Though controversial, the description of science by
Karl R. Popper (1902-2004) stands as one of the most influential achievements
of the philosophy of the twentieth century; his falsification hypothesis is a
turning point of philosophy of science and divides the field in "before
Popper" and "after Popper". As it often happens with
breakthroughs, the falsification hypothesis is easy to explain and rich in
relevant consequences: scientific theories, Popper says, cannot be proven true
because of two related reasons: first they have infinite consequences of which
experiments test only a few. If the experiment agrees with the prediction, this
increases our confidence in the theory but we cannot exclude that it will fail
to predict the next experiment. Second, the theory may correctly predict the
outcome of the experiment by chance, and the mechanisms it postulates may be
incorrect; e.g. Ptolemy’s cosmology, although wrong, correctly predicts the
apparent movements of many stars and planets. In the long run bad theories fail
their predictions and are discarded, whereas good theories are further
developed and perfected. If scientific theories cannot be proven (i.e.
verified), they can be disproved (i.e. demonstrated to be false, falsified),
for as soon as a theory fails a prediction it must be declared wrong. Thus an
asymmetry exists between truth and falsity, for only the latter, and not the
former, can be demonstrated.

Popper followed his hypothesis with rigor and built up a "negative"
epistemology, in which the scientist is committed to formulate theories and to
demonstrate that they are not disconfirmed, rather than confirmed, by
experiments. It can be argued that scientists try to propose true theories, and
to demonstrate that they are indeed true; some scientists become so involved
with their theories that they stuck to them even in the front of obvious
inadequacies. To this Popper would reply that the very same experiment that
"corroborates" the theory if its result confirms the expectation,
disproves it if the contrary occurs; and that in any case what the scientist
feels is not necessarily what a rigorous logical examination would reveal.

Popper’s negative epistemology can be used to distinguish scientific
theories, which make risky predictions and can be falsified, from
non-scientific or pseudo-scientific theories, which cannot be falsified. Indeed
Popper weighted on his balance several theories, such as Marxism and
psychoanalysis, which pretended to be scientific and found them incapable of
any testable prediction and non-falsifiable.

Karl Popper: Critical Appraisals is a collection of eleven essays that evaluate the most controversial
aspects of Popper’s philosophy of science and society. It is an excellent book
and all contributors are highly qualified; the intended audience is however
almost as qualified as the authors themselves, and it is expected that the
reader is quite familiar with most, if not all, the writings of Popper. It is
neither an introduction to Popper, nor a global analysis of his contributions.
Actually, some of the essays focus on quite specific and problematic aspects of
his thought, and his most important hypotheses are discussed to a lesser
extent, as is typical of specialized analyses.

Two crucial and related points of Popper’s epistemology are the refusal
of induction and the role of observation and experiment. These are analyzed and
criticized in several essays of this book. Popper had two objections to
induction, clearly but not ordinately formulated: we cannot completely trust
observation; and we cannot legitimately generalize from observation. The former
objection is not strictly against induction, and is usually formulated by
conventionalist theoreticians; the latter is Hume’s classical argument against
induction. Popper’s solution is radical: the experiment is not a reason of the
theory, it is only a reason to trust (or not to trust) the theory. Denying the
observation a status in the content of the theory is a bold move, more easily
defensible when considering theoretical physics than anatomy, and it is
difficult to believe that it is adopted by many researchers. If you feel
stimulated (rather than bored) by this type of enquiry, then "Karl Popper:
critical appraisals" is your book, and you will find there a thorough
analysis of Popper’s ambivalent feeling about experiment and observation, even
better than the one you find in Popper’s writings. Take the simplest empirical
description you can imagine, something like "I am reading from a computer
screen". You can substantiate your statement with other statements, but
this leads to infinite regression; to stop regression you may either establish
non-questionable postulates (this being conventionalism or dogmatism) or accept
as a proof a statement describing your sensorial perception (this being
psychologism). Most scientists are implicitly psychologists: they trust
experience and observation, and Popper concedes that sensory perceptions are in
general remarkably accurate, for this grants the organism a selective
evolutionary advantage. However, Popper thinks that hypothesis and theories
cannot be based upon undemonstrable sensory perceptions and thus he only
assigns to experience a role in justifying our belief in a theory: if we see a
theoretical prediction fulfilled, then we trust the theory. Popper’s path
between the precipices of empiricism on the one side and conventionalism on the
other is narrow indeed, and the essays in these critical appraisals are a
useful guide.

