"When the anthropologists arrive, the gods depart." --Haitian saying
The scientific method has become so influential because it works, sometimes. We constantly extol its virtues in our beliefs and daily conversations. We even rely on clinical studies to steer us back toward common sense. (A patient just told me he started napping because a study said it would make him feel more rested.) Medical science can be straight-up miraculous. It has saved our lives, enabled our children to survive into adulthood, given us tremendous surgical capability, and even made it painless. I won't be talking about any of this because so many other people already are.
I am an acupuncturist. You might think I have a vested interest in promoting so-called alternative medicine, but I don't. I don't care if you like it or "believe in it." My disinterest in selling you anything is probably my only selling point. I became an acupuncturist because it gives me a vocabulary and means to express the way I naturally move in the world.
I will not be talking about about how acupuncture works, or why I think it's silly to put it under western scientific scrutiny. (Square peg, round hole.) Nor will I address the myriad and often divergent systems of classical Chinese medicine that evolved separately over millennia. I won't talk about the problems of reducing these medical ideologies through a singular Communist lens, or about further subjecting this fabrication to our Western, Eurocentric "Enlightenment" ideology, fraught with its own set of loopholes, errors, and ulterior motives. Authors such as Paul Unschuld1 and Heiner Fruehauf  have given due attention to these complex topics, and I have discussed it briefly in the article Herbal Medicine on Its Own Terms. 
If science values neutrality, we need to do a little self-reflecting. Think of this as an elegy to scientific dogma, which we think makes us practical and rigorous, but actually decreases our capacity. I will be pointing out some of the discrepancies of modern experimental science, illustrating the importance of context, and describing the major delusions that compromise our scientific integrity.
Scientific observation has confirmed the interdependence of all life. Everything, it turns out, is joined at the hip. But for some reason, our scientific practices do not acknowledge interdependence. The act of observation changes that which is being observed, from the medium-sized science of clinical trials to the infinitesimal realm of quantum oddities. How can we not affect stuff? How can stuff not affect us?
Yet when we design experiments, we reinforce our first major delusion: that we are separate from the universe, and that the universe itself is made up of other, separate things, which can be studied in isolation to further isolate more things. Enter René Descartes, who taught us that the world is made of discrete parts. If science is the observation of phenomena over time, then the only agreed-upon result that any enduring observation yields is that things are temporary. We can even say that they do not exist at all, at least not in any solid way. I'm not talking Nihilism here. I'm talking endlessly-bursting-forth-in-every-moment life. I'm talking about the exhilaration of groundlessness with no parachute.
Phenomena exist in the way that constellations exist, patterns generated into contexts by shared hallucinations, repeated. Imagine far-flung stars hundreds of millions of light-years apart, made of an arrangement of interactions we have named "light," which is literally ancient history. What we are seeing has often already exploded, and transformed into its next phase of being. Yet we construct a relationship we can hold onto, grasping for a 2-dimensional Orion in an infinite expanse of space that is often "too chaotic to be noticed or named."  As Vi Hart says, "I wonder if a proton plus an electron making a hydrogen atom […] is any more real than these stars making Orion. I mean, is it an atom, or does it just look like an atom?" 
(Of course, these versions of the story are brought to us by science. Is science really much different from connecting far-flung dots to create imaginary pictures?)
But let's come to our senses. Western scientific mythology holds that we have five of them. Among these, vision tends to dominate. Examining something with human eyes already involves many layers of interpretation--perception of color, depth, dimensionality, contemporary connotations, beliefs about the world, visual acuity, not to mention our brain's oddly filling in visual gaps with made-up information that seems to fit, but isn't necessarily accurate. We can't conceive of most alternative ways of viewing the world because we're human beings with human eyes. We think that telescopes and microscopes alter the scale of our capacity by orders of magnitude, when really they just extend our insular view with mechanisms that resemble the human eye, to be viewed by--you guessed it--the human eye.
