The history of frameworks.

Pre-literate societies operated on dense framework architecture. They couldn't write any of it down. That's the part the modern reader skips, because the version of history that starts with the printing press makes it sound like systematic thinking is a recent achievement.

It isn't. The impulse to systematize work is older than the alphabet.

Consider what a society without writing had to transmit. How to track game across terrain that shifts with the seasons. How to read which berries kill you. How to fire pottery without cracking the clay. How to settle a dispute over property when the property in question was a goat and the dispute was between cousins. None of that is intuitive. All of it had to be passed down. And it was, for tens of thousands of years, through structures that we'd now recognize as frameworks even though no one called them that.

Apprenticeship is the clearest example. The master-to-apprentice transmission model existed across cultures and trades for thousands of years before any culture wrote down its rules. Smithing, weaving, carving, leatherwork, midwifery, tracking. Each had stages, decision points, and quality standards that traveled through generations without paper. The fact that the structures repeat across cultures that never met is the evidence. People reach for systematic transmission when the work is too complex for one demonstration.

Religious and ceremonial frameworks worked the same way. Vedic recitation, oral Torah, Indigenous ceremonial practice. Long, structured bodies of knowledge that survived through repetition and ritual rather than text. Some of these traditions preserve more accuracy across centuries than written texts do, because the transmission method itself was engineered for fidelity.

The cleanest historical hinge between this era and the next sits inside the law. When Hammurabi's code was carved around 1792 BCE, and again when the Romans produced the Twelve Tables around 449 BCE, both projects described themselves the same way. They were writing down customary law that had been operating for centuries. The decemvirs in Rome weren't inventing the rules. They were codifying what the city had been doing since long before anyone could read.

Once writing existed, frameworks could outlive their authors. That's the structural shift. A method no longer required the master to be alive and present. It could be carved, copied, smuggled across borders, and applied in places its inventor never visited.

The famous early example is the Code of Hammurabi, which sits in the Louvre as a basalt stele covered in 282 case laws. Each one written as an "if this, then that" rule. If a builder builds a house and the house collapses and kills the owner, the builder is put to death. If a man's ox gores another man's ox, the value of the dead ox is split. The structure is recognizably a rule system, organized by category, applied by consistent reasoning. It governs criminal, family, property, and commercial life across the Babylonian empire.

Worth noting that Hammurabi's code wasn't the first written law. The Code of Ur-Nammu, from around 2100 BCE, predates it by roughly three centuries. Tablets from the city of Ebla in modern Syria contain even earlier law-like material from around 2400 BCE. Hammurabi's version became the famous one because it survived, traveled, and influenced everything that came after. The first systematic legal frameworks weren't his. He just made the most influential one.

Roman law made the same move at scale. The Twelve Tables of 449 BCE became the foundation of Roman legal reasoning, which continued evolving for nearly a thousand years until Emperor Justinian commissioned the Corpus Juris Civilis, with the Codex published in 529 CE and revised in 534 CE. That's a single living legal architecture stretched across a millennium, with named jurists adding to it and editing it the whole way through. Roman law wasn't a static document. It was a framework that kept getting refined as the empire encountered situations the original drafters hadn't imagined.

Military strategy got the same treatment. The Art of War, traditionally attributed to Sun Tzu, dates to the 5th century BCE in its earliest layers, though modern scholarship suggests the text is a composite assembled across generations rather than a single-author work. Authorship aside, the thirteen chapters operate as a framework. Categories of decisions, principles for each category, examples worked through to illustrate the reasoning. It's still in print 2,500 years later because the structure travels.

Religious frameworks did the same work, and the most influential one in the Western tradition is also the most structurally clean example in the entire era. The Ten Commandments. Ten numbered, categorized behavioral rules covering the relationship between people and God, the relationship between people and each other, and the boundaries around speech, property, family, and time. The structure is mutually exclusive and collectively exhaustive across the domain it governs. It has a clear issuing authority, a defined audience, and explicit consequences for non-compliance. If you ran it through any modern framework evaluation, it would check every box. Barbara Minto would approve. The Pyramid Principle technically owes a debt to a document that predates it by roughly three thousand years.

The transmission story is what makes it remarkable. Whether you read it as literal divine revelation at Sinai or as the codification of existing Hebrew moral law, the result is the same thing the rest of this era did. Structured behavior, written down, made transmissible to people who never met the original teacher. Carried across thousands of years and billions of readers. It's arguably the most successfully transmitted framework in human history. Nothing else on the historical record competes on reach or longevity.

