Transcript PPTX file - UCL Department of Geography

Graham Bradley
Lecture 1
What is science?
 Geography and science
 Scientific explanation
 Scientific reasoning
 Francis Bacon and induction
 David Hume’s problem
 Karl Popper and falsification
 The hypothetico-deductive method
 Example: climate change
What do you think science is?
 Sceptical (up to a point)
 Based on observation, data, experimentation
 Conclusions are tentative
 Theories (models) can be tested / falsified
 Assumes a chain of cause and effect
 Explains by generalising
 Often quantitative and mathematical
But what do scientists think science is?
“Ask a scientist what he conceives the scientific
method to be, and he will adopt an expression
that is at once solemn and shifty-eyed: solemn
because he feels he ought to declare an opinion;
shifty-eyed because he is wondering how to
conceal the fact that he has no opinion to
declare”
Sir Peter Medawar
UCL Professor of Zoology 1951-1962
Nobel Prize in Medicine 1960
What are the goals of science?
1. Description
 Identification and classification of entities, events &
patterns
2. Prediction
 Use observed regularities to infer unobserved
phenomenon
3. Explanation
 Explicate the causal relations between described and
predicted phenomenon
4. Stewardship/Control (applied science/engineering)
 Apply knowledge to bring about desired outcome
What is scientific explanation?
1. Cause and effect – to explain a phenomenon is
to say what caused it

E.g. What is the cause of lake acidification?
2. Covering laws – show that phenomenon to be
explained is ‘covered’ by some general law of
nature
E.g. Darcy’s Law (groundwater flow) Q = KiA
 General law & particular facts → explanation
 Every reliable prediction is a potential explanation

