Philosophy 226f: Philosophy of Science
Prof. Robert DiSalle ([email protected])
Talbot College 408, 519-661-2111 x85763
Course Website:
Philosophy of Science in the 20th Century:
What was the “Received View”?
Science is empirical: the ultimate criterion for judging a
scientific theory is its agreement with the empirical facts.
Science is rational: scientists’ judgments in general are
influenced by empirical facts and logical inferences from them- not by “extraneous” social, psychological, or political
Science is cumulative: The achievements of scientists are
permanent possessions of science in general-- facts that can be
built upon by future scientists
“Scientific Philosophy”, a.k.a. Logical Empiricism:
In general, statements can either be true or false.
If true, this means:
If it is a logical or mathematical truth, it can be logically
derived from first principles.
If it is a statement about what there is in the real world, or
any matter of empirical fact, it can be verified by some
If false, this means either that it is logically contradictory, or
that it is contradicted by the facts.
Verification and Meaning:
A statement that cannot be verified by any empirical
observation or logical reasoning, even in principle, is neither
true nor false.
It is completely meaningless.
Example: “Nothing nothings” is neither true nor false.
It simply has no cognitive content.
Whatever content it might have is emotional rather than
(It is not a direct statement about any state of affairs. It is an
indirect statement about the emotional state of the speaker.)
Deductive logic (drastically oversimplified):
All A are B.
X is an A.
Therefore X is B.
Inductive logic
All copper we have tested conducts electricity.
X is a piece of copper yet to be tested.
Therefore X will conduct electricity.
The logic of induction (cf. David Hume)
All A observed so far are B. [i.e. All A are B]
X is an A not yet observed.
[i.e. X is not an A]
Therefore X is B.
[X is B.]
What does it take to confirm a universal generalization?
“Falsificationism” (Karl Popper)
Scientific theories are never truly verified. Moreover, to be
always verified is not a virtue in a scientific theory.
Verification and falsification are asymmetrical:
No accumulation of confirming instances is sufficient to
verify a universal generalization.
But only one disconfirming instance suffices to refute a
universal generalization.
Scientific theories are distinguished by the fact that they are
capable of being refuted. They are falsifiable.
Example: Confirming Freudian psychoanalysis
How do we know that repressed memories of infantile sexual
desires are the causes of neurosis?
These desires are revealed in our dreams, “Freudian slips,”
free associations, and other “symptoms”. They are the “latent
content” expressed symbolically.
How do we determine the true meaning of these symbols?
We interpret them-- which requires us to apply Freud’s theory.
What if the patient denies the interpretation?
The patient is “resisting,” which indicates that the
interpretation is correct and therefore disturbing to the patient’s
conscious mind..
Pseudo-science: A theory with the
empirical “trappings” of real
science, including a system of
theoretical concepts and a wealth of
corroborating evidence.
But a pseudo-science has built-in
“defense mechanisms” against
possible refutation.
The Freudian theory provides an
interpretation for every
conceiveable symptom of the
Its “predictions” therefore can never
be refuted.
Einstein’s General Relativity: If it
had failed its famous test of 1919,
no one would have taken it
But it passed the test, and
Newton’s theory of gravitation was
Albert Einstein (1879-1955)
“On the electrodynamics of
moving bodies” (1905)
“The foundation of the general
theory of relativity” (1916)
Empirical test of general relativity vs. Newtonian gravitation:
Light from a star
passing near the Sun
should be deflected.
The evidence is the
displacement of the
star’s apparent
How the observation worked:
Path of the eclipse
Actual observation
Gravitational “lensing” as we now know it
Einstein and Eddington, looking pretty pleased with themselves
“The Methodology of Scientific Research Programmes”
(Imre Lakatos. 1922-1974)
How do scientists
decide whether, or
when, their theory is
How do we explain
why scientists persist in
working on theories in
the face of counterexamples?
Lakatos: Scientific theories are not really falsifiable. They are
“research programmes” that consist of:
a “hard core” of fundamental principles that contain what the
theory really says about the world, and
a “protective belt” of “auxiliary hypotheses” that explain how the
fundamental principles apply to particular cases, and how
to deal with apparent discrepancies. These include “ceteris
paribus” (“other things being equal”) clauses that accommodate
problematic cases.
Contrary to Popper, even good theories have “defensemechanisms.”
Newtonian Gravitation Theory: Predicts that every acceleration
of every body can be traced to an interaction with some other
body, according to their masses and the distance between them.
F 
Gm 1 m 2
What to do when we observe an acceleration that has no
visible source? Is the theory refuted?
The theory demands that the missing mass be found.
1687-1727: Newton breaks his head against the problem of
the motion of the moon, which he cannot predict precisely
from his law of gravitation
1748: Clairaut solves the problem of the motion of the moon
using Newton’s theory and better mathematical techniques.
1821: A slight discrepancy is noted between the actual motion
of Uranus and the motion predicted by Newton’s law of
gravitation. Astronomers puzzle over it for a while.
1843: John Crouch Adams deduces, from Newton’s law and
Uranus’ orbit, the approximate location of a new planet.
1846: The new planet is discovered by Adams and Leverrier.
