The art of knowing.

Knowledge. Information is out there in the world in massive amounts. This is not knowledge, but it is potential knowledge. This potential knowledge is in the objective world of reality, which all knowledge seeks to describe. There is also real and vast knowledge in this objective reality, which has been invented and tested by our ancestors, and collected for us to draw upon over thousands of years. Our problem as learners is a matter of somehow transferring the knowledge and potential knowledge in this objective world into the subjective world within our minds. In this way the information and knowledge becomes our knowledge, as it is understood by us. But how do we make this transference? We can invent our own theories and and test them ourselves, but this is a very laborious way of acquiring knowledge. We can also take the theories presented to us by others and test them, but this too would be unimaginably laborious. If we had to rely on testing every theory ourselves we would actually have time to learn very little. We would most likely still be cave men and unable to progress further.

This, of course, is not the way we normally learn. Most of what we learn, is a matter of assimilating and accommodating the theories of others, into our personal models of reality, without testing them. The philosopher Karl Popper calls this the transferring of world 3 knowledge to world 2; World 3 being our cultural heritage, the body of information (knowledge?) held in common in books computers and other forms of media; while world 2 is the private and subjective world of knowledge within each individual.

Every moment of our lives we are presented with information by (so called) experts or authorities. The question is how can we know if this information is correct so we can decide whether to accept it into our model of reality or not? The problem is how can we trust the theories of others sufficiently to be willing to include them in our own personal models of reality? How can we trust this information coming from outside ourselves and untested by us? How can we discern what is true in world 3 so we can incorporate into our own personal maps of reality?

Now you might think the answer to this is both easy and obvious. But this is not the case. Do you think some information is likely to be true because it is presented in a newspaper? Well, you might be a bit skeptical of what is printed in newspapers. What about a prestigious scientific journal? Can you simply accept information because it appears in a prestigious scientific journal? If you think you can you would be dead wrong. But if we can't trust in scientific journals what can we trust in?

The short answer is that we can't trust any authority, anyone or anything. How then do we go about choosing what to accept into our personal models of reality? Well there is both good and bad news there.

In his book "Nonsense on Stilts" Massimo Pigiucci Quotes from the work of Alvin I Goldman to provide a novice with some some way of assessing the worth of a 'so called expert'. He tells us there are five criteria a novice can use to assess the worth of a possible expert as follows: 

The five kinds of evidence that a novice can use to determine whether someone is a trustworthy expert are:

1. an examination of the arguments presented by the expert and his rival(s);

2. evidence of agreement by other experts;

3. some independent evidence that the expert is, indeed, an expert;

4. an investigation into what biases the expert may have concerning the question at hand;

5. the track record of the expert.

Consonance, the best way to know. One way a novice can choose is on the basis of how well the new information fits with the rest of what he/she knows. A novice can judge the validity of the new information by how well it fits together with all the information in his/her personal model of reality. Of course this kind of validity is only reliable if the novice already knows a lot about the subject matter in question. The more a novice knows about the subject matter, the more he/she has knowledge of the field, the better he/she can judge if the new information is likely to be correct or not. In other words in order to judge if new information in a particular subject field is likely to be correct or not, a novice basically has to be an expert.

A novice can say to himself, 'Is this new information consonant with what I already know, is it even expected from what I already know?' This is most easily accomplished by a novice when the so called expert is not an expert at all and the novice may actually have some knowledge that can easily catch the impostor out. Of course the information a novice has already accepted into his/her personal model of reality, is only a better way to judge the truth of new information, if it is comprehensive, internally consistent and (well) correct itself. If what a novice knows/believes is already wrong, then of course, information may appear consonant with it, that is also wrong. In this way people can build models of reality that are in fact wrong.

In line with Goldman's first criterion, if a novice is adept at using logic, he/she can examine an expert's position in other ways. A novice can examine the opposition to our expert and assess how congruently that opposition also fits with the novice's model of reality. If the opposition to this expert fails badly in this regard this lends some credence to the expert. Likewise if the opposition succeeds in being consonant with the novice's model of reality, then this should decrease his/her confidence in the said expert.

There is also area where the position of an expert can be examined by use of logic and that is by assessing whether the expert's position is itself internally consistent. Of course if the position of the expert is internally consistent this does not make it right but if it is inconsistent then we can certainly infer that that it is wrong. This too may be applied when examining the opposition.    

The judgment of other experts. If a novice does not have a lot of information already accumulated on a subject in his/her personal model of reality, it becomes much harder to judge whether new information is expert or not. The novice finds that he/she has to fall back on the judgment of other experts. But experts vary in their expertise and in their reliability. The problem then becomes, "How can we know which experts to trust?

While it is possible that the majority of experts in a particular field can be very wrong, and in fact are every time there is a huge change happening in their field, this in no way devalues the statistical relevance of an expert being supported by the majority of fellow experts in his field of endeavor. Support from a large number of fellow experts in the same field of expertise has to be a strong form of inference of expertise, if that field of endeavor is recognized by the scientific community as a whole. In other words, while it is always possible to find someone with a PhD who will support some pseudo science, it is very unlikely that the majority of PhDs, in a scientifically accredited field, will support such things. The agreement of large numbers of accredited others in a field of endeavor must be significant when a novice is assessing the worth of a possible expert.

Independent evidence that the expert is, indeed, an expert. The individual scientist may have credentials but these credentials have to be the right kind of credentials. Not all PhDs have the same value when an expert is being assessed by a novice. For a start credentials in a different field from that, in which the expert is supposed to be expert, do not count. Also, while many universities only hand out PhDs to people who really know their stuff, other universities hand out degrees rather easily to people who have learned very little, and some so called universities are themselves unaccredited and bogus. Finally, when examining a degree to assess trustworthiness of an expert, a novice should also assess whether the field itself is accepted by the scientific community in general. Having a PhD in a field such as bible studies that is not acceptable as a science denies it any value for the novice assessing its scientific trustworthiness.

In assessing an expert in terms of his degrees a novice should therefore, check if the degree is from a respectable university that has the reputation of not handing out degrees easily. A novice should also check if the degree is in a subject that is held to a strict scientific methodology of a science acceptable to the general scientific community. Finally a novice should check if the degree, the so called expert has obtained, is in the subject for which he/she is supposed to be an expert.           

The biases of the expert. Bias is a strange phenomenon. While obvious biasing elements in the life of an expert, such as where his/her funding is coming from, should indeed, give a novice pause in assessing the trustworthiness of the so called expert, such elements are in no way a conclusive argument that the expert is even biased. On the other hand, certain institutions (such as some think tanks) are notorious for only employing people, who are in fact biased toward the think tank's agenda. When assessing an expert in terms of his biases a novice should both consider the number of biasing elements in the expert's life and whether those biasing elements themselves do or do not have a reputation for having biased (so called) experts in their pockets.    

The track record of the expert. The track record of a proposed expert is of course the amount of times he has been right in the past. Sometimes this is easy to judge where his expertise has been called upon several times in the past and the advice given or work done on those occasions proved effective. For a novice, it can be quite difficult for him/her to assess if an expert has been successful in the past or not, in any particular area of knowledge. Peers can assess one another's successes by means of how often their work is sited in other research. For a novice this is not practical. Thus novices have to be content with peer reviews of past work which they can easily confuse with popularity in news reviews. A Nobel prize in the field would of course would be a good indicator of track record, but how many experts have those?