Induction and deduction
 Deduction
 Inference, by reasoning, from general to particular: from
theory to data
 Premises: i) every mammal has a heart; ii) every horse is a mammal.
 Conclusion: Every horse has a heart.
 Premises: i) all ‘U’ shaped valleys were formed by glaciers; ii)
Wasdale is ‘U’ shaped valley.
 Conclusion: Wasdale was formed by a glacier
 Valid if the truth of premises guarantees truth of
conclusions & false otherwise.
 Conclusion is either true or false
Induction and deduction
 Induction
 Process of inferring general principles from observation of
particular cases: from data to theory
 Premise: every horse that has ever been observed has a heart
 Conclusion: Every horse has a heart.
 Premises: i) Death by cholera spatially clustered; ii) spatial clusters
around water pumps.
 Conclusion: water pumps are the source of cholera.
 Conclusion goes beyond information present, even
implicitly, in premises
 Conclusions have a degree of strength (weak -> near
certain).
Induction and deduction
The origins of science:
Aristotle (384-322 B.C.E.)
 Plato – emphasis on a priori knowledge
 Aristotle – greater emphasis on a posteriori knowledge
 Empirical inquiry of “the form within things”
 Elements: earth; air; fire; water; aether
 Teleology – nature reflects inherent purpose and
direction
 Aristotelian view remained dominant until 16thC
Francis Bacon (1561-1626)
Can induction identify causes?
 Attorney General, Lord Chancellor of England and
philosopher who inspired the formation of the Royal
Society
 Rejected many a priori assumptions of Aristotelian
view and advocated the Baconian method of inductive
inquiry:
 Identify phenomenon and rank list of things in which it occurs
 Use inductive reasoning to verify the cause of phenomenon
Rev. Thos. Bayes (1702-1761)
Formalise scientific process via probability
P ( Hypoth. | Data, I ) µ P ( Data | Hypoth., I ) ´ P ( Hypoth. | I )
Posterior
Likelihood
Prior
 Bayes’ Theorem: solves the inverse (inductive) problem i.e. gives
probability of a hypothesis being true given some data and any prior
knowledge
 THIS is how science is really done! BUT is (sort of) subjective as
requires stating priors explicitly P(H|I)
 Ignored for 200 years: replaced by “statistics” – estimate reliability of
a given set of data (compared to infinite other possible sets) in the
light of a given (null) hypothesis (model). But we don’t HAVE infinite
other data sets in practive
Eg Laplace & the mass of Saturn
P ( H | Data, I ) µ P ( Data | H, I ) ´ P ( H | I )
Posterior
Likelihood
Prior
 Laplace (1749-1827) estimated MSaturn from orbital data i.e. H is the posterior
prob(M|{data},I) where I was background knowledge of orbital mechanics etc.
 Shaded area shows degree of belief that m1 ≤ MSaturn < m2 (right to within < 0.7%)
 How do we interpret this pdf in terms of frequencies?
 Some ensemble of universes all constant other than MSaturn? Distribution of MSaturn in repeated
experiments?
 But data consist of orbital periods, and these multiple expts. didn’t happen
Best estimate of M
Degree of
certainty of M
The posterior pdf expresses
ALL our best understanding
of the problem
Karl Popper (1902 - 1994)
Can deduction identify causes?
 Impossible to verify a universal
statement which would require infinite observations
 Possible to falsify a universal statement with a single
counter-observation
 Falsifiction is deductive: if the single case is false then it
logically follows that the universal case is also false
 Popper stated some theories thought to be scientific at
the time are unfalsifiable and therefore not science
e.g. Freud’s psychoanalysis; Marxist theory
Reference: Popper, K., 1959. “The Logic of Scientific Discovery”
Example: Climate Change
What has caused global warming?
How much of the observed change is
due to natural variability and how much to
anthropogenic influences?
Natural Variability and Climate Change
 Internal mechanisms
 Ocean/atmosphere interaction e.g. ENSO
 Thermohaline circulation
 External mechanisms
 Ash from volcanic eruptions
 Variability in solar irradiance
Effect of Volcanic Ash
Variations in Solar Irradiance
Anthropogenic carbon dioxide
Competing Hypotheses
H1 The observed record is consistent
with natural climate variability only
H2 The observed record is consistent
with natural and anthropogenic forcing
Use proxy record of global temperature and
General Circulation Models to test the hypotheses
First Hypothesis:
The observed temperature record is consistent
with natural climate forcing only
Falsified
Second Hypothesis:
The observed temperature record is consistent
with natural and anthropogenic climate forcing
Corroborated
Problems of Falsification
 Most scientists are not trying to falsify theories:
 e.g. When asked ‘What if relativity had been falsified?’ Einstein replied:
‘I would have been sorry for the dear Lord as my theory is correct.’
 Tenacity – It may be good to hold on to a falsified theory
 Popper: ‘He who gives up a theory too easily in the face of apparent
refutations will never discover the possibilities inherent in his theory’
e.g. Uranus didn’t conform to Newton’s laws - the discovery of Neptune
 Corroboration – Degree of testing a theory has undergone
 Popper argued a highly corroborated theory has a greater level of
‘truthlikeness’, but the logical conditions for comparison cannot be met
 Popper was unable to provide a logical method of consistently choosing
between unfalsified theories
Gauch (2006): “Seven pillars of Science”
1.
2.
3.
4.
5.
6.
7.
Realism: physical world is real;
Presuppositions: world is orderly and comprehensible;
Evidence: science demands evidence;
Logic: science uses standard, settled logic to connect
evidence and assumptions with conclusions;
Limits: many matters cannot usefully be examined by
science;
Universality: science is public and inclusive;
Worldview: science must contribute to a meaningful
worldview.
Summary
 Scientific methods address empirical claims
 Demarcation criteria: empirical, laws, testable etc
 Scientific explanations: cause & effect, covering laws
 Scientific reasoning: inductive or deductive
 The problem of induction
 Karl Popper and falsification
 The hypothetico-deductive method
Final thought:
How do scientists choose between
unfalsified theories?
Reading:
 Okasha, S., 2002. “Philosophy of science, a very short introduction”
 Recommended introduction for the general reader. Also available as audio book.
 Chalmers, A. F., 1999. “What is this thing called science?” 3rd edition
 Recommended text for anyone with an interest in the philosophy of science.
 Inkpen, R., 2005. “Science, philosophy and physical geography”
 Introduction to philosophy of science for physical geography undergraduates.
 Gauch, H. (2003) The Scientific Method in Practice
 Gauch, H. (2006) Science, Worldviews and Education, Sci. and Edu., DOI
10.1007/s11191-006-9059-1.
 see bothon Moodle
 Montello, D. R. and Sutton, P. C., 2006. “An introduction to scientific research
methods in geography”
 Beginners guide to empirical aspects of human and physical geography research, with well
balanced introduction on the philosophy of science and its place in geography.
 Orme, A. R., 2002. “Shifting paradigms in geomorphology” Geomorphology, Vol.
47, Issues 2-4, pages 325-342
 A paper of particular interest to physical geographers.
What’s in a theory?
Natural Science – Social Science – Humanities
Sets of assumptions, ideas, arguments and conclusions
•
An analytic structure
designed to provide a
general explanation of
observations
•
A set of interpretive
principles that facilitate a
specific rational or moral
analyses