Three essays, by Alan Musgrave, Semiha Akinci and Philip Catton describe
the relationships between Popper’s theory and conventionalism on the one hand
and with induction on the other hand. I found them very interesting as they
clarify some points of Popper’s theory that I had always found quite obscure.
Popper opposed the verificationist theory of the logical positivists of his
time, who assumed that describing an experience is non-problematic; he pointed
out that between the observed fact and its description there stands the logical
barrier of psychologism that introduces in the logically demonstrable structure
of the theory the empirical and non-demonstrable step that uses the hard-wired
circuitry of our brain to convert a fact into a description. Refusing
psychologism entails the paradoxical consequence that Popper’s theory may
appear a refined version of conventionalism. To state it more clearly, we may
ask ourselves which proof we would accept of a scientific hypothesis or
prediction. If the proof we demand is an observation or an experiment, then we
are positivists and Popper accuses us of psychologism, i.e. of relying upon the
poorly known functioning of our brain for the judgment of consistence between a
fact and a statement. If the proof we demand is logical coherence with other
parts of the theory, then we are conventionalists, and Popper accuses us of
neglecting the world we try to describe. We may then ask which proof Popper
would accept, and his answer is none: a hypothesis can be falsified but cannot
be verified. However, we can provisionally trust our experience as the judge
trusts the eyewitness: we weight favorable and contrary empirical evidence and
come to a decision that is neither conventional nor arbitrary. We notice that
the problem of psychologism is particularly relevant to Popper for he conceives
objectivity as inter-subjectivity (i.e. an observation is objective if it can
be repeated by every subject); Jacques Monod defined objective an observation
that could be made by an instrument (i.e. minimizing the subject’s contribution)
and confined the problem of psychologism to a less relevant and more controlled
position. I may add that Popper in his analysis did not consider some well
established means of controlling psychologism, e.g. blind methods, as employed
in medical research. Akinci’s conclusion is that Popper’s conventionalism is
epistemological, i.e. conventions are made about the proper methods of
scientific investigation, not epistemical, i.e. related to the content of
scientific theories.

Epistemical conventionalism was formalized by the French epistemologist
and mathematician Henry Poincare’, but has a long standing tradition in
philosophy, even though I cannot believe that it has been espoused by many
scientists. I think that it was best explained by Andreas Osiander in his
preface to the Copernicus’ De Revolutionibus Orbium Coelestium:

Neque enim necesse est eas hypotheses esse veras,
imo ne verisimiles quidem, sed sufficit hoc unum, si calculum observationibus
congruentem exhibeant. (Indeed it is not necessary that these hypotheses are
true, nor verisimilar, but it is enough if the calculus we base on them is
congruent with the observation.)

Osiander was not a conventionalist: he skillfully constructed the
argument to protect his friend Copernicus from possible retaliations by the
Inquisition. Copernicus thought (as Osiander) that his hypothesis was true,
i.e. that it described the real relationships between the apparent movements of
the stars and the actual movements of the Earth; and Popper no doubt concurs with
this view and confines conventionalism to methodology.

