Welcome to my blog

Hello. I am iciclehunter and this is my diary. My job title is "osteopath", and my work is hunting for clues, detective work, problem-solving. These things involve reason and science, but are not limited by them. They also involve the eye of experience, and "hunches". Thus, some would regard my activities as those of a quack, a title I assume here with irony. I am writing this blog because I like writing. I am quite opinionated, and perhaps I suffer from a repressed need for expression. I have no particular prior "agenda"; if I have any bees in my bonnet, no doubt they will make themselves apparent by their buzzing. All names and identifying details of any people featuring in these anecdotes have been changed. Thank you for reading.

Saturday, 27 July 2013

Science isn't perfect but it's the best way we have of knowing useful things (that are not immediately obvious and beyond doubt)

Somebody recently told me about science and other important fields of knowledge, "It's all a lie!", his point being that real truth is kept obscured from the people by official versions of it. He is not alone in his contempt. I wrote previously about a woman who told me, “I hope the scientists don't discover any more about anything.” Among the general population suspicion of science and scientists is not unusual. And a degree of scepticism is healthy, so long as it does not transmutate into blind prejudice. In the two comments I have reported above I think I detect unthinking prejudice, and if I am right, this is both sad and worrying, especially so since these two statements were actually made by health care practitioners.

Science is the best way to know complex and hidden things

The number of things is quite limited about which, in my heart, I can bring myself to hold an unequivocal opinion. But here is one:

Science is the best method we have to know facts about those things in our natural universe which are not immediately obvious and beyond doubt.

What do I mean by "immediately obvious and beyond doubt"? If lightening flashes it is immediately evident to observers that there has been a lightening bolt, we don't need science to tell us one has occurred. If I doubt my eyes or my brain I can consult other witnesses. However, I do need science to  tell me about the physical nature of lightening, for example that it is a electrical discharge. I would also need science to tell me if there were anything wrong with my eyes or my brain if I saw a lightening bolt when no other observers present had seen it. There is no better method to tell me these things than science.

Science is not a complete or perfect method of knowing things, for a number of reasons:

1) There are limits concerning the kinds of things that science can investigate.

There are absolute limits, determined by the very nature of the subject matter, and relative limits, which are the constraints imposed by the level of development of our methods of investigation. An example of an absolute limit is God. God cannot be investigated by science because God is supposed to transcend natural phenomena (1). However, belief in God can be investigated, as belief is a universal human cognitive behaviour, and very much of this world.

((1) Professor Richard Dawkins disagrees. As I understand it he considers the existence of God to be a potentially testable hypothesis. In the absence of plausible evidence, he concludes that God does not exist. Professor Dawkins is a very eminent academic. Who am I to say that his reasoning is is flawed, his attitude unscientific, or that the presumptuous fellow needs a poke in the eye (perhaps from God, if He/She/It exists)? Declaration of impartiality: I do not believe in God, and neither do I disbelieve. Declaration of irrational partiality: However, Dawkins comes across as being rather too smug and sure of himself for my liking.) 

Very rare or unique events are difficult or impossible to investigate because science is concerned with finding out generalities, for which it needs to study a number of events of the same kind. We can only guess (based on our prior knowledge of nature, and therefore plausibility) what caused the Tunguska Event in Siberia. Because it happened in the past, we cannot observe it, and because it is the only event of its kind to have been described, we have nothing valid to compare it with.

A similar difficulty is studying subjective phenomena. If I see a bolt of lightening but nobody who was with you at the time saw it, it is an experience which is unique to me. If a number of people experienced similar events, one might hope feasibly to find characteristics of these people that set them apart from the rest. If with me it was a repeated experience, perhaps one could find peculiarities in my brain which could explain the experience.

Pain is a subjective experience, which presents difficulties to researchers. Indeed pain researcher J. H. Kellgren wrote in 1948 that:

"The phenomena of pain belong to that borderland between the body and soul about which it is so delightful to speculate from the comfort of an armchair but which offers such formidable obstacles to scientific inquiry."

Since 1948 science has made great strides in the investigation of pain, but then it presented, as Kellgren noted, a "formidable obstacle", because of the relative limits of science at that time. Pain is a complex, multifactorial phenomenon. Science still has difficulty with the investigation of complex multifactorial phenomena, especially those dependent of human behaviour, such as economics and health. Methods for investigating such complex systems are only partially developed and limited in their reach and validity. This is not to say they are useless, rather that any single new piece of information gained from them generally has to be considered preliminary, incomplete and of questionably reliability. In other words, we cannot discount that any such fragment of information may be an illusion.

2) Advances in scientific knowledge generally proceed slowly.