The trustworthiness of the expert is not everything. It's not enough to apply the above criteria when assessing whether the information given by someone claiming to be an expert is trustworthy information. Knowing about whether an expert is an expert or not, even if such an assessment could be accurate, does not in itself validate the trustworthiness of any information, although it is certainly a good start.

Approaching truth. Some people think that science tells us what is true or that it should. But this is not possible because science is continually in a state of improving its theories. Science is continually amending, shoring up, rearranging, tearing apart and even completely axing old theories and replacing them with newer ones that are superior in what they can predict and explain. The most we can say is that the new theories are nearer to the truth than the older ones and that while science approaches closer and closer to the truth, it can either; never reach a final truth because it is beyond the limits of human ability; or if it does reach a final truth we would have no way of knowing that it had. In the book  "This Will Make You Smarter" Carlo Rovelli explains it:

"There is a widely held notion that does plenty of damage; the notion of 'scientifically proved.' Nearly an oxymoron. The very foundation of science is to keep the door open to doubt. Precisely because we keep questioning everything, especially our own premises, we are always ready to improve our knowledge. Therefore a good scientist is never 'certain.' Lack of certainty is precisely what makes conclusions more reliable than the conclusions of those who are certain, because the good scientist will be ready to shift to a different point of view if better evidence or novel arguments emerge."

Pseudo science and faith based explanations. Although science cannot tell us what is true it can demonstrate its superiority to pseudo science and the supernatural in terms of explanation and prediction. Science has three criteria that separate it from pseudo science.

Naturalism. Firstly, science takes as given, that there is a natural cause for any effect or outcome. As far as science is concerned supernatural causes are not causes at all but rather admissions that we do not have a clue. Instead of saying we have no idea how or why something happened, we excuse our ignorance by saying god got personally involved. Pseudo science and the supernatural do not require a natural explanation, although they may sometimes give one.

Theory. Secondly, science requires that any explanation must be in the form of a theory that explains observable events and is internally consistent. In his book "Nonsense on Stilts" Massimo Pigliucci explains it as follows:

"The presence of coherent conceptual constructs in the form of theories and hypotheses is also a necessary component of science. Science is not just a collection of facts, as Francis Bacon thought. Theories are creative productions of the human mind and reflect our best attempts at making sense of the world as it is." 

Empiricism. Thirdly, science requires that a theoretical hypothesis must be empirically testable. This is the main thing that science does that makes it superior to any other kind of explanation. All scientific theories are not only tested in the research that produces them as results, but all testing must be open for others to duplicate, and is indeed duplicated before any kind of acceptance can begin. On the other hand pseudo science relies on personal testimony, coincidental occurrences and just plain faith. Unfortunately the personal testimony of a friend tends, in many minds, to be more persuasive than scientific research. However, the beauty of scientific research is that with the right equipment and expertise anyone can duplicate the research and check for themselves. This process is the backbone of science and is called empirical testing. When certain fields of science fail to do this, they can drift into being pseudo science, as were the cases of cold fusion and eugenics.  

Good science and bad science. Unfortunately scientists are human beings, with all the fallibilities of humans, and in evaluating the trustworthiness of scientific research a novice needs more and better ways of evaluating expertise than just examining the alleged experts. In this regard we need to look at ways a novice can distinguish good science from bad science and how and why good science can show its superiority over bad science in terms of explanation and prediction.

Trusting the information taught in schools and universities. The major way important information reaches us is though the institutions of education. While the process of review for textbooks, the process of construction of curriculums, and the ability of teachers to teach from these, is far from perfect, it is still the most trustworthy expertise we will encounter in our lives. Of course not all textbooks and curriculums are equally good or correct. Almost all textbooks probably have some typos or technical errors and some subjects that are taught in colleges and universities are completely unscientific and no textbook written on that subject could be trusted as being scientifically accurate. However, despite all that, textbooks are far more likely to contain accurate information than any other sort of book.

There is a vast apparatus of experts that decide which textbooks and curriculums will be used in schools and universities. There is also a massive checking apparatus of experts, teachers and students that continually vetting and vetoing the choices of the original experts. In fact the information that comes to us through schools is so well checked and rechecked and held back till experts are absolutely sure, that it actually out of date by the time it reaches the students. Later in university when students tend to get more up to date information there is far less likelihood of the information being correct. It is more up to date but also more speculative and cutting edge and thus there is a good possibility the information might be wrong. Indeed at that level students can be exposed to several studies and may contradict one another. Students are then learning, not facts, theories or even how to find such, but rather learning what tentative theories might be worth checking in the student's own research studies. Perhaps the best thing about schools and the institutions of education is how well they perform this one function.

The cosmic joke. It is through the understanding of knowledge that we try to predict what will happen in the universe and thus have some control over it. Humans have come a long way in being able to do this. We have invented a principle that ensures we will be able to do this called cause and effect. If A then B. If A occurs then so will B or if you do A, B will occur. Unfortunately, if we stop to think about this at all, we know that this cannot be all there is to cause and effect. Just because B follows A does not indicate there is cause and effect taking place. While a cause must precede an effect the condition of two events following one another in itself is insufficient to infer cause and effect. This is called the post-hoc fallacy. In his book "Everything is Obvious Once You Know the Answer" Duncan Watts puts it like this:

"If a billiard ball starts to move before it is struck by another billiard ball, something else must have caused it to move. Conversely, if we feel the wind blow and only then see the branches of a nearby tree begin to sway, we feel safe concluding that it was the wind that caused the movement. All this is fine but just because B follows A doesn't mean that A has caused B. If you hear a bird sing or see a cat walk along a wall, and then see the branches start to move, you probably don't conclude that either the bird or the cat is causing the branches to move."    

The trick is of course to some how distinguish when there is a causal relation and separate those from the others. Sometimes the principle of cause and effect works beautifully and we can predict what will happen to a very fine degree. Other times its usefulness is imprecise or nonexistent.

In a sense, this knowledge of the limits of knowledge, this knowledge of where our knowledge ends, is the most important knowledge we have. When we know, what we do not know, all manner of things become open to us. We start to see what questions to ask to find new knowledge and we begin to understand that it is better to do nothing than act on a delusion where we think we know something but actually do not.

It is important knowledge, but it is also a cosmic joke. Unfortunately the universe simply refuses to be completely open about what is happening at any one time, thus we can incorrectly perceive causation where there is none or little. The universe seems to play tricks on us. We try to find causation by discovering patterns where phenomenon always occur together. We call this correlation. Unfortunately many phenomenon that occur together do so randomly and there is no causal connection between them. Even stranger there are occurrences of correlation called coincidence where there seems to be, or that there should be a causal connection, but there is none. On top of that, when causation does work, it usually only works statistically. In other words there is only a probability that if A occurs that B will follow. B usually follows A but there are a few occasions when it doesn't. These may be causal exceptions or they may simply be what science calls outliers. Outliers are not exceptions from the causal rule, but are simply that part of the probability that did not conform.

