Home | | Science | | Share This Page |
The proper role of science.
Copyright © 2010, Paul Lutus — Message Page
(double-click any word to see its definition)
During my multi-year campaign1 to try to get psychologists to see the benefit of scientific standards, I've had many interesting conversations about the nature of science itself. During these exchanges it has slowly come to me that most people don't understand science, including many who assert an attachment to science for personal or professional reasons.
Science confusion isn't limited to psychologists. In a recent online conversation about science, a correspondent said, "Michio Kaku is a wildly popular and incredibly intelligent theoretical physicist. He believes in multiverses. I'm going to share his beliefs." When I read this, I was forced to realize there are people who turn to science, not because they understand it, but because they see it as a belief system much like religion, but more respectable and up-to-date.
There are others who want science's prestige without the perspiration, who think science means lab coats, clipboards and slickly printed journals. Then there are companies and individuals who encourage public ignorance of science, because this can be turned into a marketing strategy (by promoting the idea that science is truth). Ironically, the least effective science boosters are scientists, who for reasons of personal integrity and professional discipline don't normally get involved in science advocacy, because that stand might appear inconsistent with scientific principles.
Without oversimplifying a complex relationship, let's portray science as an algebraic equation:
The Science EquationEarnest seekers
+ the insatiably curious
+ entertainment-seekers
+ religion's walking wounded
+ philosophers
+ people who expect life to have
a reasonable explanation
≈ Scientists
+ pseudoscientists
+ politicians seeking science's "authority"
+ salesmen
+ hucksters
+ science gurus
+ unreformed priests
+ (actual science)
On the left-hand side of the science equation, along with the earnest seekers we have people who want to believe in something, who feel betrayed by religion and other fixed belief systems and who hope science will offer a certainty that religion cannot. We also have people who just want to find out things, who want to be entertained, or who seek validation that life makes sense. On the right-hand side of the equation are those who would like to exploit public ignorance of science to sell you something, or who want to entertain you by pretending that science is something it's not. Yet another term in the equation are those who try to shape public policy in irrational ways based on a grave misunderstanding of science itself.
Included in our equation, almost as an afterthought, is a small term that represents science as it really is — a doubt factory. It remains to be seen whether this term can influence the solution of the equation of which it is a part.
In this article I'll describe some common science myths, then I'll try to explain why and how science achieves what it does.
As science becomes more important in our lives, as it acquires and deserves more attention, the number of science myths increases in step. Here are some of them:
The Myth The RealityScience is a search for truth. Science's purpose is not to discover truth, but to establish which theories cannot be falsified in spite of the most sincere efforts. Science's purpose is to manufacture doubt. The goal of scientific research is to prove that a theory is true. Scientific research cannot prove a theory true, it is only able to prove a theory false. Philosopher David Hume summarized this by saying, "No amount of observations of white swans can allow the inference that all swans are white, but the observation of a single black swan is sufficient to refute that conclusion." Scientific theories must remain perpetually open to falsification by new evidence. Proven scientific theories become scientific laws. In spite of how often one hears the term "scientific law," there is no such thing. In short, there are no scientific laws, only falsifiable theories.
- Because scientific theories cannot be proven true (see above), they cannot ever become laws.
- The commonly heard expression "scientific law" is an informal way to add emphasis to an idea, but it is technically incorrect.
- A "law" is by definition something permanent and immutable, but because scientific theories can always be disproven by new evidence, the idea of a "scientific law" has no basis in reality.
- As just one example, Newton's "Law of Gravity" has been replaced by Einstein's "Law of Gravity" and, because of some theoretical problems, Einstein's "Law of Gravity" will eventually be replaced by a new "Law of Gravity" that is unknown at present.
A scientific theory is what a scientist tries to prove. Scientific theories can never be proven. A scientific theory is an idea that:
- Is consistent with existing theories or can replace them,
- Has some supporting evidence, and
- Can be falsified in practical tests.
In the context of science, here are working definitions for the terms speculation, hypothesis and theory:
This is not meant to suggest there is something wrong with speculations and hypotheses. Most scientific theories begin as speculations, and free thought is important to science. But it's important to know the difference between speculation, hypothesis and theory, and all scientific ideas must eventually be supported by evidence.
