In the previous essay, we asked what an event horizon is.
Our answer was deliberately relational.
The horizon is not a wall.
It is not a material surface.
It is not a container.
It is a feature of the causal structure: a boundary in the structure of possible future relations.
That distinction allowed us to make another important distinction.
Information becoming inaccessible is not the same as information being destroyed.
An event horizon certainly changes what can be related to what from the perspective of the distant exterior. But that alone does not establish that the information associated with an infalling system has ceased to exist.
We are now in a position to ask the question that seemed obvious at the beginning of our investigation:
What exactly is the information that falls into a black hole?
The answer may be:
nothing.
Not because there is no information.
But because information is not the sort of thing that falls.
The grammatical trap
We say:
The particle falls into the black hole.
That is straightforward enough.
We say:
The particle carries information.
Still reasonable.
Then we say:
The information falls into the black hole.
And something subtle has happened.
The grammar has transformed a relation into a thing.
The particle is an actual physical event or system.
Its informational significance is constituted by its relations to possible states, other systems, and other events.
But once we say that the particle has information, it becomes natural to imagine information as an additional object attached to the particle.
Then the particle crosses the horizon.
So the information crosses with it.
The black hole has swallowed the information.
We have now created a little drama in which two things fall through the horizon:
the physical system, and its information.
But only the first was ever a thing.
The second was a relational structure.
This may seem like a linguistic nicety.
It is not.
The entire information paradox is concerned with whether that relational structure survives the evolution of the physical system.
So we need to stop treating the structure as though it were a passenger.
What actually crosses?
Suppose a quantum system falls toward a black hole.
There is an actual physical system.
It has a quantum state.
That state describes a structured space of potential actualisations.
The system interacts with its environment.
It becomes correlated with other physical systems.
Eventually, the system crosses the event horizon.
What has crossed?
The physical system has crossed.
Its energy-momentum has contributed to the gravitational configuration.
Its physical degrees of freedom are now associated with the black-hole interior.
But what about the information?
Our relational answer is:
The information is not another entity crossing alongside the system.
Rather, the system has entered a new relational regime.
Its possible future relations to the exterior have changed.
Its correlations with other systems have been reorganised.
The structure of possible observations has changed.
The informational consequences of its state have therefore changed.
But nothing called “information” needs to have passed through the horizon.
Information belongs to relations
Recall the formulation from our second essay:
Information is the structure of distinctions and constraints within a space of possibility.
Suppose a quantum system can be in states (A), (B), (C), and so forth.
Its actual state matters because the alternatives matter.
If it is actually in (A), then the world is different from the world in which it is actually in (B).
The information is the physically meaningful distinction between those alternatives, together with the relations connecting them.
Now suppose the system crosses a horizon.
The actual state does not suddenly cease to be the actual state.
What changes is the system's relation to the rest of the causal structure.
An external observer may no longer be able to interrogate the system in the same way.
The correlations available to that observer are altered.
The possible future signals change.
The system has entered a region with a radically different causal structure.
But that does not by itself erase the distinctions among the possible states.
The distinction between (A) and (B) does not need to fall through the horizon.
The physical evolution of the system carries the consequences of that distinction forward.
The question is what form those consequences take.
A simple analogy: the sealed room
Imagine that I place one of two different objects inside a sealed room.
Object A is a book.
Object B is a photograph.
Once the door is sealed, you cannot see which object is inside.
Your information has decreased.
But nothing has necessarily happened to the distinction between A and B.
The room contains one or the other.
The difference remains physically instantiated.
What has changed is the relation between the contents of the room and you.
The information has become inaccessible, not necessarily destroyed.
A black-hole horizon is vastly more profound than a sealed room, of course.
But the logical distinction is useful.
The inability of an observer to access a distinction does not establish the destruction of the distinction.
The information paradox requires a stronger claim.
It requires that, after the black hole has disappeared, the final physical state no longer preserves the distinction between the possible initial states.
