Recently I was interested to learn more about the influence of philosophy on psychiatry. Since psychiatry is a branch of medicine which in turn is a branch of science I reasoned that the philosophy of science would be a reasonable starting point. Of course, practically speaking there are many other influences on psychiatry other than science alone but I would consider this to be an important aspect of the core of psychiatry. There are a number of branches of science that are important for psychiatry and this further adds to the potential benefits of an understanding of the philosophy of science. My initial and casual inquiries led me to ‘Against Method’ by Paul Feyerabend. Upon reading this I was moved to write a response and although I have no training in philosophy, I found that the issues that Feyerabend discussed were relevant to psychiatry but I have disagreed with what Feyerabend has written in a good many places. As Feyerabend has written on historical areas in order to draw conclusions about science it has been necessary for me to step into areas with which I am not familiar in order to more fully understand and discuss his arguments.
Preface to the Chinese Version
In the preface to the Chinese version, Feyerabend writes that
‘First world science is one science among many; by claiming to be more it ceases to be an instrument of research and turns into a (political) pressure group…’My main motive in writing the book was humanitarian, not intellectual…..The ‘progress of knowledge and civilization’ – as the process of pushing Western ways and values into all corners of the globe is being called – destroyed these wonderful products of human ingenuity and compassion without a single glance in their direction‘
I must note at this point that Feyerabend does mention the word science at the beginning of the above quotation but he then talks about something else. What he does appear to be describing is a (mono)culture and I think there is a risk that without explicit statement science and culture can be confused in his argument. If science is a process by which a deeper truth about the universe can be obtained then it follows that if western or eastern sciences are different at some point, if as they progress they better approximate the truth, then they should converge. Similarly if we are discussing different ways of describing the world – scientific and alternative – then again when they are describing the same subject there should be some analogy between the bodies of knowledge accumulated in both systems.
When writing about a revival of other traditions, Feyerabend writes
‘I have tried to show, by an analysis of the apparently hardest parts of science, the natural sciences, that science, properly understood, has no argument against such a procedure‘
Feyerabend does not qualify what he means by ‘apparently hardest parts of science, the natural sciences’ which is curious as later in the book he suggests that science is not a unitary whole. Presumably he means that the hardest parts of science are those which concern themselves with the simplest phenomenon whereby there is the least chance of error in conclusions drawn and predictions made. Nevertheless it would have been better had he qualified this.
In the introduction, Feyerabend argues that history is much richer than the history conveyed by historians and methodologists
‘Are we really to believe that the naive and simple-minded rules which methodologist take as their guide are capable of accounting for such a ‘maze of interactions‘
This however is as true of history as it is of science. The only place where simple rules are relevant is in an abstract environment governed by such rules – a game of chess for example. It could be argued that simple ‘closed systems’ such as a heating system might be a good example of a system in which simple rules apply. But even here the laws of thermodynamics would not apply on very close inspection as this is not truly a closed system and there will be some degree of exchange of heat energy with the surrounding environment no matter how small. He then outlines the beginning of an anachistic science.
In this chapter, Feyerabend writes that
‘The only principle that does not inhibit progress is anything goes‘
He argues that notable developments in science occurred because the rules of methodology were flouted. He suggests quite reasonably that there are times when rules should be ignored. In the external world, simple rule based systems have a limited applicability as reality is much more complex as argued above. In some situations they are extremely useful and have good predictive value while in other areas the reality is dynamic and the interactions incalculable. A good example of this is a snooker table. If a ball is hit into another ball, approximations can be made and reasonable predictions made. As soon as the cue ball is hit off one cushion and into a cluster of 3 balls the results become unpredictable. The laws of physics are useful in very simple situations but as soon as the situations become more complex these laws are much less useful and the laws from other branches of science become much more successful than the laws of physics. Feyerabend then suggests that arguments hinder progress and that alternatives to arguments are much more useful. In this regards he suggests that children do not learn by argument and adults should learn from their example:-
‘We certainly cannot take it for granted that what is possible for a small child – to acquire new modes of behaviour on the slightest provocation, to slide into them without any noticeable effort – is beyond the reach of his elders‘
The difficulty I have here is that child cognitive development goes through several stages. For instance Piaget has characterised these stages. It is also important to remember that the child has a brain volume that is very much less than that of the adult and there are thus several reasons why adults thinking may be different and even more flexible than a child’s thinking. There are of course exceptions but it can be argued that ‘his elders’ may be used to adopting an approach which has stood the test of time although this is conjecture.
Feyerabend then suggests that in training (presumably scientific and other forms)
‘The teaching of standards and their defence never consists merely in putting them before the mind of the student and making them as clear as possible. The standards are supposed to have maximal causal efficacy as well‘
Thus he argues that the training is expected to have a psychological effect on the student which creates the effect of a ‘voice of reason’ and that
‘He will be quite unable to discover that the appeal to reason to which he succumbs so readily is nothing but a political manoeuvre‘
It can also be argued that Feyerabends work has become quite influential and indeed has been published in book format. This influence itself can be considered a political manoeuvre since the wide dissemination of a work in such a manner can increase it’s credibility. Thus I would argue that Feyerabends own work ‘Against Method’ can be criticised on the same grounds that he uses against argument – indeed his book consists of a series of arguments.
He moves on to consider the case of Galileo who pursued what could have been considered a ‘silly cosmology’ and that the theory become clear only some time after it first came into use and continued to be used. Galileo is often used as an exemplar of science and this persistent use of Galileo’s example can lead to what is known as confirmation bias. What this means is that if someone has a hypothesis and they look back in time at something that has already happened they can look for those facts that fit their hypothesis. This means that they are selecting facts. A more rigorous approach involves defining beforehand what facts you are looking for to test the hypothesis and then examining those facts in the dataset. This can also be applied prospectively.