Philip Catton in his essay criticizes Popper’s view of science as an
eminently theoretical enterprise (indeed Popper himself wrote in The Logic of
Scientific Discovery that sciences are systems of theories, thus leaving aside
descriptive sciences like anatomy or geography). Data from these sciences can
be used to build up theories, as Catton demonstrates, but their intrinsic
theoretical content is minimal: he points out that Newton used this method, that
he called deducing from experiments. Catton’s essay demonstrates that
scientists do not think and behave as Popper; however they neither behave in
ways Popper would forbid, and surely they would concur on Popper main point,
i.e. that their hypotheses should be falsifiable, and should be rejected or
modified if falsified. Catton’s point is that experiments have not only the
negative function of testing the hypothesis, they also positively suggest and
shape the hypothesis. Why was Popper so adamant in denying the positive role of
the experiment? Again, the reason is that admitting the experiment positive
role would grant some status to induction, Popper’s bete noire, as Alan
Musgrave thoroughly discusses in his contribution.

Essentially, Popper failed to recognize that Hume’s argument against
induction only works if we assign absolute rather than probabilistic validity
to induction. Popper dealt at length with probability in The Logic of
Scientific Discovery, and distinguished between two meanings of the term in
common usage: indeed, probable may be properly used to mean that an event has
measurable chances of happening, as when we say that it is likely that any day
of august is warmer than any day of october; or we may improperly use probable
to indicate that we believe that an assertion is true, but we want not to
commit ourselves too strongly, as when we say that it is likely that Copernicus
was right and Ptolemy was wrong. Popper strongly opposed the latter use of the
term, but not the former; however, he never explicitly admitted the obvious
consequence that induction may be reformulated probabilistically. This mode of
reasoning clearly shows up in the writings of the scientists quoted by Catton
who creatively formulated deterministic hypothesis compatible with inductively
inferred regularities and probabilities; all of them were perfectly aware that
the process is not infallible.

Later in his life, Popper turned his attention to
social sciences, and to what he called historicism, the idea that some deterministic
rationale exists in social events, scientifically investigable. In Conjectures
and Confutations Popper discusses the links between scientific and social
theories in reference to his acquaintance with the heretical psychoanalyst
Alfred Adler, and his dislike of Marxist philosophy is expressed in The Open
Society and its Enemies. Popper’s position with respect to social sciences is
somewhat different from the one he takes for natural sciences. Popper thought
that man is free to choose among socially acceptable alternatives and therefore
no specific prediction can be made on his behavior, even though general
regularities in social phenomena may be recognized. Thus, he fiercely opposed
two deterministic hypotheses, Marxism and psychoanalysis; time proved him
right. The essays by Gonzalez, Shearmur, List and Pettit, Macdonald, Ryan, O’
Hear and Waldron discuss Popper’s political philosophy and its position in the
philosophy of the twentieth century. Moreover, some of these essays critically
analyze the logical relationships between Popper’s philosophy of science and of
society. Here the questions are subtler than in the case of the philosophy of
science, and any summary is bound to be incomplete. A crucial difference
between philosophy of science and political philosophy is that the former
analyzes hypotheses, the latter opinions. Opinions can be based on logic, but
ultimately they do not compete with each other in the same way as hypotheses
do, and two contrasting opinions may be both (subjectively) true, whereas two
contrasting hypotheses cannot be both (objectively) true. Often we misrepresent
our subjective opinions as objective hypotheses in order to discredit the
opinions of our adversaries. Popper tried to fight these misrepresentations,
and probably went a bit too far: indeed he judged Marxism and psychoanalysis as
false scientific hypotheses rather than as plausible but subjective opinions
improperly presented as hypotheses.

In summary, Karl Popper: Critical Appraisals is an extremely good
book, destined to a public who has already read Popper critically and wants to
be guided through the most complex parts of his thought. You will find in this
book some answers that may surprise you, and will reach a deeper understanding
of one of the greatest philosophers of our time; but do not venture through it
if you did not have in mind some questions, otherwise it will be a boring and
pedantic effort.

©
2005 Andrea Bellelli

Andrea Bellelli
has an MD and a degree in psychology, and teaches biochemistry in theMedical School
of the University of Rome, Italy.

Categories: Philosophical