Consistent with existing theories or able to replace them? Supported by evidence and falsifiable with new evidence? Speculation • • Hypothesis • • Theory • • Science is lifesaving vaccines, spacecraft and computers — science is science's results. No. Science is not a product, but a process. A vaccine is a scientific result, but it is not how science is defined. Science is not a vaccine, it is the process that led to the vaccine. An experienced, well-respected scientist is more likely to be right and should be trusted. No. Science only respects evidence, and rejects authority and expertise. On this topic Richard Feynman said, "Science is the organized skepticism in the reliability of expert opinion". This is a particularly controversial point for those not familiar with science, because there are many cases where someone tries to exercise nonexistent scientific authority. The idea of "scientific authority" contradicts the most basic principles and spirit of science.
- When Nobel Prizewinner Linus Pauling proposed that vitamin C could cure the common cold, scientists asked, "Where's the evidence?", but there was none.
- When Swiss patent clerk Albert Einstein proposed that his new relativity theory might overthrow much of physics, scientists asked, "Where's the evidence?" and it was forthcoming.
It seems the greatest amount of scientific eminence is trumped by the smallest amount of scientific evidence.
- Nobel Prizewinner Pauling's high standing in the world of science made no difference, only evidence mattered.
- Patent clerk Einstein's low standing in the world of science made no difference, only evidence mattered.
No individual can possibly test every scientific idea personally, so we are forced to trust scientists. A qualified no:
- One should never accept ideas uncritically or on faith.
- Scientific theories are expected to survive retesting by different workers in different laboratories (a process known as replication).
- Because of the replication requirement, it's reasonable to wait for, and give more credence to, a theory that has been successfully replicated.
- Many scientific theories are never replicated — such theories are inherently untrustworthy, and constitute a standing joke among scientists (see The Journal of Irreproducible Results).
- One scientist testing a hundred ideas cannot compare to a hundred scientists testing one idea.
- In science it is not a question of accepting ideas on faith, and it should never come down to trusting scientists, but of evaluating ideas based on the quality of their evidence.
Scientific theories are assumed to be true until proven false. No. This is a surprisingly common, mistaken belief about science. Scientists take the opposite position — that an idea has no standing until evidence supports it. This may seem overly skeptical until one sees the reasoning: As a result, scientists see ideas as having no merit until evidence appears. Ideas are not assumed to be false, they are simply not taken seriously until there is evidence.
- Let's say I believe that Bigfoot exists, and I will continue to believe until Bigfoot has been proven to be false.
- Critics of this idea must therefore prove a negative — they must prove that Bigfoot cannot exist. Simply saying there is no positive evidence is not good enough.
- But proving a negative is not possible in the general case, indeed this is a logical fallacy called the "argument from ignorance".
- By insisting on a negative proof, I have isolated myself from any challenge to what may be a false belief.
- In this way a requirement for negative proof erases the distinction between science and religion.
Fields are scientific because of the presence of scientists and research programs. No. Scientific fields are defined by one or more testable, falsifiable scientific theories and because research in the field addresses and potentially falsifies those theories. Again, to validate a field's scientific standing, research must address a field's defining theories. If I conduct a scientific study of Astrology and produce a scientific finding about it, for example that there are more "Leos" than "Capricorns," does this legitimate scientific result make Astrology scientific? No, because the study doesn't address Astrology's theories — Astrology's claims have not been tested.
- Phrenology is the theory that personality traits can be determined by the shape of a person's skull. There was once a field of phrenology, but the theory that defines the field is now known to be false, as a result of which there is no scientific field called phrenology (no theory supported by evidence, therefore no field).
- Astrology is loosely defined as the theory that the position of the stars and planets at the time of our birth, and on a given day, determine the course of our lives. But astrologers differ about what their field means, and all efforts to locate evidence for the theory have failed. As a result, astrology is not a scientific field.
- Physics is a field defined by its theories about physical reality. Physical theories are testable, falsifiable, and well-supported by scientific evidence, as a result of which physics is accepted as a scientific field.
Science is a suitable replacement for religion. No. Science's purpose and outlook are completely at odds with religion's purpose and outlook. Religion's purpose is to make its followers comfortable and self-assured in their innate superiority, but science (when properly understood) can only make one uncomfortable and doubtful about knowing anything for certain. Science can only accomplish what it does by encouraging doubt and skepticism, by challenging received wisdom, by being irreverent and subversive. All these traits, essential to science's effectiveness, place it at odds with the religious outlook.
This is not to suggest that science and religion cannot coexist. They coexist because (and while) they address different, nonintersecting, domains. Certain social problems arise when religious people try to challenge scientific methods and results, but this doesn't contradict the idea that religion and science can coexist, it only means religion must respect science's domain, just as science respects religion's domain.If a medicine is described as "laboratory tested," this means it's more likely to be effective. Not necessarily — drug advertisers often makes claims like this, but without bothering to say what, if anything, the laboratory tests discovered. Nothing is more common than to see an advertisement that says, "Scientifically tested!" but without adding, "The tests failed!"