All this considered we should be very careful not to accept authorities and experts just because that's what they are. Charlatans have been using this feature of the universe to convince us of things that were very wrong, for as long as man had walked the earth. David H. Freedman in his book "Wrong" calls this kind bamboozling the "Hitchcock effect" after a Hitchcock story that seemed to typify how it worked as follows:

"That story revolved around a man who receives a series of mailed predictions that all prove correct, at which point he is ready to trust the infallible predictor with his money - but as it turns out, the predictor had started off mailing various predictions to a large number of people, then focused each subsequent mailing on the increasingly smaller subset of people who had received only the predictions that happened to prove correct, until he had one victim who had by pure chance received all the winning predictions. It sounds like a far-fetched scheme, but in fact we often pick our leading experts this way - that is, we look back to see which expert among a field of many happened to call it right and then hold up that expert as having special insight. But we need to remember that if there are many experts predicting many things, some of those predictions will have to prove right, though it may be entirely a matter of luck or even bad judgment."

"There are also non-crowd-related variations on this theme. Any one expert may be able to sort through her long history of various predictions and find the few that proved correct, holding these up to our attention while glossing over the others. Or an expert can make somewhat vague predictions, fortune-teller-style and then sharpen them after the fact to make the predictions seem highly accurate."

We do not even need this to be a scam. If we believe, we tend not to notice when a predictor is wrong, but we are primed to notice when the predictor is correct. When an expert asks us to believe they are stacking the cards so we will only notice when they are right.

Karl Popper and science. Karl Popper shows that when we are learning we, first conjecture a possible solution to a problem. Then we either actively test it, as a scientist would, or accept it till the events of life seem to corroborate it or refute it. Popper not only explains that this is the way we must learn, but that other ways of trying to learn, if we try to adhere to them, such as induction, are inefficient and can lead us far away from the truth. For this reason Popper feels that following his understanding of learning as a principle is the most efficient way of conducting science. Popper explains this as follows in his book "Unended Quest":

"I suggested that all scientific discussion start with a problem (P1) to which we offer some sort of tentative solution - a tentative theory (TT); this theory is then criticized, in an attempt at error elimination (EE); and as in the case of dialectic, this process renews itself: the theory and its critical revision give rise to new problems (P2) Later, I condensed this into the following schema:
P1 -> TT -> EE -> P2, a schema which I often used in lectures."

Popper shows that in science each individual scientist creates and revises conjecture from analysis of their own sensory input. However, he also shows that the same sensory input can only be perceived through the lens of existing theory. Thus it follows that, without theories about how the world works, incoming sensory data is meaningless.

Popper further shows that no amount of corroboration can ever validate a theory, but that a single instance of deviation can invalidate it. Not only that, but the amount of corroboration does not even improve the statistical probability that scientists are correct. Methodologically theories can not be refuted unless scientist formulate them in such a way as to show just how they could be refuted. Scientists should therefore as an article of method not try to evade refutation. Instead scientists should formulate and present their theories as unambiguously and clearly as possible, to invite refutation. Thus a truly scientific theory, is one that openly exposes itself to testing.

Of course Popper's schema for learning is by no means universal in its use as the base method for doing scientific research. Indeed scientific research is often conducted in ways that seem very far from this ideal. Just how far science deviates from this ideal has been exhaustively catalogued by David H. Freedman in his book "Wrong". As Popper suggests, good science should be about inventing theories and testing them by trying to prove that they are wrong. One would expect that most of such theories would turn out to be wrong, just as a probability statistic, but that would be good science. However, the wrongness being considered here is, "How good is actual scientific testing, and can it and its experts be trusted?" Freedman does not suggest or intend to imply that science is bad, but rather he wishes to alert us to stop and think before rushing to accept new information from experts. He intends only to advise us that there is good reason to be skeptical of individual scientific papers. Here are some facts from his book that should give us pause before we accept any new research as gospel.

1. Fallibility. All scientists are fallible and can make genuinely innocent methodological errors.

2. Sloppiness. Some scientists do sloppy work that has little value.

3. Bias. Some scientists (perhaps many) are led astray by their own personal beliefs.

4. Intimidation. Some scientists are pressured into misrepresenting data by those who have power over them.

5. Corruption. Some scientists are corrupt and fake their work to increase or maintain their positions, increase their prestige, or support their expensive tastes.

6. Limits. A great deal of testing is itself flawed.

Fallibility. All scientists are fallible and can make genuinely innocent methodological errors. In his book David H. Freedman tells a story about research undertaken by Albert Einstein and Wander Johannes de Haas who measured something called the g-factor (how much an iron bar would twist in a magnetic field). They conjectured that the g-factor should be precisely 1 for each atom. After about a year of fine tuning they were able to measure a result of 1.02 using highly sensitive instruments. Unfortunately their measurements were way off and subsequent experiments consistently produced a g-factor of about twice that value. The point is that Einstein and de Haas had simply measured incorrectly. If someone like Einstein can make an error in measurement then surely any scientist could make such an error.       

Sloppiness. Some scientists do sloppy work that has little value. There are many and various ways for scientists to be sloppy with data.

1. Mismeasuring. Like Einstein all scientist can and do make mistakes in measurement, most of them much more sloppy that of Einstein. According to Freedman, scientists have been known to misread blood pressure, height and heart rhythms and have given wrong dosages and even wrong drugs. They have misrecorded the location of subject's homes. Even when they measure correctly, they may do so on people who do not properly represent the population such as the young, the old, alcoholics, drug abusers, illegal immigrants and the homeless.

   

2. Surrogate measurements. The most common way for scientists to be sloppy is by making surrogate or proxy measurements. Surrogate measurements are those where you attempt to discover changes in one thing by measuring something else. Unless a long standing cause and effect has been shown between the two things one cannot assume it to be so. This however happens very often in research for reasons of convenience. We study animals instead of humans because applying research to humans might damage or even kill them. We study the flow of blood in the brain with functional magnetic resonance imaging to try and discover what is happening in the brain. This fMRI does not tell us directly what is happening in the brain which would require cutting the brain open, but rather tells us how much blood is flowing where in the brain. It might be telling us what is going on in the brain or it might not. It is convenient not to have to cut brains open to experiment, however.

       

3. Tossing out or ignoring data. One way is to draw a line in the wrong place between data that is bad or contaminated and data that is inconvenient. One can throw out relevant data that is inconvenient. This can be done intentionally or excusably accidentally or simply carelessly as in sloppy work. There are many ways of tossing out data. If data is truly contaminated that section of the data should be done again not simply ignored or left out. Another way of tossing out data is to simply fail to submit the whole of the research for publication. If the research disproves what the researcher set out to prove, the negative finding may well be more important. 

    

4. Moving the goalposts. Another way to be sloppy is an exercise in self deception where the scientist seeks to discover some positive finding in the research after the research has refuted the theory it had set out to prove or disprove. The research is then presented as if it had set out to prove or disprove that positive finding. This is akin to footballers moving the goalposts after the ball has been kicked in order to insure the ball goes through them. In research misconduct terms this is referred to as using a retrospectoscope.

