Download - Challenges to Science Philosophy and Theory
Page 1 of 15
Challenges to Science Philosophy and Theory
Russell C. Reinsch
Arizona State University
Science and Technology Studies
May 11, 2011
Page 2 of 15
Table of Contents
Section 1 – Introduction 3
Definition of terms 4
Background 5
Section 2 – Philosophical problems for science in the 20th century 7
Demarcation: the line between what is science and what is not. 7
Falsification and Induction 7
Section 3 - Theoretical problems for science in the 20th century 9
Constructivism 10
Section 4 – Solutions in Philosophy and Theory 12
Section 5 – Conclusion 14
Bibliography 15
Page 3 of 15
Section 1 – Introduction
Science and its methods suffered from a full spectrum of extremism in the 20th century.
Scientists in the 1900’s operated with an overly austere view of what defined their discipline.
The prevailing philosophy of the time, now regarded as the ‘empiricist’ philosophy, was
principally represented by a group called the Vienna Circle. In the decades following the turn of
the century science was forced to deal with attacks directed toward the scientific method and
doubts about justifications for theories, which presented challenges to both the philosophy of
science and the social interpretations of the discipline.
The rigid and restrictive grasp of the empiricists was gradually loosened by powerful
theories put forth by philosophers that challenged conventional thinking about science, namely
by the theories championed by Karl Popper, W. V. Quine, and Thomas Kuhn. As recognition of
the qualities in these theories gained adherents throughout the scientific fields, the pendulum of
sentiment swung away from the strict views held by the Vienna Circle, more to a moderate
position, and in some ways closer to the meta-physical principles of the older centuries, like
those from Francis Bacon and Rene Descartes. (Descartes felt that even if everyone were to
agree on something, like the Ptolemaic theory of the universe, it may still be a deception).
Eventually some of the looser practitioners focused so intently on the shortcomings of the
scientific method and whether we should believe science provides true accounts of our world,
they pushed the pendulum past the point of common sense, swinging beyond the center-point of
balance and over correcting into the other extreme, a range where relativism, realism, and
constructivism postulate much different assertions about science and theory.
The thesis of this essay maintains that humans can understand reality and conceive
whether theories are adequate, by using the best parts of science, which are sufficiently
evidentiary. It allows for the belief that science is and can be empirically successful without
automatically warranting the belief that truths of theories always have to be perfect.
Page 4 of 15
Definition of Terms
Ampliative rules: Likely able to go beyond the given information; providing justification for the
inferred conclusion.
Constructivism: The constructivist concept of rationality involves conscious analysis, and
deliberate design of models or rules. The models classify individual behaviors in order to explain
general behavior. It is neo-classical, but not inherently inconsistent or in opposition to Vernon
Smith’s other form of rationality, ‘ecological’. The two are different ways of understanding
behavior that work together.
Empiricism: A benchmark era for science, the years around 1900, when hypotheses would only
be accepted under austere circumstances, where the cold hard facts having been confirmed and
verified through deductive testing, were thought to be objective observation, and involving
universal laws of nature.
Falsification: Karl Popper suggested the demarcation line for science could be found through
falsifying theories instead of trying to verify them. So scientific theories needed to contain
something that you can actually dispute; the position “Cherry pie is good,” is not falsifiable.
Induction: Considered the biggest problem for finding scientific criterion for theory choice. The
problem of induction pre-dates the 1800’s; it is deeply philosophical and tricky to comprehend.
Technically, it is a cognitive process that includes statistical laws, or conditional probability. An
interesting place to start when setting out to understand induction is with ‘the Monty Hall’
problem, where pigeons learn from experience in laboratory tests to switch doors, but humans do
not.
Realism: an overly loose interpretation of intangible, unobservable things, to the extent that they
are considered objective items of evidence in every case. Even if they are independent of
accepted concepts, they still make for empirical theory, and belief in them is still required for
coherent science. In one version of realism, the success of science put forth as the proof of its
objectivity. Science has not historically been so successful however, in fact, it has been the
opposite.
Page 5 of 15
Underdetermination: the Duhem-Quine (D-Q) theorem: D-Q has two components:
1. That there are too many unknowns for evidence to be sufficient for us to identify
what belief we should hold or what decision we should make between rival choices; thus
all theories must remain unsupported.