To restate a point made above, a scientific theory is an idea that is supported by evidence and is falsifiable in practical tests. If an idea has no evidence but is consistent with existing theories, it is a hypothesis. If an idea is not consistent with existing theories and has no evidence, it is speculation. All these categories have appropriate roles in science, but all scientific ideas must eventually be supported by evidence.
in a literal sense or manner : actually <took the remark literally> <was literally insane>
Because words don't have fixed meanings (see "Word Definitions" box on this page), my definitions for theory, hypothesis and speculation are obviously open to debate, but I think these definitions are consistent with common usage, and provisional agreement on these terms is important to the topic. However, because the meaning of scientific theory is critical to science, its definition is less open to debate — a theory must be supported by evidence and must be falsifiable.
Some fields are accepted as scientific, some are not. The distinction between the two depends on the presence or absence of testable, falsifiable theories and a few other things:
The implications of item (5) above are far-reaching and suggest that many fields commonly thought to be scientific, fields with scientists, results and scholarly journals, are scientific in name only. For example, much of the scientific work in psychology cannot confer scientific status to psychology itself on the ground that it doesn't address psychology's central defining theories. The reason for that, in turn, is because psychology doesn't have a central, clearly defined theoretical structure open to test and falsification. This defect is shared by many of the "social sciences," fields described as sciences only to confer an unearned status.
To make this point, let's compare psychology with physics.Physics Example
The Global Positioning System (GPS) is a reliable way to establish one's position on Earth's surface. The GPS system relies on satellites carrying very accurate atomic clocks. The satellites send radio signals to a user's GPS receiver, which calculates a position based on the arrival times of the satellite signals.
To launch a GPS satellite, one must have a deep understanding of physical theory:This example shows the importance of theory to a practical result, the central role played by theory in physics, and the testability and falsifiability of the theories. Specifically, the designers of the GPS system knew in advance that the atomic clocks carried by the GPS satellites needed to be adjusted to accommodate the effects of both Special and General Relativity. Upon launching the satellites, the prediction was confirmed and the system met or exceeded its design goals — and coincidentally provided another experimental confirmation of physical theory.
Because physics is a science, because it is based on tested theories, physicists can reliably say what will happen in a given situation, and more important, they can say why — in other words, they can move beyond description to explanation.
Some have argued that describing is enough to create science, and there are fields entirely based on description, but to shape a theory one needs an explanation:
This is why description alone cannot lead to science — without an explanation, there is no basis for generalization, prediction, testing, and falsification.
Psychology Examples
My use of psychology as a counterpoint isn't to argue that it's a particularly bad example of sloppy science and pseudoscience, only that I've studied psychology extensively over the past five years as part of an effort to persuade psychologists that a more scientific approach might be in their best interests.
Psychology is not defined by central, testable theories, and it is almost entirely reliant on description, not explanation. When a psychologist makes a claim — for example, "Cognitive-Behavioral Therapy (CBT) is more effective than its alternatives" — it is a description without any effort to explain. If someone were to try to explain why CBT is effective, that explanation might lead to a theory that could be tested in different, complementary circumstances, and the possibility of a practical test would create the falsifiability criterion on which science depends (see the "Falsifiability" box on this page).
As a result of psychology's theory vacuum, if someone contradicts the claim that CBT is more effective than alternatives (e.g. more effective than speaking to a sympathetic aunt), this counterevidence would make no difference to psychology or the practice of CBT. I know this to be true, because that refutation has been made repeatedly and supported by research2,3,4 but these results have had no effect, on the ground that an experimental result cannot refute a nonexistent theoretical claim.
Other serious obstacles to psychology's scientific standing are poor experimental controls and nearly nonexistent replication rates. Here is an example of poor experimental controls drawn from psychology's professional literature:"A meta-analytic review of interventions based on MI found effect sizes across studies in the small to moderate range for alcohol and the moderate range for drug use when compared to a placebo or no-treatment control group ..."
Excuse me? What is a "no-treatment control"? Well, it's a group of people who are not given any treatment. Their outcome is being compared to a group of people who are given treatment. This experimental design is typical of modern psychological research — the "control group" are people who are simply told to go home, while the experimental group receive one-on-one sessions with a mental health professional. The outcome is expected to have scientific validity.
Because of the Placebo Effect (see the "Placebo Effect" box on this page), the described study is less than pointless. It is pointless because the control group's treatment is so distinct that they cannot realistically be thought of as a control, and it is less than pointless because the result has been published as though it has scientific meaning. This study joins a huge corpus of research of similar quality that represents the present state of psychological research.
