Modality

The reading you are doing is one trajectory of many that were open when you picked up this book. You could have closed it. Skipped ahead. Set it down for a week.

The trajectories you did not take are not nothing. They are not happening either. Real as trajectory-space — the structural condition under which you could have taken them. Not real as record-history, because the record at your site is the one you entered.

The reading you are doing is the actual. The others were the possible. That you, a body with eyes and a brain, must be doing something at this moment is the necessary at the resolution of biology.

Three categories. One situation.

Late seventeenth-century philosophy asked what makes two plus two equals four necessary and the sky is blue merely contingent. By the mid-twentieth century that had become modal logic — formal systems for necessarily and possibly, which produced theorems without saying what the words meant.

In 1959 necessity was read as truth in all possible worlds, possibility as truth in some. The semantics was clean. The question became what the worlds were.

In 1986 the boldest answer. The worlds are real — complete universes, each as real as ours, causally isolated, populated. Actual is an indexical, like here. Modal realism. Expensive.

Nobody has displaced possible-worlds semantics as the formal instrument, and most of the field has moved past concrete worlds. Ersatzism: worlds are maximal consistent sets of sentences. Fictionalism: worlds-talk is useful pretence. Primitivism: modal facts are basic. Combinatorialism: worlds are recombinations of actual elements. Each keeps the semantics and pays less. The metaphysics stays contested.

One precision. This chapter does not reject the formalism. Possible-worlds models are powerful mathematics and most who use them do not believe in concrete worlds. The target is the inflation — from useful model-points to real manifolds.

The realist saw something true. The road not taken is not nothing. A world in which you did the other thing is available to be referred to.

The mistake is converting the availability of a reference into the existence of a complete second world. Nothing forces that.

One axiom runs. One now. No second actualisation, no parallel manifold. The conditions under which things might have gone differently are real — they are trajectory-space, the open paths at any coupling site. But trajectory-space is not a manifold of completed worlds. It is what is open before commitment closes the paths that did not commit.

A building with corridors. At each junction several are available. You walk into one.

Did the others exist before you walked? Yes. Do they exist after? Yes, in the same sense — corridors of this building, available to anyone who reaches that junction. They are not other buildings with their own inhabitants. And the corridor you did not walk is no longer open as your trajectory. It stays real as a path of the building and as counterfactual trajectory-space at that junction.

The realist saw the road’s realness and concluded it must be walked in some other manifold. It is real as trajectory-space. Not as another actualised path.

Possibility is real. The manifold is not.

Closing is not the annihilation of a parallel world. Closing is what non-selection is.

Ersatzism is right that no concrete worlds are needed, and cannot say what makes its sets consistent without modal facts it cannot ground. Fictionalism is right that worlds-talk commits nobody to worlds, and cannot say why the fiction works — it tracks something real. Primitivism is right that modality is real, and leaves the primitive floating. Combinatorialism is right that possibility is grounded in the actual, and cannot say which recombinations are permitted. Trajectory-space at actual sites supplies what each was reaching for.

One ground for what each was reaching for.

The actual is the now read backwards along R. What has been written and kept.

The possible is the now read forwards along trajectory-space. What is still open before commitment.

The necessary is the now read across the four conditions themselves. What they force whichever path commits. Two plus two is four because the binary break and record composition leave no room for another result. No record is erased without cost because R is what it is. The laws, because they are what the conditions produce at the resolution of phenomena.

One structure. Three directions. No domain for the possible to live in. Where do the possible worlds live is malformed. The possible obtains as trajectory-space at the writing-edge.

If you had left twenty minutes earlier, you would have caught the train.

The claim asks: in trajectory-space, at the moment of choice this morning, was leaving earlier an open path — and did the rest of the space permit catching the train? If so, true. It refers to trajectory-space at the actual moment, not to a parallel world in which you did.

This is the ground Five installed for causation. Modality inherits it.

