Common Cause Explanation
&
The Search for a “Smoking Gun”
Carol E. Cleland
Philosophy Department
Center for Astrobiology
University of Colorado (Boulder)
OVERVIEW
 Myths about the scientific Method.
 Classical experimental science and prototypical
historical science: two different but equally rational and
objective patterns of evidential reasoning.
 How evidence acquired through field work justifies
historical hypotheses: Common cause explanation and
the search for a “smoking gun”.
Part I
Myths about the Scientific method
 Inductivism: Scientists prove theories and
hypotheses by a logical process of
induction.
Myths about the Scientific method
Falsificationism: Scientists falsify
theories and hypotheses by using
empirical evidence to refute them.
The Logic of Prediction
Basic Concepts
1. Hypothesis (H):
(All C’s are E’s)
“Toy” Example
All copper expands when
heated.
2. Test Implication (I):
(If x is a C, then x is an
E.)
If sample of copper #4 is
heated, then it will expand
3. Test Condition (C):
(Heating copper sample #4)
4. Prediction (E):
(Copper sample #4 will
expand)
The Logic of Evaluating the Results
of an Experiment
 Successful Prediction
1.
2.
C.
If H, then I
I
H
 Logical Fallacy: “affirming the consequent”.
(This is just another version of the problem of induction.)
The Logic of Evaluating the Results
of an Experiment
 Failed Prediction
1.
2.
C.
If H, then I
Not-I
Not-H
 Valid Argument Form: “denying the consequent”.
(This explains the appeal of falsificationism.)
The Terrible Truth about
Falsificationism
The form of the first premise in the
previous argument is:
If H and A, then I
(where ‘A’ stands for a set of auxiliary
assumptions {a1, a2, …, an} about other
conditions, known and unknown, about the
actual experimental situation.)
The Terrible Truth about Falsificationism (continued)
 This changes the form of the
argument to:
1. If H and {a1, a2, …, an}, then I
2. Not-I
3. Not-(H and {a1, a2, …, an})
4. Not-H or not-{a1, a2, …, an}
5. Not-H or not-a1 or not-a2 or … or
not-an
(by De Morgan’s theorem)
The Terrible Truth about Falsificationism (continued)
From a logical standpoint, no observation
(whether experimental or in the field), can
conclusively falsify a hypothesis. For it is always
possible to salvage the hypothesis in the face of
a failed prediction by denying an auxiliary
assumption.
More Nails in the Coffin of
falsificationism
 Falsificationism is not only logically but also
historically flawed.
 Faced with failed predictions scientists have
historically
denied auxiliary assumptions, e.g., perturbations in the orbits
of Uranus & Mercury.
 Falsificationism is inconsistent with the practice
of scientists & training of young scientists
 Faced with failed predictions scientists typically deny and
indeed are trained to deny auxiliary assumptions rather than
the target hypothesis.
Conclusion
Neither inductivism nor falsificationism
provides a satisfactory account of any
scientific practice; the scientific method of
yore is a myth.
Part II
Differences in the methodology of classical
experimental science and prototypical
historical, natural science:
Is historical natural science methodologically
inferior to experimental science?
The structure of
Classical Experimental Science
 Focus: Is on a single (sometimes complex) hypothesis
which typically has the form of a universal generalization
(All C’s are E’s).
 Central Research Activity: Consists in repeatedly
bringing about the test conditions specified by the
hypothesis and controlling for extraneous conditions
that might be responsible for false positives and false
negatives.
The Experimental Program vs.
Solitary Experiment
 Failed predictions: do not result in the rejection of
hypotheses; they are best interpreted as attempts to
protect the hypothesis from false negatives.
 Successful predictions: Are not followed by risky tests
(in Popper’s sense); they are best interpreted as
attempts to protect the hypothesis from false positives.
 Acceptance/rejection of a hypothesis: occurs only
after a hypothesis is subjected to a series of
experiments controlling for plausible auxiliary
assumptions that could explain predictive successes and
predictive failures.
The structure of
Prototypical Historical Science
Focus: Is on proliferating multiple, rival hypotheses
to explain a puzzling body of traces of past
events (data) encountered in field work.
Central Research Activity: Consists in searching
for a ‘smoking gun’ a trace(s) that sets apart
one or more hypotheses as providing a better
explanation for the body of traces thus far
acquired than the others.
A Case Study
The Alvarez Hypothesis