‘Now when we attempt to describe and to understand developments of this kind in a general way, we are, of course, obliged to appeal to the existing forms of speech which do not take them into account and which must be distorted, misused, beaten into new patterns in order to fit unforeseen situations (without a constant misuse of language there cannot be any discovery, any progress)‘
Here Feyerabend is arguing that there are some parts of the scientific process which proceed without the use of language. Can this really be so? How do we know? This part seems paradoxical. Perhaps there are some features of the universe that cannot be described by humans because our brains have evolved under and therefore adapted to very specific environmental conditions. I would agree that there people can arrive at accurate judgements very quickly without the use of formal language. Pattern recognition is a good example. However that does not mean that these same patterns cannot be described using language after a little consideration. What Feyerabend is effectively saying is – there is something in science that can’t be described with language – but I’m not entirely clear on what that something is and by definition Feyerabend cannot be either. Unfortunately then Feyerabend might as well be invoking the use of fantasy to support his argument. I would add that if a concept is not communicable using language or at least a symbolic language then it sits outside of science in as much as science is created by a community. Even if someone did come across an invaluable truth, which helps them to make predictions in a languageless manner, it cannot properly be called a science if it cannot be communicated to another person who can test this. Instead I would argue that the person should be able to use introspection to identify within their stream of consciousness the processes leading to their judgement and to attempt thereby to translate this into language. This process should proceed even if new words need be constructed to convey these concepts, fashioned out of a relationship with known words. If instead we consider the case of a theory parts of which cannot be explained then this does not matter too much so long as clear predictions can be made by that theory. We can forgive the theory without meaning providing it is clear about relationships and therefore has predictive utility. The meaning can be attributed at another time. Although discussing mathematics, the noted mathematician Srinivasa Ramanujan was able to arrive at equations and formulas intuitively and these were later proved correct. In this case it does not matter that language was not invoked because what he wrote had predictive utility and the proofs could be established by others. This however does not negate the value of the proofs and predictive utility and indeed it could be argued that such proofs and predictions have made his work useful for the mathematical community. Indeed a relatively small number of Ramanujan’s conclusions were refuted and this process of checking and refutation made the remaining conclusions which had been confirmed all the more valuable. The formal language is part of the validation process which is just as important as the creative process.
In the second chapter, Feyerabend reveals his intentions
‘An anarchist is like an undercover agent who plays the game of reason in order to undercut the authority of Reason (Truth, Honesty, Justice and so on)‘
I would however separate honesty and justice from reason in that the former are value-based whereas reason is a process (although reason may be used to determine if honesty and justice are the case under specific circumstances). Indeed honesty is a factor that is integral to trusting relationships and so I am a little curious as to why it has been invoked here although I might simply have misunderstood Feyerabend’s intentions. In this chapter, Feyerabend writes about a rule that should be used to advance science
‘The ‘counterrule’ corresponding to it advises us to introduce and elaborate hypotheses which are inconsistent with well-established theories and/or well-established facts. It advises us to proceed counterinductively‘
I agree with this on the basis of the following. If the underlying theory is invalid or invalid under certain circumstances then resources will have been invested in this one area to the detriment of the development of other theories which are presumably valid. If the most basic of the theory’s assumptions have not been repeatedly tested then experiments will proceed further upstream and these also will ultimately represent resources that would have been better invested elsewhere. Take for example the assumptions about the ether which were incorporated into theories in the nineteenth century. The effort spent in considering the ether in specialised experiments would now be considered time misspent in arriving at the present state of knowledge. Establishing alternative hypotheses also encourages competition between proponents of hypotheses and theories. This might increase standards although this is speculation. This does however appear to be an important principle in both economics and natural selection. There is also the assumption in the above statement that there is a correct theory. This in itself can be challenged. We might say that there is no such thing as a correct theory. This might happen for a number of reasons.
1. Firstly a theory that is predictive may not be describable in terms that people can understand.
2. Secondly there may be no valid theory. In other words the natural phenomenon exists which we cannot describe theoretically.
3. Thirdly the theory may be valid but only under specific circumstances and so we might say that the validity of a theory lies on a continuum which ranges from applicable under all circumstances to applicable under no circumstances.
He then writes about
‘The second ‘counterrule’ which favours hypotheses inconsistent with observations, facts and experimental results‘
He then suggests that all theories have contradictory facts. While he goes onto justify this rule further, the existence of accepted theories with contradictory facts seems to me sufficient to justify the validity of hypotheses inconsistent with facts but does not justify the favouring of such hypotheses over others.
