1. Physics With Minimal Assumptions

“I am this thing”. That is perhaps the most basic fact I know of. It is not a conclusion based on reasoning, but something much more instinctive: “I know how it feels.“. For me, the only subjective experience is from this thing.  It’s at the very center of my perspective. Let’s call it “self”.

I learn about my surrounding by interacting with them. Via these interactions, I formed the conception of external objects and the world around me. If I study the interactions carefully, I would discover certain patterns: rules that can explain and predict them. These rules also help me to describe external objects as the interactions’ counterparties.

Then there’s a crucial realization: the first-person perspective I’m experiencing doesn’t have to be the only valid perspective. I realized external objects also interact with their environments. I can imagine thinking from a particular object’s viewpoint. So that thing would become the “self”, and I become an external object interacting with it.

Back to the rules that my interactions seem to follow. There are restrictive rules, that only work for a limited range of perspectives and interactions. And there are other, more general rules applicable to a wide range of perspectives. I should try to induce rules of this type as they give a deeper insight into how the world works. Let’s call the general rules “physical laws”, and the depictions of the world using these laws “physical descriptions”.

2. Classical Physics Is Weird

This parsimonious interpretation of physics has some implications. First of all, physical analysis has to be conducted from a prespecified perspective. It can be the natural first-person perspective of you or me or the perspective of any object. Which perspective is the reasoning to be conducted from i.e. what is the “self”, is an exogenous input, not something to be explained by physics.

Following the above, studying interactions received by the “self” (the thing at the perspective center) enables conceptualizing external objects as the actions’ counterparties. Physical descriptions of external objects are based on these interactions.

Finally, because objects are described based on their interactions with the self, the “self” itself is not describable.  I.E. the “thing at the perspective center” is not within the domain of explanation. To physically analyze something we must not take the perspective of that object, but the perspective of some other things interacting with it.

These implications fit very nicely with the interpretive challenges of quantum mechanics. For example, the special role of the “observer” is to be expected. It is simply the prespecified self, from whose perspective the analysis is conducted. “Measurement” is not a convoluted notion either, just an interaction with the “self”. Physically describing something not interacting with the observer is impossible.

With this approach, quantum mechanics seems unsurprising. Instead, the interesting question is how come classical physics can work by thinking directly in terms of an absolute “physical reality” without paying any attention to perspectives?

3. Energy Saving Approximation

Directly thinking in terms of the “absolute reality” is an intuition formed by living in the macroscopic world. In our daily life, we only conceptualize large objects and do not make precise physical descriptions.

Imagine there is a glass bottle in front of you. A camera is also pointing at it. From your perspective, some of your interaction with the environment is, directly or indirectly, affected by that bottle. Same for the camera’s perspective. Therefore the bottle can be conceptualized from either perspective by studying the interactions. Yet, even using the same physical laws, the descriptions from the two perspectives are inevitably going to be different since they are based on different sets of interactions.

However, for a coarse-grained characterization, the detail differences are swept away. For example, the photons get in my eyes bounced off from different atoms of the bottle than those photons get into the camera, therefore the two perspectives won’t give identical descriptions at the atomic level. But if we are only interested in macro-level features such as “the shape of the bottle”, then both perspectives have countless interactions giving enough information well beyond the need to deliver these rough depictions. Their description of the macro features would be virtually identical.

Furthermore, because daily objects are massive, there would be numerous interactions for any practical choice of perspectives. The interactions are effectively continuous so the objects can be described at any given moment. Effectively, for coarse-grained macro-level physical analysis, there are two important traits: perspective invariability, and object permanence. This gives rise to the idea of the “absolute physical reality” that we can reason directly about. Without needing to define which perspective is the analysis conducted from.

And we have good reason to do so. To correctly capture the descriptive differences across perspectives requires analyzing minuscule interactions highly accurately. That is perceptively and computationally very expensive yet offers marginal benefit for our survival. It is much more efficient to treat external objects as absolute fundamental existences rather than derived conceptualizations.

Nonetheless, thinking in terms of this “absolute reality” is just a useful approximation. The approximation breaks down gradually as we begin focusing on the microscopic. Where interactions are few and far between and a high level of accuracy is required.

4. The God’s Eye View and Schrodinger’s Box

Some may consider the classical way of thinking in terms of absolute reality as merely conducting the analysis from a god’s eye view. So it is no different from the perspective-based understanding of physics as laid out earlier. But that is not so.

There is nothing inherently wrong with thinking from an imaginary thing’s perspective. However, reasoning directly in terms of physical reality uses a view from nowhere. There is no self: nothing is at the perspective center. So everything is part of the external environment. Hence the domain of physical explanation includes the whole world. Furthermore, such a view is omniscient as it doesn’t need interactions to describe objects. It is an unjustified assumption to say physical laws applicable from the perspective of ordinary things can also be used from such a supernatural perspective.

