systemic errors in science

Your favorite books and links
Post Reply
Myakka
Posts: 122
Joined: Thu Sep 13, 2012 3:39 am

systemic errors in science

Post by Myakka »

http://ckraju.net/papers/Planetary-moti ... ddhism.pdf
Which is one paper of many from this author's site -- which can be found here: http://ckraju.net/index.html

I found his critiques of science and math compelling and the lack of explanation of what he is saying is the better approach somewhat frustrating. (This might be my fault because I haven't yet bought any of his books nor read everything on his site.)

Interesting questions he has raised in this article and some others I have looked at include:
Why assume that the natural motion of an object is in a straight line?
Why do we take logical proof as superior to empirical proof?
Why do we site prior sources as proof of our information rather than direct observational evidence?

(Which add to one I have been carrying around for a while in my brain:
Why do we take the data generated by a machine as superior to human observation?)

In Early Retirement Extreme thought we reject our culture's story about spending our lives grinding away at a job we despise for the reward of buying some trinkets. Perhaps the lies of our culture do not end there. Perhaps the whole of our source of information is full of deception and manipulation. And by considering the ideas of people who question the largely unconscious assumptions we carry around we can reclaim some more of our own power.

What do you think?

Dragline
Posts: 4436
Joined: Wed Aug 24, 2011 1:50 am

Re: systemic errors in science

Post by Dragline »

[quote="Myakka"]http://ckraju.net/papers/Planetary-moti ... ddhism.pdf

Interesting questions he has raised in this article and some others I have looked at include:
Why assume that the natural motion of an object is in a straight line?
Why do we take logical proof as superior to empirical proof?
Why do we site prior sources as proof of our information rather than direct observational evidence?

(Which add to one I have been carrying around for a while in my brain:
Why do we take the data generated by a machine as superior to human observation?)

[/quote]


1. Why assume that the natural motion of an object is in a straight line?

All of our evidence supports this basic model, until you get to warped space and time with black holes and other such phenomena. You can experiment yourself with a billiard table. Or you can propose an alternative model, which you then can falsify with proofs or with observational data.

2. Why do we take logical proof as superior to empirical proof?

It's not superior, just more complete. Logical proof is rare and is only dependent on accepting the initial premises. With empirical proof, you always have the original Black Swan problem, which is concluding that all swans are white because every swan you have ever seen is white (empirical proof). The premise is falsified when the unknown data becomes known. Empirical proof is always subject to data collection issues that can never be completely resolved.

This cannot happen in a true logical proof. In that case, you can only question whether the underlying assumptions are valid.

3. Why do we [c]ite prior sources as proof of our information rather than direct observational evidence?

We do both. As a practical matter, most of what we know is taken on faith in prior sources and people, though, because we do not have the wherewithal to repeat or collect the direct observational evidence ourselves. Human knowledge is built of collective observations, thoughts and records thereof.

4. Why do we take the data generated by a machine as superior to human observation?

I don't think we do necessarily. But human observation is limited by our senses and our brain's ability to use them. We know for a fact that our brains often "fill in blanks" that our senses did not actually perceive so as to not overload our thinking capacities. Both our senses and memories are faulty.

Machines are also limited in what they are designed to measure and our ability to interpret what they are measuring. But obviously an electron microscope is better at examining very small things than a human eye, for example.

As for "data generated by a machine", you may be talking about something else which could be real or imaginary. Hard to tell.

daylen
Posts: 2542
Joined: Wed Dec 16, 2015 4:17 am
Location: Lawrence, KS

Re: systemic errors in science

Post by daylen »

Didn't read any of his paper, but I will try to answer some of the questions you mentioned.

Why assume that the natural motion of an object is in a straight line?
First off, we must assume something, else there is nothing to go on. Second, if a theory were developed such that the above was not assumed and the derived laws worked to explain empirical behavior as does classical mechanics, then the theory would be just a valid as classical mechanics. The reason we do not use such a theory is because this assumption is more intuitive to us and therefore simpler (Occam's razor).

Why do we take logical proof as superior to empirical proof?
I wouldn't necessarily agree with this. Both are important, and science relies on both of them.

Why do we site prior sources as proof of our information rather than direct observational evidence?
There is no such thing as direct observational evidence.. We observe through our senses.. then record, by which time the experiment is in the past and therefore all sources are prior.

Mot
Posts: 1
Joined: Thu Nov 03, 2016 6:28 pm

Re: systemic errors in science

Post by Mot »

Data generated by a machine (think fax machine reproducing a document or a smart phone taking a video) is inherently consistent. While it may not pick up the clues that we would from context or past experience machine data tend to be more precise (note I didn't say accurate as collected data can often be biased knowingly and unknowingly). Think of the Boston bombing where police relied more on video and photographs than eyewitness reports to find the perpetrators.

Are machine data better? Wrong question. Both data collection techniques have advantages and disadvantages. Machines are good at crunching numbers and repetitive behavior. Humans are better at pattern recognition and linking abstruse thoughts to develop a new understanding.

Dragline
Posts: 4436
Joined: Wed Aug 24, 2011 1:50 am

Re: systemic errors in science

Post by Dragline »

Mot wrote:
Are machine data better? Wrong question. Both data collection techniques have advantages and disadvantages. Machines are good at crunching numbers and repetitive behavior. Humans are better at pattern recognition and linking abstruse thoughts to develop a new understanding.
This was Kevin Kelly's (WIRED magazine guy) conclusion in a recent interview where he noted that in "team chess", the superior teams had both machine and human team members.

Post Reply