I've harbored the notion that the continual interactions observable in the quantum substrate seed subjective-objective interaction all the way up to classical reality including conscious life. This impression (a sense, not a scientific understanding in my case) began when I read several papers by Carlo Rovelli about five years ago concerning information exchange between physical fields bringing about first disorder and then more complex and integrated or entangled order. Rovelli has developed a theory referred to as Relational Quantum Mechanics, and in a search last night I reached a recent extension of it in a paper shared two years ago at the FQXi site, discussed in peer comments which I will link below with another paper.
CATEGORY: FQXi Essay Contest - It From Bit or Bit From It?
TOPIC:
Relative information at the foundation of physics by Carlo Rovelli
Essay Abstract: I observe that Shannon's notion of relative information between two physical systems can effectively function as a foundation for statistical mechanics and quantum mechanics, without referring to any subjectivism or idealism. It can also represent the key missing element in the foundation of the naturalistic picture of the world, providing the conceptual tool for dealing with its apparent limitations. I comment on the relation between these ideas and Democritus.
Author Bio: Carlo Rovelli is professor of theoretical physics at the University of Aux-Marseille. His main interest is quantum gravity, but he has worked also on the foundations of quantum theory and general covariant statistical mechanics, and on the ancient history and philosophy of physics.
Download Essay PDF File
From the comments:
Conrad Dale Johnson wrote on Jul. 1, 2013 @ 15:18 GMT
Dear Carlo -
I'm very glad to see a new essay of yours revisiting some themes of Relational Quantum Mechanics. I've always considered that paper a milestone, or rather a signpost pointing a way that remains to be explored. Maybe the time is ripe... there are a few other essays in this contest - Knuth's, for example - working with the idea that all of physics concerns "the information systems have about other systems."
I very much agree with your conclusion - "The universe is not just simply the position of all its Democritean atoms. It is also the net of information that all systems have about one another. Objects are not just aggregates of atoms. They are particular configurations of atoms singled out because of the manner a given other system interacts with them."
However, I agree with Walter Smilga's comment above - in order to grasp what it means for things to have information, even in physics, we need to deal with contexts of meaning. You want to stick with Shannon's definition, as you wrote, to show "that there are meaningful notions of information and relative information in simple physics, without need to refer to semantic meaning." As in the RQM paper, here also you define "having information about" a system just as "being correlated" with it. (
Knuth's contest essay likewise uses an abstract notion of systems "influencing" each other. And even
Smilga, who wants to bring semantics into the picture, uses a very generalized notion of "semantic frames of reference.")
I don't doubt that your approach gets at something very important about the structure of physics. But the point of
my essay is that something else that's important is missed when we abstract from the specific kinds of contexts in which information actually becomes measurable.
These contexts are not mysterious - we know all about how to assemble them when we make measurements. There's nothing subjective or mental about them - the same physics we use in the lab describes how any system gets information about other systems. But there are major obstacles to formulating any realistic general definition of a "measurement-context". It's not just that such arrangements are never physically simple, but also that any way of measuring something depends on other ways of measuring other things. I argue that this complex interdependency of different ways of "observing" is really what's behind the measurement problem in QM.
In physics we're always trying to show how the underlying structure is basically simple - so the many different ways in which things actually "have information about each other" give us a picture that hardly seems as though it could be relevant to the physical foundations. Yet if we only think about abstract and generalized information-processes, we lose touch with the way information is physically present in the world.
My suggestion is that measurement can be conceived as fundamental, if we can see it as an evolving process. Though it takes a very complex interactive environment to communicate definite information about and between its subsystems, this kind of environment can exist and maintain itself for the same reason that life does, if it's the kind of system that can evolve through random selection.
Thanks again for the new essay - Conrad
Author Carlo Rovelli replied on Jul. 1, 2013 @ 16:23 GMT
Dear Conrad,
you touch something basic here. I agree that what you talk about is a central issue, and I am uncertain myself.
We certainly agree on the relevance of context, and I feel everybody would agree, at least after a good discussion clarifying what we mean. But I have tried to bring this down to good old physics. You are right that in quantum mechanics this affects the measurement issue and you are right that it affects the definition of what is a measurement context. But the central point of Relational Quantum Mechanics is to solve this issue by accepting the idea that *any* physical interaction is a measurement. When an atom in a SternGerlack apparatus is deviated by the magnetic field, the position of the atom is measuring the spin. This seems to me the only possible solution; I have never found a convincing alternative. The price to pay is of course the Relational Quantum Mechanics observation that events are indexed by the context. That is, in this case the spin is measured by the position, and does not take value with respect to a system not interacting with it. This allows interference to affect possible later interactions with position or spin. Thus, in this sense I agree with you that measurement is fundamental, but I prefer to view it as synonymous of interaction, rather than trying to view it, as you suggest to attribute it to "a very complex interactive environment".
Conrad Dale Johnson replied on Jul. 2, 2013 @ 14:36 GMT
Carlo - thanks for your response, and I get your point. In fact, I pulled out my old marked-up copy of the RQM paper and was reminded again what a thorough piece of work it is, given its limited scope. It lays out - more carefully than any work of philosophy I know of - the basic philosophical issues involved in the meaning of objectivity.
It is just the notion that events are indexed by the observer-relative context that's important to me. The world only exists from the standpoint of some observer. This isn't subjective (mental), in that anything counts as an observer. It's not solipsistic, in that communication between observers is as fundamental as observation itself - in fact, from the QM viewpoint there's no difference between these two. But as you say, it's an error to describe this world of multiple observers as if it could be envisioned "from outside", from no point of view - as if there could be well-defined information without a context to define it from a specific point of view.
This is a very radical notion, and I think it will be some time before we have the conceptual tools we need to be clear about it.
So I understand your "only possible solution" - treating any interaction as a measurement. But I would remind you of the point you make in RQM, that even the correlation between two systems is only definable from the standpoint of a third system. And the position of the atom "measures" the spin, insofar as something else observes the atom, in some context in which its position is definable over time.
There's no specific level of complexity at which interactions become measurements, or systems become observers. To that extent I agree, it's better to treat all systems as observers and all interactions as measurements. But this does not really "solve" any problem. Many different kinds of interactions are still needed to define / measure any physical information, and though I well understand your preference for "the good old physics", ultimately I think we can't set this fact aside as insignificant.
In
my essay I acknowledge the difficulty of dealing with it, and try to show how they can be addressed. In the end this points to a way of answering the basic question that's left - in my mind, anyway - by RQM: how and why do things work out so that at the macroscopic level, the quantum world of communicating observers ends up looking so much like the objective, deterministic reality of classical physics?
Author Carlo Rovelli replied on Jul. 2, 2013 @ 16:18 GMT
Ok, you definitely convinced me to read with care what you have written! I will now print it out and study it... thanks! Carlo”
I haven’t looked yet for further discussion between Rovelli and Johnson in the long thread at FXQi, and it may be that Rovelli and Johnson corresponded privately afterward. The links to the Rovelli and Johnson papers are repeated here:
Johnson, On the evolution of determinate information
http://fqxi.org/data/essay-contest-files/Johnson_CDJohnson_fqxi2013.pdf
Rovelli, Relative information at the foundation of physics
http://fqxi.org/data/essay-contest-files/Rovelli_information.pdf
and comments at
Relative information at the foundation of physics by Carlo Rovelli