaped like a stretched letter S.

It results from a reinforcing loop that dominates early and a balancing loop that takes over later.

Example: Smartphone ownership grew slowly at first, then very rapidly, and then leveled off once most people already owned one.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Same-Direction Link

A plain-language name for a causal relationship in which the effect moves in parallel with the cause, rising when it rises and falling when it falls, used to avoid confusing a plus sign with a good outcome.

Contrast with: Opposite-Direction Link

Covered in: Chapter 3: Causal Loop Diagram Notation and Loop Identification

Scale Out

The strategy of increasing a system's capacity by adding more machines that share the work, rather than replacing one machine with a larger and more powerful one.

For a graph database, this means growing its capability by adding new servers to its cluster.

See also: Distributed Graph Database

Covered in: Chapter 16: Graph Database Architecture

Scale-Free Network

A web of connections in which a few hubs have an enormous number of links while most members have only a few, so that no single number of links is typical of the whole.

Example: On the web, a few sites such as major search engines receive links from millions of pages, while most personal blogs receive only a handful.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Scale-Out Graph Architecture

The design approach of splitting a graph's vertices and edges across a cluster of machines while trying to keep frequently followed connections on the same machine, since steps across the network are much slower.

Example: A retailer stores each customer and that customer's orders on the same server so that common queries rarely leave one machine.

Covered in: Chapter 16: Graph Database Architecture

Scaling Laws

Observed relationships that describe how a system's performance changes as an input resource is increased, often following a predictable pattern for a long period before a limit bends the curve.

Example: AI researchers found that language models improve predictably as training data, model size, and computing power grow, though each gain costs more.

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Scenario Testing

A method of running a model under several different plausible futures or sets of conditions to see how the system might behave in each one.

See also: Sensitivity Analysis

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Schema

A formal description of how data is organized, including the types of records, their fields and data types, and the relationships allowed among them.

Example: A library's database design says each book has a title, an author, and an ISBN, and each loan links one book to one borrower.

Covered in: Chapter 17: Knowledge Representation and Metadata

Schema Mapping

The process of writing the specific rules that transform data from one data structure into another, based on the correspondences found between their fields.

Example: A rule that splits the billing system's single "full name" field into separate "first name" and "last name" fields for the sales system.

See also: Schema Matching, ETL Process

Covered in: Chapter 17: Knowledge Representation and Metadata

Schema Matching

The process of identifying which fields or elements in one data structure correspond to fields or elements in another structure.

Example: Discovering that the field "cust_nm" in the billing system holds the same information as "CustomerName" in the sales system.

See also: Schema Mapping

Covered in: Chapter 17: Knowledge Representation and Metadata

Search And Reuse

The practice and goal of locating data, code, or other components that already exist inside an organization and adapting them, rather than collecting or building them again from scratch.

Example: Before building a new customer address checker, a developer searches the company's code catalog and finds one another team built last year.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling · Search and Reuse archetype

Search Engine Improvement Cycle

A reinforcing loop in which more web queries generate more data about which results people click, which trains a better ranking method, which returns more relevant results and attracts still more queries.

See also: AI Flywheel, Data Advantage

Covered in: Chapter 23: AI Systems Dynamics

Second-Order Effect

A consequence that follows from the direct result of an action rather than from the action itself, and that is therefore easy to overlook when planning.

Example: A city lowers bus fares and ridership rises; the later drop in downtown parking revenue is a consequence of the ridership change, not of the fare cut directly.

See also: Ripple Effect, Unintended Consequence

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Seeking The Wrong Goal Archetype

A systems archetype in which a system works hard to improve an easily measured stand-in for success, so the stand-in improves while the real, harder-to-measure purpose stays flat or declines.

Example: A school raises its test scores by drilling test items while student curiosity and deep understanding fall.

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Self-Correcting System

A system whose own structure detects departures from a target and pushes back toward it automatically, without a person needing to notice and step in each time.

Example: A home heating system with a thermostat keeps the room warm without anyone flipping a switch.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Self-Fulfilling Prophecy

A prediction or belief that causes people to act in ways that make the prediction come true, even if it was not true at first.