It is a frustrating fact of life that science often seems to proceed too slowly for our needs. People coping with chronic illnesses read news about "breakthroughs" which could lead to significant improvements in their treatment in 5 or 10 years time. Which could be too late for them or their loved ones.

It is also a common criticism that science frequently produces a result which had seemed obvious to people since time immemorial, that it eventually catches up with "what everybody knew anyway". However, this is not a fair criticism. Lots of other things which have seemed obvious to people at some time are now known, through science, to be improbable fantasies. Things are not always what they seem, and belief is not always the same as knowledge.

3) Science does not provide absolute truths. 

Scientific knowledge, unlike religious belief, is not absolute. Science produces models and then attempts to see how well the fit with observations, how well they approximate to reality. The models are a map, they are not the territory, and it can only ever be a partial description. They can only ever approximate to reality because we will never know everything there is to know about any thing, however small. Science acknowledges this. In that way it is humble.

4) Science is categorical, that is, it deals with categories.

In order to carry out any sort of scientific investigation, it is necessary to define the things we are studying. That is, we have to consider categories. We might wish, for example, to study health in one way or another. So we will have to create the category health and define it. This is not such an easy task because health is multifaceted and has different meanings and connotations to different individuals and populations. Our definition will always have limits. The tighter we make it, the less it will be relevant to the real world of human experience, but the more complete and encompassing we make it, the less it will be amenable to the constraints of scientific rigor.

Another annoying problem is that scientific knowledge is only valid within the categories the study of which has produced it. For instance, a study may have shown that a certain treatment is effective in slowing the advancement of osteoarthritis in the knee in European people between the ages of 50 and 70. However, this does not say anything about the effectiveness of the treatment in osteoarthritis of the hand, or in African people, or in people over 70 years of age.

5) There is no such thing as a perfect experiment.

A good experiment is difficult to design. This is for a variety of reasons, including:
  • An experiment comprises several different stages or components, each of which has to be thought about in depth. The potential for oversight, error, and false assumptions is high.
  • A typical experiment takes into consideration a number of examples the thing studied. This is called a sample. The larger the sample, the more representative it is of the whole "population" of the thing studied, and the more realistic the results are likely to be. However, it may be unfeasible for logistical or financial reasons to study a large sample.
A good experiment may be difficult to design, but a perfect experiment is impossible. A perfect experiment would study the whole population of the thing studied (if this involves human beings, that would be over 7 billion people), and be devoid of all unsubstantiated assumption, not to mention oversight and error.

As a result, most published scientific studies are not of particularly good quality.

6) "Proving" something isn't easy.

I have written "proving" in quotation marks, because as mentioned above, science provides only an attempt at an approximation to reality. Nothing is ever "proven" absolutely. Nevertheless, let us accept the term "prove" for the purposes of this argument, taking it to mean "demonstrate a high likelihood of being true".

It is not easy to do because:
  • First we need a good experimental design, and the resources (money, time, trained people, space, equipment, access to a large sample population) to carry it out.
  • Then we need to effectively carry out the experiment, thrashing out any teething troubles, and resolving any problems which crop up along the way. Probably we will give the experiment a dry run with a small sample first, in order to test the experiment itself. Perhaps we will do this more than once. Maybe our experiment even requires our sample to be observed over several or many years.
  • After that we need to write up our experiment and its results and present it for publishing. To be published in a quality, peer-reviewed journal, the competition is intense. The selection process for  studies to publish is subject to the biases of the reviewers and editors, something called "publication bias".
  • Then, assuming the results are of interest, they will need to be confirmed by other researchers. That is, the study and its results need to be repeatable. The same kind of study must consistently produce the same kinds of results.
All this takes a lot of time, a lot of people, a lot of money, a lot of patience, a lot of intelligence, and a lot of good will. Not easy.

7) Science can "prove" ideas much more easily than it can disprove them.

Although proving an idea is difficult, it is by a long chalk easier than disproving one. One reason for this is that non-existence of proof is not proof of non-existence.

Let us take fairies as an example. Just because I cannot prove there are fairies at the bottom of my garden (I cannot show you one either in life, on film, or dead in a jar) does not mean they are not there. They have magical powers, so can avoid being seen, filmed or otherwise observed except by whom they chose to reveal themselves. Myself, for example. They never die, or if by chance they do, they immediately disappear from this universe. Can you prove me wrong?

