Every System Has a Pyramid

There is a 190-page monograph sitting in the course materials of the Naval Postgraduate School that most practicing engineers have never encountered. Written by Professor Robert Harney for his systems engineering curriculum, Technology Evaluation for Strategic Planning and Innovation builds a rigorous theory of how technologies emerge, mature, and die — and how to evaluate where any given technology sits on that arc. It is dense, methodical, and not available in any bookstore. It also contains one of the most useful mental models I've found for thinking about engineered systems.

Harney's central argument is deceptively simple: every technology rests on a pyramid of prerequisite technologies. If any layer of the pyramid is missing, the capstone cannot be built — no matter how brilliant the concept, no matter how much funding is available, no matter how urgently the market demands it. The pyramid is not a metaphor. It is a structural dependency that governs what is possible and what is premature.

The Pyramid of Causality

Harney identifies six layers that every novel technology requires. At the base sits knowledge — the scientific understanding that makes the technology conceivable. Above that, instruments and tools — the ability to observe, measure, connect, and manipulate at the relevant scales. Then parts and materials — the physical building blocks with the right properties. Then components — what those parts and materials are assembled into. Then domain technologies — the implementations the capstone is built from directly. And at the apex, the capstone — the novel technology itself.

The power of this model is diagnostic. When a project is failing — behind schedule, over budget, delivering less than promised — the instinct is to blame execution. More engineers, more money, more urgency. Harney's framework asks a different question: is a layer of the pyramid missing? Because if a technology the capstone rests on doesn't exist at the required maturity, no amount of execution discipline at the capstone level will compensate. You are building on an incomplete foundation.

Consider autonomous vehicles. The capstone — a fully self-driving car — has been "five years away" for over a decade. The knowledge exists (machine learning, computer vision, sensor fusion). The instruments exist (LiDAR, radar, cameras). The parts exist (GPUs, sensors, actuators). But the components and domain technologies above them — real-time edge inference at sufficient reliability, mapping at global scale — are not mature enough to support the capstone, and the regulatory frameworks around them are not settled either. The pyramid is incomplete. No quantity of venture capital will complete it faster than the prerequisite layers allow.

Every Technology Follows the Same Curve

Harney's second major framework is the technology S-curve — the observation that every technology traverses the same lifecycle: from precedents through conception, gestation, birth, exponential growth, maturation, maturity, senescence, decay, and eventually death. This isn't a vague analogy. Harney catalogs specific metrics for each stage and demonstrates the pattern across domains from vacuum tubes to aircraft carriers.

The critical insight is the growth phase. During growth, performance improves exponentially — until it doesn't. Every S-curve has a ceiling, and that ceiling is defined by the limits the technology encounters. Understanding whether a limit is fundamental (physics — cannot be overcome), technological (engineering — can be overcome by switching implementation), or societal (economics, regulation, culture — soft but real) determines whether the S-curve will resume climbing after a breakthrough or flatten permanently.

Three Kinds of Ceilings

This is where Harney's framework becomes most practically useful. He argues that the single most important skill in technology evaluation is distinguishing between limits that are real and limits that are inherited — between constraints imposed by nature and constraints imposed by convention, economics, or institutional inertia. An evaluator who mistakes a technological limit for a fundamental one gives up too early. An evaluator who mistakes a fundamental limit for a technological one wastes resources pursuing the impossible.

The misidentification of limits is responsible for more failed programs than any engineering error. Harney documents the pattern repeatedly: a technology stalls, leadership concludes it has hit a fundamental wall, funding is redirected — and five years later, a competitor finds the technological workaround that was always available. Or conversely: a program pushes relentlessly against a limit that is genuinely fundamental, burning through budget and credibility until physics wins.

What Changes When You Can See the Pyramid

Most systems engineering tools model what a system is — its components, interfaces, requirements, behaviors. None of them model where the system sits on the technology landscape. They can tell you the system has a battery subsystem. They cannot tell you that the battery chemistry is approaching a fundamental energy density limit, that the manufacturing process relies on a technology with a 3-year maturation gap, or that the regulatory framework in the target market classifies the chemistry as hazardous — a societal limit that adds 18 months to certification.

Cairn's Causality lens takes on part of that: it shows what each part of a model stands on, and how far along each of those foundations is.

Harney wrote his framework for defense acquisition officers evaluating whether emerging technologies were ready for procurement. But the underlying insight is universal: you cannot design well what you do not understand structurally. The pyramid of causality is a structural understanding of technological possibility. The S-curve is a temporal understanding of technological readiness. The limits taxonomy is a diagnostic for why progress stalls.

Together, they give engineers a vocabulary for the questions that decomposition trees and requirements databases never ask: not just "what is this system made of," but "what does this system stand on, how mature are its foundations, and what will prevent it from reaching its potential?"

The Causality lens answers the first two for any node in your model. The third, why progress stalls, still takes an engineer who can tell a fundamental limit from a technological one.