Friday, July 18, 2014

Mel Acheson: How Science Can Lose Its Way


Presentation below from one of the Electric Universe people at their recent conference, on a general topic that should be familiar and dear to our readers, and in which he identifies the importance of the creative process that is our assignment of the words and I would add the establishment of terms in the development of our theories.

"Have a pattern of sound words"  2 Tim 1:13

This is a key excerpt from the presentation:
What is evidence?  We see a meter needle pointing to the numeral 5, that's evidence, but of what? 
With the theory of electromagnetism, we can say the voltage drop across that resister is 5 volts, that's a fact. 
Between the evidence and the fact is a bridge of interpretation. Evidence tells us that something is, fact tells us what something is, the theory ties them together. 
The meter needle gives you 5, the theory gives you volts. 
So a fact is the interpretation of specific evidence of raw sensation in the light of some collection of general concepts, ie a theory. 
The theory tells us what words to use to talk about the evidence.  Without words, the evidence is meaningless and incommunicable.  Words mean, evidence doesn't. 
Theories then are the rules of assigning words to evidence. 
Facts are evidence fossilized in the words of a theory. 
This business of assigning words is a creative act. There can be more than one way to do it. 
Tullman draws the metaphor that theories and their pre-conceptions are like eyeglasses, we can't see the evidence without looking through them.  We might not realize we are wearing them unless we take them off. 
This indicates the danger of consensus science.


.



4 comments:

Peter Pan said...

A voltmeter is calibrated to show voltage. If it weren't for instrument calibration and an agreed upon system of weights and measures, a reading of 5 volts could not be defined. Theory has nothing to do with measurement.

Matt Franko said...

Bob,

The theory establishes what a volt is in the first place... The Theory establishes this term 'volt'... then measurements can be made in 'volts', if you don't first establish what a volt is, you can't take a measurement in 'volts' in the first place...


A theory is a body of knowledge... Which includes terms and their definitions...

the academe of economics doesn't work this way that's why it is so f-ed up.... No one even agrees what the term 'money' means and they use it all the time, etc... Conflating stocks and flows, etc... Metaphors like 'inflation' or 'deflation' instead of creating a standard unit, etc

RSP,




Peter Pan said...

A volt is a unit of measure. It was established as a result of scientific inquiry. In some cases, a conjecture or hypothesis was used to guide the inquiry. In this example however, electromotive force was initially observed as a property of nature. It was discovered through experimentation. The usual methodology requires that instruments be built to detect and measure the property.
A theory establishes the relationship(s) between observed phenomena. Quantifiable terms such as the volt are established and maintained through metrology.

In economics, there are terms that involve the use of statistics rather than measuring devices. That is no excuse for failing to develop precise definitions.

Tom Hickey said...

Matt is correct in terms of philosophy of science. The volt is defined operationally as Bob points out. In science definitions should be operational to be unambiguous, based on a criteria that is not itself arbitrary.

Some measurements are arbitrary in a sense, like a meter in the metric system of measurement. However, scientists are aware of this and keep a standard meter under special (near) invariant conditions in Paris. This is also what the US Bureau of Standards does.

That's not the case in economics and a reason that economics are generally practiced is pseudo-science. Again, as Matt correctly points out.

In math, the derivative is defined by the operation for calculating it.

Without criteria and operational definitions, assertions are ambiguous. A lot of what goes on in econ is arguing over ambiguity.