“The theory of gravity, which, by so many applications, has
become a means of discovery, as certain as by observation
itself, has made known to [the mathematician] several new
inequalities...enabled him to predict the return of the comet of
1759....He has been enabled by this means to deduce from
observation, as from a rich mine, a great number of important
and delicate elements, which, without the aid of analysis, would
have been forever hidden from view....”
Laplace, Mécanique céleste
Mercury’s perihelion is found to precess at a rate that does not
agree with Newton’s theory. The difference is 43” per century.
1855-1916: various hypotheses are advanced to explain this
discrepancy between theory and observation.
For example:
--is there another planet (“Vulcan”) near to Mercury’s orbit?
--does the force of gravity vary, not as
r2 ,
but as
--is there a cloud of matter near the sun that affects Mercury’s
1916: Einstein shows that general relativity predicts precisely the
missing 43”.
Lakatos: The fundamental difference is between progressive
and degenerating research programmes.
Progressive research programmes lead to novel predictions, new
problems, and new solutions.
Degenerating programmes spend their time trying to adjust after
the fact to new information, and to protect themselves from
refutation by constant adjustment.
The “pseudo-sciences” are really degenerating programmes
whose practitioners are mainly devoted to defending the
programme against contrary evidence. The progressive
programmes view contrary evidence as a challenge that will
broaden and deepen the theory.
Thomas Kuhn on Scientific Revolutions
(cf. The Structure of Scientific Revolutions, 1962.)
Paradigm: Kuhn’s idea that a scientific theory is not just a set of
theoretical principles. It is an entire world-view, consisting of:
--“Metaphysical” views about the nature of the world and the
things in it;
--methodological rules about correct scientific practice;
--a conception of what constitutes a legitimate scientific question
and what doesn’t
--a conception of what constitutes a scientific fact;
--“paradigm exemplars” of the right kind of problem to solve and
the right way to solve it.
A paradigm, therefore, determines not only a set of beliefs about
the world.
It also defines what counts as good science, and even determines
what counts as a scientific fact.
It is a conceptual framework that determines how the world
looks to those who have accepted it.
It defines not only the scientific outlook for practitioners of a
particular science, but also the scientific “form of life.”
There are two aspects to the history of any science:
Normal science: science pursued within the constraints of a
particular paradigm, without questioning its principles. The
characteristic activity is “puzzle solving,” i.e. answering
questions set by the paradigm using the methods sanctioned by
Revolutionary science: a time of decreasing confidence in the
existing paradigm (because of the accumulation of unsolved
puzzles), and conflict with alternative paradigms.
This is like a political crisis, with uncertainty, and conflict
among many views, until a new order becomes established and a
single paradigm takes a position of authority.
Some philosophical claims arising from Kuhn’s view:
The conflict among paradigms can’t be settled on any rational
methodological grounds, because each paradigm contains its
own view of rational scientific methodology.
The conflict can’t be resolved by an appeal to the facts, since
each paradigm contains a view of what counts as a fact, and will
determine how its adherents view the facts.
Different paradigms are in fact “incommensurable,” not
comparable by any neutral standard. Adherents of different
paradigms “live in different worlds,” and speak different
languages that are not inter-translatable. A change of paradigm
involves changes in the meanings of basic theoretical terms.
The replacement of one paradigm by another can’t be viewed
as progressive on any objective grounds.
Since adherents of different paradigms define the questions
differently, and accept different standards for a good answer,
the conflict between them has no neutral resolution.
A scientific revolution has to be regarded as a social and
psychological phenomenon rather than as a purely
intellectual one. For an individual scientist, the change in
point of view is more like a religious conversion than a
rational process of comparing theories against the facts.
Q uickTim e™ and a
TIFF (U ncompressed)decompr essor
are needed to see this pic ture.
Qu ickTi me™ a nd a
TIFF (U ncomp re ssed) deco mpressor
are nee ded to see thi s picture.
Qui ckTime™ and a
TIF F (Unc ompress ed) dec ompress or
are needed to see t his pic ture.
Some historical claims arising from Kuhn’s view:
Scientists with different theoretical viewpoints generally fail to
understand one another.
Competing paradigms appeal to different and conflicting sets of
facts, and proceed by conflicting methods.
The arguments made in favor of one theory cannot be fully
understood by, or persuasive for , adherents of the other.
New paradigms introduce new theoretical terms, or change the
meanings of old ones, in ways that are incomprehensible to
anyone who doesn’t already accept the new theory.
A new paradigm doesn’t explain more than its predecessor. Even
if it can explain things that the old theory couldn’t, it will
typically fail to explain many things that the old theory could
explain. (This has been called “Kuhn loss.”
The history of science is not cumulative: new theories can’t
incorporate the successes of older ones, because they have a
completely different view of what counts as success. The new
theory redefines the old theory in its own terms.
Kuhn’s list of “values” for judging scientific theories:
Accuracy: degree of agreement with the available empirical
Consistency: freedom from logical contradictions
Simplicity: lack of unnecessary complication; “unity”
Scope: Range of phenomena that fall within the theory’s grasp
Fruitfulness: Power to generate new principles, problems,
solutions, predictions, etc.
Question: Does agreement on these values imply agreement on
their application, their relative importance, etc?