Transcript Main Points from reading Kuhn

“Science” – or how to learn
Lessons from Thomas Kuhn
and
the Philosophy of Science
Economics 101, Fall 2004
Prof. Smitka
What is “Science”? -- the Received View
“The” Scientific Method
• Form the hypothesis
– Then design an experiment
– Collect data
• Test the hypothesis
– Check controls
– “support” the hypothesis, or reject it, or (typically) find weak evidence.
• in the latter two cases, iterate:
– Modify the hypothesis (and implicitly any underlying theory)
or
– when confident of the underlying theory, modify the experiment
• Such “normal science” does however underly all applied work
Chronology of the Copernican Revolution
actual science versus the stereotype
Aristotle (384-322 BC): systematizer of classical science, into which the astronomy of Ptolemy (127-151
AD) was incorporated, rather than the heliocentric views of the Heracleides (4th century BC).
Copernicus (d. 1543): late medieval cleric, doctor, lawyer, astronomer, whose final text was influential
for its mathematical framework even if its cosmology was ignored.
Tycho Brahe (d. 1601): Danish observational astronomer, who used better instruments and years of
systematic observation to qualitatively improve the data on planetary motion. He proposed a mix of
the heliocentric and geocentric models.
Johann Kepler (1571-1630): Brahe's final assistant, applied a neo-Pythagorean metaphysics of
"harmony" while correctly positing in 1609 a Copernican solar system with elliptical orbits.
Galileo (1564-1642) with his invention of the telescope around 1610 observed Jovian moons and the
phases of Venus to provide the first independent evidence of orbiting bodies. But his heresy trial in
1633 made adherence to a geocentric view dangerous.
Isaac Newton (published 1687) whose theory of gravity and of motion provided a unifying framework for
physics & astronomy; his work on planetary motion dates to 1684-85.
Popular acceptance did not come until the 1700s, and even later in continental Europe, where Renee
Descartes vigorously opposed it.
How Copernicus Won
• Few astronomers ever changed camps
• Instead, the new astronomy gathered more
disciples over time
• That process only partially matched
– advances in empirical work (Brahe, Galileo)
– advances in theory (Kepler, Newton)
• Scientific change was thus a social process
What is “Science”?
The Empirical Reality
• Science is a system of Apprenticeship
• Cf. learning the ropes at W&L
• Science is a Social Endeavor focused on a set of problems
• Cf. learning the ropes at W&L
– Jargon
– Varying standards of “good” work or “truth”
– Conservative (in the sense of resistant to change)
• Learning is organized in that manner
– solving standard exemplars and using them to approach new problems
– learning by doing, learning by association
– ultimately the gradual accumulation of discretely learned problems
• It is false to distinguish “hard” from “social” science
– hypothesis testing is not the core feature of science
– even English is science, with jargon, models, an apprenticeship system
Disciplinary Matrix
• Socialization / organizational identity
– Jargon [separates “us” from “them”]
– Professional organization [where “science” happens]
– Journals [peer recognition, communication]
– Personal networks [tied to all of above]
• Shared
– Symbolic generalizations ["tools" – graphs, equations]
– Exemplars [standard examples & problems]
– Stylized facts [complementary to and not independent of theory]
– Models [formal “theory”]
How economists work
• Data mining, not always consciously
– Same data reused, with general of sense of what works
• Theories almost never rejected
• “Supply & Demand” is never tested -- it’s a tool, not a theory
– Work without statistical significance never sees the
light of day, biasing judgments of what “works”
– Don’t reject the theory, revise functional forms…
– Core theories, once accepted, die slowly if ever
• In macro, the “liquidationist” theory of business cycles
• So for an economist, what is “truth”?
Desiderata for a Theory
Plus notes on economists’ “truth”
•
tractable (you can actually do the math, important in economics)
•
accurate (in tension with simplicity, and not important in economics
with its poor quality data and unobservable concepts)
•
orthodox (builds upon rather than displacing existing work -generally important in economics)
•
fruitful (generates new insights, not central in economics)
•
general (explains lots of phenomena, not central in economics)
•
important ("key" questions or policy, often central in economics)
•
beautiful (a low priority in economics)
•
clever (the primary consideration in economics; we attack and
“solve” puzzles, ideally with a new trick or “tool”)
Adapted from Thomas Kuhn, who elaborated a smaller set of criteria.
Sum: Learning Matrix
• We should always associate:
– Exemplar
– Jargon
– Tools / solution method
– Theory / model
• You should watch for:
– “Keying” terms that guide towards the exemplar
• Always ask, “Where have I seen that before?”)