If our sight were infrared and five-dimensional, for example, or our point of view included 360 degrees, or if we were blind, we would be observing something else entirely. Don't get me wrong. Limitations are great. It's so nice to wake up in the morning and not have to figure out whether we're a human being or a lamp. But can you imagine how many ways we've burdened phenomena by the time we have put our (non-existent) "objects" of study through scientific analysis? Who knows how various natural phenomena experience reality, but it's far-fetched to believe that nature exists on the terms that human beings create for it.
aka, The Decline Effect
Perhaps some of these limitations also help explain the decline effect. Jonah Lehrer wrote an article about the decline effect in the New Yorker that describes how scientific facts erode over time. Joseph Banks Rhine first used the term in the 1930s to illustrate a disturbing trend in his findings: results of well-conducted scientific research tended to receive decreasing support over time. In other words, when scientists repeated the same experiment under similar conditions, the results became increasingly uncertain. 
This brings us to our second grand delusion: That phenomena, including "facts," are permanent. Replicating findings is a cornerstone of modern research, so we can see how problematic the delusion of permanence can be. In a solidly-conducted clinical trial, researchers found that second generation antipsychotics worked better than first-line antipsychotics in people with schizophrenia.  The study was documented, confirmed, and widely accepted. Then suddenly, subsequent trials showed steep, incremental decreases in the validity of the initial findings. Apparently, this happens all the time in many fields. Scientists have even found discrepancies in the law of gravity.  Yet we still use second generation antipsychotics, and the law of gravity is found today's textbooks.
Some people assume that the decline effect can be chalked up to "regression to the mean." This means as the experiment is repeated, early statistical flukes get cancelled out. But the brightest people in data analysis, such as researcher Jonathan Schooler, say that the data sets that end up declining seem statistically solid. He explains, "The odds of them [sic] being random are typically quite remote, like one in a million. This means that the decline effect should almost never happen. But it happens all the time!"  Schooler refers to the problem as "cosmic habituation" to describe the decrease in response that occurs when individuals get used to certain stimuli. For example, the overprescription of antibiotics has compromised their effectiveness.
Scientific trends offer another explanation. The rise and fall of intellectual fads occurs frequently. Following scientific fashion has been our modus operandi for a several hundred years. Liu Ming calls our religious adherence to science "scientism," a mythology in its own right. Scientism is an engine of doubt that runs on disproving itself. Here is a 16th century example: The Earth is the center; the Sun is the center. And a 21st century example: the brain is the seat of memory; the brain is not the seat of memory.  Science fads show up in the results of clinical trials, as early peer-reviewed processes tend to confirm positive findings. When scientific incentives shift, however, the most notable results are those which disprove the original theory.
This leads us to publication bias. Journals usually only publish "significant" findings over insignificant ones. Significant, in statistical terms, means that the results of a study have less than a 1/20 probability of being due to chance. Refusal to include subsequent studies on exciting new research "breakthroughs" keeps the decline effect hidden from public view. It is a problem among scientists, which most scientists don't want to acknowledge. Publication bias is a form of data selection that skews every field of thought contained in research journals. Richard Palmer, a biologist who has noticed and studied the decline effect in his field, explains, "We cannot escape the troubling conclusion that some--perhaps many--of our cherished generalities are at best exaggerated in their biological significance and at worst a collective illusion nurtured by strong a-priori beliefs often repeated." 
Acupuncture research conducted in different countries shows us a noticeable example of how our beliefs inform our results. While many Asian countries acknowledged acupuncture as effective and incorporate into their medical infrastructures, people in the west tend to view it with skepticism, at best. These differences of opinion, which millions of people uphold in their respective countries, also show up in scientific research.
Between 1966 and 1995, China, Japan, and Taiwan conducted 47 acupuncture studies. Every single trial concluded that acupuncture was an effective treatment. During the same years, the United States, Sweden, and the UK held 94 clinical trials, and only 56% of these studies found any therapeutic benefits.5 Without realizing it, scientists confirm their preferred hypothesis and disregard what they don't want to see. (We all do this, by the way.) Another reason for conflicting results could be that Eastern researchers design studies that better reflect the medical worldview they're analyzing. Studying acupuncture in a country that developed acupuncture could be less susceptible to colonial douchery--er, dominion.
