Guest post: Philosophy of Science.

Being systematic, rigorous and quantitative, science is a good way to pursue knowledge. However, it is limited to a specific philosophical framework. The philosophy of science defines its limits, methods and presuppositions. Being the foundation of critical thinking, it’s a good way of investigating and testing our beliefs about the natural word, and more broadly what humans can know. Because humans are plagued with flaws in cognition, perception and memory, a system of methods was developed to compensate for those failings.


Epistemology

We turn now to Epistemology, the philosophy that deals with the nature of human knowledge, but also sets the limits of what’s knowable. There are many questions science cannot answer because they are scientifically unanswerable, as opposed to factual questions. Although many might not think in these terms, it’s hard to avoid doing philosophy in science.

Empiricism states that knowledge comes from sensory experience. John Locke is considered the first empiricist. He was concerned about the boundaries of human knowledge, and claimed that thought comes from experience, that experience is the essence of thought and stated:

“Nothing is in the mind that was not first in the senses.”

Locke recognized the limitations of sensory experience, and therefore the limitations of human knowledge.


The Epistemological Limits of Science

Science only deals with scientific claims, any claim can be scientific if it is practically testable and falsifiable. If the notion is testable it involves science. Science works within the philosophical framework of science. If all of the assumptions of science are correct, they should produce some objective, positive results overtime.


Pragmatism

William James and John Dewey developed a comprehensive philosophy based on a “scientific” response to experience. For them, universal knowledge of what lies beyond experience is impossible. Dewey understood Pragmatism as prospective empiricism, and that science should be a pragmatic belief system. Pragmatism was a philosophy built on an experience-based conception of scientific knowledge.


Logical positivism

Later came the logical positivists. For them, science should be adequately reconstructed in empiricist terms, as a linguistic kind of thing with a set of sentences with certain properties. This is the best way to relate evidence and meaning, they thought, and what makes science special. This logical positivism approach to science became known as rational reconstruction.

An empirical interpretation of such language was need such as “There is an object X, and object has property P.” However science isn’t just about observations and properties, but also explanations. Observations and properties can be listed but we also need to elaborate a theory and make predictions, and this goes beyond experience. If science must limit itself to experience, how can it also go beyond experience?

How can we say object X is fragile, without violating empiricism and enter the realm of metaphysics? Fragile is not an observation term, but a disposition term. To solve this problem we have to place object X under certain test conditions. “X” must be defined as “fragile” in the condition that if we were to strike it, “X” would break”. This is a counter-factual conditional and cannot be defined in terms of the logical or observational vocabulary. Such conditionals depend on how the world would behave if it were different that it presently is, that is, under experimental conditions.

We need to move away from the observational level in order to explain phenomena and generate predictions about the world and increasingly complex phenomena. Thus, vocabulary must keep expanding, interpreting new terms and creating statements linking those terms.

The structure of a scientific theory is like a mathematical theory, and general laws serve as axioms, such as Newton’s laws. New statements provide empirical meaning that directly connect to observations. Sentences or statements that does not hook up to other statements are called isolated sentences, they don’t generate new predictions, and don’t add anything to the theory. Another way logical positivists avoid metaphysics is by refusing theories or statements about unobservable reality. A statement is only meaningful if it can be verified observationally.

For the logical positivists, the connections between theoretical terms are crucial for deriving observations, not for describing reality. Many statements in a scientific theory are inferences, saying that it is all right to infer this from that, that’s the instrumental conception of scientific theories. Science doesn’t make statements that go beyond observation, but make statements about patterns in experience.

A scientific theory is an axiomatic system, all relevant observation statements are deduced from a set of logically consistent, true theory statements, and upon rules for correlating theory and observation.

Logical Positivism persisted from the 1930s into the 1960s, offering hope that a classically objective theory of scientific knowledge was possible, but in the end, Logical Positivism failed. Carl Hempel criticized the separability of theory and observation statements, while W. V. O. Quine dismissed the distinctions between analytic and synthetic sentences, and between theory and observation statements. The verifiability criterion didn’t resist because some non-verifiable terms were inevitable in meaningful theories.


Methodological naturalism and assumptions

The methods of science are called methodological naturalism, in simple terms it underlines that material effects must have material causes. The material includes all macroscopic matter, all forms of energy, dark matter and dark energy recently discovered. This philosophy follows natural laws, therefore it cannot invoke miracles or supernatural phenomena as an explanation for any experiment or observation outcome. Science requires that we follow the methods that assume there is nothing beyond the natural world, and does not rely upon any supernaturalism.

For example, one scientific assumption is that of an objective reality. Without that assumption it’s impossible to investigate how the universe works. Another assumption is that the world is predictable and, therefore, knowable.