In Chapter 3, Feyerabend argues that
‘The consistency condition which demands that new hypotheses agree with accepted theories is unreasonable because it preserves the older theory, and not the better theory‘ p.24
This in itself is reasonable and it is here that the issue of falsification could be raised although this veers away from Feyerabend’s intentions. I would argue that it is the theories which lie at the base of dependent hypotheses that require special testing and efforts at falsification given the importance they assume within the hierarchy. I would go further and add that a ‘tree’ of explicit assumptions should be made corresponding to the underyling theory, branching subtheories and dependent hypotheses such that this path can be easily visualised in connection with a particular study. Since later assumptions are contingent on earlier assumptions (regardless of whether they are part of theories, subtheories or hypotheses) these assumptions can be given a weighted value according to their position in the hierarchy and by this reasoning it is the earliest assumptions which should be most challenged. Note that in choosing the assumptions to challenge it is an error to think that the underlying theories are robust and have stood the test of time because they might not have been as rigorously tested as later dependent hypotheses. One has only to look at drug trials to see the dramatic differences in drug trials in contemporary times compared to the earliest research showing the efficacy of Lithium or Penicillin for example. Instead Feyerabend argues that there is a tendency to consider hypotheses which are consistent with the main theory and not those that are inconsistent with this theory. This is the essence of what he refers to as the ‘consistency condition’. Further he states that the emphasis is on the identification of facts which can challenge the main theory. The essence of his argument is that resources are directed to what I will refer to as a winner-takes-all theory. A winner-takes-all theory is one that is focused on to the exclusion of all of the other theories. His argument focuses on the need for variety in the theories that are examined and it is here that the reasoning can be seen behind his assertion that hypotheses should be sought with are incompatible with ‘well-established theories’. However if we take his arguments to their logical conclusion then it should be asserted that all possible theories should be generated along with their corresponding hypotheses for due consideration. The obvious rejoinder to this is that if there are an extremely large number of possible theories then there may be insufficient resources to be allocated to each. This leads to the principle of prioritisation of resources for competing theories unless sufficient resources are available for each. This in turn leads me to the following conclusion:-
Feyerabend has been using a dichotomous framework in his approach to this topic and makes no mention of the resource issues that are important in the real-world scenario. The alternative and perhaps more realistic viewpoint is that this should be considered in a probabilistic manner and with due consideration for resources. Thus if there are a finite number of resources R and a finite number of competing theories T1, T2… and each has a corresponding probability of being correct p(T1), p(T2) then it follows that resources could be allocated according to the formula
Rx = R.p(Tx)
where x ranges from 1 to n and n is the number of theories.
This rests on the assumptions that
a. Resources can be clearly delineated and expressed quantitatively. This should be possible through task analysis and it should be recognised that in a number of cases numerical markers may afford approximations.
b. That there is a consensus on the probability of each theory being correct. Again, due consideration can be given to the method of choice.
There is also the caveat that a certain proportion of the resources should be invested in the generation of new theories which would mean a small correction to the above formula. Additional modifications could be made to the resource allocation according to practical impact of the theory and this itself can be further refined. He summarises some of the views of John Stuart Mill on the process by which scientific theories become popular with time. In the course of his discussion on this, I was somewhat surprised to read the following statement by Feyerabend:-
‘An empirical theory such as quantum mechanics or a pseudoempirical practice such as modern scientific medicine with its materialistic background can of course point to numerous achievements but any view and any practice that has been around for sometime has achievements‘ p.30
It is unfortunate that Feyerabend has not qualified this statement but has merely referred to a rather vague ‘pseudoempirical practice’ and applies this term to ‘modern scientific medicine’. Given the potential ramifications of such a statement, I would expect a rigorous treatment of his supporting arguments but instead there are none. As this ‘throwaway statement’ appears in a book about the philosophy of science, I will suggest that this statement is out of keeping with many other aspects of the book and the lack of supporting reasoning or evidence suggests that this is ‘pseudophilosophy’. To be more explicit, by pseudophilosophy I am referring to a statement or collection of associated statements which masquerade as philosophy but which lack supporting evidence or reasoning or where it is present it does not stand up to close scrutiny of even the most basic kind. To take this further I would argue on the basis of the current example that ‘pseudophilosophy’ allows prejudices to be mistaken for philosophy.
In this chapter, Feyerabend starts by stating that all ideas are capable of furthering our knowledge ‘however ancient or absurd’. He advocates a pluralistic methodology which involves all theories. He is thus taking a step back from the consideration of the facts and hypotheses and looking at the bigger picture – the underlying theory.
‘Now this politically enforced dualism has led to most interesting and puzzling discoveries both in China and in the West and to the realization that there are effects and means of diagnosis which modern medicine cannot repeat and for which it has no explanation‘ p.37.
He also gives the example of a herb which is analysed for pharmacological activity upon which the active ingredient is removed. He argues that it is possible that the herb in its entirety is responsible for the herb’s actions rather than the individual agent. This hypothesis seems reasonable but would need to be supported by the empirical evidence. He also argues that
‘Does this not mean that they must learn one particular set of views to the exclusion of everything else? And, if a trace of their imagination is still to remain, will it not find its proper application in the arts or in a thin domain of dreams that has but little to do with the world we live in?‘ p.38
Again this needs to be supported by the evidence. There are many instance of the ‘imagination’ and other aspects of creativity being used to good effect in science.
‘According to Hume theories cannot be derived from facts’. This is a very brief statement and is not further explained. I presume that it refers to the induction problem in which Hume suggests that predictions cannot be made on the basis of past experience. Nevertheless this statement appears to be incorrect. Newton’s in deriving the laws of gravitation utilised Kepler’s laws of motion which in turn were derived from an observation of the motion of the planets in the solar system. Furthermore Darwin accumulated a vast amount of evidence from different species before proposing his theory of evolution (natural selection). Surely then a theory can be derived from the facts and furthermore there are a number of ways in which this can take place including pattern recognition and induction, iterative hypothesis testing and unstructured methods. These theories are more robust when it is large numbers of data points that are being considered rather than individual data points.
He then further argues that a theory is not consistent with all of the facts and therefore theories should not have to be accepted on the basis of them being ‘consistent with the available and accepted facts’. I would argue that this is an oversimplification of the case. Theories are evidently not an all-or-nothing phenomenon but instead exist on the balance of probability explaining a proportion of the available facts. Therefore I would argue that Feyerabend has not sufficiently presented his case for this rule to be dropped. There is still a case for theories being selected which fit with the facts. Although it seems reasonable to select also those theories which do not fit with the facts. However such theories should have some basis in facts, no matter how few the facts they fit with, as otherwise there would be no reason for them to bear any relation to reality which ultimately is the purpose of a theory.