That is a very different proposition.
But the black hole evaporates
Here we reach the real difficulty.
If the black hole simply remained forever, we could say:
The information is inside.
That would be an unsatisfying answer from the perspective of an outside observer, but it would not necessarily be a violation of unitarity.
The information could remain encoded in the complete quantum state of the black-hole interior.
The problem becomes acute because black holes are expected to evaporate.
If the black hole eventually disappears, what remains?
Suppose two different initial states formed two black holes.
Call them:
Black hole A
and
Black hole B.
If their Hawking radiation is exactly the same thermal state, then the final exterior states do not distinguish the two initial conditions.
The relational distinctions present initially appear to have disappeared.
That is the genuine information problem.
The question is no longer:
Did information cross the horizon?
It is:
Does the final physical state preserve the distinctions that distinguished the possible initial states?
This formulation is much closer to the heart of the matter.
Information does not have to come back out
This also changes the way we should think about the common phrase:
The information must come back out.
Why?
Because information is not a substance that went in and therefore has to return by the same route.
If the information is relational, then preservation does not require literal transport.
A relation can be transformed.
A correlation can be redistributed.
A distinction can become encoded in a different set of degrees of freedom.
A physical system can preserve information without preserving the original representation of that information.
Consider an ordinary computation.
A computer begins in one state.
It undergoes a sequence of transformations.
The final state may look nothing like the initial state.
Yet, if the computation is reversible, the final state preserves enough structure to reconstruct the initial state.
The information has not remained in the same physical form.
It has remained in the relations among states across the transformation.
This is a more useful model for black-hole information than the picture of information as a conserved fluid.
Information can be transformed
Suppose, then, that the black hole evaporates.
The initial state has a complicated relational structure.
The horizon forms.
The interior and exterior degrees of freedom become related in new ways.
Quantum fields produce Hawking radiation.
The black hole loses mass.
Eventually the black hole disappears.
There are now two broad possibilities.
Either the final radiation and remaining degrees of freedom preserve the distinctions among the possible initial states.
Or they do not.
If they do, then information has been transformed.
It may no longer resemble the information associated with the original matter.
It may be distributed across subtle correlations among many quanta.
It may be extraordinarily difficult to recover.
But it has not been destroyed.
If they do not, then we have genuine information loss.
The distinction is stark.
And importantly, neither outcome requires information to have been a thing at any stage.
The relational meaning of unitarity
This gives us a new way to think about quantum unitarity.
The usual statement is that quantum evolution preserves information.
But what does “preserves” mean under our ontology?
It means that the evolution maintains the relevant structure of distinctions and correlations.
An initial state is transformed into a final state through a lawful mapping.
The final state remains sufficiently related to the initial state that the distinctions among possible initial states have not simply been erased.
Unitarity therefore need not be understood as the conservation of an information substance.
It is the preservation of relational structure under quantum evolution.
This is a much more natural formulation for our purposes.
It also makes the black-hole problem sharper.
The question becomes:
Is black-hole evaporation a transformation that preserves the relational structure of the initial quantum state, or one that destroys it?
That is the question we actually need to answer.
What happens to the correlations?
The word correlation now becomes central.
Suppose the initial state of a collapsing object contains enormous numbers of correlations.
The positions of its constituents.
Their momenta.
Their spins.
Their quantum phases.
Their interactions.
Their entanglements.
Their relations to the surrounding environment.
The informational content of the system is not simply a list of properties stored in individual particles.
It is the vast structure of correlations among degrees of freedom.
When the system collapses, those correlations do not necessarily vanish.
They evolve.
Some become correlations among interior degrees of freedom.
Some may become correlations between the black hole and its surroundings.
Some may become encoded in the Hawking radiation.
The information problem therefore becomes a problem of correlation evolution.
Which correlations survive?
Which are redistributed?
Which become inaccessible?
Which, if any, are genuinely destroyed?