Another reason is statistics. Statistical analysis is an essential part of much scientific research. We have to demonstrate the "statistical significance" of any result, that is, the probability that it did not come about by pure chance. To do this we have to determine a "confidence level", which is a probability, commonly expressed as a percentage, above which we can be reasonably sure that the result is not due to chance. Thus, if we select a 95% confidence level as "reasonable", and our experimental results achieved 97%, we could accept the idea (technically called a "hypothesis") we are testing and refute the possibility that the result is casual. (The selection of the confidence limit is arbitrary, but the higher it is, the more likely that the results are realistic).

However, if our results achieve only 93% confidence level, we cannot say our hypothesis is wrong. All it means is that we have not achieved our pre-selected "reasonable" confidence in it. But in fact, that 93% only leaves a 7% probability that the result is due to chance, so it could still be a true result even so. We have not disproved it. The corollary is that in order for a hypothesis to achieve acceptance as a good approximation to reality, science really requires quite high standards.

8) Science is easily misused, misrepresented and misunderstood.
    To an extent science assumes integrity and competence in those who do it, publish it, and report on it.
    Unfortunately, that trust is often betrayed. Sometimes scientists make up their results for money or fame or outside pressure. Sometimes publishers will only publish certain kinds of studies, or studies with certain kinds of results. Headlines are more about selling newspapers than encapsulating essential information. Frequently the journalists who present science to the public are insufficiently knowledgeable about it to present it faithfully. Public and private bodies pay scientists to support their propaganda. This the story of a good tool being misused through ignorance, incompetence and dishonesty.

    9) You can only find what you look for.

    As they say, "Ask a stupid question and you'll get a stupid answer". The answers science gives are only as useful as the questions posed. You will never find out about an aspect of a phenomenon which you never even think to examine.

    Alternatives to science

    Despite all these limitations, I still believe that science is the best means we have of knowing useful things (that are not immediately obvious and beyond doubt). Let us take a brief look at some of the alternatives.


    Instinct is what guides animals to eat the right foods, drink water, and select sexual partners. It provides us with rapid systems of rapid recognition and response in times of danger and in contexts of competition with our peers. It has been essential to survival for much of human evolution. But it is concerned with basic, immediate survival, and cannot by its own means provide access to deeper knowledge about the workings of the universe.


    Sensory experience is unreliable. That is, you cannot be sure that what you think you have observed means what you think it does. The problem comes with the interpretation of the sensory experience. There are lots of ways in which this interpretation may be unconsciously biased. One common way in which this happens is "seeing" what one already expected to see.

    Many people in the medical professions, myself included, have justified their practices on the basis of personal experiences over a longer period of time. The kind of thinking I am talking about is exemplified in the thought, "Many people have got better after I have treated them in such and such a way. Therefore, it is effective". However, the interpretations we make about our life experience, such as the aforementioned one, are subject to a host of cognitive biases and logical fallacies, rendering them unreliable. Some of them may be valid, others not at all, and others partially so.

    My own view is that decisions and practices based on experience form an invaluable component of medical practice, so long as they do not disregard solid science or logic, and so long as a high awareness of their fallibility is maintained.


    People may believe things for rational or quite irrational reasons. Most often people will believe what they like to believe for no other reason than it feels right to them. There are over 7 billion people in the world each holding a unique set of beliefs. They cannot all be right, unless we accept that belief has only subjective value i.e. that a belief represents the truth only to the believer, having no external validity. This is not a good basis for knowing generally applicable and objectively acceptable facts about our universe.


    Intuition is the ability to "know" things without the use of reason. It has been related to the ability to quickly interpret subtle sensory cues at a pre-conscious level. It is not always reliable, but becomes more reliable the more experience one has in a particular area. In highly intuitive people engaging in an area in which they have a lot of knowledge and experience, intuition may be a valid method of decision-making in the here and now. However, it cannot be used to derive either general principles or to discover the detailed workings of things. Moreover, I would hazard a guess that people claiming particularly developed intuitive powers are no more likely actually to possess them than those who make no such claims. I have come across a few such people for whom the borderline between intuition and fantasy is as blurred as the horizon across the sea on a misty summer morning.


    Reason is an excellent method of working things out, if one is in possession of all facts and premises, and can avoid succumbing to logical fallacies. This becomes difficult when we are reasoning about partially understood complex systems. Many practices have been adopted in the medical field because by reason they ought to be right. Except that relevant facts have been unknown or disconsidered and make them wrong in real life. The necessary relevant facts to make reasoned judgments about complex issues very often come from scientific research.

    The more complicated the system of "knowledge" built up by pure reason, the greater the likelihood of errors within it. Alternative medicine is jam-packed with complicated systems which some guru or another thinks ought to be right, and are widely proclaimed as the truth by their disciples. Personally, I prefer my models to be as simple as possible. I have had no reason to believe, so far, that this renders them less functional.