Feyerabend then goes onto challenge the quality of evidence for instance referring to the sensory apparatus thus
‘The sensory impression, however simple, contains a component that expresses the physiological reaction of the perceiving organism and has no objective correlate‘
Again, it can be argued that our senses and perceptions are highly effective and sufficiently so to enable us to move through the day walking, eating, working, playing and basically interacting with the world in a predictable way. In some cases our senses play tricks on us and these illusions have been closely studied and provide a source of amusement. These cases how represent a fraction of our sensory experiences and our ability to recognise illusions as illusions suggests there is a deeper element to our perception which chimes with our internal model of the external world. What Feyerabend leads up to is the case of Copernicus who’s theory was in stark contradiction to the observable facts of the day which would have obstructed the acceptance of his theory. However this example of celebrity science is but a single retrospective example which may represent a confirmatory bias. In order to use induction to generate a rule as Feyerabend seems to be doing in this case, I would prefer to see further data from different branches of science to support this rule. Again if we turn to Darwin’s theory of evolution, we can ask – does this theory contradict the sensory information that Darwin was furnished with prior to the construction of his theory? It is difficult if not impossible to know as we are talking about an historical case. We no longer have access to Darwin’s thought processes except through his writings. Nor for that matter do we have access to the thought processes of Copernicus. However Darwin once wrote the following
‘I am turned into a sort of machine for observing facts and grinding out conclusions‘
Another curious point here is that Feyerabend is arguing from the experience of Copernicus – the same experience that he maintains Hume argues cannot be used to generate theories. So if experience cannot be used to generate theories, why is Feyerabend using this to generate rules? If there is a rule for science then why should there not be this same rule for philosophy? In any case Feyerabend goes on to assert that counterinduction is a ‘fact’ and a ‘much needed move in the game of science’.
Chapters 6 and 7
In Chapter 6, Feyerabend gives the example of Galileo who says that he does not explicitly state each part of the chain in his reasoning. Although Feyerabend provides us with a ‘fact’ to support his theory this again could result from confirmatory bias. Similarly using the same line of reasoning that Feyerabend proposes about facts and theory, why is it necessary for Feyerabend to support his rule with a fact. In any case, the cited example does not exclude the possibility that Galileo had a detailed chain of reasoning and indeed another author has suggested that theories are constructed by an elaborate line of reasoning which may occur out of conscious awareness (see Godel, Escher and Bach). In this chapter he describes natural interpretations which involve the combination of the sensations and the responses to those sensations. He says that this can be treated in one of two ways. In the first method suggested by Bacon, the natural intepretations should be broken down into their constituent parts much like the peeling the layers of an onion. In the second method proposed by Kant they ‘have been regarded as a priori presuppositions of science’. Feyerabend then argues that natural interpretation was used to confirm the Copernican theory of motion and that Galileo’s trick was to subtly introduce an alternative theory and to enable his audience to confirm this theory by means of natural intepretation. He also replaced natural interpretations which contradicted the Copernican theory with those that did not. He further develops this idea in chapter 7. In this chapter, Feyerabend states that Galileo has introduced the law of circular inertia and the idea of relativity of all motion. He asserts that
‘Galileo uses propaganda‘ (p.65)
I found it difficult to follow Feyerabend’s discussion for several reasons.
1. Feyerabends discussion occurs in the form of an essay whereas on close reading there are a number of assumptions which are being stated. These could have been presented differently so as to produce a more coherent structure.
2. I found myself dependent on Feyerabend for his interpretation of Galileo and Copernicus which is relevant given points 3 and 4
3. Feyerabend selectively quotes from Galileo’s works.
4. There are difficulties in understanding Galileo’s works.
a. The works referenced by Feyerabend are translations of the original Italian.
b. The language used by Galileo may be peculiar to that period of history and therefore a contemporary reading may lead to misinterpretations of the original meaning. Feyerabend himself hints at this when he writes that
‘They speak the language of real motion in the context of 17th century everyday thought. At least, that is what Galileo tells us‘ (p.59)
c. Galileo’s writings must be interpreted in the proper historical context as there would be expected to be profound cultural differencs in comparison with contemporary culture.
In this chapter, Feyerabend reiterates some of the points from the previous chapter before moving onto the telescope. While at the beginning of the chapter Feyerabend states that
‘However, he offers no theoretical reasons why the telescope should be expected to give a true picture of the sky‘
He goes on to suggest that Galileo had written about some calculations he had made to confirm the accuracy of the telescope but Feyerabend is unable to identify these and instead suggests that Galileo verified the telescope’s accuracy through experience. He further supports this assertion by reference to a translated quote from Galileo in which he writes that the accuracy has been verified with reference to one hundred thousand observations on stars and other objects. The essence of this chapter consists of an interpretation of a specific aspect of medieval European history in order to support the assertion that Galileo’s reliance on the telescope, an extension of his senses was based on experience and not on theory. However it could be argued that relevant evidence of Galileo’s calculations has not remained. Feyerabend also comments on the writing of Galileo about the nature of light thus
‘Even if we consider such utterances with the care that is needed in the case of a whimsical author like Galileo…..‘ p.82
and he further quotes Professor E. Hoppe thus
‘Galileo’s assertion that having heard of the Dutch telescope he reconstructed the apparatus by mathematical calculation must of course be understood with a grain of salt for in his writings we do not find any calculations…..‘ p.83
Feyerabend thus takes time to discredit Galileo both in his scientific endeavours and in his skills as an author and yet it is from a translation of these same writings that Feyerabend bases the central argument in his work.