Falsifiability

Non-material causes cannot be falsified; therefore, they fail to meet a necessary criterion for science - they are not constrained in any way, they have no limits, they do not follow the laws of nature of material cause and effect. In short, constraint is necessary for falsifiability.

Bertrand Russell explored the relationship between testable and untestable claims and created an analogy known as Russell's teapot: a hypothetical teapot proposed that is orbiting the Sun between Earth and Mars to make the point that not all claims that cannot be proven false should be accepted as true.

Therefore, one cannot prove the teapot doesn’t exist, but is it reasonable to conclude that it does exist?

This example is used to make the point that the burden of proof for any scientific claim lies with those making the claim. The inability to prove something false is not sufficient justification for the claim.

This also raises the issue of proving a negative in science. While it’s not possible to absolutely prove the non-existence of something, given what we know about the laws of science and nature we might be able to demonstrate that something is impossible.


Pathways and methods of science

An observation leads to a hypothesis, and the hypothesis is then tested by an experiment. After the results, the hypothesis is refined. Then, the experimental phase is repeated.

Observation → hypothesis → experiment → refine → repeat

However, science can follow many pathways, for example, a hypothesis may come before a single observation.

Scientific methods develop a model of how the world works. Hypotheses and theories must explain nature and generate predictions about observations. Scientific theories are explanations for a pattern of observations, when there is more than one explanation that can account for data we already have, an experiment can be a way to separate them. A theory is only useful if it makes predictions that are different than other existing theories.

The highest form of criticism is: “not even wrong”. This is a common phrase used in scientific criticism, meaning that the theory does not make specific predictions and is therefore useless, an idea that can’t be tested or refuted isn’t even wrong.

All conclusions in science are provisional and subject to further evidence and theories. There’s no metaphysical certitude, scientists and critical thinkers need to be comfortable with uncertainty.


Occam’s razor

Another way to separate two competing theories that account for the same data, if there is no empirical method, is a philosophical rule of thumb called Occam’s razor, or the rule of parsimony. Occam’s razor states that the theory with the fewest new (ad hoc) assumptions is preferred to make the theory work, in principle that theory at least should be preferred until eliminated. Otherwise, we could endlessly generate ad-hoc theories to explain any given phenomenon.


Paradigm shifts

Science is a process of refinement. You can have paradigm shifts, resembling Kuhn’s paradigm shift model, but as science progresses, the process of refinement becomes more apparent.

For example, Einstein’s relativity did not prove Newton’s classical ideas wrong, it was a refinement to Newton and not a total replacement.

In many sciences, while refinements are possible, fundamental knowledge will not be overturned, because they are established to a high degree and have solid scientific knowledge.

For example, electrons do not spontaneously decide to change their fundamental properties.

Hard science should be considered so by its rigorous methods, objective outcomes and systematic observations, and not by the subject matter.


Empiricism

Berkeley, George. Three Dialogues between Hylas and Philonous. New York: Oxford University Press, 1998

Hume, David. A Treatise of Human Nature. New York: Oxford University Press, 2000.


Pragmatism

Dewey, John. Experience and Nature. New York: Dover Publications, 1958.


Logical positivism

Ayer, Alfred Jules. Language, Truth and Logic, 2^nded. New York: Dover Publications, 1952.

Balashov, Yuri, and Alex Rosenberg, eds. Philosophy of Science: Contemporary Readings. pp. 132–140. New York: Routledge, 2002.

Boyd, Richard, Philip Gasper, and J. D. Trout, eds. The Philosophy of Science. Cambridge, pp. 71–84. pp. 299−315. MA: MIT Press, 1991 Science.


Science

Asimov, Isaac. Asimov’s New Guide to Science. London: Penguin Books, 1993. Asimov explains the process of science using classic historical examples, offering the human dimension of how science works.

Klemke, E. D., Robert Hollinger, and David Wyss Rudge ,eds. Introductory Readings in the Philosophy of Science. Amherst: Prometheus, 1988. The standard introductory text to the philosophy of science; better suited for a classroom than casual reading.

Hines, Terence. Pseudoscience and the Paranormal. Amherst: Prometheus Books, 2002. Hines takes a no-nonsense scientific look at paranormal belief in popular culture.


Blogs

Novella, Steven. “New Scientist on Miracles.”NeuroLogica Blog. http://theness.com/neurologicablog/index.php/new-scientist-on-miracles

Novella, Steven. “Science and Faith.” NeuroLogica Blog. http://theness.com/neurologicablog/index.php/science-and-faith

Novella, Steven. “The Context of Anecdotes and Anomalies.” NeuroLogica Blog. http://theness.com/neurologicablog/index.php/the-context-of-anecdotes-and-anomalies


For more information on Sérgio Fontinhas, see Big Fitness Project.