A walk through the garden · Step 5 of 19 · What is there

Why trust tomorrow?

Two readings of one now. No gap between them, because they were never two places.

Branch of philosophy: Knowing (epistemology), and philosophy of science. How you can know what comes next, and what science has found when it finds a law.

Where the question lives

You reach for the railing on the stairs. It holds.

You did not stop first. You did not calculate, or weigh the odds, or hedge. You reached, it held, and you took the next step.

The cup holds the water. The floor holds your weight. The word on this line is followed by the next word. The sun rose this morning.

Every one of those was an expectation. You went from what has been to what will be. Philosophers call that induction. It works almost without fail. And for nearly three hundred years, nobody has justified it without going in a circle.

That is this step’s question. Why trust tomorrow?

Three more sit under it. What is time? What is a law of nature? What is a cause? Each has its own long history. This step walks them together.

What philosophy said

Will tomorrow be like yesterday?

The Scottish philosopher David Hume gave the problem its classic form, in 1739. “It has always happened” does not give you “it will happen”. Fire has burned every time. Tomorrow it might not. There is no contradiction in saying so. The step from past to future is a habit of mind. The habit cannot justify itself.

Four replies followed.

Make it probable. What has held in the past is more likely to hold again. But the rule that lets you say so is itself an induction.

Make it practical. Charles Sanders Peirce, and later Hans Reichenbach, said we must use it to survive. That explains the use. It does not show the use is right.

Make it formal. Bayesians build an exact system for how strongly you should believe something as evidence comes in. But the system takes for granted that evidence tracks what comes next. That is the very thing in question.

Admit defeat. Induction cannot be justified. Live without a foundation. Inside the old framing, that is the honest answer. For those who accept the framing cannot be made to work, it has stayed the default.

Then, in 1955, the American philosopher Nelson Goodman made it worse. Call an emerald grue if it is green when seen before some future date, and blue after. Every emerald ever seen has been grue, exactly as it has been green. So why expect green tomorrow, and not blue? Goodman’s answer was entrenchment: green is the word we have always projected. That moves the question. It does not close it.

What is time?

Behind induction sits time.

Isaac Newton said time is a container. It flows evenly, by itself, and events are placed in it. Gottfried Leibniz said no. Time is an ordering of events, one relative to another. The debate ran for two centuries.

In 1908 the British philosopher J. M. E. McTaggart argued that a moving now is incoherent. So, he concluded, time is unreal.

Albert Einstein’s special relativity, in 1905, had already shaken the container. Two observers moving past each other can disagree about whether two distant events happened at the same moment. The block-universe reading took this furthest: all events laid out in four dimensions, with no special now. It is the closest the tradition has come to what this step will show. It is still a container. Four dimensions instead of three. And it has no now at all. The 420 Code keeps one now: the place where the writing happens.

Every container version, from Newton’s to the block, treats time as coming first. Events happen in time. Leibniz came closer. For him, time is an ordering. But he never said what it orders.

What is a law of nature?

Drop an apple. It falls.

Drop it in another room, another continent, another century. It falls. The universe does not need to be persuaded to repeat itself. The repetition is the question.

Newton, in 1687, gave the same equations for the moon and the apple, the tides and the planets. What is a law, that it holds everywhere, always? Two answers have led since.

Regularity. Hume again. A law is a pattern that has been seen to hold. Nothing more. This sees something real. Laws are read off the world. Nobody sees the necessity: you see one billiard ball strike another and the second move, not the “must”. But regularity cannot say why the pattern goes on, beyond “so far, so good”. And it cannot tell “apples fall” from “every coin in my pocket is silver”. Both are patterns in what has been seen.

Necessity. Laws are more than what we happened to notice. Armstrong said they are relations of necessitation between universals. Lewis, coming from Hume’s side, said a law is whatever the best summary of all the facts includes. This sees the direction regularity misses: the second ball had to move. But it cannot say where the necessity lives. Or it makes laws features of our descriptions, not of the world.

Regularity sees the reading, not the source. Necessity sees the source, not where it sits.

What is a cause?

Strike a match. The head lights.

Something happened that would not have happened without the strike. The tradition called it causation, and spent three hundred years looking for what makes the connection tight.

Hume looked hard. Strike, then flame. Many times. Always in that order. He did not see a binding force. He saw constant conjunction: two things that always come together, one first. The mind supplied the connection. The world held only the pattern.

He saw something real. No logic binds the strike to the flame. You can imagine the strike without the flame. But regularity cannot do anything else. It cannot tell causes from accidents: day follows night, but night does not cause day. It cannot explain “if you had not struck it, it would not have lit”. And it cannot handle two assassins, where the first bullet kills and the second would have.