   

5. Correlation. Just because one factor, behavior or condition correlates with another does not necessarily imply there is a causal relation between them. Correlations have the same validity, whether they are presented as the findings of science, or as random occurrences, or as coincidental patterns that we call superstitions. If two factors always or nearly always occur together it is possible one may be the cause of the other. There are however other possibilities:

   

   1. One other possibility is that there is a causal relation between the two but that it is only a part cause. The fact is that there can be multiple causes. The syndrome known as schizophrenia was thought at one time to have been solely caused by a dynamic that occurs in families called a double bind. These days psychologists are more inclined to a theory that schizophrenia is caused by genetic programming. However, it is more likely that both of these phenomena may be part causes and that there may be other cases as well.

   2. Also another possibility is that the relation between the two is not causal but simply a predisposition for something to occur. They may simply be risk factors. To add to the confusion, it is possible that a number of different predispositions, while each on its own is not causal, may if acting together, become a cause.

   3. Yet another possibility is that the correlated factors may both be caused by a third factor. In his book David H. Freedman provides an example: "It may be true that a lack of sleep is linked in some way with obesity, but it's a big jump from there to conclude that if someone starts getting more sleep, they'll lose weight. It may be, for example that people who sleep less also loosely tend to be people who exercise less, or eat less healthfully, or have a hormone disorder, or are depressed - in which case it could be any of these factors, rather than sleep levels, that needs to addressed in order to affect obesity. That would mean the link to sleep is pretty much incidental, mostly useless, and misleading."

   4. It is also possible that we may mistake the direction of flow of causation. What we think is a cause may be an effect and what we think is an effect may be the cause. There is a clear correlation between what people eat and how fat they are. We believe that eating a lot of food or fatty food causes people to get fat. But if we look at what fat people eat we will probable find they eat lean food and diet food. Causation is flowing the other way. The fact that people are fat causes them to eat lean and diet food.

   5. There is even the possibility that causation might flow in both directions. We call this a chain reaction. A cause creates an effect but that effect becomes the cause of other effects which turn cause even more effects and so on. 

   6. Finally there is a possibility is that the correlated phenomena simply occur together by chance without there being any causal link. 

   

   David H. Freedman has this to say about it in his book "Wrong":

   "We hear about these 'people who do this are most likely to do that' studies all the time... But they're among the most frequently misleading of all research studies, and for a simple reason: so many interconnected things are going on in people's lives that its often nearly impossible to to reliably determine that one factor is the main cause of some behavior, condition or achievement.      

Bias. Some scientists (perhaps most) are led astray by their own personal beliefs. We each perceive the world through the lens of our beliefs. It colors what we see and all the data from our other senses. It determines how we analyze that data and and the meaning we give to that data. Both Popper and Kelly focus on how this is both essential to, and causes problems of accuracy for learning. David H. Freedman in his book "Wrong" comments on the work of Thomas Khun who arrived at the same sort understanding:

"Thomas Khun, the MIT science historian who famously gave the world the phrase 'paradigm shift', argued in the early 1960s that what scientists choose to measure, how they measure it, which measurements they keep, and what they conclude from them are all shaped by their own and their colleague's ideas and beliefs."

How is it that, while most scientists fall into the many traps of bias, some scientists seem to excel at being fairly unbiased and manage to come up with useful findings again and again? Such people could be said to have good biases. Jack Cuzick of Cancer Research UK says, "Some people have a good nose for sniffing out right answers." This does not help the non expert in deciding who to trust because to the non expert, good biases and bad biases look exactly the same. It is likely that biases can be mitigated somewhat if scientists formulate and present their theories as unambiguously and clearly as possible, to invite refutation, and then conscientiously attempt to disprove them. Of course if biases are in fact good biases they will inevitably produce good results more than once and probably often.

Intimidation. Some scientists are pressured into misrepresenting data by those who have power over them. Some of this is just common sense. If a scientist works for a company and makes a finding that is not in the company's interest the company is obviously going to put pressure on them not to publish. If a set of data tends to put the company in a poor light that company is going to pressure the researcher to toss out or ignore that set of data. Imagine scientists at tobacco companies reporting finding that tobacco could be a factor in causing cancer.

Whistle blowers are not appreciated or tolerated in most walks of life and this is very true of science. David H. Freedman commented about this kind of pressure in his book "Wrong":

"Gerald Koocher, the Simmons College dean who studies research misconduct, has gathered online more than two thousand anonymous accounts of research misconduct that wasn't otherwise reported. 'I wasn't surprised when I got a lot of people saying, 'I was afraid my boss would fire me if I blew the whistle on what he was doing,'' he says. 'I was more surprised to get people saying, I caught my research assistant fabricating data, so we fired them or moved them out of the lab, but we didn't report it because we were afraid our grant money wouldn't be renewed.'

...Nicholas Steneck, the ORI researcher, confirms the plentiful evidence showing the reluctance to report misconduct. 'Almost every time I speak to a group, at least one or two students or young researchers will come up to me afterward and say, 'This is what's going on. What should I do?'' he told me. 'Or they'll say, 'I'm not going to do anything about it until after I leave the lab' - but why would they report it after they've? It's almost signing your own career death warrant to blow the whistle."

Corruption. Some scientists are corrupt and fake their work to increase or maintain their positions, increase their prestige, or support their expensive tastes. Here is what David H. Freedman says about this in his book "Wrong": "Most of us don't like to think of scientists and other academic researchers as as cheaters. I certainly don't. What could motivate such surprisingly non trivial apparent levels of dishonesty? The answer turns out to be pretty simple: researchers need to publish impressive findings to keep their careers alive, and some seem unable to come up with those findings via honest work. Bear in mind that researchers who don't publish well regarded work typically don't get tenure and are forced out of their institutions." Also given the number of scientists in the world statistically there are likely to be quite a few who are just using science to feather their nests. There have been quite a few famous hoaxes uncovered, where data and evidence have been knowingly and intentionally tampered with, by famous scientists. One famous incident was the the story of the missing link.

The Piltdown Man (often referred to as the missing link) is a famous Anthropological hoax concerning the supposed finding of the remains of a previously unknown early human by Charles Dawson. The hoax find consisted of fragments of a skull and jawbone reportedly collected in 1912 from a gravel pit at Piltdown, a village near Uckfield, East Sussex, England. Charles Dawson claimed to have been given a fragment of the skull four years earlier by a workman at the Piltdown gravel pit. According to Dawson, workmen at the site had discovered the skull shortly before his visit and had broken it up. Revisiting the site on several occasions, Dawson found further fragments of the skull.

The significance of the specimen remained the subject of controversy until it was exposed in 1953 as a forgery. Franz Weidenreich examined the remains and correctly reported that they consisted of a modern human cranium and an orangutan jaw with filed-down teeth. Weidenreich, being an anatomist, had easily exposed the hoax for what it was. However, it took thirty years for the scientific community to concede that Weidenreich was correct.

The Piltdown hoax is perhaps the most famous paleontological hoax in history. It has been prominent for two reasons: the attention paid to the issue of human evolution, and the length of time (more than 40 years) that elapsed from its discovery to its full exposure as a forgery.

Another famous case of a scientist faking his results was the work of Woo-Suk Hwang. Woo-Suk Hwang is a South Korean veterinarian and researcher. He was a professor of theriogenology and biotechnology at Seoul National University. He became infamous for fabricating a series of experiments, which appeared in high-profile journals, in the field of stem cell research. Until November 2005, he was considered one of the pioneering experts in the field, best known for two articles published in the journal Science in 2004 and 2005 where he reported to have succeeded in creating human embryonic stem cells by cloning. Both papers were later editorially retracted after they were found to contain a large amount of fabricated data.