This shows one of the drawbacks of a historical analysis when investigating the process by which science is performed. If instead a living scientist had been selected, then the scientist could have been interviewed directly in order to confirm or refute the above assertions and the quality of evidence to support the arguments in the case of important scientific discoveries would be expected to be of a higher quality. The quality should be higher given the wide variety of media that are available for disseminating such discoveries as well as the process behind them, the means for storing items from these media and the larger population from which may be drawn a larger group of people focused on responding to and reporting on important developments in science.
Here Feyerabend continues the argument that there was little reason for Galileo to suppose that the telescope offered a ‘true picture’ of the sky and focuses on why Galileo’s drawings of the moon bore no resemblance to the lunar landscape that is seen today.
‘Having been influenced by Wittgenstein, Hanson and others, I was for some time inclined towards the second version but it now seems to me that it is ruled out both by physiology (psychology) and by historical information‘ p.96
He argues that the telescope provided results that could be refuted by naked eye observations and that at the time, there was no way of determining which of the telescope’s findings were correct and which incorrect. He notes also that the telescopes of that time would have been very different from those currently used.
Having no training in astronomy, nor access to the original writings of Galileo, I turned instead to an online resource ‘The Galileo Project’ and at the following page (http://galileo.rice.edu/sci/observations/moon.html) identified a statement suggesting that Galileo had observed that shadows were differentially cast across the lunar landscape according to the lunar cycle and that he had intended to return to describe the lunar landscape more comprehensively. Nevertheless he had apparently not done so. It is not unreasonable to suppose that if he intended to return to this area that he considered the original sketches insufficient. In any case, Feyerabend seems to side-step the critical point of Galileo’s observations which challenged the model in which all objects orbited the earth. Thus again returning to the Galileo Project at the following page (http://galileo.rice.edu/chron/galileo.html) there are a number of significant events that precede the trials – 1604-1605 he witnesses a supernova explosion and concludes that there is change in the ‘heavens’, 1609 Kepler publishes his first two laws of planetary motion, in 1610 his observations lead him to conclude that Jupiter has three orbiting satellites. In addition to all of this Galileo was lecturing in mathematics (including lectures on motion), lectures on astronomy, advances a theory about gravity, experiments on the relationship between tension and pitch in a string, invents a pump for raising water, builds scientific instruments, experiments with the pendulum, invents the thermoscope and produces a work on hydrostatics. Assuming that all of the above is accurate then it is reasonable to draw a number of conclusions
1. Galileo has demonstrated both a depth of knowledge and breadth of knowledge in closely related areas (to planetary motion)
2. The witnessing of a supernova explosion did not require a telescope but challenged the notion of a ‘constant heaven’
3. Kepler had already published his first two laws of planetary motion
4. Galileo had closely observed the motion of satellites around Jupiter
In addition to this, I would argue that on the basis of the above achievements it is entirely likely that Galileo utilised a combination of a vast body of accumulated knowledge in mathematics, contemporary physics and experimental science in combination with hard work in order to arrive at his conclusions. I infer hard work on the basis of the considerable list of achievements that were attained in numerous areas. The emphasis on hours of study for attaining extraordinary achievements has been discussed at some length in a work by Malcolm Gladwell titled ‘Outliers’ (see review here). Thus I would speculate that the combination of long hours of study and accumulated knowledge has resulted in a significant intuition in the relevant area and this intuition might not have been easy for Galileo to formulate explicitly. Additionally bringing forth all of the relevant material at the time may have been sensitive – it is difficult to know Galileo’s reasoning or indeed the events on the basis of Feyerabend’s discussion of a translation of a text that is nearly 400 years old.
In Chapter 10, Feyerabend writes that Copernican theory and astronomy were in trouble because of refutations and it survived because
‘but that being in harmony with still further inadequate theories it gained strength, and was retained, the refutations being made ineffective by ad hoc hypotheses and clever techniques of persuasion‘ p.105
While he quotes from others to support his argument, I would suggest in response the following. Assuming that in the above quote, he is referring to selected observations of the telescope being used to support Copernican theory, I would challenge both parts of this. Firstly although Copernican theory might have been in trouble, every theory can be challenged by those who are so interested to do so. It is therefore not unusual that a theory and successful one at that would have its critics and whether it is considered to be in trouble or not is a matter for judgement. On the second point, we know in retrospect that the telescope is a successful technology and observations made with the telescope have provided many important scientific discoveries. At the time, he argues that it did provide some observations in support of the Copernican theory and it is reasonable to assume that he selected those observations which fitted with the theory – what we might refer to now as confirmation bias. He then argues that other techniques such as ‘persuasion’ were used to support the argument. Even when the supporting dataset is watertight however, persuasion is an important skill and represents the way that not just science but any activity with a social aspect is performed. I would thus classify this as a distinction between the process of doing science and communicating science.
In Chapter 11, Feyerabend argues that a new theory will have little empirical evidence. Therefore there will be a move away from a dominant theory with abundant empirical evidence to one with very little. During this period an alternative with little empirical support would need to be reinforced by alternative means predominantly tools of persuasion:-
‘I suggest that what Galileo did was to let refuted theories support each other, that he built in this way a new world-view which was onl+y loosely (if at all!) connected with the preceding cosmology (everyday experience included) that he established fake connections….and that whenever possible he replaced old facts by a new type of experience which he simply invented for the purpose of supporting Copernicus‘ p. 121
While he is more complimentary of Galileo a little later and indeed praises his techniques of effective communication, I would question the above. Firstly as already argued above, I don’t think it is reasonable to say that he was dealing with refuted theories for the reasons given above. Secondly I would argue that the evidence has not been adequately presented to support the assertion that he has ‘established fake connections’ or that he ‘invented’ experience to support Copernicus.