2. A small theory can never be isolated and tested by itself; if a small theory appears
to fail a test, then the entire corporate body, or the test, or the scientist must be called into
question, but not the small theory.
Background of Philosophy
As described in the introduction, science held to an extremely narrow concept and rigid
interpretation of scientific procedure at the beginning of the 1900’s. The indisputability of facts
were paramount virtues of clear cut reasoning and exacting rationality. Only unmistakable
evidence could be used in investigations to discover rules and laws. Laws for prediction and truth
are what distinguished science and the activities of science were above this line of demarcation.
This overly strict philosophy hampered practitioner’s efforts to understand the world around
them. Skeptics and critics of empiricism claimed that the true nature of testing is limited, as
theories do not ever find perfect “truths;” and that empiricism failed to detect this very deviation
between itself and reality.
Background of Theory
After the Renaissance, human knowledge developed to the point where it established
itself as a full or authentic partner to reality. Humankind came to trust that any subject could be
credibly understood if the activities of science and technology followed systematic discovery of
evidence. Intellectual communities received increasing support, gradually replacing the old
world way of using the senses as inputs and then haphazardly constructing a belief from there. In
this way science and technology eventually became institutionalized in the twentieth century. At
the apex of scientific heyday, The Vienna Circle permitted only the narrowest of definitions of
Page 6 of 15
what constituted a valuable hypothesis. Scientists or the layperson could accept them or not,
there was no middle ground. Nor was there any need to postulate about hidden entities, the Circle
did not want the rules of the universe to have to continue into an infinite string of explanations.
Popper advocated an innovative way to identify the products of science, and argued that
scientific inferences do not use induction. His theory loosened up the structure of what
constituted the infamous demarcation point.
Kuhn wrote that everything is relative to the culture or time period in which the
circumstance exists, and that the one thing that we do know for sure is that science will be re-
written in the future. Kuhn proposed that the context of time breaks the line-of-decent model
from old science as the foundation for newer science; that two different periods of science are
not comparable, and he acknowledged the existence of subjective elements within science.
From there we viewed science’s dependency on theory: that science can never escape its
relationship with theory, because even the laws of science will change over time or at least be
conceived differently from one society or another. From this outlook, science is dependent on
theory as a set up or precursor for the scientific method. In light of this dependency, social
scientists highlighted various troublesome issues in scientific elements, such as conflicting
evidence, partial evidence, and weird evidence, and used these issues to critique the scientific
method.
Larry Laudan proposed splitting the action of problem solving from the concept of the
solution. In this perspective effective problem solving remains a rational activity, while what
counts as a solution is allowed to be relative, and in this way Laudan found an answer to a major
problem for determining acceptability of a theory.
Page 7 of 15
Section 2 – Philosophical Problems for Science
How to justify belief is the most important epistemic problem for scientific investigation
(it also happens to be equally problematic for induction). Science entered the 1900’s with a pre-
existing problem of induction, stuck like a thorn in its side that it carried around for hundreds of
years. David Hume explicated the problem in his mid-nineteenth century works, and it has been
seen as the major obstacle for science ever since.
A second serious challenge for science surfaced as more attention turned to the fact that
every theory or at least some parts of theories are eventually found to be inadequate or wrong.
In a third challenge, we came to face the fact that scientific methodology, like all contexts
that involve humans as the practitioners, is an activity that works in ways that we do not exactly
understand. Although the empiricists in the Vienna Circle attempted to deny it, science in
practice involves social aspects that are subjective, and a general method for obtaining ‘correct’
conclusions through objective investigation will not always follow some universal recipe for
getting to an explanation of the world. Every person has a unique set of principles, we can each
look at the same data and come to different conclusions, and science has proven to be unable to
escape this ‘problem.’
Demarcation: In order to establish a solid baseline for the reputation of scientific
methods, the demarcation line stood as the separation between science’s concrete evidence and
everything else below it for Rudolph Carnap, Carl G. Hempel, and the Vienna Circle. They were
very committed to observation and measurements that could be used to formulate laws with
predictive power, and it was these bullet-proof rules that were the backbone of their model of
science. Empiricists were especially enamored with the predictive power of a rule or law.
Falsification and Induction: Popper’s solution for demarcation suggested we not worry
about confirmation, and instead focus on falsifying a theory. Popper argued that since we are
limited by finite sets of observations, anything can technically be confirmed using induction,
though he did not feel induction was used in true scientific critique, only deduction.
