Free Capitalist Network - Community Archive
Mises Community Archive
An online community for fans of Austrian economics and libertarianism, featuring forums, user blogs, and more.

Is science prediction?

This post has 10 Replies | 3 Followers

Top 200 Contributor
Male
Posts 429
Points 7,400
NonAntiAnarchist Posted: Tue, May 7 2013 8:38 AM

I don't think so. It can be a feature of science, but I don't think it is at the core.

From thinking about it just for a minute, I'd say (good) science is something like true premises + internal consistency + predictability. Some times we maybe will have to settle for imperfect premises when we model things, but I think a correct premise is always preferable to an incorrect premise  in the realm of science.

Curious to see what everyone else thinks.

  • | Post Points: 50
Top 50 Contributor
Posts 2,679
Points 45,110

Science is about making sense of reality by determining cause and effect relationships. People can use this knowledge to explain past events or make predictions about the future. The more complete the knowledge, the more accurate the explanations and predictions become. I don't know if you have read The Relativity of Wrong by Isaac Asimov, but it is somewhat relevant to your question and it's a great (short) read anyway.

  • | Post Points: 20
Top 500 Contributor
Posts 228
Points 3,640
Blargg replied on Tue, May 7 2013 11:24 AM

I see science as a particular approach to better determining what's "out there". It is a tool that we use to overcome our natural tendency to form beliefs and seek only to confirm them, similar to how mathematical proofs give confidence beyond what we could do through thinking alone. Science gives us a ratchet effect, solidly falsifying things so that we don't waste further time on those things.

  • | Post Points: 5
Not Ranked
Posts 5
Points 80
jfrega replied on Tue, May 7 2013 4:36 PM

There are four kinds of science:

1. Predictable, observable, repeatable, controllable - Typical laboratory operations

2. Predictable, observable, non-repeatable, non-controllable - Ellipses, planetary motion

3. Non-predictale, observable, non-repeatable, non-controllable - Earthquakes, random particle motion

4. Non-predictable, non-observable, non-repeatable, non-controllable - Action of intelligent beings, human action, purely random occurances, evaluation of testimony, ex post facto application of scientific methods to evidence collection.

Thanks to Stanton Friedman

http://youtu.be/cDbhSQaqYXM

  • | Post Points: 20
Top 50 Contributor
Male
Posts 2,439
Points 44,650
Neodoxy replied on Tue, May 7 2013 4:55 PM

Relevant to any discussion on the nature and study of science:

At last those coming came and they never looked back With blinding stars in their eyes but all they saw was black...
  • | Post Points: 5
Top 50 Contributor
Male
Posts 2,493
Points 39,355
Malachi replied on Tue, May 7 2013 5:50 PM

gotlucky:

Science is about making sense of reality by determining cause and effect relationships. People can use this knowledge to explain past events or make predictions about the future. The more complete the knowledge, the more accurate the explanations and predictions become. 

gotlucky answered, I just want to embellish a bit. explanation and prediction are two sides of the same coin, they both speak to a logical (cause/effect) relationship in the real world.

Keep the faith, Strannix. -Casey Ryback, Under Siege (Steven Seagal)
  • | Post Points: 20
Top 200 Contributor
Male
Posts 429
Points 7,400

Why does quantum mechanics supposedly yeild incredilby accurate predictions, yet lack scientific rigor in terms of establishing cause and effect, consistent definitions, etc. (as I've heard some on this board claim - I lack the expertise to judge the science in any way tbh).

I understand an understanding of cause and effect yeild greater predictability. But is the opposite true? Are there not mathematical relationships and algorithms or whatnot that are used soley for prediction? What would these be considered a category of? technology?

I've heard Milton Freidman claim that even models with false premises can be useful in terms of prediction. But is this science? It seems to me it's not, but does the mainstream disagree? What is the mainstream position on what constitutes 'science'. 

Thanks for the replies, guys. I will be sure to check out anything linked here. Any of you know a good place to start (philosophers, texts, etc) in terms of getting familiar with the philosophy of science?

  • | Post Points: 50
Top 50 Contributor
Posts 2,679
Points 45,110

One of the points Asimov was making was that wrong theories can still have usefulness, which is why certain wrong theories have lasted sometimes for centuries. This doesn't mean that we should just use wrong theories for the hell of it. The point is that scientists should continually refine these wrong theories to make them less wrong and therefore more useful. However, it will rarely be useful to use a wrong theory when there is a better theory available. So yes it is science, but that didn't always make it correct.

  • | Post Points: 20
Top 200 Contributor
Male
Posts 429
Points 7,400

Right, makes sense. I haven't read the paper, so I'll do that before I ask anything else. Thanks.

  • | Post Points: 5
Not Ranked
Posts 77
Points 1,150

NonAntiAnarchist:

I understand an understanding of cause and effect yeild greater predictability. But is the opposite true? Are there not mathematical relationships and algorithms or whatnot that are used soley for prediction? What would these be considered a category of? technology?

Statistical models are often used to make forecasts. Statistical models are used when underlying causal relationships between independent and dependent variables are not known. Correlations between variables and patterns are used to construct the models. As an example, we observe that people consume more energy on hot and cold days. Is this a causal relationship? Not really. These is an underlying casual relationship you can hypothesise to justify the use of the correlation in your model.

  • | Post Points: 5
Top 50 Contributor
Male
Posts 2,551
Points 46,635
AJ replied on Mon, Oct 21 2013 5:30 AM

NonAntiAnarchist:
Why does quantum mechanics supposedly yeild incredilby accurate predictions, yet lack scientific rigor in terms of establishing cause and effect, consistent definitions, etc. (as I've heard some on this board claim - I lack the expertise to judge the science in any way tbh).

Because they use tricks like renormalization and other fudges to match the empirical data. It's not as if they started out predicting some specific, unknown data then when the data appeared it turned out to match theory, which would be impressive. It's totally finessed.


Random example among many others: http://milesmathis.com/gf.pdf

  • | Post Points: 5
Page 1 of 1 (11 items) | RSS