Now let's move onto Evidence-Based Medicine (EBM), which is the self-proclaimed gold standard for medical decision-making. EBM treats individual patients based on the outcomes of huge medical trials.  First we need to be able to understand the language that these studies use. (If you don't care, skip the next three paragraphs.)
Clinical trials call a result significant if its "p-value," or probability value is less than or equal to 0.05. We can take two groups of people, a control group and an experimental group, and measure the difference between them, with the p value being the threshold that indicates whether or not the results are supposedly important. Let's pretend the control group gets usual allopathic care for [X disease], and the experimental group receives acupuncture for [X disease]. We are looking for a significant difference in the health of the two groups after the trial.
If, after the trial, the group receiving acupuncture is measurably healthier than the control group, this finding would support acupuncture as a valid treatment option for the disease. Or, if it turns out that the control group fared better, acupuncture would be dismissed as an ineffective treatment for the disease. Often, results describe something in between, in which case we will probably never hear about them. Significant-sounding results might make for more interesting reading, but it is not a very scientific practice.
Hillary Roberts, PhD and Steve Hickney, PhD wrote a brilliant piece on the pitfalls of EMB, which I will plagiarize--I mean, paraphrase--below. They explain, "Evidence-based medicine fosters marginally effective treatments based on population averages rather than individual need."  Large trials detect small differences. For example, if we have two groups, and measure the blood pressure of individuals from each group, we will find no significant difference between them. If we take a hundred people from each population, we get a low level of significance (p < 0.05), but if we take a thousand, we now find a very highly significant result. However, the magnitude of the small differences in blood pressure remains the same in each case. So, we can demonstrate a highly significant result with very little meaning. In a large trial, highly significant effects are often clinically irrelevant. 
EBM proponents can make their findings sound even more impressive by using relative rather than absolute values. A common example is the statement, "[X drug] halves your risk of developing cancer." This sounds like a great case for taking the drug. However, the reported 50% reduction may lessen your risk by just one in ten thousand from two in ten thousand. These are the absolute values to which the statement refers. Such a small benefit is negligible, but it sounds important when expressed in relative terms. Hickney and Roberts use the analogy, "Buying two lottery tickets doubles your chance of winning the lottery."  This is technically true, but the odds are minuscule either way.
Imagine the government wants to give a new pair of shoes to every American based on the average American foot size. We achieve a mean result, and impose it as a standard, which we apply to the whole population. This describes a typical clinical trial. Obviously, this solution would not work for many people. As Roberts and Hickney point out, "Individual responses to medical therapies vary by at least as much as their shoe sizes, yet despite this, EBM relies on aggregated data."  This is an error in logic, since group statistics cannot predict an individual's response to treatment.
Lastly, EMB has strict rules regarding which data they use in the final crunch. Statistics that fall outside thetypical data spread get tossed. Pretty ironic for a model that denounces data selection. In this way, clinical trials provide snapshots of decontextualized information forced into a significant-seeming pattern. Given these circumstances, achieving similar results in a repeat trial is unlikely. The very nature of large scale studies makes any attempt at replication unlikely, as they are difficult, expensive, and time consuming. These factors combined really take a toll on the overall credulity of Evidence-Based Medicine.
In Cybernetics, the principle of "requisite variety" holds that the solution to a problem has to contain the same amount of relevant information as the problem itself.  This is not so good in cases of large-scale clinical trials, which typically do not take into account different causes and expressions of a given health concern. Violating the principle of requisite variety is why researchers in most trials cannot predict who will respond to which treatments, and why results are so mixed. They simply do not take enough relevant information into account. A cold compress will remedy elevated body temperature due to heat exhaustion. However, a cold compress would do nothing to treat elevated body temperature due to infection. This person would need a different treatment altogether.
