When
I speak of science I do not speak as a scientist or even
as a science
educator. I speak, rather, as a philosopher
with
a modest background, interest,
and teaching experience in the topics of the history and philosophy of science.
I also speak as a concerned citizen who has been active in promoting science
education. Finally, I speak as one living and teaching in the state of Kansas
where, as recently as 1999-2000, a cultural battle was waged, especially at
the level of the state board of education, over the very definition of science.
My judgments on the subject of science are tempered by a deep respect for
the arduous efforts and the achievements of scientists—there are scientists
and science educators in my immediate family. I am also mindful of (if unimpressed
by) what Steven Weinberg calls “the unreasonable ineffectiveness of
philosophy” (Weinberg: 169). In any event, my task here is not to defend
philosophy or to add to the body of knowledge regarded as scientific but
simply to reflect on the nature and practice of science.
Science
as Disciplined Objectivity
I shall define science as a special kind of disciplined objectivity, a phrase I borrow from my former professor, Don Crosby (Crosby: 48). At least part of what I am aiming at with this broad
definition can be expressed in
the form of a parable. Three farmers come upon a turtle stuck on a fence
post. The first said, “Well it sure didn’t get there by itself. I’ll
bet that ol’ prankster Jimmy Watkins put it up there.” The second
said, “Maybe that plank in the grass was leaning against the post. The
turtle crawled up, the plank fell down, and the critter got stuck.” The
third farmer said, “An invisible spirit must have put it up there.” The
other two looked at their friend in puzzlement. As if in explanation the third
said, “Well what are the odds that you’d find a turtle on a fence
post?”
The first two farmers offer hypotheses that are in principle testable against
experience. They could find Jimmy Watkins, for example, and ask about his
whereabouts. They could find out whether he has a predilection for putting
turtles on fence
posts as a practical joke. Or, they could test whether turtles are capable
of climbing up an inclined board and why this might be of interest to the
turtle. The third farmer, however, merely restates the problem. An invisible
spirit
may have put the turtle on the post, but the fact that it is unusual to find
a turtle on a fence post is no evidence that an invisible spirit put it there.
It is that very unlikelihood that poses the question, but does not answer
it.
In my definition of science I refer to objectivity. What I mean
by objectivity is that science aims at truth. Actually, I’m not entirely sure that “truth” is
the right word. One of my friends is a retired physics professor at the university
where I teach, Dr. Bruce Daniel. Once I asked him a question about the Big
Bang. I was having difficulty expressing my question and so I said, “Well,
Bruce, I’m just looking for the truth here.” His reply was memorable: “Truth
is a theological concept. I’m a physicist and we don’t care about
the truth, only what works.” This brought me up short. As a philosopher
I am aware of the lively controversies about the nature of truth (see, for
instance, Kirkham 1992). I had always assumed, however, that scientists wanted
to get a “true” picture of the world, whatever “true” might
mean. Until I can hone the point further, perhaps I should say that science
aims at something that is at least a close cousin to truth. After all, those
farmers want to know how that turtle got stuck on the fence post.
I qualified the word “objective” in my definition with the adjective “disciplined.” What
I mean by this is that science has its characteristic methods. I am of the
opinion that there is no such thing as the scientific method that
is or should be practiced by all scientists. There are statistical methods,
non-statistical
methods, hypothesis testing, data gathering, exploratory research, and even
anecdotal approaches in science. The hypotheses proposed by the first two
farmers demand very different methods. In one case, the farmers would need
to investigate
the whereabouts of the Watkins boy and his recent activities. In the other
case, they would need to investigate the behavior of turtles—or even
this particular turtle—and their (or its) abilities and motivations to
climb up inclined planes of the appropriate angle.
