Kevin T. Kelly
Department of Philosophy
Carnegie Mellon University
Don't always believe what an author in a firestorm says he always meant.
Communities exist independently of paradigms.
Education
Affiliation
Circulation of drafts.
Paradigms are solved problems that community members share.
Normal science requires the right kind of paradigm, not any old paradigm (social scientists sought to enforce paradigms after reading Kuhn).
Subject matters do not pick out communities. Paradigms belong to communities, not subject matters.
Revolutions are changes of paradigm in a community. If the community is small, the revolution may appear small from the outside.
What is shared are theories, but philosophers have ruined the term.
A neutral term is "disciplinary matrix".
symbolic generalizations (laws)
can function partly as definitions and partly as generalizations.
ontology and heuristic models
values
prediction
precision
problem-solving effectiveness
coherence
plausibility
simplicity
compatibility
"values shared but not applied the same way", e.g., simplicity. (Maybe better to say not shared?)
exemplars (what were previously called paradigms)
solved problems encountered in shared scientific education
inclined plane
derivation of Kepler's laws
calorimeter
Problem solutions are prior to the understanding of laws and theories.
f = ma is a law schema that changes form in different applications. (Because the force f changes!)
f = ma signals the gestalt to be sought in a given situation. (Come on. It is a generalization that is instantiated in different applications. That's what generalizations are for).
Problems are solved by analogy. (Logic doesn't tell you how to get a proof, it only tells you that you have one).
Solving problems provides a new way of seeing the situtaions. (Right, the law has been applied to them).
Scientists solve problems by analogy without applying the laws directly. (If the laws are known, this is a valid heuristic for finding solutions. If the laws are not known, this is a way of finding potential solutions with high prior probability).
Knowledge founded in exemplars without general rules or principles connecting them.
Different worlds = different sensations in the same stimulus environment due to different conditioning.
Learning is accomplished by presenting examples that members of the group see as being the same (i.e., learning by example).
Conditioned sensory responses to stimulation are like knowledge:
modified through inquiry
survival value
transmitted through education
No direct access to such knowledge.
Seeing electrons and other theoretical entitites is an overused metaphor that should be dropped.
Cloud droplets are interpreted as particles. (Taking back the idealism).
Still, expert and novice have different processes of interpretation.
Philosopher's wild interpretation: can't speak the same language so can't choose new paradigm for good reasons.
All he meant:
can't cast scientific method into the form of a logical proof!!!!! (David Hume already said that).
reasons function as values.
can agree on values without agreeing on application. (No need to punt to ethics so soon. They could also have different hunches or degrees of belief).
No decision algorithm that must lead all to the same decision. (But there may be a decision algorithm that doesn't).
How do all the community members end up on one side?
Persuasion
How can persuasion survive incommensurability?
Incommensurability:
Meanings shifts: metals were compounds, now are elements.
Holism: meaning fixed by usage, practice, and other changed concepts.
Incommensurability: parties cannot define differences of usage.
Centrality: incommensurability is worst on just the points where the two sides disagree.
Some positive considerations:
Impinging stimuli (not basic experiences) are the same
Neural wiring and environment up to revolution are the same.
So ordinary experience is the same.
Parties to debate become translators (Quine reference).
Learn to predict what other would say
What history of science is supposed to do.
Persuasion = Vicarious experience of merits of other's viewpoint.
Conversion doesn't necessarily follow persuasion.
Levels of change
Pre-translation persuasion: "I don't know how they do it but I better find out". Beginning graduate student.
Persuasion: seeing how they do it through translation. Foreign to normal science, unsettling.
Conversion (going native): seeing what they see, talking how they talk. Not voluntary. Perhaps impossible for the older generation.
Relativism
Meaning and usage are relative
Value of puzzle solving is not relative in science
Objective progress:
No progress toward truth (anti-realism)
No theory-neutral conept of "really there" to survive revolutions
Normative/Descriptive
is is mixed with ought
circular arguments are OK
success is ===> ought
Scientific debate cannot be