Incorporating adequate relevant information in large-scale trials would be next to impossible. But requisite variety is integral to a healthy patient-practitioner relationship. Acupuncturists, for example, create custom treatments based on what a patient is presenting at the time of her appointment. Our real-time methods of analysis acknowledge human beings as complex, dynamic systems. Cybernetics and General Systems Theory, two scientific models that arise from our own western lineage, give us ways to examine living systems (biomes, physiology as a whole) rather than discrete units (molecules in a drug, genetic markers).
Open systems are alive, self-regulatory, and comprised of whole parts. We can apply systems thinking to any discipline, including medicine. A good medical practitioner is a good regulator; she will continually assess the interplay between change (toward balance or decline) and homeostasis (the self-regulation that allows for stability). In this light, one-to-one relationships between practitioners and patients, or even smaller clinical trials, help us determine the most accurate treatment for a health concern because they provide more context.
The Pharmaceutical Industry
The pharmaceutical industry plays a huge role in health-related clinical trials, including low-tech therapies such as acupuncture. Professor of neuroncology Mark R. Gilbert, conducted a large double-blind study with 600 brain cancer patients who were randomly assigned to two evenly-balanced groups, those who got the drug Avastin along with standard treatment, and a control group, who received standard treatment and a placebo. Some patients who took Avastin significantly beat the average, and some did not respond at all. The trial was unable to discover the "responders" or examine what might have accounted for the difference. 
The New York Times article, "Do Clinical Trials Work?" describes 26 randomized, controlled studies of high quality that were followed up by larger trials, in which the initial finding was wholly contradicted in three cases. In another six cases, the later trials found the benefit to be less than half of what was first reported. Journalist Clifton Leaf asks the question, "Are the diseases of individuals so particular that testing experimental medicines in broad groups is doomed more frustration than knowledge?" (Hint: yes.)
The following is a bit of a worst-case scenario, but it does happen. For a drug to go on the market, it has to pass three rounds of clinical trials. Large drug companies often have a lot of influence in initial clinical victories by ghost writing the research. Pharmaceutical companies solicit academics to publish their highly-constructed findings. In extreme cases, drug companies pay for trials by contract research organizations, analyze the data in-house, have professionals write manuscripts, and pay communication companies to usher them into publication in the best journals. 
The resulting articles affect conclusions found in medical literature, and are used in promoting drugs to doctors. After a drug achieves these early statistical wins, it endures more equivocal second-round results, due to many of the factors that I have described so far. Finally, they roll the dice in the third round, the round that determines whether a new drug will be introduced to the market. Drug companies do not imagine that the drug will suddenly work, but that the trial will make it appear to work. 
It does not take a conspiracy theorist to realize that if large pharmaceutical companies have a vested interest in promoting drugs, they also have an interest in suppressing low-cost, non-pharmaceutical therapies (such as acupuncture). Large-scale trials are extremely unlikely to be repeated in the field of acupuncture in the first place, simply because there are no financial incentives for validating acupuncture as an effective treatment. Though political backdrops are not always visible, we need consider them among the many factors that distort the landscape of published research.
Clifton Leaf looks for solutions to remedy this billion-dollar gambling industry. He has a good idea, limited only by a reductionist approach. He suggests designing small clinical trials that enroll only those who have the appropriate genetic markers or molecular signature. This is how Genetech developed the breast cancer drug Herceptin, which targets tumor cells that have an abundance of a certain type of protein.  Certain molecules can destroy certain pathological cells. We use this knowledge routinely by isolating pathogens (cancer cells, bacteria) and seeing which drugs eradicate them most effectively. A building-blocks approach to scientific inquiry certainly narrows the population applicable for a study, but not necessarily in a way that will enable more appropriate therapies. We are still missing the bigger picture. Though it can be helpful to study the sum of our parts, we are not the sum of our parts, and neither are our diseases.