The Sermon on the Mount, several centuries later, is itself a worked example of framework methodology that the rest of this article doesn't otherwise cover. Jesus didn't replace the law. He reframed it. "You have heard that it was said, but I say to you" is a systematic methodology for upgrading a framework while preserving its foundation. Existing rules about murder get extended into rules about anger. Existing rules about adultery get extended into rules about intent. The framework moves up a layer, from behavior to motivation, without invalidating what came before. That's a methodological move. Modern strategy literature would call it abstraction or principle-level reframing. The Sermon on the Mount did it first.

The Beatitudes operate as a structured values hierarchy. The Two Great Commandments operate as a compression framework, reducing the entire law to two governing principles that everything else derives from. Whether or not you share the underlying theology, the methodological architecture is plainly there, and ignoring it would be the same kind of editing-for-comfort that this article isn't otherwise interested in.

Buddhist precepts and Quranic structure carry similar architectural weight in their own traditions. Each became a transmissible artifact independent of any one teacher, which is what allowed those frameworks to travel across continents and centuries.

The pattern across Era 2 is consistent. Authorship was sacred, anonymous, or attributed to a king, a sage, or a god. The framework belonged to an institution or a tradition, not an entrepreneur. Hammurabi credited his code to Shamash. Moses credited his to Yahweh. Sun Tzu, if he existed at all, was a sage figure rather than a brand. The framework wasn't something anyone owned. It was something a civilization carried.

That changes in Era 3.

The shift that defines the modern era of management thinking is when frameworks became proprietary. Authored by named individuals, owned by named firms, sold to clients, and rented out by the engagement. The framework was no longer civilizational architecture. It was professional services revenue.

Frederick Winslow Taylor wrote The Principles of Scientific Management in 1911. Time-motion studies, standardization of industrial work into measurable units, the conversion of craft labor into a system that could be analyzed and optimized. Taylorism was the moment when systematic methodology became its own commercial category. Frank and Lillian Gilbreth extended the work in the same period. The factory floor became a framework laboratory.

McKinsey & Company opened in Chicago in 1926, founded by James O. McKinsey, a University of Chicago accounting professor. The firm spent its early years as an accounting and management practice. Marvin Bower, hired in 1933, did the work of turning it into the cultural institution it became. By mid-century McKinsey had built the playbook that the rest of the industry would copy. Recruit the smartest graduates from the best schools, train them in proprietary methodology, charge clients for the methodology and the people who applied it.

The Boston Consulting Group followed in 1963, founded by Bruce Henderson with $500 in first-month billings. Bain & Company spun out of BCG in 1973, when Bill Bain and a handful of colleagues departed after losing patience with what they saw as the firm's report-and-leave model. Henderson didn't take it well. The Big Three of strategy consulting consolidated into something close to its modern form between 1926 and 1973, three firms running on the same fundamental business model. Sell methodology by the engagement.

The frameworks themselves became branded artifacts. The BCG Growth-Share Matrix, with its four quadrants of Stars, Cash Cows, Question Marks, and Dogs, emerged from collaborative work inside BCG between 1968 and 1970. Alan Zakon sketched the early version. The team refined it. Henderson published it as "The Product Portfolio" in BCG's Perspectives in 1970, which is how it got attached to his name. Single-author attribution flattens what was actually a firm product, but the matrix became the most recognizable strategy artifact of the 20th century.

The McKinsey 7-S Framework came out of a two-day working session in San Francisco in 1980, with Tom Peters, Robert Waterman, Anthony Athos, and Richard Pascale. First published that June in Business Horizons. Strategy, Structure, Systems, Skills, Staff, Style, and Shared Values. Peters and Waterman built In Search of Excellence on top of it in 1982, which sold three million copies and made the framework a household concept among American managers.

Michael Porter, then a young associate professor at Harvard Business School, published "How Competitive Forces Shape Strategy" in Harvard Business Review in March 1979. The Five Forces. Threat of new entrants, supplier power, buyer power, threat of substitutes, rivalry among existing competitors. Porter expanded the framework in Competitive Strategy the following year and built an academic and consulting career on top of it.

Barbara Minto joined McKinsey in 1963 as the firm's first female MBA hire, working out of the Cleveland office before moving to London in 1966. While editing reports for McKinsey consultants in the late 1960s, she developed the MECE principle, which stands for mutually exclusive, collectively exhaustive. Her book The Pyramid Principle, first published in 1985 and revised in 1996, became the structural backbone of how strategy consultants write and think. She is sometimes underweighted in the public history of McKinsey. She shouldn't be.