As Aristotle understood, one picks the method most appropriate to the subject
matter one is studying. To take less fanciful examples: figuring out extinction
of learned responses in rats in a maze requires the mazes, the rats, and
ways of measuring their responses. But figuring out whether judges and juries
tend
to accept or deny free will or something else in their approaches to the
law may require questionnaires. It would be unhelpful to run the judge and
juries
through mazes and one cannot give a questionnaire to a rat. Also of relevance
is that scientists aren’t always good at following what they understand
as proper method. An article by Stephen G. Brush was published in Science many
years ago with the curious title, “Should the History of Science be Rated
X?” (1974). Brush’s point was that much of what goes on in science
that leads to fruitful discoveries is not part of what most of us would recognize
as proper procedures in science. An excellent example is the feud between the
nineteenth century American paleontologists Edward Cope and O. C. Marsh. According
to Richard Milner, Cope and Marsh would “try most any underhanded trick
to beat the other to a new discovery” including dynamiting fossil beds!
Yet, their rivalry advanced knowledge in paleontology through the discovery
and naming of numerous new genera and species of fossil animals and by stocking
several major museums with dinosaur skeletons (Milner 1990, pp. 94-95).
What
Kind of Disciplined Objectivity?
So, science is a disciplined objectivity, but what kind?
Well, a “special
kind”; but what is “special” about this kind of disciplined
objectivity? What I have in mind is that science offers explanations and
these explanations should, ideally, have certain characteristics. I identify
six
of them:
1. Scientific explanations should aim at testability by empirical
observation. Scientific explanations should be verifiable (or
confirmable) by experience but this is not sufficient, for many hypotheses
are confirmed
by experience
without being true. Data are generally capable of multiple interpretations.
One might say that different curves can be drawn through the same set of
data points. Karl Popper famously added—anticipated by Blaise Pascal as early
as 1647 (Pascal 1989)—that falsifiability is the sine qua
non of scientific
explanations. That is to say, they must be vulnerable to contrary evidence.
I think that Popper is correct, but it is not necessary that a hypothesis
pass every test of falsifiability. One can always fudge the data or hand
off the
problem to one’s graduate students for future reference. For example,
one of the more compelling arguments for a stationary earth was the lack of
an observed parallax. Yet, scientists accepted the movement of the earth around
the sun fully two centuries before the parallax was measured in the 1830s.
Or again, Newton, could not explain why the night sky is dark. This was only
possible with the discovery of the red shift in the stars, whose importance
was recognized only in the twentieth century by Edwin Hubble. In both of these
cases there was sufficient evidence to lead reasonable people to set the problem
to one side until a solution could be found, but they provide real examples
that a single disconfirming instance, important though it may be, does not
necessarily spell the death of a theory.
If a hypothesis, to be considered scientific, must be falsifiable, it does
not follow that scientists are merely about the business of looking for
disconfirming instances of theories. This is not what scientists do, nor
is it the only
thing they should do (cf. Gardner). Albert Einstein, in his general
theory of
relativity, first proposed in 1915, predicted a greater bending of light
in gravitational fields than what Newton had predicted. In 1919, the two
theories
were put to the test when separate expeditions—one to Sobral in Brazil
and one to the Island of Principe off Africa’s west coast—measured
the bending of light from the Hyades star cluster around the sun during a solar
eclipse. The results confirmed Einstein’s theory and disconfirmed Newton’s.
Alfred North Whitehead was present when Frank Watson Dyson, the Astronomer
Royal, announced the results. According to Whitehead, “The whole atmosphere
of tense interest was exactly that of the Greek drama: we were the chorus commenting
on the decree of destiny as disclosed in the development of a supreme incident” (Whitehead
1967: 10). The tense interest was as much in knowing whether Newton’s
theory failed as it was in knowing whether Einstein’s theory succeeded.
In this case, confirmation was as important as disconfirmation.
2. Scientific explanations should be consistent with other things we know
about the world, including other scientific explanations. Ideally, there
should be
what William Whewell and more recently E. O. Wilson called consilience of
explanations, a coming together of mutually supportive conclusions. When
this consilience
is lacking, there are problems for science. In the nineteenth century Charles
Darwin’s evolutionary theory met with great success and the scientific
community was converted within a few decades. However, the great physicist,
Lord Kelvin, had provided an argument from the cooling of the earth that seemed
to demonstrate that there was not enough time for evolution to have occurred.