Hwang has admitted to various charges of fraud and on May 12, 2006, he was indicted on embezzlement and bioethics law violations linked to faked stem cell research. The Korea Times reported on June 10, 2007 that The university had expelled him (he was dismissed on March 20, 2006) and the government rescinded its financial and legal support. The government has subsequently barred Hwang from conducting human cloning research.

Limits. A great deal of testing is itself flawed. It is flawed by the limits of its very nature. David H. Freedman concludes that there are four different basic study designs that have varying degrees of trustworthiness but none of which is is completely trustworthy. They are:

1. Observational studies: These are the least trustworthy.

2. Epidemiological studies: These can be more trustworthy if large and well executed.

3. Meta-analysis or review studies: These can be even more trustworthy if carefully executed.

4. Randomized controlled trial studies: These are the most trustworthy if large and carefully conducted.

1. Observational studies: These are the least trustworthy. They consist of researchers observing how a small group of subjects respond under varying conditions. This could be physicians observing patients, technicians observing volunteer subjects, bribed subjects like criminals offered a reduced sentence, or it could be an animal study. Because the test subject samples are small and unlikely to be representative individual studies of this sort must be suspect and hardly ever conclusive. They also suffer from confounding variables and researcher bias.     
2. Epidemiological studies: These can be more trustworthy if very large and well executed. These studies involve following a large group of people (as many as tens of thousands) over months, years or even decades. Such studies suffer from two important drawbacks. Unlike the small studies where different variables can be controlled for and the subjects can be observed every minute of the test, these studies involve subjects that cannot be observed all the time and where variables cannot be properly controlled for. Such studies often have to rely on subject self reporting for their data. Also the type of research elements that such studies tend to look at, often involve such small changes that the slightest imprecise measurement can play havoc with their results. Such studies also have to be suspect and are very unlikely to be conclusive.     
3. Meta-analysis or review studies: These can be even more trustworthy if carefully executed. These studies consist of data taken from many previous studies which is combined and reanalyzed. These tend to be untrustworthy because the studies they are composed of tend not to be simply duplicate studies. If 50 people did exactly the same study the meta-analysis of those studies would be straight forward and trustworthy, but the fact of the matter is that they not exactly the same. Meta-analysis studies are usually made up of studies, that are not only trying to prove different things, but have controlled for different factors. If a study is left out because something wasn't controlled for it will distort the meta-analysis if that factor is unimportant to the findings. On the other hand if a study is included where a factor that was important to the findings was not controlled for that too will distort the meta-analysis. Also obviously it is easy to be biased or corrupt in this sort of analysis where what is included and what is left out can be of crucial importance. Also these studies can be distorted by researcher's failure to publish many studies. What is more it has been shown mathematically that meta-analysis  based on data from studies that were unreliable in themselves, while more likely to be reliable than the original studies, are still more likely to be wrong than right. So these types of studies must also be suspect and inconclusive.      
4. Randomized controlled trial studies: These are the most trustworthy if large and carefully conducted. Nevertheless randomized controlled trials, or RCTs cannot be automatically trusted simply as a matter of course. Controlled means that there are at least two groups in the study. Controlled means that there are at least two groups in the study, typically in medical trials, one of which gets the treatment under study, while the other gets a placebo. In non medical trials the second group simply experiences no intervention by the researchers while the first group does. Randomized means subjects are randomly assigned to one group or the other, to avoid confounding variables, and usually neither the subjects nor the researchers know who is in which group until all the data are gathered making it a so called double-blind study to avoid bias. David H. Freedman in his book "Wrong" tells us that RCTs can if fact go wrong in any number of ways:

   "For one thing, randomization of large pools of people does little to protect against most of the other problems with studies we've looked at, including shaky surrogate measurements, mismeasurement, unreliable self reporting, moving the goalposts, tossing out data, and bad statistical analysis. As with epidemiological studies, large RCTs often traffic in exceedingly small effects. What's more, RCT findings are usually averages for results that often vary wildly among different individuals, so that the findings don't really get at what's likely to happen to you."

Albert Einstein

The failure of journals and journalists to provide us with the truth. If scientists cannot be trusted to give us expert advice, then the journals and journalist through which their expert advice is transmitted to us is doubly suspect. David H. Freedman in his book "Wrong" explains:

"But more often the media simply draw the most resonant, provocative and colorful - and therefore most likely to be wrong - findings from a pool of journal-published research that already has a high wrongness rate.

Generally even the most highly respected science journals and their editors want to grab our attention. They want studies that are groundbreaking, shocking, and interesting. That means almost that the very least they are looking for studies that have positive findings. Why would anybody want to read about a theory that has been disproved. David H. Freedman in his book "Wrong" provides some information:

"Research by Dickersin and others suggests that on average positive studies are at least ten times more likely than negative studies to be submitted and accepted for publication. That might well mean that if one mistakenly positive study is published, on average only two of the nineteen studies that correctly ended up with negative results will be published. The seventeen others will probably go into a file draw, so to speak, or if they're submitted for publication they'll probably be rejected for having ended with negative results that simply confirmed what everyone suspected was true anyway."

Important scientific journals tend not to print negative findings. These can have very distorting impact, where theories are refuted and studies are invalidated, and nobody knows. This creates incredible waste where these theories and studies are refuted over and over again without anyone being aware that it had all been done before.

What can we believe? So if all testing is methodologically inconclusive and possibly suspect, how can we ever trust any scientific findings? There are a number of indicators that help in this, but again none which is conclusive or permanent. Indeed as Popper has shown it is impossible for science to be completely conclusive. We tend to, and indeed should tentatively accept particular scientific ideas as being true, because there is consensus in the scientific community over a reasonable long period of time. However, we should also bear in mind the words of Bertrand Russell, "Even when the experts all agree, they may well be mistaken." David H. Freedman in his book "Wrong" quotes polymath Charles Ferguson who perhaps should have the last word on listening to experts. He explains:

"The point isn't that you should always do what experts say, but rather that making giant sweeping decisions without listening to them at all is really dumb."

Perhaps laypeople simply can't be expected to know. It is kind of sad but we are badly equipped to try and determine which experts are right if we are not expert in the field ourselves. Harvard Law School professor Scott Brewer concludes, "...laypeople simply can't be expected to figure out which experts to believe, no matter what technique they employ." Evolution and our genetic disposition almost guarantees that we will tend to ignore harsh probable truths and believe comforting lies.

David H. Freedman in his book "Wrong" provides some tips for detecting the likelihood of the grosser elements of scientists and tests fallibility, and provides some indicators of scientific accuracy, thoroughness and credibility, as follows:

Typical Characteristics of Less Trustworthy Expert Advice.

1. It's simplistic, universal and definitive. We all want certainty, universality and simplicity. In his book "Wrong" David Freedman points out that if you are given a choice of following the advice of one of two doctors, we will always prefer to follow the advice of the one who seems most sure in what he is telling us. A doctor who tells us that its hard to tell exactly what's wrong and that the treatment he is recommending usually doesn't work, but does work slightly more often with people like yourself, is difficult to accept as being good advice. Despite the fact that very little in science turns out to be universal, we also tend to drawn in by claims that something is universally applicable. Like wise if something is simple to understand we are also attracted to it. If we know these features are biasing us to trust some expert advice, we can allow for this, and allocate it to a less trustworthy position.   