Example: Rumors that a bank is failing lead customers to withdraw their savings all at once, which then causes the bank to fail.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Self-Organization

The process by which a system forms its own order or structure through local interactions among its parts, without direction from a central controller or outside designer.

Example: A flock of birds forms a coordinated, shifting shape with no leader, because each bird simply keeps a set distance from its nearest neighbors.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Self-Organizing System Structure

A system's built-in ability to create new roles, groups, and procedures from within in response to changing conditions, without waiting for an outside party to redesign it.

Example: Because the web runs on open standards, new kinds of websites and apps appear constantly without anyone redesigning the internet.

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Self-Regulation

The general capacity of a system to adjust its own behavior to maintain function without needing an outside controller to step in each time.

See also: Homeostasis, Self-Correcting System

Covered in: Chapter 7: Feedback Resilience and Robustness

Semantic Layer

A part of an information architecture that translates underlying technical structures into consistent, business-friendly terms, so that a measure means the same thing no matter which system it comes from.

Example: "Customer lifetime value" is calculated the same way whether a report is run by the sales team or the finance team.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Semantic Web

A vision and set of standards, led by the World Wide Web Consortium, for publishing data online in machine-readable form with shared meanings so that computers can combine and reason over it.

See also: RDF, Ontology, Linked Data

Covered in: Chapter 15: Graph Theory Fundamentals

Sensitivity Analysis

A modeling technique that changes one input or assumption at a time and measures how much the model's results change in response.

It reveals which assumptions matter most and therefore deserve the most careful checking.

Example: A school budget model shows that a 5% change in enrollment moves costs far more than a 5% change in utility prices, so enrollment forecasts get extra attention.

See also: Model Assumptions, Scenario Testing

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Sensor (Control)

The component of a control system that measures the current state of a variable and reports it so that it can be compared with a target value.

See also: Actuator (Control), Control System

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Set Point

The target value that a goal-seeking system tries to reach and maintain, against which the system's current state is continually compared.

See also: Goal-Seeking Behavior, Condition

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Seven Bridges Of Konigsberg

A puzzle solved by mathematician Leonhard Euler in 1736 that asked whether a walker could cross each of the seven bridges of the city of Königsberg exactly once.

Euler proved it impossible by treating land areas as points and bridges as connections, an insight that founded graph theory.

See also: Graph (Data Structure)

Covered in: Chapter 16: Graph Database Architecture

Shadow IT

Technology, software, or cloud services used by employees or departments without the knowledge or approval of the organization's central technology group.

Example: A marketing team stores customer lists in a free online spreadsheet tool that the IT department does not know about.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Shallow Leverage Point

A place to intervene that is easy to identify and quick to change, such as a number or a physical capacity, but that usually has weak effects because the rest of the system absorbs the change.

Example: Raising a library's late fee by 25 cents changes little about how many books come back on time.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Shared Metrics

Measures of success that several teams or departments agree to track and are jointly responsible for, so that each has a reason to support the others.

Example: Sales and support are both judged partly on customer renewal rates, so they work together on unhappy accounts.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Sharpness Of Transition

The width of the range of an underlying condition over which a system's abrupt shift from one state to another takes place, from razor-thin shifts at a single point to gradual shifts spread across a wide range.

See also: Phase Transition

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Shifting Dominance

The change over time in which of a system's competing feedback loops has the strongest influence, causing the system's overall behavior to switch patterns, such as from rapid growth to leveling off.

Example: A startup grows quickly while word of mouth drives sign-ups, then levels off once most likely customers have already joined.

See also: Loop Dominance, S-Curve

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Shifting The Burden Archetype

A systems archetype in which reliance on a symptomatic solution weakens the system's ability to use a fundamental solution, so the system grows more and more dependent on the quick fix.

Example: A developer who uses an AI tool for every coding task stops practicing debugging and becomes less able to catch the tool's mistakes.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Shifting the Burden examples

Shortest Path

The route between two vertices in a graph that crosses the fewest edges or has the lowest total edge weight.