    Under this title come both the intuitive "eureka" moment and divine illumination. I have little to say about these. The first provides major advancements in knowledge relatively infrequently. The second is claimed by many. You believe them or you don't (see "belief" above).


    Chinese whispers is not the best way to attain reliable knowledge.


    I have a particularly low regard for "experts". In a whole host of fields of knowledge, from medicine to economics to finance to psychology to education, expert authority has consistently demonstrated that it knows less than it likes us to believe. All manner of ineffectual or destructive theories and practices have held reign under the tutelage of authority.


    This is an interesting one. There is an argument I have a lot of sympathy with that a process of natural selection down through the ages produces many ideas and practices in folk culture that are accurate and/or effective. Through this means an organic body of traditional knowledge arises.

    However, there are some problems with this. Firstly, in the living world, natural selection of species also allows for the generational reproduction of traits which have no impact on species survival (neither positive nor negative) but which are physically linked (in the chemical structure of the genes) with other traits which do provide an evolutionary advantage. So a lot of inconsequential stuff gets passed on down with the useful stuff. I would be surprised if this did not happen with folk lore, too.

    Secondly, it seems beyond doubt that some traditional practices are inaccurate and/or harmful (except as means of maintaining a cultural identity). Natural selection is not the only process involved in the generation of a body of traditional knowledge. Others are the influence of myth, legend, and spiritual belief. In brief, traditional knowledge is a mixture of useful practical knowledge, myth and superstition. Only science can sort the wheat from the chaff.


    Orthodoxy equals tradition multiplied by authority. Nothing more need be said.

    Although some of these methods (in particular reason, intuition, experience) are invaluable when used within their own domains and limits, only science can provide a broader and deeper level of rational understanding. Rational, left-brain, understanding is not all there is. There is also intuitive, right-brain, understanding. Yet in the practical world, when we need to do jobs and get predictable results, the first is insuperable, especially when aided by experience and intuition.

    Science is frequently criticised for its limitations. It would be absurd to deny the utility of a tool because there are limitations to its use or performance. What tool doesn't? Moreover, a limitation can also constitute a cutting edge. Science is unrivaled at answering rational questions. Any tool must be used within its natural domain. The natural domain of science is the world of the rational, where it is supreme. It is not supreme in the world of the irrational, for instance in artistic expression, explaining myth, describing sensory qualities, creating dreams and fantasies, expressing emotions.

    Science and osteopathy

    My profession is osteopathy, and as an "alternative" medical profession it has received criticism for being unscientific and for not being "evidence based". In actual fact it is evidence-based to some extent, but it is the nature, quality and reliability of the evidence that is considered to be poor by accepted standards.

    Andrew Still always insisted that osteopathy be scientific, in contrast to what passed as medical practice in his day. However, what Still regarded as science consisted of clinical experience and the application of reason. We know today that our interpretation of our experience and our reason are both highly fallible. This is why they have been superseded by scientific method.

    Osteopathy has traditionally been a discipline based on Still's take on science, and guided by a set of fundamental "principles" enunciated by him. It's body of "knowledge" has derived in the main from pure reason, authority, belief, tradition and experience. It is heavily based on the sciences of anatomy and physiology, but beyond these it has shown a tendency to cherry-pick its science.

    The push by national health providers towards "evidence based medicine" has polarised attitudes within osteopathy. On the one hand it has encouraged osteopaths to be more science-literate and to engage in scientific research. On the other, some osteopaths firmly believe that scientific method, as currently applied to medicine, cannot answer questions about the application of osteopathy in real life.

    Personally, I remain firm in the beliefs that:

    1) Although Still's methods of demonstration would be considered highly unreliable by today's standards, nevertheless, his insistence in osteopathy's basis in science, and the requirement for claims to be demonstrable must stand.

    2) Methods used for drug trials are not appropriate for osteopathy.

    3) Acceptable levels of proof in osteopathy are not the same as those for drugs, because the resources available to obtain proofs are far far inferior. It would foolish to refuse a potentially highly useful on the grounds of impossibility to achieve a "gold standard" developed to fit an entirely different context.

    4) Science is powerful enough and scientists ingenious enough for methods to be developed to validly and effectively investigate complex interventions such as osteopathy.

    5) Osteopathy should be informed by science, not enslaved to it. Lack of proof is not proof of lack:  if extensive, appropriate, good-quality studies are lacking, methods validated by experience cannot be consigned to the scrapheap. However, if such solid science has consistently failed to demonstrate an effect, those methods should be abandoned in the contexts in which they are failing.

    6) Two things that are doing no end of harm to the osteopathic cause are (a) dogma, and (b) the proliferation of unsubstantiated theorising.

    No comments: