“An old story tells of a visitor who encounters three stonemasons working on a medieval cathedral and asks each what he is doing. ‘I am cutting this stone to shape,’ says the first, describing his basic actions. ‘I am building a great cathedral,’ says the second, describing his intermediate goal. ‘And I am working for the glory of God,’ says the third, describing his high-level objective. The construction of architectural masterpieces required that high objectives be pursued through lesser, but nonetheless fulfilling, goals and actions.”

John Kay, Obliquity

All efforts, from daily personal projects to global collaborative endeavors, fit in a webbed hierarchy of abstraction.

Understanding the full hierarchy of an effort is critical to accomplishing it, along with its higher-level objectives in the long-term. Not understanding it can result in bad planning, mismanagement, and failed expectations.

Ladder of Abstraction

“…the most powerful way to gain insight into a system is by moving between levels of abstraction.” — Bret Victor

The Ladder of Abstraction is a mental model originally applied to writing by S.I. Hayakawa in his book Language in Thought and Action.

The ladder represents a top-level concept or domain, with each rung a subset of the one above it. The rungs move from abstract at the top, to concrete at the bottom. The lower down, the more detailed and specific. The higher up, the broader and more abstract the concept.

The model is very simple, and can be applied to almost any discipline with a hierarchy of nested groups. This includes applying it to efforts.

Nested Efforts

First of all — what do I mean by effort?

An effort is the active search for the best outcome of an objective. It encompasses both the objective and the pursuit of that objective — both of which are not fixed and can evolve over time. The objective always has some boundaries, but otherwise can be very broad (“solving climate change”) or very narrow (“double next-month’s sales volume”).

All efforts are multi-scale and nested.¹ This means we can put them on a ladder of abstraction, each rung with an objective or method that’s a prerequisite of the one above it. Lower-level goals are nested in higher-level purposes. Good project managers do this intuitively when breaking an objective down into tasks and sub-goals, mapping their dependencies.

Because efforts can have many dependencies and relationships aren’t just one-to-one, they exist in more of a webbed hierarchy of abstraction than a ladder. A simple one-dimensional ladder of abstraction is just a slice of the larger hierarchy.

Here’s an example of a hierarchy of abstraction for the efforts relating to Covid-19:

Why/how Chain

What’s the best way to determine the hierarchy of abstraction for an effort?

A simple way to move up and down the ladder is the Why/How Chain. To move up, ask “Why?”; to move down, ask “How?”. Many know this technique from the Toyota Production System’s method of asking 5 Why’s to find the root cause of an issue.

You can start by finishing the phrase: In what ways might we ___? This method can work on almost everything, from large-scale efforts to small-scale jobs-to-be-done:

• ⬆️ Why? To make my home look good.
• ⬆️ Why? To hang a picture.
• ❇️ In what ways might we drill a hole?
• ⬇️ How? Use a drill.

In the Covid-19 example, you could start at whatever level is most relevant to you.

• ❇️ In what ways might we provide better medical care for COVID patients?
• ⬆️ Why? To stop people from dying and reopen the economy.
• ⬇️ How? Protect medical workers from getting sick.
• ⬇️ How? Source and distribute PPE.
• ⬇️ How? Contact regional manufacturers.

There will always be multiple “how”s, which is the essence of breaking a goal down into sub-goals. There can be multiple “why”s as well, especially the further you go down the ladder. But high up in the hierarchy the whys and hows become more and more vague. This means you have to approach them in a completely different way.

Abstraction = Obliquity

Knowing where an effort fits on the hierarchy is the first step. Now we need to understand how the different levels of scale need to be treated.

This is where John Kay’s concept of obliquity, from his book of the same title, comes in.

To solve a problem obliquely is to solve it through experiment and adaptation. In general, the bigger the scope and complexity of an objective, the more indirect the path is to achieve it.

The ladder of abstraction is a proxy for obliquity. The higher on the ladder, the more adaptive the problem should be solved. John Kay: “High-level objectives — live a fulfilling life, create a successful business, produce a distinguished work of art, glorify God — are almost always too imprecise for us to have any clear idea how to achieve them.” In the process of making progress on these objectives, we don’t only learn how improve, but “about the nature of the objectives themselves.” You’re wayfinding, rather than following a prescribed path.

The lower on the ladder, the more direct. “Directness is only appropriate when the environment is stable, objectives are one-dimensional and transparent and it is possible to determine when and whether goals have been achieved.”

The following table compares the different aspects of both ends of abstraction.

	Direct (concrete)	Oblique (abstract)
Objectives	Clear and simple	Loosely defined and multidimensional
Intentionality	Most outcomes are intended	Outcomes arise through complex processes with no simple cause and effect
Interactions with others	Limited and predictable	Dependent on many variables, including interpretation of them
Options	Range of available options is fixed and known	Only a subset of options are available from successive limited comparison
Risks	Can be described probabilistically	Uncertain: Range of what might happen is not known
Consistency	Insists on consistency: always treating the same problem the same way	Consistency is minor and possibly dangerous — rare that same problem is encountered twice
Adaptation	Conscious maximization of objectives	Adapt continuously to changing circumstances

Consistency is vital when you’re low on the ladder, not so much higher up. “The oblique decision maker, the fox,” John Kay remarks, “is not hung up on consistency and frequently holds contradictory ideas simultaneously.”