Two-pronged hypothesis: 1) impact; 2) extinction.
Initially many different explanations for the endCretaceous mass extinction: pandemic, evolutionary
senescence, climate change, supernova, volcanism,
and meteorite Impact.
Discovery of an iridium anomaly (“smoking gun”) in
K-T boundary sediments narrowed it down to two
possibilities: volcanism and meteorite impact.
Discovery of extensive quantities of a rare form of
shocked mineral subsequently cinched the case for
impact over volcanism.
A Case Study: The Alvarez Hypothesis (cont)
 Paleontologists weren’t convinced: They
agreed that there had been a meteorite impact
but many doubted that it caused the endCretaceous extinctions.
 The discovery of extensive pertinent fossil evidence
(especially small organisms such as foraminifera and
ammonites, and fern spores and angiosperm pollin) on
either side of the K-T boundary was pivotal in
changing their minds, providing the needed smoking
gun for the second prong (mass extinction) of the
hypothesis.
Lessons from the Alvarez hypothesis: The
evaluation of historical hyotheses is:
 Not grounded in prediction:
 Historical predictions are not ‘risky’ in Popper’s sense; too
many highly plausible extraneous conditions (e.g., iridium
poor meteorite, geological processes of concentration and
dispersal, unrepresentative samples of K-T boundary)
capable of defeating them.
 Predictions are typically vague, e.g., Ward’s ‘prediction’
about Cretaceous ammonites; they serve more as roadmaps
for looking for a smoking gun than predictions.
The Evaluation of Historical Hypotheses (cont.)
 A hypothesis may be rejected on the basis of
evidence that does not refute it, e.g., the
contagion hypothesis for the end-Cretaceous
extinctions.
 The acceptance of a hypothesis does not
require a successful prediction, e.g., the iridium
anomaly was not and could not have been
predicted or retrodicted.
The Evaluation of Historical Hypotheses(cont.)
 Grounded in explanatory power:
 Hypotheses are accepted and rejected in virtue of
their power to explain (vs. predict) puzzling bodies of
traces discovered through field work.
 The Alvarez hypothesis explains an otherwise
puzzling association (correlation) among traces better
than any of its rivals. It is for this reason that it is
viewed as ‘confirmed’ and its rivals are no longer
seriously entertained by scientists.
Part III
Common cause explanation and the
search for a smoking gun
Common Cause explanation

Reichenbach’s Principle of the Common Cause: seemingly
improbable associations (correlations or similarities) among
traces are best explained by reference to a common cause.
C

E1
E2
E3
E4
Presupposes that the temporal structure of causal
relations in our universe is such that most (not all) events
form causal forks opening from past to future (leave many
traces in the future).
Common cause explanation (cont.):
But is there any reason to believe the
principle of the common cause is true?
YES!
The Asymmetry of Overdetermination
 A time asymmetry of causation:
Most local events & structures overdetermine their
past causes (because the latter typically leave
extensive and diverse effects)and underdetermine
their future effects (because they rarely constitute
the total cause of an effect)
 Much easier to infer an ancient volcanic eruption than a near future
volcanic eruption.
The Asymmetry of Overdetermination (cont.)
 Physical source is controversial but it characterizes all wave
(radiative asymmetry)and particle (2nd law of thermodynamics)
phenomena above the quantum level; an objective and pervasive
physical feature of world.
 Physically (vs. logically or strictly metaphysically) grounds the
Principle of the Common Cause and the methodology of
historical natural science: the Search for a smoking gun.
An illustration: The colors of
dinosaurs
Asym of OD Asserts that the present is filled with
overdetermining traces of the past; hence
one can never completely rule out finding a smoking gun
for any scientific hypothesis about the past. The
methodology of historical field work is based upon
this possibility.
Conclusions
1. Historical Scientists exploit the overdetermination of
the past by the localized present by searching for a
smoking gun to discriminate among competing
hypotheses; the asymmetry of overdetermination
guarantees there are likely to be many such telling
traces. The problem is recognizing them for what they
represent.
Conclusions
2. Experimental scientists try to circumvent the
underdetermination of the future by the localized
present by constantly testing for false positives and
false negatives that might yield misleading confirmations
or disconfirmations of their hypotheses; the asymmetry
of overdetermination guarantees that this is always a
threat.
 There are no records of the future.
Conclusions
3.The methodology of historical science is
different from that of classical
experimental science but it is not inferior;
each practice is designed to exploit the
differing information that nature puts at
its disposal.
References
 “Common cause explanation and the search for a smoking gun” in Baker, V.
(ed.), 125th Anniversary volume of the Geological Society of America
(forthcoming).
 “Prediction and Explanation in Historical Natural Science,” British Journal of
Philosophy of Science 62 (2011), 551-582.
 “Philosophical issues in natural history and its historiography” in Tucker,
A. (ed.), Blackwell Companions to Philosophy: A Companion to the
Philosophy of History and Historiography. Oxford: Blackwell Pub.
(2009), pp. 44-62.
 “Methodological and Epistemic Differences Between Historical Science and
Experimental Science,” Philosophy of Science 69, (2002), pp. 474-496.
 “Historical science, experimental science, and the scientific method,” Geology
29, (2001), pp. 987-990.