The second family says a cause is more than a sequence. Some speak of powers, or necessitation. Mackie gave a careful logic: the strike is one needed part of a set of conditions that is enough together, with the oxygen, the dry head, the friction, the warmth. Lewis said A causes B if, in the closest possible world where A does not happen, B does not happen either. Precise. And paid for with real possible worlds, each as real as this one.

This family is right that “because” is not “and then”. What none of its versions says is what the connection is.

What the 420 Code sees

The book does not answer the problem of induction. It dissolves it: it shows the question was asked wrongly.

The problem of induction and the problem of time rest on one premise. That past and future are two separate domains, with a gap between them that has to be bridged. Take the premise away, and both problems go together.

The one now

At step 2 you met the Actualization State, AS: the now, where the break is made, held and read. Here it does its full work.

The Actualization State is the whole of what the axiom produces, read as one. It is not something local, at your spot. It is the one now the structure has. Every place where a record is being written or read is a site, and every site is a local reading of the one now.

Three things follow.

The now is singular. One, not one per site. What each site reads is a cross-section of it, at the resolution the site can reach: the grain at which it reads.

What has been written is in the now, as what it carries. Not held in a container behind you.

What can next be written is in the now too, as its trajectory-space: the paths still open from here, given what has happened so far. Not waiting in a container ahead of you.

That is the hinge.

The past is what Record, the third condition, has kept: records written and not undone. The future is what the now permits next.

Two readings of one now. No gap between them, because they were never two places.

They are not the same kind of reading. The past reading finds what is written and cannot be unwritten. The future reading finds what is permitted and not yet decided. That one-way difference is what Record produces. It is why induction runs forward. The past is fixed because it has been written. The future is open because it has not been written yet.

Philosophy called that the direction of time, and treated it as a separate problem. It is the same fact.

Time is what Record does

Before the break, perfect symmetry holds no record of itself. No before. No after. Nothing to order.

The break produces records. Records cannot be undone, and each one carries the direction of its writing. A record being written is what makes before and after different at all.

So time is the direction Record keeps. Not a line that records are placed along. The direction of writing itself, read from the one now. The book’s short name for Record here is holding. Time is what holding is.

This is the inversion. Events are not in time. Events, written one way and kept, are what produce the thing you read as time. What time orders is records.

The container is a model. It is real as an experience. It is not what time is.

Why does the now feel special? Because it is where the reading happens. The only place it can.

Induction is reading, not leaping

Now the problem of induction changes shape.

In the old framing, induction was a leap from one domain to another. The leap needed a justification the framing could not supply.

There is no leap.

The regularities of the past are not patterns that happened to hold some number of times. They are consequences of the four conditions. Matter behaves as matter because what matter is constrains what it can do next. Fire burns because what fire is constrains what it does when it meets fuel.

Induction is reading those constraints from your site, and projecting what the now permits next under them. When you expect fire to burn tomorrow, you are not jumping a gap. You are reading that fire goes on being what the conditions make it.

When induction succeeds, the constraints held. When it fails, induction did not fail as a method. Your picture of the constraints was too coarse. The structure was doing something at a finer grain than your reading caught.

The justification the tradition wanted is not available, because it was looking at the wrong layer. What is available is the fact: the conditions produce regularities, and induction reads them.

Two objections press.

“That the conditions will hold tomorrow is itself an induction.” No. The four conditions are not a pattern seen many times. They are what must be true for any record to exist. You found them from the one record in hand, at step 2. Tomorrow’s records, when they exist, will meet them. Not because anyone inspected tomorrow. Because any record, anywhere, is what the conditions produce. That is going from a record to what the record needs. It is not induction.

“And grue?” The constraints govern particular quantities. Energy. Momentum. Charge. Spin. Grue is not at that layer. It splices two conditions together across a date. Someone built it. So the line is not natural against artificial. It is governed against constructed. The test is sharp. A term is governed if it can be stated without an either-or that switches at a date, or between sides. Induction tracks energy. It tracks momentum. It does not track an either-or with a date in it.

A law is what the structure does

Now the apple.

The book dissolves this question too. It does not choose between regularity and necessity. It shows the layer both missed.

A law of nature is a constraint the four conditions put on every record, read at the resolution of one kind of happening. It is not added to the world. It is what the world is.

The book’s picture is a building with load-bearing walls. The walls do not stand because a blueprint says they must, and they obey. They stand because of what they are: these materials, this shape, under these forces. The blueprint records what the walls do. If the blueprint says load-bearing and the wall is papier-mâché, the building falls. Not because the wall disobeyed. The conditions were not met. The blueprint is downstream. The walls are upstream.

A law is the same case. The universe does not consult the law before acting. The universe is the structure. The law is what we read when we examine what the structure does.