Unfortunately, we cannot simply deny that we use induction. Wesley Salmon writes that with
Popper’s falsification, we would be stuck in a situation with infinite conjectures; and, according
Page 8 of 15
to Salmon, Popper’s ideas when closely examined contain circular runarounds. Summarized by
Scott Scheall at Arizona State University, “we cannot use a conjecture's degree of corroboration
as a measure of its reasonableness as a basis for prediction. To do so would be to let induction in
through the back door and we would again be saddled with the problem of induction. In other
words, a conjecture's degree of corroboration tells us how well it has performed with respect to
past predictive tests, but it tells us nothing (logically) about how it will perform in future tests.”
Thus Popper’s falsification and its contingent sub premises of conjecture and
collaboration, and demarcation, fail to detail a demarcation for science or formalize the scientific
method much if any better than past attempts. Laudan brings final clarification to the discussion
however, noting that we never have sufficient justification to need an assertion to be true in a
perfect sense in order to accept it; justification for induction is simply not required.
Page 9 of 15
Section 3 - Theoretical Problems for Science
As described above, Popper proposed falsification as the solution to the problem of
induction and multiple philosophers clarified that falsification is not the solution. The D-Q
theory of underdetermination also shows falsification is not a work around for the problem of
induction. D-Q declares the procedures one would use to falsify theories are ambiguous, and
second, that we can only falsify an entire corporate body, not a single small theory in isolation.
Later theorists then expanded on weaknesses identified by D-Q, interpreting D-Q as showing that
rules or “as-if” rationalities are impossible.
In his attempt to loosen the overly strict grip of empiricist philosophy on science and
provide guidance when deciding on what theory to follow, Kuhn championed the idea that
demarcation is only relevant within normal science, and what makes a theory scientific is the
absence of debate over theories; whenever critiques are silent, is when we are experiencing
science. Kuhn saw two distinct periods of scientific activity, with the period of what he referred
to as normal science making up the supermajority of time, and only during the very rare
revolutionary periods would Popper’s falsification be useful for demarcation. He also saw any
challenge to a theory as necessarily directed at the scientist, not at a paradigm itself. Kuhn agreed
with D-Q in this respect, but whereas D-Q underdetermination considers paradigms as more or
less static and permanent, for Kuhn, neither the standards of evaluation or conditions in the field
are permanent, they are always changing.
Changing scientific evidence causes problems for anyone who wants to adhere to a
particular theory. Imagine a person makes a decision to eat fish for the omega 3 acids that are
good for the heart, or decides to exclude fish from his diet because of the mercury content, based
on the existing knowledge and theories on food science. To then hear of a new study that has
determined that those same omega 3 acids are now apparently bad for the prostate, and trans fats
that were thought to be bad for the heart, are what is good for your prostate, calls the whole
paradigm of food science into question.
People operate with some kind of personal philosophy either to believe in no theory at all,
or some theory in particular, and might at this point find themselves with a freezer full of fish
Page 10 of 15
that they no longer wish to eat, because science has decided “healthy eating may be a much more
complicated matter than nutritionists previously realized.” (The Week, 2011).
The D-Q principles (that advise theories must remain unsupported), and Bacon’s analysis
that almost nothing is a full treatment of a subject for everyone (and that there is no single
question on which all people can agree on the answer), and various misinterpretations of the
critiques of empiricism & Popper, combined and led to unlicensed promotions of constructivism,
realism, or relativism by Bruno Latour, Paul Feyerbend, and several others.
Laudan corrects the D-Q/Kuhn inseparability of paradigm pyramid structure by replacing
it with a web structure, and weakens DQ to the simulacrum of rendering it moot. Laudan also
liberalized the standard view of paradigms as static systems; he explained that they are always
comparative, subject to change, and dependent on circumstances of context. Determining
whether certain criteria are more important than others is not a straightforward process, but we
have no reason to consider unbalanced concepts like relativism while we still have common
sense at our disposal. Laudan also clarifies how induction is really not such a big problem when
ampliative rules of evidence can be incorporated.