2. It's supported by only one study, or many small or less careful ones, or animal studies. As a rule the more studies the better the more careful the studies the better and if we are trying to learn about humans it is obviously better to conduct the tests directly on them rather than on animals. If we know these features indicate this is less trustworthy expert advice, we can allow for this, and allocate it to a less trustworthy position.

3. It's groundbreaking. As David Freedman points out, "...most expert insights that seem novel and surprising are based on a small number of less rigorous studies and often just one small or animal study. That's because big rigorous studies are almost never undertaken until several smaller ones pave the way, and if there had already been several studies backing this exciting finding, you probably would have heard about it then and it wouldn't seem so novel now. If we know novel and surprising indicate less trustworthy expert advice, we can again allow for this, and allocate the advice to being understood to be less trustworthy.

4. It's pushed by people or organizations that stand to benefit from it's acceptance. As David Freedman also points out, "All experts stand to benefit from their research winning a big audience, of course, and that's well worth remembering. That's especially true when the research is coming out of or being directly funded by individual companies or industry groups whose profits may be impacted. Corporate sponsorship doesn't mean a study is wrong, but there's simply no question it sharply raises the risk of serious bias..." If we know that promoters are of necessity biased and may even be tempted to lie, what they promote, should be understood to be untrustworthy expert advice, and we can make it so in our own minds.

5. It's geared toward preventing a future occurrence of a prominent recent failure or crisis. As David Freedman also points out, "This is the 'locking the barn door' effect: we're so irked or even traumatized by whatever has just gone wrong that we're eager to do now whatever we might have done before to have avoided the problem." Just because we are itching to follow this advice now, does not make it any better expert advice than it was before, it is in fact less trustworthy, because we want to follow it. We should allow for this, and allocate the advice to being considered by us less trustworthy.

Characteristics of Expert Advice we should ignore.

1. It's mildly resonant. You've heard the old saying that a little knowledge is a dangerous thing. When we have a little knowledge it often feels like we know a lot. Expert advice can just sound right to us, because we misjudge what we know. We think we know a lot when we know only a little. It fits with our view of the world, but we don't actually have enough information about that field to make the informed choice needed. As David Freedman points out: "...it appeals to our common sense, it's amusing, it makes life easier for us, it offers a solution to a pressing problem. Too bad none of that improves the chances of an expert conclusion being true."

2. It's provocative. Again as David Freedman points out: "We love to hear an expert turn a conventional view on it's ear: Fat is good for you! Being messy can be a good thing! We're tricked by the surprise, and at the same time it may ring true because we're so used to finding out that what we've all been led to believe is right is actually wrong. The conventional view is indeed often wrong, or at least limited, but look for good evidence before abandoning that view."

3. It gets a lot of positive attention. Yet again as David Freedman points out: "The press, the online crowd, your friends - what do they know? The coverage drawn by an expert claim usually has more to do with how skillfully it has been spun or promoted, combined with its resonance and provocativeness, rather than how trustworthy it might be."

4. Other experts embrace it. David Freedman points out that while other experts should not be completely ignored they need to be put in perspective: "...communities of experts can succumb to politics, bandwagon effects, funding related biases and other corrupting phenomena. It's also often hard for laypeople to to tell if most of the experts in a field do in fact support a particular claim - press reports may be presenting a biased  sampling of experts. ...In that light, the immediate wide support of a community of experts for a new claim might be seen as a warning sign rather than a recommendation. More trustworthy, by this reasoning, would be the support that gradually builds among experts over a longer period of time."

5. It appears in a prestigious journal. Prestigious journals do not publish research because it is likely to be thorough or correct. David Freedman quotes D. G. Altman who studies bad research practices: "There are many factors that influence publication, but the number one factor is interest in the study topic." Freedman continues: "What makes a study's results important or otherwise interesting? There are no hard-and-fast rules, but editors and researchers tend to speak of results that break new ground, or that might have impact on what other researchers study, or that have important real-world applications such as drugs for a major illness. Its also widely understood in the research community that, all things being equal, journals much prefer to publish 'positive' findings - that is, studies whose results back the study's hypothesis." After all who wants to read about theories that have been refuted. The research that is published is that which is novel, groundbreaking, provocative and positive.

6. It's supported by a big rigorous study. David Freedman points out that while big rigorous studies are generally more trustworthy this should not be allowed to distract you from the possibility that it can very easily still be very wrong: "No study, or even group of studies, comes close to giving us take-it-to-the-bank proof. When several big rigorous studies have come to the same conclusion, you'd be wise to give t serious consideration - though there may still be plenty of reason for doubt, perhaps on the grounds of publication bias (the dissenting studies may have been dropped somewhere along the line), sponsorship corruption (as when a drug company is backing all the studies to bolster a product), measurement problems (as where questionable markers are involved), flawed analysis (as when cause and effect are at risk of being confused), and more."

7. The experts backing it boast impressive credentials. David Freedman points out that: "We've seen that some of the baldest cases of fraud oozed out of Ivy League campuses, world class hospitals and legendary industrial labs - where competence and standards may be sky-high, but so are the pressures to perform, along with freedom from close oversight. If experts can cheat, they certainly can succumb to bias gamesmanship, sloppiness and error. And they do all the time." Of course it should be understood that people who have impressive credentials are generally more trustworthy than those that do not, but that in no way should incline us to accept and trust in the advice of those highly credentialed researchers. Nor should it incline us to dismiss the work of those with lesser qualifications. Even laypeople can sometimes be right when the experts are wrong. Real world observations while being less meticulous than scientific research, can sometimes be more relevant.     

Some Characteristics of More Trustworthy Expert Advice.

1. It doesn't trip the other alarms. David Freedman explains that: "Knowing now the characteristics of less trustworthy advice, we can obviously assume that expert advice not exhibiting such traits is likely to be more trustworthy. In other words we ought to give more weight to expert advice that isn't simplistic...," universal or definitive. We should find more trustworthy, that research which has been satisfactorily duplicated many times, that has the support of large careful studies, that avoids conflicts of interest, that isn't groundbreaking and that isn't a reaction to a recent crisis. 

2. It's a negative finding. David Freedman continues: "As we have seen, there is a significant bias every step of the way against findings that fail to confirm an interesting or useful hypothesis - no one is going to to stop the presses over the claim that coffee doesn't stave off Alzheimer's disease. There isn't much reason to game a disappointing conclusion, and anyone who publishes one or reports on it probably isn't overly concerned with compromising truth in order to dazzle readers."

3. It's heavy on qualifying statements. David Freedman continues: "The process by which experts come up with findings, and by which those findings make their way to the rest of us, is biased toward sweeping under the rug flaws, weaknesses and limitations. What can experts, or the journals that publish their work, or the newspapers and television shows that trumpet it, expect to gain by hitting us over the head with all the ways in which the study may have screwed up? And yet sometimes journal articles and media reports do contain comments and information intended to get us to question the reliability of the study methodology, or the data analysis, or how broadly the findings apply. Given that we should pretty much always question the reliability and applicability of expert findings, it can speak to the credibility of the experts, editors, or reporters who explicitly raise these questions, encouraging us to do the same."