Example: A map app finds the quickest drive home, and a professional networking site shows whether a stranger is a second- or third-degree connection.

See also: Weighted Graph, Graph Algorithms

Covered in: Chapter 16: Graph Database Architecture

Side Effect

Any additional consequence of an action beyond the one it was intended to produce, which in quick-fix situations is often what eventually makes the original problem worse.

Example: A stronger pesticide kills more insects this season but also kills the birds that ate those insects.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

Silo Busting

The deliberate work of breaking down barriers between isolated departments or systems so that information, people, and goals connect across the whole organization.

Example: A company creates shared customer data, joint goals for sales and support, and regular cross-department meetings.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Single Source Of Truth

The practice of storing and maintaining each important piece of data in one authoritative location, so that every system and report draws on the same values.

Example: When a customer changes an address, it is updated in one place, and every department sees the new address.

Covered in: Chapter 18: Data Management and Governance

Single View Of Customer

One consistent, unified representation of a buyer's identity, assembled from every system that holds records about that person or organization.

Example: Matching a shopper's online account, store loyalty card, and call-center records into one profile.

See also: Customer 360, Entity Resolution

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Skipping Levels Mistake

The error of attempting a deep intervention without first building the supporting changes at the intermediate depths of the leverage points hierarchy, such as declaring a new paradigm without the rules and structures that would let people act on it.

Example: A company declares a "data-driven culture" but gives most employees no access to data and no training in reading it.

See also: Level Confusion Mistake

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Small-World Network

A web of connections in which most members are not directly linked, yet any two can be reached in surprisingly few steps because tightly knit clusters are joined by a few long-range shortcuts.

Example: The popular idea of "six degrees of separation" suggests that any two people on Earth are linked by a short chain of acquaintances.

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

The conditions in which people are born, grow, live, work, and age, such as income, education, housing, neighborhood safety, and access to food, that strongly shape how well and how long they live.

Example: Two people with the same illness recover differently when one lives near a clinic and a grocery store and the other does not.

See also: Health Disparities

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

A reinforcing loop in which a person's clicks and viewing time train a platform's ordering software, which then shows more of whatever held that person's attention the last time.

Example: Watching two cooking videos leads the app to fill the screen with cooking videos, which leads to watching still more of them.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

A lasting pattern of relationships, roles, norms, and interactions among people, such as a family, a school, a community, or a whole society.

Example: A school's unwritten rules about where students sit at lunch shape who becomes friends with whom.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Speculative Bubble

A period in which the price of an asset rises far above its underlying worth because buyers expect to resell it at a still higher price, followed by a sharp collapse when that expectation fails.

Example: Tulip bulb prices in the Netherlands soared to extraordinary levels and then crashed in 1637.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Stakeholder

Any person, group, or organization that affects, is affected by, or has an interest in a system, project, or decision.

See also: Stakeholder Analysis

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Stakeholder Analysis

A structured method for identifying every person or group affected by a system or decision and mapping each one's interests, influence, and likely response.

Example: Before closing a school, a district lists students, parents, teachers, bus drivers, nearby businesses, and neighbors, noting how each will be affected.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Star Schema

A data warehouse design in which one central table of measurable events, such as sales, is linked to several surrounding tables that describe those events by time, product, place, or customer.

Example: A central Sales table connects to Date, Store, Product, and Customer tables, so managers can total sales by any of them.

Covered in: Chapter 18: Data Management and Governance

Steady State

A condition in which the measured variables of a system stay roughly constant over a period of time, regardless of which processes keep them there.

See also: Dynamic Equilibrium

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Stock

A quantity of material, money, people, or information that has built up in a system over time and can be counted or measured at any single moment.

It acts as the system's memory of past flows and can change only through inflows and outflows.

Example: The water in a reservoir, the money in a savings account, and the number of employees at a company can each be counted at any moment.

See also: Flow, Accumulation

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Stock And Flow Diagram

A diagram that draws accumulations as boxes and the rates that fill or drain them as pipes with valves, making quantities and their changes explicit and measurable.