He correctly saw that evolution requires vast time scales, but he put the age
of the earth at between 20 and 100 million years old. Most evolutionists tried
to cram all of evolutionary history into that time scale, but it was an exercise
in making a procrustean bed for the data. However, by the end of the century,
Marie Curie was burning her fingers on radiation. With the discovery of radioactivity,
a new source of heat was found, making the cooling of the earth an unreliable
argument for a younger earth. Another example of consilience is the support
given to Darwinian evolution by Mendelian genetics. Darwin had only a very
unsatisfactory explanation of the inheritance of characteristics and the mechanism
of mutation. A better explanation was provided by genetic theory which was
originally developed by Gregor Mendel working on problems unrelated to the
immediate concerns of Darwinian evolution. These cases, in addition to being
examples of consilience, also illustrate the way that science can become a
cumulative enterprise, as ever more adequate explanations become available,
we begin to draw up a more satisfactory view of the world.
3. Scientific explanations should change our ignorance to knowledge. Scientific
explanations provide more than verbal solutions to problems. This is often
manifested in the technology that is associated with the theoretical side
of science. Benjamin Franklin not only gave us the beginnings of a theory
of electricity,
he gave us the lightning rod which does real work in controlling the powers
of nature. A more recent and even more dramatic example of technologies
tied to scientific advance is the harnessing of the power of the atom in
the form
of weapons of terrible destructive potential and in the form of nuclear
reactors to provide energy for daily life. A final example: research in
genetics holds
out the promise of diagnosis, treatment, and cure of a variety of diseases
and disorders. One could fill an encyclopedia with examples like these.
The prestige enjoyed by science is in large measure a function of its fruitfulness
in helping us to manage our lives and shape our environment.
4. Scientific explanations should lead to new and unexpected results. Science
has heuristic value. When perturbations were found in the orbit
of the seventh planet, astronomers were puzzled because this wiggling of
the planet didn’t
match predictions from Newtonian physics. Two astronomers independently came
up with the same hypothesis. Reasoning backwards, they argued that a large
planetary object outside of Uranus’s orbit might cause the anomaly. Sure
enough, when they pointed their telescopes to the sky, there was the planet
that we now call Neptune. In this case, the discovery was a result of using
the very physics that had been put in question. This story illustrates another
point about scientific discovery. Discovery requires recognition.
Neptune, as it turns out, was already on the star charts, but no one had
recognized
it as a planet. This is like me overhearing one of my supervisors saying
to another, “You’ll be lucky to get him to work for you.” Not
recognizing the irony in his voice I suppose him to be praising me. If I don’t
realize that he is actually criticizing me, should I be insulted? Arguably,
the answer is that I should not. The insult “works” only if I recognize
it as an insult.
5. Scientific explanations should be transparent to the scientific
community. In other words, one must always be willing to put one’s data forward
and invite others to replicate what one has found. The standard way in which
this is done is to vet one’s results through a peer-review process. In
the late 1980s, Martin Fleischmann and Stanley Pons announced in a news conference
that they believed they had achieved cold fusion. The announcement was met
with much excitement because of the promise of an inexpensive and abundant
supply of energy. However, subsequent research by other scientists around the
world failed to support the Fleischmann and Pons results. Some scientists went
so far as to accuse the two of shoddy research and wishful thinking. I have
no opinion on those accusations, but the fact is that the promise of cold fusion
never materialized. This example not only illustrates the importance of Popper’s
emphasis on falsifiability, it also demonstrates Popper’s point that
the objectivity of science is guaranteed most by the fact that science is a
communal activity. The lonely scientist working in his or her
lonely lab is at best only the beginning of science. That lonely scientist
must
go out of
the lab and defend his or her results in the face of what is found in other
scientific labs.