4. It's candid about refutational evidence. David Freedman continues: "Claims by experts rarely stand unopposed or enjoy the support of all available data. (A saying in academia: for PhD, there's an equal and opposite PhD.) Any expert, journal editor, or reporter who takes the trouble to dig up this sort of conflicting information and highlight it when passing on to us a claim ought to get a bit more of our attention. But don't be impressed by by token skeptical quotes tossed into the media reports in the name of 'balance'; nor by brief toothless, proforma 'study limitations' sections of journal articles that listlessly toss out a few possible sources of of mild error; nor by the contradictory evidence that seems to been introduced just to provide an opportunity for shooting it down. The frustration of on-the-one-hand-but-on-the-other-hand treatments of an expert claim is that they may leave us without a clear answer, but sometimes that's exactly the right place to end up. And, once in a while, watching the negative evidence take its best shot leaves us recognizing that the positive evidence actually seems to survive it and is worth following."

5. It's provides some context for research. David Freedman continues: "Expert findings rarely emerge clear out of the blue - there is usually a history of claims and counterclaims, previous studies, arguments pro and con, alternative theories, new studies under way, and so forth. A finding that seems highly credible when presented by itself as a sort of snapshot can sometimes more clearly be seen as a distant long shot when presented against this richer background, and it's a good sign when a report provides it. Andrew Fano, a computer scientist at the giant high-tech consultancy Accenture, it to me this way: 'The trick is not to to look at expertise as it's reflected in a single, brief distillation but rather as the behavior of a group of experts over a long period of time.'"

6. It's provides perspective. David Freedman continues: "Expert claims are frequently presented in a way that makes it hard to come up with a good answer to the simple question 'What if anything, does this mean for me?' We often need help not simply in knowing the facts but also in how to think and feel about them." [What we need is some perspective about whether we should be inclined to act on it or not.] "...Such meta-findings might take the form of 'Though the effect is interesting, it's a very small one, and even if it isn't just a fluke of chance, it probably doesn't merit changing behavior'; 'Chances are slim this will apply to you.' ...Experts, journal editors, and journalists might reasonably argue that their audiences shouldn't need these sorts of reminders, that the fact should speak for themselves, that it's not their place to interject such semi-subjective commentary. Well fair enough, but I'd assert that those who go ahead and do it anyway ought to be rewarded with a higher level of trust, in that it demonstrates they're willing to sacrifice the potential impact of an expert claim to help our the rest of us in knowing what to make of it. The need for such explanation is particularly acute when research involves statements about about probabilities and risk, which the public has a terrible time interpreting (as do many experts)."  

Why do we want to believe the experts? There seems to be some evolutionary value in having special people who we understand to be more knowledgeable than ourselves in particular areas of knowledge. This is obviously an advantage to humans as a species. The problem is not the experts nor our willingness to believe they may know more than us about something. It rather the uncritical and un-skeptical way in which we tend to accept what they have to say. It is the the way we deify experts and grant them the certainty of almost omniscient powers of foresight. David H. Freedman in his book "Wrong" calls this the "Wizard of Oz effect":

"We're brought up under the spell of what might be called the 'Wizard of Oz effect' - starting with our parents, and then on to teachers, and then to the authoritative voices our teachers introduce us to in textbooks, and then to mass experts whose words we see our parents hanging on in the newspapers and on TV, we're progressively steeped throughout our upbringing in the notion that there are people in the world who know much, much more than we do, and that we ought to take their word for whatever it is they say is so."

What we really need, is not to simply accept expert advice because it is expert, but rather use it as a stepping off point for investigating further. There is some evidence that children that have grown up with the Internet may be more open to this skeptical approach.

Expert opinion and this site. On this site a great deal of expert opinion has been presented. Just as David H. Freedman says of his book "Wrong" this site must admit that a balanced view covering all points of view has not been presented or attempted here. Also it is likely that this site is riddled with factual and conceptual errors of which we have no awareness. Perhaps some errors have snuck in because something has been misread or misunderstood. This site is guilty of all the things that Freedman complains about. However, whatever errors and biases have crept into this site it seems likely that they are of insufficient magnitude to compromise the overall arguments of this site. The arguments and messages of this site can be taken to have good possibility of being valid on the grounds that this is a massive amount of expert opinion, that is all in basic agreement, and that this agreement has taken place over a vast amount of time. These experts agree with the basic premise of this site and have done so over a long period of time.

Knowing. Popper's schema for learning has other consequences for knowing. Not only is it difficult to find ways of trusting that knowledge is correct, but the tentativeness of knowledge implied by Popper's ideas makes it difficult to state that you know something. The things that we truly know cannot include theories that can be replaced by other theories at any moment. The mere fact that things have always happened does not make them certain. We believe the sun will rise in the East and set in the West every day, but there are unlikely circumstances in which this would not be so. At the North and South poles the sun does not rise or set. Also, some astronomical calamity could knock earth out of orbit etc. What we know is limited to universal agreement, in that it is how concepts are defined. We know that 1 + 1 = 2 because 2 is defined as 1 + 1. We know the rules of logic hold true and that we can use deduction to arrive at truth. However, these truths are also open to question because they always have to include some given truth. For instance it is a given that, 'All men are mortal', though that is true only until we discover or create an immortal man. So in the end what we 'know' is very little. Despite this we live in a world that is very predictable. There is a high probability that most of what we believe to be true is true or will occur as we believe.       

In his book "On Being Certain" Robert Burton suggests that knowing has little to do with the logic of error elimination or with careful observation and is simply an emotional feeling that occurs because evolution has found it advantageous to reward some simple associations. Burton says:

"The message at the heart of this book is that the feelings of knowing, correctness, conviction, and certainty aren't deliberate conclusions and conscious choices. They are mental sensations that happen to us."  

Certainty. In the book "This Will Make You Smarter" Lawrence Krauss points out:

"The notion of uncertainty is perhaps the least well understood concept in science. In the public parlance, uncertainty is a bad thing, implying a lack of rigor and predictability. The fact that global warming estimates are uncertain, for example, has been used by many to argue against any action at the present time.

In fact, however, uncertainty is a central component of what makes science successful. Being able to quantify uncertainty and incorporate it into models is what makes science quantitative rather than qualitative. Indeed, no number, no measurement, no observable in science is exact. Quoting numbers without attaching an uncertainty to them implies that they have, in essence, no meaning."

Kathryn Schulz in the same book brings to our attention the startling idea that not only is scientific knowledge uncertain but that all knowledge is uncertain. The title of her contribution is "The Pessimistic Meta-induction from the History of Science." What does that mean? She explains:

"Here's the gist: Because so many scientific theories from bygone eras have turned out to be wrong, we must assume that most of todays theories will eventually prove incorrect as well. And what goes for science goes in general. Politics, economics, technology, law, religion, medicine, child rearing, education: No matter the domain of life, one generation's verities so often become the next generation's falsehoods that we might as well have a pessimistic meta-induction from the history of everything.