It adds the numeric detail that a causal loop diagram leaves out.

Example: A bathtub drawing shows the water in the tub as a box, the faucet as a pipe flowing in, and the drain as a pipe flowing out.

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Stock And Flow Model

A quantitative model, usually run on a computer, that represents accumulations and the rates that fill and drain them with equations in order to simulate how a system changes over time.

Example: A public-health department simulates how many hospital beds will be needed during flu season by tracking patients entering and leaving care.

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Streetlight Effect

The tendency to search for answers or solutions where searching is easiest rather than where the answer is most likely to be found.

The name comes from an old joke about a person looking for lost keys under a streetlight because that is where the light is.

Example: A company tries to improve customer satisfaction by fixing only the issues its survey tool can easily count, ignoring deeper frustrations customers mention in calls.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Structural Leverage Point

A place to intervene that changes the feedback loops and information pathways governing how a system responds to its numbers, rather than the numbers themselves, making change harder but longer lasting.

Example: Posting a live dashboard of website errors where every developer can see it changes how quickly problems get fixed.

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Structure Drives Behavior

The principle that the patterns a system produces over time come mainly from its underlying arrangement of stocks, flows, feedback loops, and delays, rather than from individual events or people.

Example: When different employees in the same role keep making the same mistake, the workflow design is a more likely cause than the individuals.

See also: Iceberg Model, Structures Layer

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Structures Layer

The third level of the iceberg model, made up of the rules, feedback loops, physical arrangements, incentives, and information flows that produce the patterns above them.

Example: "Quarterly financial reports run on the same server as payroll and overload it."

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Suboptimization

The behavior that results when the goals of one part of a system win out at the expense of the goals of the whole system.

Example: A sales team hits its targets by promising custom features that overload engineering and delay the product for every other customer.

Covered in: Chapter 7: Feedback Resilience and Robustness

Subsystem

A smaller group of interacting parts that works as a complete unit of its own while also serving as one component inside a larger whole.

Example: A hospital's emergency department has its own staff, rules, and flow of patients, yet it depends on the lab, the pharmacy, and admissions to do its job.

Covered in: Chapter 1: Foundations of Systems Thinking

Success To Successful Archetype

A systems archetype in which two or more parties compete for limited resources and the one that gets slightly more early on gains the ability to win still more, while the others fall further behind.

Example: A project that gets a small early budget boost shows better results, which earns it an even larger budget the next year.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Success to the Successful examples

Supply Chain

The connected sequence of organizations, people, activities, and resources involved in moving a product from raw materials to the final customer.

Example: A smartphone's parts come from dozens of countries before being assembled in one factory and shipped to stores worldwide.

See also: Supply Chain Graph, Bottleneck

Covered in: Chapter 27: Systems Thinking Across Disciplines

Supply Chain Graph

A connected data model of vendors, parts, factories, shipping routes, and customers that lets an organization trace how a disruption at one point will affect products and orders further along.

Example: When a flood closes a chip factory, a carmaker immediately sees which models and dealers will be affected.

Covered in: Chapter 26: Knowledge Graph Applications and Data Architecture

Sustainability

The capacity of a system to keep performing well over long periods despite changes in its environment, without using up the resources or damaging the conditions it depends on.

Sustainable systems remain resilient when unexpected changes hit either their structure or their flows of material and information.

Example: A fishery that catches fewer fish each year than the population replaces can continue indefinitely.

Covered in: Chapter 7: Feedback Resilience and Robustness

Sustainable Growth

An increase in size or activity that stays within what its supporting resources and conditions can replenish or support, so that it can continue over a long period.

Contrast with: Unsustainable Growth

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Symptom

A visible sign or effect of a problem that points toward, but is not the same as, the underlying condition that is producing it.

Treating only these visible signs gives quick relief but allows the problem to return.

Example: Constant overtime on a software team is a visible sign; deadlines set without any input from engineers may be the deeper cause.