For those unfamiliar with human studies, here is a list of experimental designs ranked in descending order by their probability of producing useful science:
Prospective studies, studies in which groups are randomly selected from a representative population and experimented on:
Retrospective studies, studies in which the experimental groups are selected from within the population based on their past histories:
The problem with retrospective studies is that there is no meaningful way to draw reliable conclusions based on them. For example, let's say a study is meant to determine whether a group that takes vitamins is more intelligent than one that doesn't. We can't just sign people up for a prospective study and give half the subjects vitamins and the other half sugar pills — that would be unethical. So we must use a retrospective experimental design, one in which the subjects are drawn from the population, based on their pre-existing behaviors — some who take vitamins, some who don't — and try to decide how this affects intelligence.
Most readers will see the problem with this design — for groups drawn from the population at large, some of whom take vitamins, and some who don't, how are we to determine whether the experimental outcome is a cause or an effect? Did the subjects take vitamins because they are intelligent, or are they intelligent because they take vitamins?
This example shows the key problem with retrospective experimental designs — one cannot reliably separate causes and effects. And the number of "studies" that try to draw scientific conclusions from observed public behaviors is staggering, and the undeserved attention these studies get is no less than a scandal. Here's an example:Study: Pot Smoking Increases Risk of Psychosis5. A quote: "Of those who smoked pot for more than six years, 3 to 4 percent went on to develop a psychotic disorder before the age of 21. By comparison, lead study author John McGrath estimates that around 1 percent of people worldwide suffer from psychotic ailments."
To avoid the pitfalls of a retrospective study, the researchers briefly considered asking teenagers to smoke pot, but quickly saw the problems with this approach. This means the classic problem endemic to retrospective studies is present in this study — a confusion of cause and effect (people inclined to use drugs might be more prone to mental illness). Well below the scare headline are comments like this: "... every study on the issue thus far has been imperfect: Despite controlling for variables like family history or childhood trauma, researchers were hard pressed to conclude that marijuana use caused psychosis, and not the other way around." Or both these factors might be correlated with a third unexamined possibility.
This quote is also noteworthy: "Armentano also points to studies that have found no connection between pot and psychosis, like a systematic review out of the United Kingdom just last year. 'I don't see any Reuters headlines on those,' he said, noting that it's much easier to get research dollars for studies into the adverse effects of marijuana."
So, if the scientists behind this study openly acknowledge that a cause-effect relationship cannot be established, how does the article merit the headline, "Pot Smoking Increases Risk of Psychosis"? The answer is that this is typical of modern-day psychological research. Psychology cannot become truly scientific until there is a defining, falsifiable theory to which all such work refers, and until psychologists begin explaining instead of describing. Until this happens, no amount of psychological research can grant psychology the status of a science.
I chose a drug study for this example because such studies are especially difficult for practical reasons, and because the topic is controversial, bias becomes a risk at every level — from acquiring grant money, to choosing subjects and protocols, to the interpretation of results.
Consider the implications of this example. On one hand we have an experimental design that cannot produce a reliable result and a study that ends up producing an ambiguous outcome. On the other hand, we have public information machinery that can be relied on to mischaracterize a study's outcome and as a side effect make science look like a tool for confirmation of popular sentiment.
Describing versus Explaining
The distinction between describing and explaining is critical to science and the shaping of theory:
Here are some examples that show the advantage of explanation over description:
One day at the seashore, the water level suddenly drops and recedes from shore in an unprecedented way, uncovering many fish suddenly stranded, flopping about on the sand, waiting to be picked up.
Joe, a teenage driver, notices that his car needs 80 feet to stop when traveling at 40 miles per hour. Joe wants to know how much stopping distance he will need at 80 miles per hour. (For simplicity, we neglect reaction time in this example.)
A group of fishermen is rescued from the North Sea, where they have been immersed in cold water for 90 minutes. They are severely hypothermic. What is the best aid strategy?
To summarize, fields that describe natural phenomena are not sciences unless they endeavor to explain what they describe. The explanations can lead to a theory, and it is testable theories that define a scientific field. This means one may find an organized, distinct field, with scientists doing legitimate scientific research and publishing in scientific journals, but unless the research addresses that field's theories, the research cannot contribute to the field's scientific status.
No testable theory → no science → no scientific field — regardless of the number of clipboards and white lab coats.