So both old answers were half right. Regularity was right: no necessity brought in from outside. Necessity was right: no mere coincidence. Both missed the layer beneath both.

The chapter shows it on three well-established laws of physics. Energy is conserved. Energy appearing or vanishing with no record elsewhere would be change with no source, or loss with no trace, and Record and Constraint forbid that. The speed of light is Constraint, measured in units we chose. Heat flows from hot to cold, and that direction is Record, named at the layer of heat. The exact amounts need more physics.

The same reading reaches quantum mechanics and general relativity, the two great theories a century of work has tried to unite. The chapter’s claim is that they were never two. Each reads the same four conditions at a different resolution. “The unification was always the framing’s job, never the universe’s.”

One number sits where those theories meet: α, the rate you met at step 2. The chapter reads it as the actualisation rate at the layer of electricity and light: how often a possible distinction becomes a recorded one. α is measured, not derived. The chapter holds that the measured value is forced: the structure allows no other.

A cause is a record on the move

The book dissolves this question as well. There was never a separate connection to find.

A cause is a record written at one site that, travelling at a bounded rate, changes the paths open at another site.

Strike the match again. The strike writes a record: friction, heat, the head reaching the temperature where it catches. That record travels at the rate Constraint permits. Across the head, into the air, through the chemistry of burning. At the flame, what can happen next has changed. The path where the match burns is open. The path where it stays dark is closed.

No connection was added. The propagation, the record travelling, was the connection.

Causation is not a fifth condition. It is Constraint, read at the layer where one record changes another.

Hume is kept, and moved. No logical necessity: he was right. But not nothing either. The connection is structural. A bounded rate. A definite path. A limit on what can follow what. That is enough, because the world does not run on logic. It runs on the four conditions.

The hard cases fall with one distinction. In the world, many things lie behind the flame: the strike, the oxygen, the dry head, the warmth. All one network of propagation. In explanation, you pick one part and name it the cause, against a contrast: the strike, as against an unstruck match. Two assassins fire, and the first bullet kills. The cause is the chain whose record actually reached the death. The second only would have.

And the “if”? If you had not struck the match, it would not have lit. Philosophers call an “if” like this a counterfactual: what would have happened, had things gone otherwise. You do not need other worlds for that. Remove the strike, and the chain that opened the flame’s path goes with it. The cause is the record whose removal changes the paths. Laws work the same way. The same axiom runs on the changed situation. Constraint does not bend in the “if”. Record does not fail.

One arrow

Now the two halves of this step meet.

Cause to effect, never the reverse. Past to future, never the reverse.

One arrow. Two names, because it is read from two positions. From inside one propagation, the earlier record constrains the later. That is causation. From across many propagations at once, earlier records sit on the past side and later ones on the future side. That is time.

Both are Record. Records do not unwrite. That is why cause and effect cannot be reversed. It is the same reason past and future cannot. A later event cannot unwrite a record already written.

The speed limit under all this is not only argued. It is measured. No test has ever found a record-carrying influence moving faster than light. And where one event can reach another, every observer agrees which came first. That is where the one arrow holds. Where nothing can pass between two events, there is no causal arrow between them to call time.

One case looks like an exception. Particles prepared together and carried apart give results that match in a way no local story can produce. Bell proved it in 1964. But no signal crosses between them. The book reads the two measurements as two readings of one structure, set up when the particles were prepared. Nothing travels between them.

What it costs you

You came for a guarantee that tomorrow will be like today. You did not get one. There was never a gap for a guarantee to cross.

You give up the river. Time does not flow past you, and you are not sliding along it.

You give up laws as commands. No law was issued. No rule was obeyed.

You give up the glue between events. The travelling was the connection.

And when a prediction fails, you cannot blame the method. Your reading was too coarse.

Sit here

Still growing

Why the now feels like a moving point, why memory and expectation feel so different, and why time seems to run faster or slower with attention: Chapter 11 names these and hands them on to other work. Why α has the value it has, and no other, is owed. The chapter holds that the value is forced. It names the switch that would break it, RES-4.4. In its own words: “The claim is sharp. The derivation is owed.” Most of the other constants of physics are owed in the same way. And some readings of quantum physics let a measurement now shape a state prepared earlier. The strongest of them are in real tension with Record. Chapter 5 does not wave them away. It carries them forward as open work.

Where it would die

Side paths

• This question in the canon — who asked it before, and what the 420 Code does with each answer.
• The Problem of Induction and the Nature of Time — Ø Dissolutions, Chapter 11. Watch · The full book — pages 286 to 313.
• Laws of Nature — Ø Resolutions, Chapter 4. Watch · The full book — pages 140 to 175.
• Causation — Ø Resolutions, Chapter 5. Watch · The full book — pages 176 to 220.