Contrast with: Root Cause

Covered in: Chapter 1: Foundations of Systems Thinking

Symptom Relief

The short-term comfort a quick fix provides, which is real but, in the shifting-the-burden pattern, replaces the harder work of building the ability to address the underlying cause.

Covered in: Chapter 10: Fixes That Fail and Shifting the Burden

Symptomatic Solution

The more formal name for an intervention aimed at the visible sign of a problem rather than its origin, used especially when describing the structure of an archetype.

Contrast with: Fundamental Solution

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary

System

A set of interrelated and interdependent parts that interact over time and together form a whole with behavior of its own that none of the parts produces alone.

Systems can be natural, such as a forest or the human body, or human-made, such as a school district or a software platform.

Example: A school is a system: students, teachers, schedules, buses, and budgets all affect one another, so changing the bell schedule ripples through lunch periods, bus routes, and after-school jobs.

See also: Subsystem, System Boundary, Systems Thinking

Covered in: Chapter 1: Foundations of Systems Thinking

System Boundary

The real or chosen dividing line that separates the parts and relationships included in an analysis from everything treated as lying outside it.

Where the line is drawn decides which causes can be seen; a line drawn too tightly hides important feedback.

Example: Studying traffic on one highway without the nearby side streets draws the line too narrowly to notice that drivers simply shift their routes when the highway is widened.

See also: Environment (System), Open System

Covered in: Chapter 1: Foundations of Systems Thinking

System Delay

A lag that comes from the physical structure or processes of a system itself, such as travel, construction, or growing time, rather than from slow reporting of results.

Example: A newly planted apple tree takes several years to bear fruit, no matter how closely the farmer watches it.

See also: Feedback Delay, Conveyor Model

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

System Mapping

The process of identifying the parts, relationships, and boundaries of a system and recording them in a diagram, often as a group activity.

Doing this work together surfaces the different mental models people hold and the gaps in what anyone knows.

Example: An IT department holds a workshop to trace how customer data moves among its sales, billing, and support applications.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

System Pushback

The specific compensating response, appearing elsewhere in a system, that arises because an intervention changed something the rest of the system depended on.

Example: When a school bans phones during class, students start using smartwatches to message each other.

See also: Policy Resistance

Covered in: Chapter 7: Feedback Resilience and Robustness

System Theory

The interdisciplinary field that studies the general principles shared by all kinds of organized wholes, from cells to societies, and how their parts interact.

It supplies the shared ideas that let lessons learned in biology, engineering, economics, and management transfer from one field to another.

See also: Systems Thinking, Systems Dynamics

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Systemic Inequality

Differences in outcomes among groups of people that are produced and maintained by the structure of a society or organization, such as its rules, feedback loops, and resource flows, rather than by individual choices alone.

Example: School funding tied to local property taxes gives wealthy neighborhoods better-funded schools, which keeps property values and taxes high there.

See also: Cumulative Advantage, Algorithmic Bias

Covered in: Chapter 11: Tragedy of the Commons and Success to the Successful

Systems Archetype

A recurring pattern of feedback loops and delays that appears across many different kinds of systems and produces a predictable, recognizable pattern of behavior.

These patterns are to systems thinking what design patterns are to software; they link psychology, economics, biology, urban planning, technology, and government because the same structures turn up in all of them.

Example: The same "fixes that fail" pattern appears when a person drinks coffee to fight fatigue caused by poor sleep and when a company cuts training to hit a budget.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary · Archetypes overview

Systems Dynamics

The field and modeling method, founded by Jay Forrester at MIT, that uses stocks, flows, feedback loops, and delays to simulate and understand how systems change over time.

Example: Researchers in this field have built computer models to study city growth, supply chains, and climate policy.

See also: Stock And Flow Model, Dynamics

Covered in: Chapter 5: Stocks, Flows, and System Dynamics

Systems Leadership

A way of guiding change that brings together stakeholders from across a shared system, helps them see the whole rather than only their own part, and creates the conditions for a joint solution to emerge.

See also: Collective Impact, Stakeholder

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Systems Map

A visual representation that shows the main parts of a system and the relationships among them, used to make a group's shared understanding visible and open to discussion.