When seen correctly, science is an efficient tool for validating skepticism, and those ideas that survive science's dispassionate filter are reasonably trustworthy. This definition may seem at odds with the public view of science as a search for truth, but it is accurate.
It's important to understand that modern science is the product of a long consensus-building process by people who wanted a reliable idea filter, one that can efficiently separate ideas worthy of attention from a vast sea of nonsense. Most of this design process took place in the years before mass marketing had even been imagined.
It's important to emphasize that science is not an ideology, rejects authority and expertise, and is by most measures a subversive activity. This means those who pay for science must want science's results badly enough to tolerate a process that stands in opposition to all conventional norms of loyalty or tact. But there's an escape strategy — those who want science's validation but who don't want the trouble caused by real science and scientists, can build a plausible imitation of science.
The present science drama is not located within science itself, but outside, where powerful social interests want science's imagined validation for ideologies and products. It is here that the chasm between the public's understanding of science, and science itself, is most obvious.
Here are some examples where a desire for science's status and validation produce the outward appearance of science with little or no substance:
Source
|
Goals/Strategies
|
Comments
|
The Discovery Institute (Intelligent Design advocates), and by extension, Christianity's struggle against science |
As outlined in the Discovery Institute's Wedge Document (PDF), "To defeat scientific materialism and its destructive moral, cultural, and political legacies", "To replace materialistic explanations with the theistic understanding that nature and human beings are created by God" and to "reverse the stifling materialist world view and replace it with a science consonant with Christian and theistic convictions". | As explicitly stated, this program demands a version of "science" that lacks an essential property of science — the spirit of free inquiry. Not surprisingly, the "science" that results can only confirm what was expected at the outset, because other outcomes are excluded on ideological grounds. Conclusion: not science. |
The drug industry, commonly known as "Big Pharma" |
The creation and marketing of medical drugs, preferably but not necessarily with scientific validation for safety and efficacy. Strategies include:
|
It's easy to see that, when an industry funds hundreds of studies and can prevent the publication of negative results, a handful of statistically unlikely results, or outcomes resulting from fraud, will end up being the entire available literature for a given drug. This problem is made worse by Big Pharma's practice of contracting with medical doctors and scientists for lucrative public speaking engagements and ongoing research programs. These programs reliably catch individuals not obsessively preoccupied with ethics.
Doctors have the right to prescribe available drugs in any way they see fit. This means "off-label" drug applications are perfectly legal, if doctors deem them to be in the interests of their patients. So drug companies will get FDA approval for a particular use of a drug, then spend millions promoting "off-label" uses to increase sales, even though the "off-label" uses may never have been studied for safety or effectiveness. Conclusion: dubious science at best. |
General Advertising |
Goals:
|
Advertising has an astonishing reputation for cynicism about the public's knowledge of science, a cynicism more than validated by results. It seems the public can be persuaded by anything remotely resembling science:
Just once I would like to see an advertisement that says, "I am a scientist, and I recommend that you go out, get an education, and make up your own mind." Conclusion: science held in contempt. |
Human psychology |
A complex, sometimes self-defeating strategy:
|
Psychology occupies an uncomfortable position among intellectual disciplines, somewhere between religion and science. Religion is willing to moralize and presumes to tell people how to improve their lives, while science (medicine) doesn't moralize or advocate, instead (in principle) limits itself to identifying objective ailments and offering treatments validated in impartial research.
Psychology rejects being described as religion, but is willing to moralize — an individual who likes coffee "too much" can be described as mentally ill and in need of improvement ... that is, treatment. Psychologist Gary Greenberg describes8 Cognitive-Behavioral Therapy9 (clinical psychology's mainstay treatment) as “a method of indoctrination into the pieties of American optimism, an ideology as much as a medical treatment.” On the other hand, psychology would like to see itself as a science, but rejects scientific discipline. Clinical psychologists routinely ignore scientific findings10, while theoretical psychologists seem unable to produce scientific results trustworthy enough to impose discipline on clinical practice. Conclusion: dubious science at best. |
Economics |
A straightforward strategy:
|
The goals of economics have so far not produced anything resembling science or even consistency. The reasons are easy to understand:
|
Governmental science funding and support |
Publicly stated goals:
|
It's difficult to imagine contemporary science without government sponsorship, but it also easy to see its drawbacks:
|
In summary, science's true purpose is to manufacture doubt, but most of science's advocates misrepresent science as a source of certainty. Because of the pervasive influence of media and by default, science advocates' view of science become the public's view of science. The remedy is obviously an overhaul of education with a renewed emphasis on reason and critical thinking, but that is easier said than done.
Home | | Science | | Share This Page |