But the real power of solving an oblique problem lays in adaptation: “If the environment is uncertain, imperfectly understood and constantly changing, the product of a process of adaptation and evolution may be better adapted to that environment than the product of conscious design. It generally will be.” There is no map, so instead you have to wayfind and look for clues in front of you, making your way with the tools you have at hand.

Keep in mind again that this is a scale — it’s rare that an effort would completely check all the above boxes for either Direct or Oblique. The point is that efforts always fall somewhere on the scale and that this determines the best methods to pursue them.

Consequences of Misplaced Directness

I’ll try to keep this section short, as whole books have been written on the consequences of misplaced directness. See Nassim Taleb’s Incerto for example.

Attempting to approach a large, complex effort too directly almost always leads to failure — or at the very least a failure to meet initial expectations. Directness is only appropriate when the objective is one-dimensional and the path to achieve it is known.

• The design and planning of the city of Brasilia serves as a good example. Engineering-wise it was incredibly impressive — a large, modern city built from the ground up between 1956 and 1961.

  The intention was to create a new Brazilian capital from scratch that was truly unique and modern, paying special attention to cars and traffic flow. (This was the same time period the U.S. began building out the intercontinental highway system.)

  As time went on, unforeseen circumstances in the messiness of the real world intervened. Overpopulation drove traffic congestion, slums, and general inequality. The focus on form over function from the top-down design caused alienation and poor quality of life. This is exactly why any such an effort can’t be planned with precision. Not only are the details of the true goal not understood, but the methods to achieve it involve unpredictable complexity. They’re in the world of “extremistan” as Taleb would say.

  “The structures in this artificial capital are impressive,” read an FT article on one of the architects, “yet few want to walk its barren streets. Politicians leave as soon as possible to return to grittier, but livelier, Brazilian cities.”

  The Brasilia master plan was partly based on architect Le Corbusier’s misplaced utopian vision of creating the ideal city. Corbusier’s work also included the Indian city of Chandigarh with similar consequences. This was, in the words of John Kay, “the hope that rational design by an omniscient planner could supersede practical knowledge derived from a process of adaptation and discovery.”

Many overfunded startups suffer from the same fate. When you have access to seemingly unlimited resources, it’s easy to be fooled into thinking you can build your exact vision into reality. But these visions generally exist in a complex world of human culture, desires, and economic feedback loops.

• Quibi — the $1.8 billion funded short-form media startup — is case in point. Ultimate success or failure remains to be seen but results have yet to come close to expectations. They had the resources to build a particular product and business model into reality — a reality where most customers don’t seem to want what they’re selling. A destination was chosen on a map that couldn’t be seen.
• Magic Leap has an amazing vision of seamless AR glasses to enable a digital layer on top of the real world. The actual objective in this case might actually be the right destination. But the complexity of the problem means the approach still can’t be direct. Currently their product has seen minimal success as they struggle to find a sustainable business model.
• WeWork seemed to start at the right level of abstraction, then an ambitious “visionary” founder is given unlimited funds, and a direct approach is applied to a still-oblique problem.

Summary: The Right Strategy for the Right Level

All efforts, and the efforts within them, can be placed on a ladder of abstraction. The higher up you go, the less concrete the objectives and less straightforward the methods to achieve them.

Where the effort falls on the scale is critical to the strategies for making progress on them. Direct, methodical approaches are only appropriate at lower, smaller-scale levels. This is where it’s good to plan details, use processes, and keep things consistent.

You can still have a grand, abstract vision. You just need to wayfind to get there: working from the bottom up, adapting and evolving the path while shaping and refining the details of the goal. Keep things flexible, adaptive, and opportunistic at the top.

---

Footnotes

1. Organizations: Given the definition of an effort above, what about coordinated groups of efforts or goals? ↩︎

   In the book The Origin of Wealth, Eric Beinhocker describes organizations as “goal directed, boundary-maintaining, and socially constructed systems of human activity. . . . There is a boundary distinguishing the inside world from the outside world, and the goals of the organization drive activities that lower entropy inside the organizational system.” This is the best description I’ve come across given its abstract nature. But I’d like to propose simpler, yet still compatible definition.

   An organization is a group of people pursuing one or more ongoing efforts, generally with the same high-level objective. This means an organization can be anything from seven-person hunting parties, to fleets of exploratory vessels, to philanthropies, to a multinational conglomerate.

---

This essay is one of an ongoing series of frameworks and stories on building. Check out the intro, Build Series, and a collection of all the essays here.