Example: A school team draws a diagram linking attendance, family income, bus routes, and grades to see why some students fall behind.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Systems Thinker Role

An organizational responsibility for applying feedback-loop, archetype, and leverage-point skills to guide how an enterprise knowledge graph is scoped, adopted, and evolved over time.

See also: Graph Systems Thinking

Covered in: Chapter 19: Enterprise Knowledge Graphs

Systems Thinking

A problem-solving approach that views a complex situation as a set of connected parts that influence one another over time, paying attention to feedback, delays, and the whole rather than to isolated pieces.

It helps explain why well-meant fixes often fail and where small changes can produce large effects.

Example: Instead of blaming one teacher for low reading scores, a systems thinker asks how class size, home reading time, library hours, and testing schedules interact.

See also: Linear Thinking, Holism, Feedback Loop

Covered in: Chapter 1: Foundations of Systems Thinking

Systems Thinking Life Skill

The framing of feedback loops, delays, and leverage points as a general-purpose habit of mind that anyone can apply to personal decisions, relationships, and routines, not only to professional work.

Example: A student notices that staying up late leads to tiredness, then more caffeine, then staying up late again, and decides to break the loop at bedtime.

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Tacit Knowledge

Know-how that a person holds through experience and practice but finds hard to write down or explain, such as judgment, intuition, and hands-on skill.

Example: An experienced nurse senses that a patient is getting worse before the monitors show any change.

Contrast with: Explicit Knowledge

Covered in: Chapter 24: Knowledge Systems and Economic Complexity

Taxonomy

A hierarchical classification scheme that arranges terms or items into parent-child categories running from general to specific, so that each narrower category sits inside a broader one.

Example: Animal, then Mammal, then Dog, then Beagle is a path from a broad category to a narrow one.

Covered in: Chapter 15: Graph Theory Fundamentals

Technical Debt

The future cost of extra work created when a quick or easy solution in software or infrastructure is chosen now instead of a better approach that would take longer.

Example: Copying the same code into ten places saves an afternoon today but means ten separate fixes every time a bug is found.

Covered in: Chapter 20: Organizational Silos and Silo Busting

Technology Adoption

The process by which individuals and organizations begin using a new tool or innovation, typically spreading from a few early users to the majority in an S-shaped pattern over time.

Example: Online video calls spread from a few tech-savvy offices to nearly every school and business within a few years.

Covered in: Chapter 23: AI Systems Dynamics

The God Graph

An informal, cautionary name for the unrealistic goal of building one all-encompassing connected data model covering every entity and relationship an organization knows about, all at once, rather than growing it step by step from a proven start.

Example: A company spends two years designing a model of "everything" and cancels the project before it answers a single business question.

See also: Microstrategy

Covered in: Chapter 19: Enterprise Knowledge Graphs

Theory Of Change

A written or diagrammed explanation of how and why a planned set of activities is expected to lead, step by step, to a desired long-term result, including the assumptions behind each step.

Example: A tutoring program expects weekly sessions to improve reading, which builds confidence, which raises attendance, which increases graduation rates.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Thermostat Balancing Loop

The classic example of a goal-seeking loop, in which a heating or cooling system compares room temperature with a set point and acts to close any gap, holding the room near its target.

See also: Balancing Loop, Set Point

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Threshold Effect

A sudden, large change in a system's behavior that occurs once a variable crosses a particular level, after a long period in which gradual changes produced little visible response.

Example: A lake can absorb fertilizer runoff for years with little change, then turn green with algae almost overnight once nutrients pass a certain level.

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Throughput

The rate at which material, energy, work, or information passes across a system from entry to exit during a given period of time.

The rate is usually set by the slowest step in the process, not by the average speed of all the steps.

Example: A coffee shop that serves 60 customers per hour has a throughput of 60 orders per hour, no matter how fast any single barista can work.

Covered in: Chapter 1: Foundations of Systems Thinking

Thurstone's Law

A method, developed by psychologist Louis Thurstone in 1927, for turning many subjective pairwise preferences into one consistent numerical ranking by treating each choice as a noisy reading of a hidden preference scale.