Life is not static. Time is an integral part of any living system. No experiment can be repeated exactly because variables differ astronomically, perhaps even literally. Many scientists actually consider time (hour, day, season, year, etc.) as an inextricable factor in their fields of inquiry. In my own ancestral literature, the Celtic sagas, people often asked the Druids for the best day, or even the best hour, to conceive a child or embark on a raid. These questions clearly demonstrate that our ancestors practiced the observation of phenomena over time. Their stone structures alone attest to this, as they archive extremely precise astronomical motions, and function as viable calendars thousands of years later. I imagine that clinical trials, too, can yield different outcomes depending on when they are administered. Researchers consider inevitable scalar changes that impede replicability, but they do not typically include time among these.
Time affects our daily activities, and plays a vital role in many non-western sciences. This is especially true of acupuncture, as it co-arose with a calendar system called the tong-shu. Before the Communist takeover, Imperial China employed thousands of calendar experts whose analyses informed all major decisions.  Many doctors practiced medicine and gathered herbs in the context of correct timing. People still use these calculations, which the Chinese classic, the Nan Jing calls Wu-Yun Liu Qi. For the sake of communication, I like to call it biodynamic acupuncture.
Practitioners needle the hourly "open points" if diagnostically applicable to structure a treatment around where the healthy life-force is most concentrated and available during a given hour. We also take the "meridian clock" into account, which has hourly associations with different sets of organs, and gives another framework for proper treatment. Some practitioners even administer treatments for chronic diseases only at certain times of day. For example, if a patient has Crohn's disease, which affects the large intestines, a practitioner could administer treatment between the hours of 5 - 7 am, the hours that correspond to the large intestines. As you can see, this is a very different model from Evidence-Based Medicine. Each treatment is alive, contextualized, and perfectly tailored to each client, assuming practitioner competency.
Adherents of western science do not need to adopt ancient Celtic or Chinese ways to engage in dynamically-aware medicine. Western medicine has begun to use these principles, which they call Chronobiology. Allopathic doctors in China have found, for example, that side effects of chemotherapy are greatly reduced in people with liver cancers if chemotherapy treatments are given in between 1-3 am (which also happens to be "Liver time." 
Toward Respect and Inclusion
Without romanticizing or making too many generalizations about indigenous cultures, I'd like to share a quote that speaks to one strand of indigenous science. Hawai'ian elder and ancestor Hale Makua explains, "Humility/reverence is the purpose of science and the goal of education." I love the idea of a science of respect and inclusion. Imagine scientists consciously engaging interdependence!
As we have seen, humility and reverence are not often the causative agents for clinical trials. Ulterior motives play a large role in the infrastructure of medical science. Even if our hearts are in the right places, our means do not often match the end. Now we study chemicals, cells, genes, and other discrete units on unwilling participants with nonhuman physiology. This does not demonstrate respect and equality among all life forms, much less reverence.
Today, researchers conduct experiments on animals that were deliberately given toxic drugs or infected with disease-causing agents for the sole purpose of improving the lives of human beings. In even less justifiable cases, we incite suffering just to determine how or whether a therapy works. Here is our third mass delusion and scientific impediment: We believe that we are not only central to the universe, but also more important than any other life form. This is simply not true; I don't know how else to say it. If we are truly scientific, our concept of self includes all phenomena. The notion of self-preservation naturally expands to respect all of life in our scientific decisions.
Some elementary errors in logic also arise from these research methods. First, we apply animal research findings to human beings. That's the whole point, right? But it doesn't always work. An unfortunate example of this occurred in the early 20th century when a scientist fed rabbits a high fat, high cholesterol diet, which gave them atherosclerosis. Of course it did. Rabbits are designed to eat vegetables. Nothing about their physiology is equipped to handle lots of fat and cholesterol. He then generalized these findings to human beings, and published the results.  Fat and cholesterol quickly became four-letter words, to be avoided at all costs. This huge misunderstanding around human fat and cholesterol consumption is still very alive in our contemporary medical culture. Now the pendulum is swinging back, and new studies show the health benefits of (good) fat and cholesterol.  Observe the decline effect.