However, due to intuitions formed in daily life, we tend to use the god’s eye view even when the approximation breaks down. Doing so can result in perplexing conclusions. Schrodinger’s cat is perhaps the best-known example.

Let’s assume the ideal box cuts off all interactions between the inside and the outside. For things inside the box, there are numerous interactions available to describe the cat. From such perspectives, the cat is clearly either dead or alive. For things outside the box, no interaction gives any information about what’s happening on the inside. From these perspectives, the current physical state of the cat is indescribable. The superposition of alive/death is not the absolute physical state of the cat from a god’s eye view. It is merely a prediction of what kind of interactions someone outside the box would expect.

If the experiment uses an ordinary box that blocks some but not all interactions, the conclusion would be different. In this case, the scale of objects does make a difference. Because a cat is massive and “alive or dead” are coarse-grained characterizations, it’s almost certain that there would be enough interactions carrying information necessary to make that description, even for someone outside but reasonably close to the box (or think from the perspective of “all things outside”). We might not analyze all the interactions to make the description. But the cat is in a definitive physically describable dead or alive state from such perspectives.

5. Some Metaphysical Stance

One thing that needs to be cleared is this interpretation does not deny the reality of the world or endorse solipsism. On the contrary, it takes the reality of the world as a primary postulate. It believes other things’ existence is as real as my own. Therefore their perspectives are as valid as mine. It simply suggests we have no way to reason about reality directly. The classical notion of “absolute physical reality” does not cut it. We are bounded to learn the true nature of the world through perspectives and interactions. They are the innate spectacles we cannot shake off.

It is also noteworthy that this interpretation puts physics and consciousness in an alternative reasoning framework. As discussed, physics cannot describe the thing from whose perspective the analysis is conducted. Its scope of explanation does not include the “self”. On the other hand, subjective experience and consciousness exclusively apply to the self.  (i.e. From my perspective, the only subjective experience that exists is that of my own.) So physics is released from the duty of explaining subjective feelings or closing the explanatory gap. As it is ontologically impossible.

Closely related to consciousness, free will is also a feature inherently applicable to the self. One can only contemplate choices or make decisions from a first-person perspective. In contrast, physically analyzing an external object to deduce its output does not involve the notion of free will at all. (This is actually the cause of Newcomb’s Problem as I have written earlier.) Due to this separation, physics cannot be used to deny the existence of free will.

Finally, since this interpretation does not consider consciousness as a physical feature but inherent to the prespecified perspective center, it is a form of panpsychism. It is not suggesting everything’s subjective experience feels similar to mine or yours. There is no way to know that.  But a thing can be considered conscious as long as we choose to reason from its perspective.

New to LessWrong?

New Comment
10 comments, sorted by Click to highlight new comments since: Today at 5:59 AM

Interestingly,  physics in the 20 century moved in the direction of more and more observer-centered perspective: first, special and general relativity, then quantum mechanics and later anthropic principle.   

Yes, and the observer-centered perspective is accompanied by the rejection of the notion of "absolute reality". Like relativity to absolute spacetime. In my opinion, the interpretive challenges of QM are nothing different. "Observer" is simply anything's perspective one wishes to conduct the physical analysis from. 

My original motivation for this topic was trying to solve anthropic paradoxes, which are surprisingly closely connected to quantum interpretations and the metaphysics of science and consciousness. 

The "God's eye view" or "view from nowhere" reminds me of the position of the eliminativists in philosophy of mind. They assert that a physical world does exist, but that we are mistaken in thinking we have 1st person experiences like redness and painfulness.

Doesn't the concept of “absolute reality” just mean that when observer 1 observes object A from a particular perspective P in a specific point in time t, it observes the same as if this would done by observer 2? If you were looking at the bottle form the perspective the camera took at the time that it did it, you would have seen the same. So there there is an absolute reality.

I probably didn't understand your point.

"Absolute objectivity" is treating physical objects as the foundation. I.e. Observers 1 and 2 should give the same depiction of the bottle, on pain of being wrong, because they are describing the same physical object.  It attempts to get rid of perspectives and reason directly in terms of this physical objective reality by taking a "view from nowhere". It fits our usual intuition of scientific objectivity: "think about how things really are instead of how they appear to be from various viewpoints". 