More than likely his prolonged periods of observation with the telescope would help him to develop an intuition for his observations allowing him to discount those findings which resulted from aberrations in the telescope technology that he was using. This is speculation but so also in my opinion are the above remarks by Feyerabend. Feyerabend does have some suggestions which I would say are reasonable. Thus he argues that the dominant scientific theories of the time occur in a historical context where there are implicit assumptions and social factors that modify the way in which the theory is perceived and used. However, here again these are generalisations without the supporting evidence base. The remaining arguments are covered in the response to Chapter 9.
Chapter 12 is brief and Feyerabend essentially argues that the same approaches discussed previously can be used ‘in other fields as well’. At this point, he refers to the mind-body problem but does not discuss it any further.
In this chapter, Feyerabend looks at the two trials of Galileo by the church. He argues that Galileo didn’t have sufficient evidence to support his defence of Copernicus at that time – that the ‘scientific’ or as it was then referred to – philosophical evidence was considered insufficient. However, Feyerabend does not delve further into why this evidence was considered insufficient. Curiously Feyerabend writes about one trial that
‘It had no special features except perhaps that Galileo was treated rather mildly, despite his lies and attempts at deception‘
Such a strong statement is not qualified except by a rather curious footnote which reads ‘the reaction of an admirer is characteristic:’ This absurd pretence..’ ‘which I do not think justifies the above. He goes on to repeat the previous assertions
‘demanded strong empirical support while the Galileans were content with far-reaching, unsupported and partially refuted theories‘
Again, I would argue that this statement is unjustified. Essentially Feyerabend is invalidating Galileo’s observations with the telescope predominantly on the basis of a poorly sketched lunar landscape which he tells us does not match with the contemporary understanding of the lunar landscape. Feyerabend goes onto argue that Galileo’s publication would potentially have disrupted wider society at the time. However this discussion seems to be outside of the main scope of the work which is to address the use of method in science and so I will not elaborate further on this.
In this chapter, Feyerabend argues that the revolution in science had not begun with Galileo and then discusses the work of both Copernicus and Maestlin. In both case, he attempts to demonstrate that both astronomers used erroneus assumptions in arriving at their conclusions.
In the case of Copernicus, he quotes from Commentariolus in looking for an alternative to equants which were essentially constructs that enabled the empirical data to be reconciled with a view of the planets which involved circular motion and he again quotes from Copernicus in stating that this was to retain the rules that the ‘ancient greeks’ had set out.
Again, this is outside of my field and so I have sought preliminary information online although I have not had first sight of the source material (this is tertiary source), Copernicus was said to have been aware of previous suggestions that the earth moved by both Philolaus and Aristarchus. Again from the same tertiary source material, Aristarchus was apparently attacked for his views. At the time of writing, the Stanford Encyclopedia of Philosophy states that there have been misinterpretations of the quote from Commentariolus (Interestingly, their passage has subtle differences from the passage quoted by Feyerabend. Although the differences are not relevant to the argument, they do demonstrate that even these brief passages of text can be interpreted differently). The authors go on to suggest that Copernicus did not choose to utilise circular movements in his model for aesthetic reasons, to produce a simpler system than Ptolemy or under the influence of Hermetism or Neoplatonism. Instead they suggest that it was made on the basis of a close analysis of the motion of the planets and cite another source (which I have not seen)(Swerdlow and Neugebauer, 1984). Here again therefore there is some evidence against Feyerabend’s argument that Copernicus based his model on an erroneus assumption and instead this lay with empirical observation.
Without going into the arguments he raises against Maestlin, it is now possible to challenge Feyerabend’s conclusion that the revolution was not accounted for by ‘sophisticated empiricism’ as he writes that
‘This is certainly not true for Copernicus…As we have seen, Copernicus thought the Ptolemaic system to be empirically adequate‘
However it has been suggested by Swederlow and Neugebauer above that it was a more considered empirical observation of the planets that led to his conclusions.
Feyerabend starts this chapter by asserting that there there is no place in science for critical rationalism as suggested by Popper. He suggests that
‘criticism starts only after the discoveries are made‘ p.147.
Nevertheless, I would argue that when a theory is sufficiently robust it allows for theoretical as well as experimental predictions. In other words, once a theory is well understood and supported by the evidence base, it becomes possible to advance subtheories – theories that are contingent upon the original theory and which constitute a systematised group of hypotheses. This may exist in theory before it is explicitly tested and therefore criticisms can arise at this stage before the formal process of empirical testing. The result of all of this is that observations and theories do not necessarily develop together simultaneously but can be separated distinctly if the above holds.
Again a little later, Feyerabend suggests that
‘a determined application of the methods of criticism and proof which are said to belong to the context of justification would wipe out science as we know it – and would never have permitted it to arise‘ p.148.
He suggests that there is a creative process much akin to an artistic process. That there is some similarity is not unreasonable particuarly when various scientists have described the methods by which their insights have arisen, the most notable case being Einstein’s thought experiment on the speed of light and his theory of general relativity. Even this example however is too simple and does not describe the knowledge and previous work of Einstein in leading to this conclusion. Most likely the ‘creativity’ arises after significant preparatory work. It is reasonable though to suggest that people will use different approaches and that these approaches will be valid and will sometimes bear similarity to those processes used in the arts. Nevertheless this again would benefit from empirical support otherwise they remain as highly speculative suggestions.