6. Ideally, scientific explanations are the best available on a given body
of evidence. Science is open ended without being relativistic or merely
subjective. The qualification “given the body of evidence” is essential. The
body of evidence is constantly changing as new facts emerge from the various
arenas of scientific investigation. For this reason, scientists must constantly
be prepared to revise their estimates of the worth of a given theory in light
of the newest evidence. Many years ago, I used to listen to Paul Harvey on
the radio while I drove to work. In those days, one of Harvey’s favorite
hobby-horses was the mercurial character of scientists, always changing their
minds about what is true. This displayed Harvey’s ignorance of the nature
of science. It is a virtue to change one’s mind when the evidence demands
it. It is a vice to stubbornly cling to one’s favored theory even when
the weight of evidence is against it. For this, and other reasons, I teach
my students that the expression “scientific proof” is something
of an oxymoron. The proof that there is no largest prime will
stand for all time. But scientific evidence is always just short of proof
in this
sense.
Is
Science Naturalistic?
The astute reader will have noticed that I have yet to
list as a characteristic of scientific explanations that
they be naturalistic. This is because I
don’t
know what to say about this. There is surely something right about this since
no current well-confirmed scientific theory steps outside the bounds of the
natural world to embrace supernatural causes. In the 2005 Dover trial, Judge John E. Jones ruled that the theory of Intelligent Design is not science---this, in part, because it "violates the centuries---old ground rules of science by invoking and permitting supernatural causation" (Jones: 64). On the other hand, Darwin is
one of my heroes and he treated the theory of special creation of his day as
a rival scientific hypothesis to his own theory of descent with modification.
Darwin speaks variously of "the theory of creation" (Darwin: 197,372,393,471,473,474,478), "the theory of independent creation" (Darwin: 355), and "the theory of independent acts of creation" (Darwin: 203, 478). He didn’t argue that special creation wasn’t scientific,
he seems to argue that it was an inferior scientific hypothesis, that phenomena inexplicable on the theory of creation is fully explainable by his own theory. Reasoning along similar
lines, Philip Kitcher speaks of intelligent design as “discarded,” “dead,” or “failed” science
(Kitcher 2007). (Intelligent design also suffers from a severe paucity of peer-review
studies.)
It is well-known in philosphical circles that the so-called demarcation problem has not been solved; that is to say, philosophers of science have not succeeded in drawing a sharp line between science and pseudo-science or even more generally between science and non-science. Crosby says, "there are no obvious, clearly discernible lines of demarcation between the natural sciences, on the one hand, and all other kinds of thought on the other hand" (Crosby: 48). In a simiar vein, Kitcher says, "We cannot seem to articulate that essential line of demarcation" (Kitcher: 11). Indeed, there may be no essential line but only a more or less fuzzy border between science and other forms of disciplined objectivity. Of course, even if the border is fuzzy, it does not follow that there is no genuine difference between science and non-science. No one supposes that bald men do not exist simply because there's no precise way to specify when a man is bald.
If, however, supernatural explanations have, in the past, been considered scientific, it is difficult to understand how in our day science could be anything but naturalistic if its explanations are, as I have argued, testable, mutually supportive, productive of real solutions, heuristic, transparent, and open-ended. The third farmer’s explanation
of the turtle on the post was that an invisible spirit put it there.
In what way might this be satisfactorily tested? Are there other lines
of evidence
pointing to invisible spirits with a propensity to place turtles on posts?
What would be the use of such a theory? Would it lead to the solution
of other problems or puzzles? Can one investigate such things in a way
that is open
to critique from one’s peers? Is it an open-ended hypothesis, revisable
in light of further evidence? My point is not that these questions must
necessarily be answered in the negative, but it is difficult to see how
they could be.
One thing, however, is clear. If the explanation of either of the other
farmers were verified, then the third farmer’s hypothesis would
become superfluous. Even if neither of the naturalistic hypotheses could
be verified, it is difficult
to imagine why anyone would give any credence to the hypothesis of the
invisible spirit.
Victor Stenger, an emeritus professor of physics and astronomy, argues
that allegedly supernatural phenomena can indeed be tested scientifically.