Many real races for dominance are decided by such small personal choices, and the ranking they produce then feeds a compounding loop.

Example: Asking thousands of users "Which of these two apps do you prefer?" can produce a single ranked list of all the apps, which app stores may then use to decide what to promote.

Covered in: Chapter 12: Named Laws, Technology Archetypes, and Complexity Modeling

Time Delay

The general category name for any gap between a cause and its effect, whatever its source, covering both lags built into physical processes and lags in noticing results.

See also: System Delay, Feedback Delay

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Tipping Point

The critical moment or level at which a small additional change pushes a system out of one stable pattern and into a new one that is often hard to reverse.

Example: Once enough of a student's friends join a new messaging app, the rest of the class quickly follows and the old app empties out.

See also: Threshold Effect, Bifurcation

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Top-Down Constraint

A boundary or rule imposed from a higher level of a system that limits which locally generated behaviors are possible without dictating the specific outcome.

Example: An open-source project's license and contribution guidelines do not write any code, but they rule out whole categories of contributions.

Contrast with: Bottom-Up Behavior

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Tragedy Of The Commons Archetype

A systems archetype in which many users each gain individually from using a shared, limited resource, so their combined use exceeds what the resource can sustain and it becomes depleted for everyone.

Example: Each herder adds one more cow to a shared pasture because the benefit is personal while the cost of overgrazing is shared, until the pasture is ruined.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Tragedy of the Commons examples

Training Data

The collection of examples used to adjust a machine learning model's internal settings so that it can recognize patterns and make predictions.

The model can only be as fair and accurate as these examples allow.

See also: Algorithmic Bias, Data Set

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Transformative Leverage Point

A place to intervene that changes the shared worldview or goal of a system, meaning the mostly unexamined beliefs that make its rules and structures seem like the only sensible arrangement.

Example: A company shifts from seeing each customer as revenue to maximize to seeing each customer as a long-term relationship to sustain.

See also: Mindset Or Paradigm, Paradigm Shift

Covered in: Chapter 13: Leverage Points: The Iceberg Model to Structural Change

Transformative Maturity Level

The fifth and highest stage of this book's systems-thinking maturity scale, at which systems thinking is part of the organization's identity, decisions favor whole-system results, and the organization actively reshapes its wider industry.

Example: A company shares its data standards openly with suppliers and competitors to improve the whole industry's supply chain.

See also: Transformative Leverage Point

Covered in: Chapter 21: Capability Maturity Model for Systems Thinking

Transparency (AI)

The degree to which information about how an AI system was built, trained, tested, and used is openly available to the people affected by it and to those who supervise it.

It differs from explainability, which concerns the reasons behind one particular output rather than the system as a whole.

See also: Explainability

Covered in: Chapter 22: Artificial Intelligence and Machine Learning Foundations

Underestimating Resistance

The error of failing to plan for the pushback that a deep intervention will provoke, so that a well-designed change fails because the opposition was never anticipated.

See also: Resistance To Leverage Change

Covered in: Chapter 14: Leverage Points: Rules, Paradigms, and Emergence

Underinvestment Archetype

A systems archetype in which a system holds back spending on capacity until demand clearly justifies it, but demand never grows enough because the capacity to serve it was never built.

Example: A regional airline refuses to add flights until a route is always full, but travelers choose other airlines because the flights are too infrequent.

Covered in: Chapter 9: Named Archetypes and Limits to Growth · Growth and Underinvestment examples

Undirected Graph

A graph in which edges do not point either way, so that each connection applies equally to both of the vertices it joins.

Example: A map of two-way streets, where every road between two intersections can be driven in either direction.

Contrast with: Directed Graph

Covered in: Chapter 15: Graph Theory Fundamentals

Unintended Consequence

An outcome of a purposeful action that the people taking the action did not plan for or foresee, whether helpful, harmful, or simply surprising.

Example: Adding a "like" button was meant to spread small bits of praise, but it also created pressure among teenagers to chase approval.

