The Possible and Impossible Reframe
Describe a system by which transformations it permits, not by what it will do next
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 68%
Constructor theory swaps the prevailing framing for a possibility-based one. Instead of taking a physical system in isolation, feeding in initial conditions and evolving equations forward to say what it will do, you ask what can be done to it, and what can never be done to it no matter what you bring. Getting the system boundary right is the load-bearing move: a fully charged Tesla can roll at a hundred miles per hour, but charging it does not count as something done to the Tesla, because charging depletes another resource, so the real subject is a Tesla plus some electricity. Once boundaries are complete, you can state prohibitions that hold universally, such as a Tesla and anything cannot be made to exceed the speed of light. Deutsch distinguishes principles, which make statements about laws not yet discovered, from ordinary laws, and argues the resulting framing is more timeless and more explanatory.
Origin
Deutsch developed constructor theory with his colleague Chiara Marletto as a response to the initial-conditions-and-laws-of-motion picture of fundamental physics, which he argues is not time-symmetric and cannot express most scientific theories. Extracted from Naval.
Core principles
- 01Ask what can be done to a system, not only what the system will do.
- 02A transformation must be stated over the complete system, including every resource it consumes.
- 03Prohibitions on what is impossible carry more content than forecasts of what happens next.
- 04Principles constrain laws that have not been discovered yet.
- 05A framing that depends on a particular starting time explains very little.
How to run it
- 1
Flip the question
Replace 'given this state, what will happen?' with 'what transformations can be performed on this, and which cannot?'. This is the whole reframe in one move.
- 2
Close the system boundary
Include everything the transformation consumes. If your proposed operation depletes another resource, that resource is part of the system, not part of the background.
Pro tip Charging the Tesla is the test case: it looks like something done to the car, but it is really something done to a car plus electricity.
Watch out Boundary errors are the standard failure. They make impossible transformations look possible because the cost was hidden outside the frame.
- 3
Enumerate the permitted transformations
State what the complete system can be made to do, for example transporting a person from Oxford to London given a car and electricity.
- 4
State the prohibitions
Write the transformations that cannot be brought about no matter what else you supply. These carry the strongest content, because they hold against any resource you might add.
Pro tip Everyday statements often already have this shape: 'you cannot go faster than light' is simple and universally understood, while expressing the same content as a prediction is very difficult.
- 5
Promote the durable ones to principles
Mark the prohibitions that should constrain laws you have not discovered yet. A principle about the speed of light must also bind any new particle that turns up later.
- 6
Check for time-dependence
Reject any formulation whose content depends on a particular starting moment. Deutsch's objection to the prevailing framing is precisely that it privileges one date and evolves outward from it.
In the wild
Deutsch works the reframe through a car. A fully charged Tesla can roll along at a hundred miles per hour, so that is something it can do. Charging it, though, does not count as something that can be done to the Tesla, because you would then have to talk about a Tesla and some electricity, and what can be done to that combination is transport a human from Oxford to London. Once the system is complete, you can state a prohibition that holds universally: a Tesla and anything cannot be made to travel at one point one times the speed of light.
→ A crisp, resource-honest statement of possibility and impossibility replaces an equation-driven forecast, and captures content that is very hard to express as a prediction.
Deutsch and Marletto argue that existing laws can either be re-expressed in constructor-theoretic terms or are approximations to laws that can be. Thermodynamics is the showcase: at present it is only expressed approximately in conventional terms, but constructor theory can state it exactly, because the laws include the things you might need to bring in order to do something. The prohibition then says that no matter what you bring, you will not be able to violate the second law.
→ A law that conventional physics only approximates becomes an exact statement about which transformations are impossible.
Common mistakes
Leaving consumed resources outside the frame
If a transformation depletes something you did not include, you are describing a different system. Every claim about what is possible then becomes unreliable.
Confusing a principle with a law
A law describes a known system; a principle constrains laws that have not been discovered. Collapsing the two loses the reframe's main advantage.
Smuggling in a start date
Any framing that requires knowing conditions on a particular day inherits the time-bound weakness the reframe exists to escape.
Is it for you?
Best for
Researchers and systems thinkers who need a framing that states what is permanently possible or impossible rather than forecasting a trajectory.
Not ideal for
Short-horizon operational forecasting where a trajectory from current conditions is exactly what is required.
From the transcript
“what Constructor Theory does is say we have the system what can we do to it what can be done to it”
“charging it doesn't count in Constructor Theory as something that can be done to it”
“the real law about the constancy of the speed of light is a Transcendent law it's what I call physical principle rather than the law”
From the episode
The Deutsch Files II