If we treat perspectives to be axiomatic then the underlying reason for the same depiction by distinct observers is different. For example, some of my interactions with the environment are affected by some cause. By analyzing these interactions using some theories, I conceptualize this cause as an object: "the bottle". Note, that not all of the affected interactions are directly between me and the bottle. I could also have interactions from the camera that are affected by the bottle. (reflections from the lens, etc) As long as the theories I used are self-consistent, analyzing the direct interactions and indirect interactions should give non-conflicting depictions of the cause (i.e. the object). For objects that affect lots of interactions, i.e. macroscopic, rough depictions from direct and indirect interactions would effectively be the same. Therefore if we imagine using the same theories from the camera's perspective, it would, effectively, give the same description of the bottle. 

So the bottom reason for why Observers 1 and 2 must give the same physical depiction is: 

1.  Correct physical theories ought to be self-consistent and could work from different perspectives. (uncontroversial)

2.  Observers 1 and 2 are interacting with each other, and these interactions are affected by the object they are describing.

No2. holds up pretty well in our daily life. That's what gives us the intuition of an invariant absolute reality. But if there are no interactions between the observers; if the cause(object) has no effect on these interactions; or even if the effect is too insignificant to carry enough information, then the physical descriptions from different observers have no reason to be the same. There is simply no way to say. 

I still don't get it, sorry.

No2. doesn't seem needed to me. You can observe at different times so observers do not interact. You may not see exactly the same, yes.

Probably my objection comes from that I think that "there is simply no way to say [whether the object is the same]" does not imply that absolute reality does not exist. But it is highly probable that I just don't understand something in your reasoning (eg. I don't know what "treat perspectives to be axiomatic" means).

Perspectives here do not just mean spatiotemporal locations. More importantly, it means which thing you are. e.g. you are experiencing the universe from the perspective of a particular human being named mikbp. 

Treating perspectives to be axiomatic means any physical description has to be based on the perspective of something. We cannot think from a god's eye view, and directly describe the world as it is. As in think directly in terms of the "absolute reality". 

If there are no interactions between two observers, then how can either of them say they are giving the same descriptions for an object? To each of them, the existence of the other observer is physically moot. There is ontologically no comparison at all, let alone saying their description are the same or different.

how can either of them say they are giving the same descriptions for an object?


They cannot. But why is it relevant? The fact that they don't know it does not mean that their description is not the same. In addition, different observers may get different descriptions (eg. an infrared camera does not record the same as a normal camera). That does not change the object observed, just what we know about it. As long as you don't think you know everything about the objects you observe, this is fine. The more we know, the more physical laws we can infer.

Taking the God's eye view is restricted to specific problems and it just implies that you know all you need to know for that problem. Of course that's not something one can do all the time.

You can say even when neither of them can compare the descriptions it still means their descriptions are the same. But from what perspective is this statement made? It is from a god's eye view that directly thinks in terms of reality.  

A self-consistent theory only means from any perspective, analyzing the interactions affected by an object cannot give conflicting descriptions of said object. If the only interaction upon you from the camera is an infra-red photo of the bottle, and from your direct interaction with the bottle you concluded it's red, and from the infra-red photo you concluded it is green, then there's something wrong with your theory that needs rectifying. 

Besides, applying the same theories from the cameras' perspective does not necessarily mean an infrared camera and a normal camera would give different (macro-scale) descriptions of the bottle. Sure the cameras give different photos. But the interactions they receive from the bottle are not that different. The infrared camera still gets bombarded by photons of all energy levels. They don't just pass through the camera unaffected without interacting with it. And physical descriptions are based on them. If instead there is an infrared camera that let photons of higher energy levels pass through without interacting, as if they don't exist to the camera, then it would give different descriptions. 


Look, what I am arguing is that a foundational "absolute reality" is an additional assumption based on our intuition formed in the daily environment. Thinking directly in terms of this absolute reality is problematic outside of our intuitive environment, e.g. high speed or microscopic so we should get rid of this assumption however convenient it proved to be in the past. 

I remembered Thomas Nagel said we get the idea of scientific objectivity through 3 steps. 

  1. Realize (or postulate) that my experiences are caused by actions upon me from the environment. 
  2. Realize (or postulate) that what causes actions upon me can also act on other things. They can exist on their own.
  3. Realize (or postulate) an "true nature" that is independent of any perspective. 

People in general regard science as the study of this "true nature" using "a view from nowhere". I am simply arguing Postulate 3 is taking it too far. 

Postulate 1 expands the scope from subjective experiences to actions, i.e. subjective experiences are not special. Postulate 2 expands the scope from my first-person perspective to anything's perspective, i.e. my perspective is not special. Using only these two is not only more parsimonious, it also fits very well with the increasingly observer/perspective-dependent trend of physical discoveries. 

Okay, I disagree :-)