Later in the chapter Feyerabend makes a remark about how the scientific process will affect the scientists themselves and makes the suggestion that it will have a detrimental impact on their personality traits and further quotes from Kierkegaard
‘that my activity as an objective [or a critico-rational] observer of nature will weaken my strength as a human being?‘ p.154
There are two rejoinders to this. Firstly and in a rather light-hearted manner, much of Feyerabend’s discussion consists of a rational attempt to justify a multiplicity of methods in science and since science occurs in the natural world (for we are part of that world) his comments would apply just as much to the pursuits in this book. The second and more significant rejoinder is that this is a fairly broad statement and ultimately quite vague as there are numerous branches of science, multiple activities within science and a wide variety of scientists of different temperaments and working for various periods of time within their fields. Feyerabend (and Kierkegaard) are firmly in the domain of psychology and as such their assertions should be more precisely stated and supported by empirical evidence particularly as there are so many factors involved as well as uncertainty about the characteristics (in their statements) which are being commented on. Feyerabend has also missed the opportunity to comment on the real issue here which is the value system.
Much of the criticism that I have in this chapter is the same. Thus Feyerabend goes on to say that ‘institutions, ideas, practices’ start from activities such as play rather than starting from problems. The very fact that ideas and practice can start from problems suggests that this question is best answered by empirical study rather than speculation. He then argues that
‘a strict principle of falsification…would wipe out science as we know it and would never have permitted it to start‘ p.154.
Again this is a difficulty in adopting dichotomous thinking. Facts that challenge theories do not necessarily need to eradicate these theories. The impact of the facts can be assessed and the likelihood in which the theory is invalidated by them judged. This again is a probabilistic function rather than an absolute.
He goes on to compare new and old theories and suggests that the ‘imagined’ content of older theories shrinks as they are displaced by new theories. The inference here is that the older theories contain an essence of truth and so in the cycle of history different truths are interchanged and the associated wisdom is lost. Well I would agree that it is a pity that the thoughts built up on these older theories become lost in time but I challenge the central underlying assumption that an older valid truth is replaced by another equally valid truth. Science is young. There has been insufficient time to see if there will be another fundamental change. Systematic historical writings date back a few thousand years and most of the writing that is considered (at least in this book) starts with the greeks 2000+ years ago. The paradigm shift which is being discussed is a single shift which occurs with Copernicus/Kepler/Galileo in the middle ages. On the basis of this understanding (added to by later theoreticians and scientists) it is possible to calculate the trajectory of an interplanetary satellite. The repeated launching of satellites and spacecraft is testimony to the power of such predictions. If there were another paradigm shift – would it interfere with the current process by which scientists are able to predict the trajectories of satellites? The answer must surely be no, other than to perhaps refine the accuracy. Thus the body of science in this particular area has achieved a practical utility which serves society’s needs in that one area. Feyerabend implicitly assumes that there could be another paradigm shift in which the current theories will be torn down and will be subsumed by another theory. Perhaps there will be, but even if a new theory were put in place, in its application to satellites it would be difficult to see how it wouldn’t ‘collapse’ into the laws that are currently successfully used in which case they would essentially remain invariant. Thus there is a distinction between the practical utility of a theory and the ‘idealistic’ truth of a theory (even this term (idealistic truth) however would be better replaced with probabilistic truth).
Thus again, I do not think that Feyerabend has sufficiently defended his initial assertion that critical rationalism is invalid.
In this Appendix, Feyerabend looks at the issue of alternatives to science. He refers to recent developments in China and states that
‘they restored important parts of the intellectual and emotional heritage of the Chinese people and they also improved the practice of medicine. It would be of advantage if other governments followed suit‘. p.163.
I would argue that this is a political statement rather than a philosophical one. Thus in 1998, China’s scientific output was stated to be roughly 20,000 articles per year but by 2009 this output was approximately 120,000 articles per year. Similarly Feyerabend earlier in the book commented on how ‘western values’ were being pushed into China (the implication being that science is playing a role in this). However it is curious that he makes no mention of the scientific heritage of China. For instance in the area that his central discussion takes place, he neglects to mention the works of Master Shi Shen (4th Century BC) in astronomical observations, Lord Gan (4th century BC) who constructed a star catalogue and the astronomical observations of Shen Kuo in the 11th century. These astronomers created large bodies of empirical evidence. These areas of scientific inquiry that have been essentially ignored by Feyerabend suggest that scientific activity has arisen in independent cultures but would also support the hypothesis that it has a biological correlate (this evidence however is only sufficient to point in that direction. It would require a close examination of historical evidence to exclude confounders. Nevertheless other more suitable methods would be better suited to investigating this question). I am tempted to go slightly astray here and provide the audience with some rather interesting footage that I captured of some Lemurs examining a novel object in their environment. The lemur lineage is estimated to have diverged from the human lineage some 63 million years ago (Dawkins, 2003). Nevertheless the footage that I have obtained demonstrates lemurs investigating the object (a camera stand) utilising all of their senses and even pushing the camera stand repeatedly. While such an explanation of this footage is highly speculative and it may have no significant meaning, the exploratory behaviour reminds me in some capacity of the exploratory work that is done in science in obtaining a better understanding of the world around us. The clip serves as a starting point for looking at how an informal science may have an underlying biological basis that transcends culture and even species although a formal written science as demonstrated by humans is another matter.