Covered in: Chapter 7: Feedback Resilience and Robustness

Unpredictability Of Emergence

The recognition that a complex or chaotic system can produce real, describable large-scale patterns even though its exact moment-to-moment path cannot be forecast in detail.

Example: No forecaster can say what the temperature will be in a given city 90 days from now, yet everyone can confidently predict that summer will be warmer than winter.

See also: Emergence, Chaos Theory

Covered in: Chapter 6: Growth Patterns and Nonlinear Behavior

Unsustainable Growth

An increase in size or activity that exceeds what its own limiting resources or conditions can support without being damaged, and that therefore cannot continue indefinitely.

Contrast with: Sustainable Growth

Covered in: Chapter 9: Named Archetypes and Limits to Growth

Urban System

The interconnected network of people, buildings, transportation, utilities, businesses, and services that make up a city and its surrounding region.

Example: Adding housing near a new train station changes traffic, school enrollment, and sales at local shops.

Covered in: Chapter 27: Systems Thinking Across Disciplines

Variable (CLD)

A named quantity shown in a causal loop diagram that can rise or fall, especially one recalculated fresh each period, like interest earned, rather than built up over time like a bank balance.

See also: Node, Stock, Condition

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Vector Database

A database that stores items as long lists of numbers representing their meaning and retrieves the items whose number lists are most similar to a request, rather than finding exact matches.

Example: A support chatbot finds help articles about resetting a login even when a user types "I can't get into my account."

Covered in: Chapter 25: Systems Design, Emerging Technology, and Practice

Vendor Lock-In

A situation in which a customer depends so heavily on one supplier's products or data formats that switching to another supplier becomes too costly or difficult.

See also: Open Standard, Path Dependence

Covered in: Chapter 20: Organizational Silos and Silo Busting

Venture Capital Funding Cycle

The boom-and-bust pattern in which investor enthusiasm for a technology sector drives waves of investment that inflate company values and expectations, followed by a correction when promised returns arrive more slowly than expected.

Example: Investors poured money into internet startups in the late 1990s, and many of those companies failed when the dot-com boom collapsed in 2000.

Covered in: Chapter 23: AI Systems Dynamics

Vertex

A single point in a graph data structure that represents one thing, such as a person, product, place, or concept, and that can be joined to other points by edges.

Vertices are also commonly called nodes.

See also: Edge (Graph), Node

Covered in: Chapter 15: Graph Theory Fundamentals

Vicious Cycle

A reinforcing loop in which a harmful change sets off responses that make the original harm worse, so the situation deteriorates further with each trip around the loop.

Example: A student falls behind, feels discouraged, studies less, and falls even further behind.

Contrast with: Virtuous Cycle

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Virtuous Cycle

A reinforcing loop in which a beneficial change sets off responses that strengthen the original benefit, so the situation improves further with each trip around the loop.

Example: A well-documented open-source library attracts users, who contribute fixes and better documentation, which attracts even more users.

Contrast with: Vicious Cycle

Covered in: Chapter 4: Feedback, Delay, and Loop Dynamics

Weighted Graph

A graph in which each edge carries a number, such as a distance, cost, time, or strength, that describes the size of the relationship it represents.

Example: A road map where each road between two towns is labeled with its length in miles.

Covered in: Chapter 15: Graph Theory Fundamentals

Wicked Problem

A challenge with many interconnected causes, conflicting stakeholder views, and no clear stopping point, where every attempted solution changes the nature of the challenge itself.

Example: Homelessness has this character: housing costs, mental health, jobs, and local politics interact, and each new program shifts the situation in unexpected ways.

Covered in: Chapter 2: Mental Models and Systems Analysis Tools

Winner-Take-All Dynamics

A pattern in which a small early lead is amplified by reinforcing loops until one competitor captures most or all of a market or resource.

Example: Once one online auction site had the most buyers, sellers flocked to it, which brought still more buyers, leaving rivals with almost nothing.

Covered in: Chapter 8: Systems Archetypes: Cross-Cutting Vocabulary