The
examples he cites, however, such as the efficacy of prayer, could better
be viewed as
paranormal phenomena. It is now acknowledged that extra-sensory perception,
psycho-kinesis, and clairvoyance are amenable to scientific investigation.
Stenger points out that, after more than a century of study, the results
are disappointing for paranormal enthusiasts (Stenger: 89-93). Had the
results been otherwise, would scientists claim to have discovered something
about
the supernatural or would they search for an enlarged theory
of the natural capacities
of human beings? We may safely assume that there is more to nature than
current science knows, and some of what we do not know might seem to
us to have a
supernatural quality. But why would it not be our concept of the natural
itself that would
need to be expanded? For example, David Ray Griffin, a philosopher, is
more sanguine than Stenger (or me) about the results of parapsychology.
But his
defense of the paranormal supports the point I am making: it is an enlarged
view of the natural, not an appeal to the supernatural, that would be
required to accommodate these phenomena (Griffin).
What does it mean to say that science is naturalistic or that it is constrained
by a methodological naturalism? At the very least, it means
that supernatural entities and powers cannot be part of its explanatory
apparatus.
It also means
that science assumes that events take place without violating laws of
nature. These senses of naturalism are shared by the discipline of history,
including
the history of science. If one assumes naturalism for the purposes
of doing science or history—or any activity, such as plumbing or highway construction—it
does not follow that there are no supernatural causes or no interruptions of
natural law. (Nor should one automatically assume that the activity of a supernatural
agent implies, even in the case of alleged miracles, interruptions of natural
laws; see Alston: 212). It should be obvious that a methodological
constraint does not permit one to draw substantive conclusions about what lies
outside of the constraint. If I am on a jury and the judge orders me to disregard
a piece of evidence, then I am methodologically constrained to ignore that
evidence. It does not follow that that evidence is unimportant or irrelevant.
It is possible that it could be the decisive reason for swaying my judgment
in another direction. Nevertheless, by the procedures of the law, I am obliged
to abide by the judge’s order.
Does
Science Have Limits?
Let me close with a few remarks on the limits of science.While some might consider it to be unscientific to raise the question of the limits of science, I believe I am in very good company. It is, after all, the central question that Peter Medawar, a Nobel laureate in Medicine, raised in his book, The Limits of Science (1984).
Are there things that science cannot explain? Perhaps,
but it is not an easy thing to
know what these are. At the very least we can say that there are two
extremes
to be avoided.
One extreme is to suppose that science tells us nothing about the world.
The other extreme is to suppose that we can know nothing about the
world apart
from science. The dramatic successes of science should be enough to
lay the first extreme to rest. These very successes,
however, tempt some
people to
adopt the second extreme. According to this view, science alone is
the arbiter of what is true and false. This is a philosophy
called scientism and
should be distinguished from the activity of science itself. Scientism
is a philosophical
thesis about the scope of science; it is not part of the body of knowledge
that we call scientific. (An example of one version of scientism is
the philosophical school of thought known as logical
positivism.) Of course,
scientism is itself
either true or false; but if scientism is not a deliverance of science,
then by its own standards, it cannot be known to be true. I am happy
with this
conclusion, but it must surely be unacceptable to anyone who wishes
to defend the thesis
of scientism. More’s the pity.
It should go without saying—though often it does not—that the acceptance
of science as important or even as our primary way of knowing the world does
not entail the acceptance of scientism. Likewise, the rejection of scientism
does not entail the rejection of science. Scientism reminds me of the story
of the boy who lost his toy in the back yard but was found by his mother looking
for it under the street lamp in the front yard. “Why are you looking
for your toy in the front yard if you lost it in the back yard,” inquired
the mother. “Because,” the boy replied, “the light is better
here.” The light is often better in science, but the world is bigger
than it may know. I think that history tells us truths about the world, about
the human past, but it is not clear to me that history is scientific. Unlike
physicists, for example, historians don’t seek to locate events in a
causal sequence governed by natural law. Nevertheless, history, like empirical
science, is a kind of disciplined objectivity; it is simply a different kind,
one that must take into account human motivations and purposes. I am also fond
of my own sub-discipline in philosophy called metaphysics. By the very meaning
of the word, “metaphysics,” or “beyond nature,” there
is the suggestion that there are questions too big for science to answer. Some
of these questions involve the very nature of science itself, the subject of
this brief paper. One leaps to the “meta-level” with such questions
and thereby one leaps outside the arena of science proper.