Constructivism
Constructivism runs into problems in social studies because social theories are
composites; they put construct parts into wholes, and schemes of relationships that are
interpretations, but they are not able to do more than that. The constructed models leave out
some of the parts. They are schemes that connect distinct, single things by using relationships
that we understand; they create wholes, but this does not make them factual. We report on them
using terms like ‘New York City’, that do not have sharp, precise definitions, because they may
have a variety of properties. A problem for the prominent social science of economics is that it
cannot distinguish how people go from a starting point and through practice in self regulated
systems, to finding equilibrium points in personal exchanges, without the use of consciously
constructed models. The structured model does not predict the higher level of cooperation or
reciprocity that takes place in the market. Studying behavior, we see people use their
Page 11 of 15
unconsciously learned experience when they need to make spontaneous moves; they dynamically
figure out what car insurance to buy or how to evaluate university ranking matrixes, either
without, or together with the existing instructions in the constructed schemes; so the schemes
often have little legitimate purpose or are redundant.
Page 12 of 15
Section 4 – Solutions in Philosophy and Theory
Laudan fixed the problems introduced by loose interpretations of D-Q by clarifying that
science is neither so static nor inseparable as D-Q posits, and he split the over used concept of
“theory” into big and little theories; where the big ones function as tools, and the little
constituents do the solving of problems. Thanks to Laudan’s perspective we have a quality
picture of the formality of the scientific method and clarity on how we can choose between
theories.
People need to be able to understand reality, and conceive whether theories are true and
whether evidence is real. This can be more difficult if the particular subject of discussion or
observation involves something as invisible the chains of bondage in the Stockholm Syndrome.
At the opposing poles of an ongoing argument over whether to believe in invisible entities before
they have been technically verified, realists and empiricists hold firm beliefs on when an
“unobservable” can be considered real. Bas Van Fraassen gives agnostic discourse on particles
that are too small to see and he notes that the best available explanation is often good enough as
a representation of the truth; but most importantly, he recommends an approach of taking
unobservables on case by case bases. Decisions on invisible particles and unobservables are
important, when we consider situations involving forensic science testimony has the power to put
people in prison based on DNA and other evidence that jurors may not fully appreciate. Jurors
are often expecting science to be responsible for solving the case, when in fact forensic evidence
is occasionally found to be invalid (Begley, 2010).
In a 1998 scientific paper published by the esteemed medical journal The Lancet, author
Andrew Wakefield linked the childhood vaccine MMR to an increased risk of autism in children.
Thirteen years later, after much debate, scientific exploration and reexamination, and a plethora
of class action lawsuits, the link has been discredited and the author vilified for both “bad
science” and for perpetrating a fraud. But the damage caused by the claim is hard to undo.
Despite scientific evidence to the contrary, many people still believe that childhood vaccination
is a confirmed major cause of autism. While it is acknowledged that vaccines can, on rare
occasions, cause severe side effects, the U.S. Institute of Medicine rejects the link between
vaccination and autism.
Page 13 of 15
Common sense dictates we not get hung up on distinction between truth and what is
useful; we can commit to a level just short of literal truth and accept the concept of
approximation as weak, but a necessary value for scientific claims. The position for science to
move forward is to just be the best at solving problems. Adequacy is fine for this; it is reliable
and economic, like the neighborhood play at second base in baseball. Scientists can referee
cognitive practices from this position and judge questions of when invisible entities are ok,
because they can observe when entities are used in, or for good theories.
Page 14 of 15
Section 5 - Conclusion
Science is simply a belief, like religion. No one size fits all regulations or broad views
work for the man on the street; life is not a carrot or stick situation. Science remains the best
alternative we have for knowledge and description of the world, and the social aspects of
scientific practice and concrete evidence are both factors for determining preferences. It we do
not try to take either one too far, technology will continue to pull science into balance, and we
might find we have both the carrot and the stick.
Tension remains between followers of the Darwinian doctrine and followers of religious
doctrines because of differences on conceptual grounds. A young person may try to decide
between Darwin and St. Peter; or between industrial progress and environmental protection. Are
they to throw their hands up? No; they can understand reality and conceive whether theories are
true and whether evidence is real, with help from empirically successful science and technology.
Page 15 of 15
Bibliography
Begley, S. 2010. But It Works on TV! Forensic ‘science’ often isn’t. Newsweek: Science. Pg. 26.
Curd, M. & Cover, J. A. 1998. Philosophy of Science: The Central Issues. New York, Norton & Company.
The Week. (2011) Health scare of the week. News: Health & Science. The Week: The Best of the U.S. and International Media, pg 21.