I have been working on Rational Observer Theory since the end of 2012, and it is the major reason that I have published virtually nothing since 2017. I was fortunate to receive tenure in 2017, although the timing was mostly coincidental. The theory still remains unpublished, as does related work on the measure of evidence. But since it is nearing completion, I provide a partial preview of it here.
I have focused the website almost entirely on the philosophy and mathematics that underlie ROT rather than the theory itself. I have tried to provide enough information to make the ideas interesting and persuasive, but not enough to make them entirely coherent. A natural and unfortunate consequence is that I make some claims about ROT and the measure of evidence that for now appear poorly justified.
The Central Claims of Rational Observer Theory
Rational Observer Theory proposes that matter is evidence, and therefore it predicts/infers. A thing/observer is a configuration of particles, and its internal and present configuration is evidence for its future configuration. It is also evidence for the external configuration of other particles.
All the relations within Rational Observer Theory, between internal and external configurations, and between present and future configurations, are rational.
Comparison to Other Physical Theories
Rational Observer Theory is more general and comprehensive than other theories. It is a precise model of both reality and what local observers know about reality, and thus includes both ontology and epistemology. Although it is a physical model it also incorporates and addresses aspects of mental phenomena.
I will argue that ROT is more coherent and systematic and better justified than other models with respect to philosophy, mathematics, and physical science. Parts of my argument have been previewed on other pages, especially Science’s Problem with Reality and The Measure of Evidence.
Rational Observer Theory Is Favored by Reason Alone
I will specifically demonstrate that among geometric models of reality, ROT is the one favored by reason alone, before or without considering empirical evidence. Furthermore, even though ROT has not yet been shown to explain any empirical observations, I will prove that it is more strongly supported by our total evidence, empirical and non-empirical combined, than other geometric models, including the standard models of physics.
I expect this last claim to appear unbelievable. How could a model that has yet to accurately predict any empirical observations be better supported by evidence than our standard physical models, which make amazingly accurate predictions (except at extremes)?
It turns out that it is quite useful to be able to measure evidence. It is a well kept secret that science has not had anything even close to an accepted or widely recognized mathematical measure of evidence. Scientists seldom measure evidence, and even more rarely do they measure it correctly.
Rather they measure the accuracy of prediction, and that is all that matters for their predictive models. External reality is not their primary concern, as I discuss in Science’s Problem with Reality. Without the philosophy and mathematics necessary to deal with external reality, and with technological applications being a priority, it is natural that science has achieved predictive accuracy by accumulating in its models a great diversity of metaphorical epicycles. Getting rid of the epicycles increases the non-empirical prior evidence (from reason alone) for a simple model like ROT so greatly that it more than compensates for the empirical evidence that supports our standard models.
Testing Rational Observer Theory
ROT makes numerous predictions and should be tested empirically. However, the mathematics must be complete before we can know what those predictions are. The first mathematical results have only been achieved recently and are yet to be published.
Comparison to My Theory of the Brain
My general theory of the brain (Fiorillo, 2008; Fiorillo et al., 2014) is closely related to the more general and quantitative Rational Observer Theory, and could be considered part of it. I expect that it can be shown to follow from ROT. ROT should allow for the precise quantification of neuronal evidence and prediction, at least in principle. My original theory proposed a semi-quantitative description of neuronal evidence and prediction that I believe is correct, but it proposed a specific method of measuring evidence (calculating probabilities) that was naive and incorrect (Fiorillo, 2008).
Why do I call it “Rational Observer Theory?”
There are many apt names I could have chosen for the theory. I chose “rational” and “observer” in part because it is obvious to most people that these are important concepts. A second reason is that there has not been any substantive physical model of observers or rationality. ROT is such a model. There has been a lot of confusion about both terms, which I discuss on other pages, particularly Science’s Observation Problem, The Neuron’s Point of View, Nature is Rational and The Irrational Brain Hypothesis.