Good scientists understand this. They recognize that they are part of a long process of approximation. They know they are constructing models rather than revealing reality. They are comfortable working under conditions of uncertainty - not just the local uncertainty of 'Will this data bear out my hypothesis?' but the sweeping uncertainty of simultaneously pursuing and being cut off from absolute truth.

So knowledge in the end is not what is 'correct' or 'true' but rather what highly likely of being correct as far as we know so far. Knowing as 'certainty', for the most part, is an illusion. Very little is certain and the best theories in science are just that, theories. Our knowledge is about what has been found to give the best approximation of reality and it can be superseded at any time. It is often what some call working knowledge. It is what gets the job done, what is good enough to do what we are trying to do. Good knowledge then is always about probabilities and not about certainties. In the book  "This Will Make You Smarter" Carlo Rovelli explains further:

"Every knowledge, even the most solid, carries a margin of uncertainty. (I am very sure what my name is ... but what if I just hit my head and got momentarily confused?) Knowledge itself is probabilistic in nature. a notion emphasized by some currents of philosophical pragmatism. A better understanding of the meaning of 'probability' - and especially realizing that we don't need (and never possess) 'scientifically proved' facts but only a sufficiently high degree of probability in order to make decisions - would improve everybody's conceptual toolkit. 

However, human beings are drawn to certainty and away from ambiguities, inconsistencies and probabilities. We want the doctor diagnosing what ails us to seem certain and not tell us about probabilities. Why humans are this way may be genetically or socially transmitted. In his book "On Being Certain" Robert Burton suggests that this desire could well be socially transmitted. He says:

"I cannot help wondering if an education system that promotes black and white or yes or no answers might be affecting how reward systems develop in our youth. If the fundamental thrust of education is "being correct" rather than acquiring a thoughtful awareness of ambiguities, inconsistencies and underlying paradoxes, it is easy to see how the brain reward systems might be molded to prefer certainty over open open-mindedness. To the extent that doubt is less emphasized, there will far more risk in asking tough questions. Conversely, we like rats rewarded for pressing the bar, will stick with the tried-and-true responses."

In his book "On Being Certain" Robert Burton further suggests that our proclivity to want to be certain is the cause of many of societys ills and which can and has led us down many dangerous and dark paths. He believes we should try to avoid this and suggests the following:

"Certainty is not biologically possible. We must learn (and teach our children) to tolerate the unpleasantness of uncertainty. Science has given us the language and the tools of probabilities. We have methods for analyzing and ranking opinion according to their likelihood of correctness. That is enough. We do not need and cannot afford the catastrophes born out of a belief in certainty. As David Gross Ph. D., and the 2004 recipient of the Nobel Prize in physics said 'The most important product of knowledge is ignorance.'"

What is knowledge? Knowledge is a survival strategy put in place by evolution to to give us humans the advantage we have needed in order to survive as a species. It started out as memory that could be retrieved from our large brains when we needed it. Like many mechanisms that have evolved, its purpose has changed, but unlike most mechanisms it has kept changing and may continue to change as time goes by. The invention of writing meant that knowledge could be recorded and transmitted to others without a human's memory being involved. Knowing was still important but less so than before. With the invention of printing knowledge changed again as knowledge that was previously understood by just a few, became available to many. Knowing still had importance but because knowledge was spread fairly evenly through the species knowing became even less important than before. Now in the era of the world wide web knowledge and what we understand it to be is changing again. 

Do we still need to know? As has been shown above, not only is it not possible to be certain and thus know about anything, but the assumed knowing we experience is just a feeling, an emotion, and thus an illusion. The question then arises, 'Do we really need to know anything?' On the one hand, if we think that knowing means certainty then we really no longer need to know. On the other hand if by know we mean that we believe there is a very high probability of something being true, then, 'Yes we still need to know.' But even these probability assessment of actions to outcomes are only needed at the time just previous to the action and to an extent do not need to be held in memory any more. The problem is that real knowing is not just information that can be absorbed at the time of action. What happens is, that incoming data combines with knowledge held in memory at the time of action. This knowledge held in memory is a map or model of reality and is what enables understanding. Incoming information adds to it and sometimes alters it structurally enabling the apropiate action.  

Who knows? Knowledge then, for the most part, no longer needs to known by individual people. The survival strategy of knowing has been replaced by a more important survival strategy. The new survival strategy is knowing where to find information when you need it. This strategy has been becoming more important as knowledge has changed through the centuries. It first became important with the invention of writing and more important when printing appeared. Now with computers the world wide web and mobile phones, knowledge can be assembled (found) quickly and easily any time any where. We have already built many tools for retrieving information and consolidating it into knowledge. Search engines and social networks have begun this massive undertaking but more and better tools will be forged as time goes by. 

Has knowledge become too big to know? Before the advent of computers and their processing power not only was there no possibility of processing large amounts of data but the means of gathering large amounts of data did not exist for the most part. Now the opposite is the case. Now masses of data about every conceivable variable that might be significant in any experiment exists in ever increasing quantities. The processing of this mass of information by human brains is no longer possible. The most we can now do is construct computer models and let those models run to process the information for us. In his book "Too Big to Know" David Weinberger explains:

"The problem - or at least the change - is that we humans cannot understand systems even as complex as that of a simple cell. It's not that we're awaiting some elegant theory that will snap all the details into place. The theory is well established already: Cellular systems consist of detailed interactions that can be thought of as signals and responses. But those interactions surpass in quantity and complexity the human brain's ability to comprehend them. The science of such systems requires computers to store all the details and to see how they interact. Systems biologists build computer models that replicate in software what happens when millions of pieces interact. It's a bit like predicting the weather, but with far more dependency on particular events and fewer general principals.

Models this complex - whether of cellular biology, the weather, the economy, even highway traffic - often fail us, because the world is more complex than our models can capture. But sometimes they can predict accurately how the system will behave. At their most complex these are sciences of emergence and complexity, studying properties of systems that cannot be seen by looking only at the parts, and cannot be well predicted except by looking at what happens."

"With the new database-based science, there is often no moment when the complex becomes simple enough for us to understand it. The model does not reduce to an equation that lets us then throw away the model. You have to run the simulation to see what emerges."

Some knowledge has to be stored in our brains. Despite all the above, of course, some knowledge needs to remain in our memories, in out meat sack brains. Nothing would make any sense at all without our maps or models of reality, and these are made up of the facts and theories we have acquired over a lifetime. Also, a large amount of data is necessary in each individual brain in order for those individual people to be able to take part in creative activity.

The future of knowing. Of course science will always require a fair amount of knowing that there is a high probability of some things happening in particular circumstances, and the people who are creative in the various scientific domains will need to know this knowledge most of all. Creative people need to know because if they do not have most of the knowledge in their domain the ideas they originate are less likely to be new and unique. However, as time goes by, science will become more and more about the systems that even the scientists that study them do not understand or know. Similarly the ordinary person will also need to know less about everything everything else and more and more about where to find information when they need it.        

"I can live with doubt and uncertainty and not knowing I have approximate answers and possible beliefs and different degrees of certainty about different things...it doesn't frighten me." Nobel  laureate Richard Feynman  

Our changing understanding of what knowing is. With the gathering importance of systems research 'knowledge' or our understanding of what knowledge is may be in the process of changing and with it the very idea of knowing. When people speak of knowing in the future they may mean something far different to what people have meant by it in the past.