That is a far more relational formulation than the image of a packet of information disappearing into a hole.
The interior is not a storage box
There is another consequence.
If information is relational, then the black-hole interior should not be imagined simply as a storage container.
The phrase “the information is inside the black hole” sounds as though there were a box containing a collection of files.
But the physical state inside the horizon is itself dynamically evolving.
Its degrees of freedom are interacting.
Its correlations are changing.
The geometry is changing.
The horizon is changing as the black hole evaporates.
There is no static container in which information waits patiently for retrieval.
There is a relationally evolving physical system.
The relevant question is therefore not:
How much information is stored inside?
but:
How is the informational structure represented in the evolving relations among the degrees of freedom?
This formulation is harder to visualise.
But it may be closer to the physics.
A change in causal accessibility
The horizon does, however, produce a genuine transformation.
Before crossing, the infalling system can have future-directed causal relations with the distant exterior.
After crossing, it cannot.
This means that the information associated with the system becomes inaccessible through those causal channels.
From the perspective of the exterior, the relational structure has been partially hidden.
But hidden is not destroyed.
This distinction is especially important because it allows us to separate two questions that are often conflated:
Can the information be accessed?
and
Does the information still exist in the complete physical state?
The first is a question about causal accessibility.
The second is a question about the global relational structure.
The information paradox arises only if the answer to the second becomes no.
The role of the observer
This also gives us a more nuanced understanding of observers.
A distant observer has access only to certain relations.
An infalling observer has access to others.
Neither perspective necessarily contains the complete relational structure of the physical process.
This does not mean that information is subjective.
It means that information is always information within a network of physical relations.
A correlation between two systems is physically real even if a particular observer cannot access it.
Likewise, the fact that an exterior observer cannot reconstruct the state behind a horizon does not imply that the state has ceased to have the relevant internal structure.
The observer's epistemic limitation should not automatically be turned into an ontological claim.
This is one of the places where relational ontology provides a useful discipline.
It forces us to distinguish:
what exists in the physical relations
from
what a particular observer can reconstruct from the relations available to them.
The danger of the word “lost”
Perhaps this is why the word lost is so dangerous.
If I lose my keys, the keys still exist.
They have simply ceased to stand in the expected relation to me.
If a message is lost in transmission, the message may or may not have ceased to exist as a physical pattern.
If a correlation is destroyed, however, something more fundamental has changed.
The relation itself has ceased.
So when physicists say “information loss”, we need to know which of these senses is intended.
If it means:
the information can no longer be recovered by an exterior observer,
then the horizon is sufficient to explain the problem.
If it means:
the information has been redistributed into correlations that are extraordinarily difficult to reconstruct,
then quantum mechanics has no obvious reason to object.
But if it means:
the final physical state contains no relational structure capable of distinguishing the initial alternatives,
then we have something much more serious.
That is the version of information loss that threatens unitarity.
The possibility of disguised preservation
This also suggests why information preservation might be difficult to recognise.
Suppose the initial state is:
A.
The final state is:
B.
At first sight, A and B may look completely different.
We might therefore say that the information has been lost.
But if B is related to A by a reversible transformation, then the information may be entirely preserved.
The representation has changed.
The relational structure has not.
This is precisely why the idea of information as a substance is misleading.
Substances are preserved by remaining somewhere.
Relational structures are preserved by maintaining the appropriate mappings and distinctions.
A black hole could therefore radically transform the physical representation of information while preserving the information itself.
Indeed, if unitarity is correct, something like this must happen.
The challenge is to understand how.
The black hole as an information transformer
Perhaps, then, we should replace one metaphor with another.
The black hole is not an information destroyer.
Nor is it necessarily an information container.
Perhaps it is an information transformer.
A collapsing star takes an extraordinarily complicated set of physical relations and subjects them to a radical gravitational reorganisation.
The horizon changes causal accessibility.
The interior evolves.
Quantum fields produce radiation.
The black hole loses mass.
