The E wrote:DDHv wrote:The closest I know at present to hard economics is the Austrian school of economics. Does anyone know another school that is doing better at predicting
I just have to pick this particular thing up again just because it's so funny.
So, DDHv, all throughout this thread, you are absolutely adamant that the only true science is science confirmed through experimentation and observation.
Here's what Ludwig von Mises, one of the austrian school's most influential figures has to say on that subject:
The subject matter of all historical sciences is the past. They cannot teach us anything which would be valid for all human actions, that is, for the future too ... No laboratory experiments can be performed with regard to human action ... Its statements and propositions are not derived from experience. They are, like those of logic and mathematics, a priori. They are not subject to verification or falsification on the ground of experience and facts.
(The above being a paraphrasation of his arguments
here)
And then we come to the austrian school's predictive power: It is nil, or at the very least, not significantly better than competing theories. It seems to me that they're not much better than a stopped clock is when it comes to telling time.
To improve clarity: the basic axioms of science are; there is a universe; the universe is reasonable. I add one: all of us are going to get some things wrong.
Idea beginnings don't matter. IIRC, Singer is reported to have produced the idea of a needle with the hole at the tip instead of the other end from a dream. Theory testing must be empirical, although it might be filtered through several layers of theories. Humans have a strong tendency to get into thinking ruts. It is important to study the ideas and tests of people with different biases than ourselves, as they are more likely to locate bad assumptions in our theories. Assumptions tend to come from some bias.
Since Euclid, geometry has a strong theory basis - the non euclidean geometries don't change that, they just extend the straight metrics of Euclid into closing curved and spreading curved geometries. My bias from an engineering training notes that the Euclidean form is useful for local extents, while the closing version (great circles, crossing, if you've read "Babel 17") is needed for describing a whole earth geometry. IIUC, the general theory of relativity states that mass tends to cause spatial metrics to shift toward the closing form: One observation suggesting that it is correct was the shifting of apparent position of a star during an eclipse of the sun. In cosmology, one suggestion is that the universe as a whole has a spreading metric.
The whole point is that what counts is not whether a conclusion agrees with a given theory, but whether high confidence methods can be used in testing it. If not, the influence of bias will probably overcome the influence of evidence. All humans have some bias.
One point often neglected in the irreducible complication discussion is that of instincts. If a poisonous life form doesn't have the correct instinct to use a given mechanism, it becomes a negative factor, not a positive one. Without an instinct to defend the hive, a bee's stinger, poison production, storage sacks, and the muscles to move the poison through the sting just absorb metabolic energy in producing them. Ditto for any other poisonous life form.
Examine the pacific golden plover, which summers in Alaska, and after a several day flight, winters in the Hawaiian islands. If it stays in Alaska, it dies. If it leaves before storing enough fat, it dies. If it flies alone instead of the V formations, it dies. It arrives with roughly a 10% reserve, so if the navigation is poor, it dies. Doesn't selection require at least some survivors? These things suggest that if grand evolutionary theory is correct, the pacific golden plover doesn't exist as described
Richard Lenski's high confidence experimentation with e. coli, produced a solid example of micro-evolution, a novel citrate metabolic pathway. Dawkins, a salesman for grand evolution, trumpeted that new genes had been produced.
Bount, ZD et al. Genomic analysis of a key innovation in an experimental Eschericha coli population, Nature. 2012:489:513>20
and
Quandt,EM et al.Recursive genomewide recombination and sequencing reveals a key refinement step in the evolution of a metabolic innovation in Escherichia coli, PNAS. 2014:111:2217>22
show the actual action at roughly generations 31500 and 33000 respectively to be a transposition of one gene and a mutation of a promoter region. These were in 1 out of roughly 10^13 total organisms.
Early stage comparison of chimp and human genomes was hampered by human DNA contamination and other problems. Even so, the estimated 1.2% of difference suggested would come to roughly 36 million DNA base pair differences. Given the evidence from micro-evolution experiments and observations, it is quite unlikely that enough individuals existed in the suggested time span to make this difference probable. Later results suggest differences in the 10>20% range.
There are more basic problems: at present belief in abiogenesis and eukaryogenesis are both based on faith in grand evolution, not empirical science. Those who have this faith should work on experiments and observations, but should not describe as proven what has many empirical problems. My current bias in this is: variation within kinds, and wide moats between kinds. This is a faith position as is grand evolution. The ID people at least have a reasonable reason to believe in a reasonable universe, instead accepting that science axiom by faith alone.
My bias is an engineering one: to return to economics, there are many feed backs. One problem is that many unexpected consequences have generational time spans. Oscillation occurs when feedback has amplification and a time span longer than the natural resonance period. You may recall that Nikolai Kondratiev was executed because his wave theory suggested that socialist economies were subject to cycles - the soviet big shots didn't like that idea at all. Talk about shooting the messenger
The currently accepted Keynesian economic theories make assumptions. Examination of places which used these over five decades or more suggests something is wrong with the basics. One probable reason is that running the control loop through a centralized decision point in these economies lengthens the control time span, increasing oscillations. Local decisions in a decentralized economy are faster. Engineering theory says this will make for more stability.
The Austrian school of economics is unlikely to be totally correct: we haven't had a Hari Seldon yet! Like geometry, it has several formally worked out theoretical patterns. Comparison of these with real world results just
may move us toward a better set of theories. My current prediction is: a near correction in US stocks, followed by a big rise for one>three years without a solid economy, then a major crash similar to the 1930s. Some people think that manipulation by big shots wanting control is the current problem, rather than their economic ignorance.
Hard science is a series of successive approximations of our models (theories) to reality. This is only possible by comparing results to reality, even indirectly. Ignoring evidence and logic by people whose assumptions come from different biases is an excellent way to not see any weaknesses in your own theories
PS. "Captain Empirical" by Sam Nicholson is a good SF read