SWOT analysis is harder to attribute. Albert Humphrey at the Stanford Research Institute is the most commonly cited origin, though scholarly consensus on the attribution is genuinely mixed. Harvard Business School faculty in the 1950s and 1960s, Igor Ansoff, and others have all been credited. By the 1980s SWOT was widely used as a standard strategic-planning tool. Whoever invented it, the framework got absorbed into the consulting toolkit and became one of the most widely deployed frameworks in business history.

What ties these together isn't the specific frameworks. It's the structural reality of the era. Frameworks had brand owners. They had price tags. They depreciated when the engagement ended, because the methodology stayed proprietary inside the firm. The McKinsey moat in Era 3 was access. You could read the books, but the real methodology lived in the partnership track.

That's the moat that Era 4 dissolved.

The current era doesn't begin with ChatGPT, though that's the cultural marker most people anchor on. ChatGPT launched on November 30, 2022, and reached 100 million users by January 2023, which made it the fastest-growing consumer application in history at that point. Claude followed in March 2023. GPT-3 had been around since June 2020 but stayed mostly inside developer circles.

What actually defines Era 4 isn't the chatbots. It's the moment when frameworks became transmissible to a non-human reader who can apply them at scale across thousands of decisions.

Anthropic launched Agent Skills on October 16, 2025, with the open-standard release on December 18, 2025. Skills are organized folders containing instructions, scripts, and resources that an AI system can dynamically load when a relevant task comes up. The architecture uses progressive disclosure. Only the skill name and short description load at startup, costing maybe fifty to a hundred tokens. The full instructions load only when the skill activates. Microsoft, Cursor, Goose, Amp, OpenCode, and most of the serious coding-agent ecosystem adopted the standard within weeks. OpenAI quietly shipped structurally identical architecture in ChatGPT and the Codex CLI.

The architecture matters less than the structural shift it represents. For the first time, a written framework can be read and applied by something that isn't human. That hasn't been true at any prior point in this 4,000-year history. Hammurabi needed judges. The 7-S Framework needed consultants. The Pyramid Principle needed analysts trained to use it. Every prior era's frameworks required a human reader who was already trained on the underlying domain.

A skill folder doesn't need that. The reader, in this case, comes pre-trained on language, reasoning, and a vast amount of domain context, and the skill itself supplies the procedural specifics. The same framework can govern a human's decision-making and an AI agent's behavior. That's the part that's genuinely new.

It also clarifies a question that's been muddying the conversation for the last year. Are skills the same thing as frameworks?

They aren't, and the distinction matters. Skills are an Era 4 product term, specifically Anthropic's term for narrow task instruction sets organized in folders. They're useful. They're real. They're a meaningful piece of infrastructure. But they're a subset of what frameworks have been across the 4,000-year arc. A framework is a behavioral standard for a category of work. A skill is an instruction set for a specific task. The framework governs the skill, not the other way around.

Anthropic's own documentation describes skills as "procedural knowledge" packages, which is a narrower term than frameworks have historically carried. The honest reading is that skills are one expression of frameworks in the current technical environment, not a replacement for the larger concept.

What's also different in Era 4 is who gets to author. In Era 3, framework authorship was effectively gated by institutional access. Becoming a McKinsey partner, getting a Harvard Business School appointment, founding a consulting firm. In Era 4, the gate is gone. A solo practitioner with a methodology and a markdown editor can produce a framework architecture that AI agents will execute as reliably as any consulting deliverable. The publishing infrastructure is free. The execution layer is provider-agnostic. The reader scales.

Four eras, with the time scales accelerating. Era 1 covers tens of thousands of years of oral transmission. Era 2 covers about four thousand years of written codification, from Ur-Nammu through Justinian and beyond. Era 3 covers about a hundred years of professional commodification, from Taylor through Porter. Era 4 covers about five years and is still being built.

The arc isn't a story about progress replacing the past. Each era added a capability the prior one couldn't deliver. Writing made frameworks portable. Professionalization made them refined. AI-readability makes them executable at scale by a reader that isn't human. The earlier capabilities don't go away. Apprenticeship still works. Books still work. Consultants still work. The new layer just sits on top.

What does change is the moat. The McKinsey moat in Era 3 was proprietary access to methodology. That moat hasn't disappeared, but it has been disrupted in a way that creates real space for individual practitioners and open-source methodology libraries. A framework that lives in a public repo, readable by any agent and any human, doesn't compete with the consulting model on price. It competes on transmissibility. The framework that travels wins.

The forgetting is understandable. The current technology cycle is loud and the historical context isn't on the homepage. But the practice itself is older than the city, and the people writing frameworks today are doing what humans have been doing since long before the alphabet. The work hasn't changed. The reader has.