Complexity describes a state where many interconnected parts interact in ways that make the whole system difficult to predict or reduce to simple rules. The word comes from the Latin “complexus,” meaning braided or entwined, combined with the English suffix “-ity.” At its core, complexity refers to how things are woven together in ways that create behavior you wouldn’t expect from looking at individual pieces alone.
That basic idea branches into surprisingly different meanings depending on the field. A biologist, a computer scientist, and a business strategist all use the word “complexity” to mean related but distinct things. Understanding those differences helps clarify what people actually mean when they call something “complex.”
Complex vs. Complicated: A Key Distinction
The most useful thing to understand about complexity is how it differs from being merely complicated. These two words are often used interchangeably in casual conversation, but they describe fundamentally different situations.
A complicated problem can be hard to solve, but it follows reliable rules and recipes. Think of a tax return with dozens of forms, or the algorithm that places targeted ads on a social media feed. These systems have many moving parts, but if you understand the rules, you can work through them step by step and arrive at a predictable answer. Complicated systems adhere to a comprehensive set of rules, so solving them is a matter of applying the right model to the situation.
Complex problems are different. They involve too many unknowns and too many interrelated factors to reduce to rules and processes. A technological disruption like blockchain, or a competitor with a completely new business model like Uber or Airbnb, are complex problems. There’s no algorithm that tells you how to respond. The system behaves in ways that can’t be fully predicted even when you understand all the individual components. Weather is complex. A supply chain spanning dozens of countries is complex. A pandemic spreading through interconnected cities is complex.
This distinction matters practically. If you treat a complex problem like a complicated one and try to solve it with rigid processes, you’ll fail. Complex systems require a more nuanced, adaptive approach.
Properties of Complex Systems
Scientists who study complexity have identified several defining features that appear across wildly different systems, from ant colonies to economies to the human body.
- Emergence: The system produces behavior that can’t be predicted by studying individual parts in isolation. The defining characteristic of a complex system is that robust order at the system level arises out of disordered interactions at the microscopic level. A flock of birds creates coordinated patterns even though no single bird is directing the group.
- Nonlinearity: Small changes can produce disproportionately large effects. A tiny shift in initial conditions can send the whole system down a completely different path.
- Self-organization: Complex systems develop internal structure spontaneously and adaptively to cope with their environment. No central controller is required. Instead, distributed control among many agents leads naturally to organized patterns.
- Adaptation: The system learns and evolves over time. Its agents change their behavior based on experience, and the system as a whole co-evolves with its environment.
The classic shorthand: in a complex system, the whole is more than the sum of its parts. That phrase gets overused, but it captures something real. You can’t understand a city by studying each resident individually. The interactions between residents, infrastructure, businesses, and culture produce something that only exists at the system level.
Complexity in Biology
Biologists have spent decades trying to pin down exactly what makes one organism more complex than another. The intuitive answer, that humans are more complex than bacteria, turns out to be surprisingly hard to measure.
Early attempts focused on counting genes, but that metric ran into problems quickly. Some relatively simple organisms have far more genes than humans. So researchers expanded the definition to include the number of different cell types found in an organism, the number of distinct morphological traits, and the number of developmental pathways involved in building the organism. All of these approaches share a common thread: complexity is defined as the number of distinct parts composing an organism.
More recent frameworks split biological complexity into two dimensions. Horizontal complexity is the number of different kinds of parts at any given level, such as how many different cell types an organism has. Vertical complexity describes how those parts nest into hierarchies: cells assemble into tissues, which assemble into organs, which form organisms. A creature with many cell types (high horizontal complexity) organized into deeply nested layers (high vertical complexity) registers as highly complex by both measures.
Biologists also look at modularity, the tendency of certain structures to vary independently of one another, and integration, the tendency for multiple traits to change in a coordinated fashion. These properties describe how tightly or loosely the parts of a system are connected, which directly shapes how the organism can evolve.
Complexity in Computing
In computer science, complexity has a precise and narrower meaning: how do the resource requirements of a program scale as the problem gets larger?
Time complexity measures how much longer a program takes to run as you feed it more data. If you double the amount of data and the program takes twice as long, that’s linear time. If it takes four times as long, that’s quadratic time. If the time doesn’t change at all regardless of data size, that’s constant time. These categories are expressed using “Big-O notation,” where O(N) means the time grows proportionally with the data size, O(N²) means it grows with the square of the data size, and O(1) means it stays flat.
Space complexity works the same way but measures memory usage instead of time. The key insight is that computer scientists don’t care about the exact number of operations a program performs. They care about the pattern of growth. A program that handles 100 items easily but chokes on 10,000 items has a complexity problem, and knowing whether it scales linearly or quadratically tells you whether throwing more hardware at it will help.
Complexity in Information Theory
Information theorists measure complexity in yet another way: by asking how much information an object contains. The foundational concept here is Kolmogorov complexity, which defines the complexity of a sequence of data as the length of the shortest computer program that could generate it and then stop.
Consider two strings of 1,000 characters. One is “ABABAB…” repeated 500 times. The other is a random jumble of characters with no discernible pattern. The first string can be described very briefly: “print AB 500 times.” The second string can’t be compressed at all. You’d essentially need to list every character. The random string has higher Kolmogorov complexity because its shortest description is nearly as long as the string itself.
This concept can be thought of as absolute information content: the amount of information that would need to be transmitted between two people who share no prior knowledge about the message. It provides an objective way to quantify how much genuine information something contains, stripped of redundancy.
Complexity and Disorder Are Not the Same
One common misunderstanding is equating complexity with disorder or randomness. In physics, entropy measures the degree of disorder in a system, and it always increases over time in a closed system. But complexity doesn’t follow the same trajectory.
Complexity increases at first as a system evolves from a simple, ordered state. But as entropy approaches its maximum and the system reaches full disorder, complexity actually drops. A gas that has spread evenly throughout a sealed container is maximally disordered (high entropy) but displays low complexity. There’s nothing interesting or structured happening. The most complex states exist in between: not perfectly ordered, not fully random, but somewhere in the middle where structure and unpredictability coexist.
This is why the most complex phenomena we observe, life, ecosystems, civilizations, all exist in that middle zone, constantly taking in energy to maintain themselves far from equilibrium.
Cognitive Complexity
In psychology, cognitive complexity refers to how many distinct concepts a person can hold and use when processing information within a given domain. Someone with high cognitive complexity can look at a social situation and see it from multiple dimensions simultaneously, considering competing motivations, cultural context, and emotional undercurrents rather than defaulting to a simple narrative.
People with higher cognitive complexity tend to show greater flexibility in developing alternative solutions to problems. They process information in a more nuanced way, which is associated with analytic thinking, a style characterized by working to understand the world in a way that’s somewhat distanced from immediate emotional reaction. This doesn’t mean complex thinkers are emotionless. It means they’re more likely to sit with ambiguity and consider multiple interpretations before reaching a conclusion.