Generally speaking, it is unrealistic and unreasonable to require a
theory or a discipline to account for itself. As I have said, the philosophy
of science is not itself science. One need not be a philosopher of
science
to practice
good science. On the other hand, it is not unrealistic or unreasonable
to require that a theory not render impossible the possibility of its
own
discovery
or
the possibility of other ways of knowing. For example, Stenger, the
professor of physics mentioned above, claimed, without a hint of irony,
that empirical
evidence is more and more pointing to the inefficacy of minds. Science,
he claims, is showing that this form of epiphenomenalism is true (2007,
p. 84).
Perhaps Stenger only intends to draw attention to the lack of evidence
for an immaterial mind substance. But this is not precisely what he
says and
his proposal has the earmark of a scientist sawing off the limb on
which he sits
to do his science. Science itself is a purposive, goal-driven, intentionally
activated enterprise, a fact born out by even a passing familiarity
with the history of science. When carried to its logical conclusion,
epiphenomenalism
entails that the scientist’s own contributions to physics had nothing
to do with the mental processes that he had in formulating them, including
the arguments he gives on their behalf. As Whitehead said, “Scientists
animated by the purpose of showing that they are purposeless constitute an
interesting subject for study” (1971, p. 16). Interesting subject for
study indeed. Examples like this—there are many others—convince
me that the question of the scope and competencies of science should be of
continuing concern to the community of scientists (and to non-scientists),
especially as the deliverances of science more and more inform public policy
decisions at all levels of government.
Alston, William P. (1989). Divine Nature and Human Language: Essays in
Philosophical Theology. Ithaca, New York: Cornell University
Press.
Brush,
Stephen G. (1974). “Should the History of Science
Be Rated X?” Science, vol. 183: 1104-1172.
Crosby, Donald A. (2002). A Religion of Nature. Albany: State University of New York Press.
Gardner,
Martin (2001). “A Skeptical Look at Karl Popper,” Skeptical
Inquirer, v. 25, n. 4 (July August): 13-14, 72.
Griffin, David Ray (1997). Parapsychology,
Philosophy, and Spirituality: A Postmodern Exploration. Albany: State University
of New York Press.
Jones, John E. (2005). Tammy Kitzmiller, et al. v. Dover Area School District, et al., Memoradum Opinion, Case No. 04cv2688.
Kirkham, Richard L. (1992). Theories
of Truth: A Critical Introduction. Cambridge, Massachusetts: MIT Press.
Kitcher, Philip (2007). Living
with Darwin: Evolution, Design, and the Future of Faith. New York: Oxford University Press.
Medawar, Peter (1984). The Limits of Science. UK: Oxford University Press.
Milner,
Richard (1990). The Encyclopedia of Evolution: Humanity’s
Search for Its Origins. New York: Facts on File.
Pascal,
Blaise (1989). “Letter to Father Noel,” reprinted
in Pascal Selections, edited by Richard H. Popkin. New York:
Macmillan: 49-55.
Stenger, Victor J. (2007). God,
the Failed Hypothesis: How Science Shows that God Does
Not Exist. Amherst, New York: Prometheus
Books.
Weinberg,
Steven (1993). Dreams of a Final Theory: The Scientist’s
Search for the Ultimate Laws of Nature. New York: Vintage Books.
Whitehead, Alfred North (1967). Science
and the Modern World.
New York: Free Press. (Originally published 1925.)
Whitehead,
Alfred North
(1971). The
Function of Reason. Boston: Beacon
Press. (Originally published 1929.)
|