Kevin T. Kelly
Department of Philosophy
Carnegie Mellon University
Paradigm tested in isolation = normal science puzzle solving: paradigm must win.
Paradigm choice = persuasion to adopt new paradigm during crisis: must have alternative available.
Paradigm choice comes closest to philosophy of science.
Verificationism:
Absolute verification: hopeless due to the problem of induction.
Probabilistic verification: seek the most probable theory.
Objection:
must consider all possible experiments and all possible theories.
no scientifically neutral language across paradigms.
hence, probabilistic comparisons are all paradigm-specific.
Falsificationism (Karl Popper): no verification: seek the boldest theory that has not yet been refuted.
Objection:
all theories have anomalies.
ranking anomalies by severity is tantamount to verificationism.
Unified view:
Falsificationism describes the onset of crisis.
Verificationism is about pairwise paradigm competitions.
Unified view is oversimplified.
incommensurability: no theory-neutral language to do it in
disagreement about relevant problems (explaining shininess of metals)
different standards of solution (unexplained forces)
different meanings for old paradigm concepts (curved space)
different paradigms make different worlds (theory-laden data). (pendulum, curved space).
Laws obvious to some cannot be demonstrated to others.
No logical, piecemeal process of change.
Must change all to change any. "Gestalt switch".
How do scientists switch paradigms?
Mostly, individuals don't.
Copernicanism spread slowly.
Cartesians rejected Newton for 50 years.
Priestly never adopted Oxygen theory.
Kelvin never adopted electromagnetic theory.
Till death do us part: Max Planck: "a new scientific truth does not triumph by convincing its opponents... but rather because its opponents eventually die, and a new generation grows up that is familiar with it"
Not stubborn resistance to proof.
Rather, the price we pay for the depth and precision of normal science.
But the community does change, so some arguments are relevant.
Persuasion (rhetoric) rather than proof.
No single argument
Sun worship (Kepler)
personality
nationality
precise solution of crisis problems ("crucial experiments")
Copernicus' calendar reform.
Newton's unification of terrestrial and celestial mechanics
Lavoisier and weight relations
Einstein's unification of electromagnetism and mechanics.
Bohr atom.
precise success on novel problems
Wave optics on diffraction when polarization produced crisis.
Copernicus and phases of Venus
Fresnel spot.
Einstein on Mercury's perihelion advance.
simplicity and aesthetic considerations
Attract a few scientists before precise solutions are found.
Essential for scientific change, for without commitment prior to success, new paradigms could never be developed.
future promise
new, undeveloped paradigm will eventually succeed better than the old one already has.
matter of faith.
Original developers drawn by aesthetics.
Succeeding generation drawn by problem solutions. No single argument persuades them all.
Holdouts cease to be scientists but are not irrational.
Questions
Why does science progress while art, politics and philosophy do not?
Usual answer: scientific method. Bad answer!
Progress ===> Science
Fields that don't debate the nature of science are secure about their own paradigms.
Art used to succeed, but then it was inseparable from science. After dropping representation as a goal, it ceased to progress and ceased to be a science.
Science ===> apparent progress
Normal inquiry is always progressive to paradigm members
Fields that lack success have competing paradigms.
Even scientists doubt progress during revolutions (Newton vs. Descartes, Einstein vs. Bohr).
Normal science ===> real progress
Normal science exploits economy of scale to solve more sophisticated problems.
Insulation of normal science from the broader community enables scientists to select problems likely to admit of solution (social science vs. physics).
Rigid education is efficient at generating crisis. So long as a few original newcomers suggest new ideas during crisis, the loss of normal science "drones" is not costly to progress.
Revolutionary science ===> real progress?
History is written by the winner.
Select only that history that contributes to your own paradigm.
Nothing like art museums for old theories.
Successes of old theories are forgotten.
Orwell's 1984.
Might makes right?
The right might makes right.
Nature of scientific authority
solution of problems about the behavior of nature.
problems of detail.
solution standards of paradigm respected.
political power must remain irrelevant.
i.e., rules of normal science.
These characteristics are good for problem solving effectiveness.
Scientists will not move to a paradigm unless they think it will enhance problem solving effectiveness.
Progress: more problems with deeper solutions.
Science grows in depth, not in breadth (old problems are banished).
Growth in breadth = production of new subdisciplines.
No convergence to truth
Evolution from primitive state to deep problem solutions.
No progress toward truth or anything.
This may solve the problem of induction.
Evolutionary metaphor
What shocked the world was that Darwin eliminated goals from evolution.
The crisis-revolution cycle produces highly adapted scientific knowledge. (adapted to what?)
Articulation and specialization, as in biology.
No goal of truth.
Why does the evolutionary process work?
Why does consensus always reappear after revolution?
Why do paradigm changes yield better paradigms?
What must the world be like for science to succeed?
No answer necessary: if the world makes progress possible, the evolutionary method just described will find it.