But openness is not enough. A counterfactual does not rewrite the whole world. It makes a minimal intervention at the antecedent — alter the departure-time record — and holds the background fixed as far as the intervention permits — the timetable, the route, the traffic. Then it reads what follows.

That is the structural replacement for what the tradition called closeness between worlds. Not a distance across manifolds. A discipline about which records are held fixed.

Three classes. Some counterfactuals are false because the consequent is not reachable under the minimal intervention. Some need careful analysis because the downstream space is complex. And counter-possibles — impossible antecedents — are the most open case. The worlds account calls them vacuously true. Here they are an investigation of what the structural conditions themselves would permit if they were different. Not closed in this chapter.

Logical necessity is what committed distinction forces at the statement layer. If P is fixed as P, not-P is not the same record. Non-contradiction and valid inference are readings of that. Not a logical realm apart from the world.

Mathematical necessity is the same ground at the resolution Two installed. Two binary breaks composed with two more under the counting policy give four, with no degree of freedom. The angles of a Euclidean triangle sum to two right angles because, given those axioms, the structure leaves no room for another result. Not true in all worlds. What the conditions produce when records relate to records at that resolution.

Conceptual necessity — all bachelors are unmarried — is what language-records force at the resolution of meaning. Not a fourth domain. A later chapter.

Metaphysical necessity is what the conditions force at the resolution of identity and kinds. Hesperus and Phosphorus name one record-lineage. Discovered empirically, and once discovered, locked — one lineage cannot be two while remaining itself. That is Seven. Water is necessarily H₂O at the chemical-kind resolution. At the ordinary-language resolution the referent shifts, as the ship’s did, and the necessity attaches differently.

Nomic necessity — that the laws could not have been otherwise — is the most contested, and the position here is careful.

The laws are the four conditions read at resolution. Same conditions, same resolution, same laws. Different laws would need different conditions. A world with the same conditions cannot have different laws.

The claim is bounded. It is not that every constant has been derived in closed form — Four named the debts. It is that lawhood itself, and any law derivable from the four conditions, is necessary by derivation, not by stipulation. Where a parameter is not yet derived, its modal status is open.

Necessity at the structural floor. The floor is the four conditions. What they force is necessary. What follows from them follows necessarily. Why they are the floor is the corpus’s own question and is not closed here.

Every choice you have made had open trajectory-space at the moment of choosing. After commitment the other paths are closed for your record. They are not happening anywhere in your history. They were real before, as trajectory-space.

The sharpest anchor is quantum mechanics.

A silver atom prepared spin-up along one axis is sent through a field aligned on another. It registers at one of two positions, each with probability exactly one half. Run it many times and the split approaches one half to high precision. Run it once and one outcome registers.

Before measurement, a structured space of possible outcomes with exact weights. After, one definite record. What becomes of the unselected term is where the interpretations disagree — collapse, branching, hidden variables, updated information — and this chapter does not choose. It commits only to the operational fact every interpretation respects. A structured space before. A definite record after. The unselected outcome is not an additional ordinary record in the observer’s history unless an interpretation supplies one.

That is the physics-facing instance of trajectory-space, exact and tested since the 1920s.

The lived case has the same form. Before a decision, neural activity shows competing policies; one stabilises, the others decay. The phenomenology of choice is trajectory-space wide, then narrowed. Not a proof. A convergence.

The realist saw that the road not taken is not nothing, that counterfactuals are not empty, that possibility is structurally real, and that fictions and constructions are too weak to ground it. All correct.

What he got wrong was the category. Possibility does not need to be actualised somewhere else to be real. It is real as the openness of paths at the moment of choice.

Not a rejection. A relocation.

What wide means at high resolution — the next chapter and the formal papers. Modality and time — trajectory-space faces forward from the edge, R faces back, and their interaction is what the now is; open. Deontic necessity is what override makes available; epistemic possibility is what the site’s record permits given the evidence. Which quantum interpretation is right — open, and empirical.