The correlations are redistributed.
Eventually the original representation is gone.
But the crucial question remains:
Has the relational structure that distinguished the initial possibilities survived the transformation?
If yes, then the black hole has transformed information without destroying it.
If no, then we have genuine information loss.
The paradox is therefore not about the disappearance of a thing.
It is about the possible disappearance of a distinction.
That is a much more profound problem.
A distinction can disappear
We should not, however, make our ontology do too much work.
Relational ontology does not guarantee that information is preserved.
A physical relation can genuinely cease to exist.
If two systems stop being correlated, that correlation is gone.
If a physical process irreversibly destroys a distinction, then the informational structure associated with that distinction can genuinely disappear.
So we cannot simply say:
Information is relational, therefore it cannot be destroyed.
That would be a philosophical sleight of hand.
Relations can change.
Relations can break.
Relations can be erased.
The question is whether the fundamental dynamics permit the complete destruction of the relational structure encoded in a quantum state.
That remains a physical question.
Our ontology changes the terms in which we ask it.
It does not decide the answer in advance.
The deeper problem: identity across transformation
There is, however, another issue lurking here.
Suppose the black hole evaporates completely.
What would it mean to say that the information in the final radiation is the same information as that associated with the initial collapsing matter?
The phrase “same information” seems harmless.
But it invokes identity across transformation.
Under an object-centred ontology, we might imagine information as a thing that persists through changing representations.
Under relational ontology, identity must be established through the preservation of relevant relations.
The final state does not need to resemble the initial state.
It needs to retain the structural distinctions that make it possible, in principle, to distinguish the initial alternatives.
This is a very different conception of identity.
The question becomes:
What relational continuity makes two radically different physical states count as informationally equivalent?
That may turn out to be one of the deepest questions in the entire paradox.
The information that falls in
We can now return to our title.
What is the information that falls in?
Strictly speaking:
there isn't any such thing.
A physical system falls in.
Its state evolves.
Its correlations evolve.
Its causal relations change.
Its quantum potential is transformed.
Its informational structure participates in that transformation.
But information does not need to be imagined as a separate entity crossing the horizon.
The phrase “the information falls in” is therefore a useful shorthand only if we remember what it conceals.
It conceals a transformation in a relational structure.
And once that is made explicit, the question changes.
We no longer need to ask:
How can information get back out?
We need to ask:
How can the relational structure that distinguished the initial possibilities remain represented in the final physical state?
That is a much more interesting question.
And it leads naturally to the next stage of the paradox.
Evaporation
Suppose the black hole evaporates.
At first, the situation seems simple.
Matter falls in.
The horizon forms.
Hawking radiation emerges.
The black hole loses mass.
Eventually, the black hole disappears.
But if information is relational, the disappearance of the black hole is not the disappearance of a container holding information.
It is the disappearance of an entire gravitational regime.
The relational structure changes again.
The horizon shrinks.
The causal relations change.
The degrees of freedom associated with the black hole interact with the radiation.
The final state may contain correlations that are extraordinarily difficult to see.
The question is whether those correlations are sufficient to preserve the distinctions present in the initial state.
And here we reach the point at which the original paradox becomes unavoidable.
Because if the black hole disappears completely, there is nowhere obvious for an independently existing interior to remain.
The information cannot simply be said to be “still inside”.
The container is gone.
So what happens to the relations?
That is the real question.
And it is the question we will confront in the next essay.
The black hole evaporates.
When it does, what must remain for us to say that the information has been preserved?
Perhaps the answer is encoded in the radiation.
Perhaps it is distributed across correlations that look thermal locally but are not thermal globally.
Perhaps our notion of locality itself must change.
Or perhaps something deeper has to give.
The next essay will therefore ask not:
Where did the information go?
but something more exacting:
What must remain invariant for information to count as preserved?
Once we ask that question, the paradox may finally reveal what it has been asking us all along.
No comments:
Post a Comment