Another error occurs when researchers confuse causation and association. This is a symptom of reductionist thinking. If scientists study the brains of humans or animals when an acupuncture needle is inserted, they may attribute the effects of acupuncture to changes in the limbic system , or increased cannaboid secretion , or development of new synapses . In other words, they will find physiological changes in the areas they are studying because acupuncture is a whole-systems approach that yields systematic effects. Studying physical responses in isolation shows only a singular association, and certainly not causation. (This can show up in statistics in very funny ways: We can accurately say that 75% of heart attacks occur within 24 hours of taking a shower. But eliminating showers does nothing to promote heart health.)
Just like western science, indigenous science relies upon direction observation. Tests exist to ensure validity, and data are used for forecasting and generating predictions. For example, the principle of Triangulation requires scientists to verify information by waiting for its content to be legitimized three times in three different ways. Let's say I have a dream that a deceased family member wants me to travel to Ireland, and then I receive an invitation to come stay with the people of Ireland, and finally, a friend offers me all of his frequent flyer miles with no knowledge of the prior events. Triangulation keeps subjective science practical, and ensures correct timing of action.
Native scientists receive training in various specializations, such as herbalism, weather observation, navigation, mental health, and time keeping. Subjective science acknowledges our role in inquiry, and requires technologies such as feeling a sense of balance, or trusting our instincts. We in the west consider these technologies unreliable, perhaps because we no longer rely on them.
Unlike western science, the data from indigenous science are not obtained by an attempt to control the forces of nature. Instead, both the methodology and its purpose are geared toward accommodating nature. Dr. Apela Colorado, Frank-Oneida elder and one of my personal teachers, offers several other distinctions. Here they are, in her own words:
- The indigenous scientist is an integral part of the research process, and there is a defined process for insuring this integrity.
- All nature is considered to be intelligent and alive, thus an active research partner.
- The purpose of indigenous science is to maintain balance.
- Compared to Western space-time notions, indigenous science collapses time and space with the result that our fields of inquiry and participation extend into and overlap with past and present.
- Indigenous science is concerned with relationships. We try to understand and complete our relationships with all living things.
- Indigenous science is holistic, drawing on all the senses, including the spiritual and psychic.
- The end point of an indigenous scientific process is a known and recognized place. This point of balance, referred to by my own tribe as the Great Peace, is both peaceful and electrifyingly alive. In the joy of exact balance, creativity occurs, which is why we can think of our way of knowing as a life science.
- When we reach the moment/place of balance, we do not say we have transcended--we say that we are normal. Always we remain embodied in the natural world.
- Humor is a critical ingredient of all truth-seeking, even in the most powerful rituals. This is true because humor balances gravity. 
We have seen some ways in which our made-up rules fail us, and foster practices that damage life. When our commonly held beliefs and behaviors lead away from balance rather than toward it, then we can simply examine our old, harmful delusions, and create new ones that are fun for everyone. This will take some time. But don't worry--the decline effect reaches everything, endlessly. When we take a look at something, it appears to be very solidly one thing, at least at first. But as we keep looking, we see that it's not what we once thought. In fact, it is difficult to pin down in any way at all. Not only does the decline effect describe the results of scientific inquiry, it describes scientific inquiry itself--an everchanging body of assumptions subject to change.
Mathemusician Vi Hart asks, "Can you see this without reading it?"  The mythology of modern science embodies some of our society's constantly-traversed delusions, preventing us from seeing phenomena without assigning them meanings that hold little value beyond the scope of our own conjecture. Yet to be a decent health care practitioner requires good pattern recognition (as a means of diagnosis) and pattern formulation (as a means of treatment), just as to be an effective sea voyager requires knowledge of the constellations. Maybe it is as important to dismantle patterns as it is to make use of them. Freedom from these constraints can exist only marginally, but it is the effort that dissolves the dogmas of bad science. The prospect of a scientific model that builds interdependence into its methodology really appeals to my sense of self-preservation. Inquiry borne of this inclusion is bound to respect and reflect the life that it seeks to understand. Now that's a science I can get behind.