The last few years I dismissed Kant for what some say is his negation of causality, or, his claim that knowledge cannot be gained from sensual perception alone but only through perception and subsequent interpretation by our brains – which is a non-empiricist process. In other words, for Kant reality is not what we see but what we interpret that we see.

Turns out it is really quite simple when drawn as a schematic. Now you can google “Immanuel Kant original illustrations” and he just didn’t make any. So I made one.

Notice that while Kalman filters are in practice used to cope with ordinary simple signals – a PLL is an example of a Kalman filter for a practical purpose – nothing stops the Kalman architecture from processing arbitrarily high level symbolic protocols – that’s just a change of data types.

So. Why did philosophers of yore write tomes full of impenetrable gibberish when a simple schematic would have done. Makes you wonder.

And: By analyzing and mixing the data flow architectures of common philosophical schools we can now create a continuum of architectures that seamlessly blend empiricism into transcendentalism and back…

…by replacing “Model” with “world view” or “Weltanschauung” we might grasp how it comes that different people have WILDLY different views of the world while receiving the same stimuli.
(most notably socialists vs. free marketers)

Cognitive dissonance then results from overly frequent needs to “correct” the model – or refuse to correct it. In extreme cases this leads to rejection of new stimuli.

The better your model the less you suffer from cognitive dissonance.

Correcting the model can lead to increasing numbers of epicycles – the model becomes more and more complex to maintin it. Examples: Ptolemaean vs. Copernican cosmology; and, in modern times: Dark Matter, Dark Energy as modern epicycles to save a model that is clearly not working.

Cognitive Dissonance in science therefore manifests itself in the invention of new physics – which is good for the scientists because it gives them reasons to publish more papers. They become mechanistic maintenance operators in a giant Kalman model.

1. Subjective Perception doesn’t negate causality.
2. https://en.wikipedia.org/wiki/State_observer underpins the Kalman filter.
3. Kalman filter is about combined estimation i.e. a vector of samples drawn from multiple sources.
This produces a wide perspecive over the subject matter being observed,
thus allowing to infer about its true nature.
Since those sources are derived ffrom common phenomenon, their outputs of concern
are a-priori correlated. This correlation is what the Kalman filter seeks to estimate,
in order to “clean” that input vector from distortion and noise
i.e. in order to generate an estimate of the properties of the sheer reality of concern.
It is a case of a https://en.wikipedia.org/wiki/Functional_(mathematics), quote:
“it is a function that takes a vector as its input argument, and returns a scalar.”
4. Behavioral sciences made similar progress to the Kalman filter at about the same time,
quesioning the authenticity of mental observation made by patients: http://www.davidjweiss.com/functionalmeasurement.htm

“3. Kalman filter is about combined estimation i.e. a vector of samples drawn from multiple sources.”

Not necessarily, Yoav. In simple cases, one source – as in the PLL example.

“This correlation is what the Kalman filter seeks to estimate,
in order to “clean” that input vector from distortion and noise
i.e. in order to generate an estimate of the properties of the sheer reality of concern.”

Exactly: ESTIMATING SHEER REALITY – (that’s what Kant is all about; or Plato or other idealists for that matter). Now, what does a sensor give us – say a camera ? A 2dimensional image of a 3dimensional reality. The model of the Kalman filter then becomes an estimate of 3dimensional reality. Internally, a 2dimensional image projection is computed and compared with the sensor (camera) input; but the model maintains a 3dimensional world model as state. The user of the Kalman filter then has access to the full internal state of the Kalman filter; not only to a 2dimensional projection.

Concerning our own vision, we have such an automatic 3D model construction built in – even one-eyed persons still maintain such an internal model without even thinking about it – using clues like known object sizes – all while only receiving 2D images of the world.

Notice that in all of this, the sensor input can have LESS information than the internal state of the model of the K. filter. Information gets EXTRAPOLATED.

…by replacing “Model” with “world view” or “Weltanschauung” we might grasp how it comes that different people have WILDLY different views of the world while receiving the same stimuli.

(most notably socialists vs. free marketers)

Cognitive dissonance then results from overly frequent needs to “correct” the model – or refuse to correct it. In extreme cases this leads to rejection of new stimuli.

The better your model the less you suffer from cognitive dissonance.

Correcting the model can lead to increasing numbers of epicycles – the model becomes more and more complex to maintin it. Examples: Ptolemaean vs. Copernican cosmology; and, in modern times: Dark Matter, Dark Energy as modern epicycles to save a model that is clearly not working.

Cognitive Dissonance in science therefore manifests itself in the invention of new physics – which is good for the scientists because it gives them reasons to publish more papers. They become mechanistic maintenance operators in a giant Kalman model.

LikeLike

1. Subjective Perception doesn’t negate causality.

2. https://en.wikipedia.org/wiki/State_observer underpins the Kalman filter.

3. Kalman filter is about combined estimation i.e. a vector of samples drawn from multiple sources.

This produces a wide perspecive over the subject matter being observed,

thus allowing to infer about its true nature.

Since those sources are derived ffrom common phenomenon, their outputs of concern

are a-priori correlated. This correlation is what the Kalman filter seeks to estimate,

in order to “clean” that input vector from distortion and noise

i.e. in order to generate an estimate of the properties of the sheer reality of concern.

It is a case of a https://en.wikipedia.org/wiki/Functional_(mathematics), quote:

“it is a function that takes a vector as its input argument, and returns a scalar.”

4. Behavioral sciences made similar progress to the Kalman filter at about the same time,

quesioning the authenticity of mental observation made by patients:

http://www.davidjweiss.com/functionalmeasurement.htm

LikeLike

“3. Kalman filter is about combined estimation i.e. a vector of samples drawn from multiple sources.”

Not necessarily, Yoav. In simple cases, one source – as in the PLL example.

“This correlation is what the Kalman filter seeks to estimate,

in order to “clean” that input vector from distortion and noise

i.e. in order to generate an estimate of the properties of the sheer reality of concern.”

Exactly: ESTIMATING SHEER REALITY – (that’s what Kant is all about; or Plato or other idealists for that matter). Now, what does a sensor give us – say a camera ? A 2dimensional image of a 3dimensional reality. The model of the Kalman filter then becomes an estimate of 3dimensional reality. Internally, a 2dimensional image projection is computed and compared with the sensor (camera) input; but the model maintains a 3dimensional world model as state. The user of the Kalman filter then has access to the full internal state of the Kalman filter; not only to a 2dimensional projection.

Concerning our own vision, we have such an automatic 3D model construction built in – even one-eyed persons still maintain such an internal model without even thinking about it – using clues like known object sizes – all while only receiving 2D images of the world.

Notice that in all of this, the sensor input can have LESS information than the internal state of the model of the K. filter. Information gets EXTRAPOLATED.

LikeLike