In chapter 16, Feyerabend argues that comparing scientific views with non-scientific views is not possible in a number of cases. He remarks that certain theories are understood and accepted well enough to have developed their own associated languages. This seems reasonable. Feyerabend discusses the view of Whorf who asserts that language contains a view of the world that
‘influences thought, behaviour, perception‘ p.164
Feyerabend includes 2 diagrams which are examples of visual illusions. He argues on the basis of the effects elicited that we are able to change our perceptions ‘at will’ and that our ‘mental sets’ are culture dependent. After choosing lines to focus on, our perceptual apparatus completes the remainder. Since we can choose the lines initially we can also ‘choose’ our corresponding perception. He also gives the example of after-images and uses these examples to suggest that alternatives to science cannot be considered simultaneously and that there must be a set-shifting from one to the other and the inference he makes is that these separate approaches shape our perception of the world. He goes on to examine different styles of art in history. He examines ancient styles of art including Egyptian and Greek art. He draws parallels between ancient greek art and literature and argues that there is sufficient evidence to suggest that ancient Greeks perceived the world differently to the way that his audience does. About Egyptian art in particular he writes
‘in the ‘frontal’ art of the Egyptians…..are explained by Loewy…Natural, physiological reasons are given for the archaic style and for its change‘ p.170-171
However from an online resource (digital egypt at UCL – http://www.digitalegypt.ucl.ac.uk/art/whatisaeart.html) about Egyptian art, it is written
‘Egyptian canonical art has a different function, which has nothing to do with representing reality‘
Instead there is an underlying philosophy that constrains the Egyptian art and thus it is not surprising that while the prevailing art fits within cultural norms, there are small variations where a more realistic individual style emerges. All of this challenges Feyerabend’s inferred argument that ancient peoples had a different physiology which caused their art to differ. If we inferred a different physiology on the basis of artistic styles then it could be argued that in the 20th century the human race experienced several dramatic changes in physiology which caused them to express the world differently in the forms of Cubism, Dadaism, Surrealism and Pop Art. However this would be an erroneus conclusion. His argument is that we cannot understand the competing view unless we step into the world of the other civilisation. Nevertheless it can also be argued that if there is an alternative to science that offers us useful knowledge about the world but without it being a science then it should be possible to obtain measurable predictions from this approach. These can be tested against predictions by the relevant body of scientific knowledge. Thus it becomes possible to sample and test this alternative without needing to step entirely into the other world view. In this way Feyerabend’s assertion that
‘the kind of comparison that underlies most methodologies…breaks down when we try to compare non-scientific views with science‘ p.164
can be challenged. Feyerabend takes us from the perceptual world of the ancient Greeks to the modern perceptual world drawing a parallel with science and taking the opportunity to note that there are several forms of logic, that logic does not unify all of the sciences, that it is not integral to the process of discovery before identifying a need for anthropological studies.
He discusses the benefits of anthropological field work in science so as to better understand real-world science and here I agree with him. He is also critical of epistemologists because
‘they never inquire, or even find it worth inquiring, whether the conditions are satisfied in this real world of ours‘ p.197
However, it is with the results of such an analysis that Feyerabend’s arguments would have been more robust. Indeed I would argue that without the results of such field work, Feyerabend’s suggestion of the irrelevance of logic in the scientific process is speculative and therefore on the basis of his discussion alone, there would be equal weighting to both this hypothesis and the counter-hypothesis if it were judged on the basis of empirical evidence. Looking from a wider perspective however, it is likely that in addition to logic, pattern recognition, serendipity, systematic trial-and-error and expertise would all play a role in this process. Although I would disagree with the material he uses for his approach, I would agree that research is needed in this area. However Feyerabend does not expand on the scope of such research other than to say that this would be anthropological and to outline how the basics of this approach might work. It is clear that careful consideration would need to be given to which branches of science are to be studied.
In this appendix, Feyerabend disagrees with Whorf on some points about his assertions about language. For instance he remarks that animals do not have language (in the sense that Whorf means it) and yet their world is not shapeless. He also reiterates that theories in the way scientists use them are different from the transformed theories that have been clarified utilising logic and other tools and which is then studied by philosophers.
In this chapter Feyerabend argues that rationalism is a tradition. However, I would argue that if by rationalism he means the application of reasoning then there is evidence that species such as chimpanzees, rooks and crows are capable of this. For instance, wild chimpanzees have been observed to use ‘anvils’ with cleavers to break food down into smaller portions. A bonobo is observed using a rock to break down food in a video at this link. A rook is seen to use two stones in sequence to release food from a test tube in a clip at this link. While it could be possibly be argued that chimpanzees and bonobos are engaging in traditions handed down by other members of their group, it is more difficult to explain the rook’s behaviour in this way. It seems highly unlikely that rooks would teach their young to transfer stones between test tubes of differing sizes in order to obtain food. It seems more likely that the rook is able to reason that certain actions will result in the release of food. It can be further concluded if the above holds that reasoning is a biological trait and this would be consistent with adaptation to the environment in which various species utilise a knowledge of cause-and-effect to anticipate events in the environment and to exert control over the environment in order to meet the needs of underlying biological drives. It further follows that if reasoning is a biological trait then the ‘tradition’ of rationalism is simply the repeated use, refinement of and shared development of an inherent trait. Furthermore, if rationalism is an inherent trait then we would also expect to see it utilised in other ‘traditions’. Furthermore it would not be unreasonable to speculate that the foundations of science lie in an inherent ability to identify and utilise causal sequences in the environment upon which layers of complexity have been added over time through the shared communication and development of this understanding in the scientific community by means of language. If this were the case then it would help to explain the cultural association between science and logic which Feyerabend has taken some effort to challenge. The suggestion in this discussion that there is an inherent biological trait underlying science is however speculation.
Feyerabend considers the observers and participants in a tradition referring to the Hegelian dialectical framework that can be used by observers. He discusses participants and then looks at rationalism in more detail suggesting that it arises in two main forms – naturalism and idealism. He is again dismissive of medicine in another off-hand comment and I found this particular comment unnecessary. Again I would qualify this as an unhelpful prejudice and another example of pseudophilosophy. Feyerabend goes on to discount idealism by asserting that the adherent soon realises that expectations do not meet real world results. However such a statement in itself does not offer support for his assertion and either supporting reasoning or evidence is needed to justify this. Feyerabend also asserts that the use of standards in both idealism and naturalism results in the
‘barrenness of the practice they engender‘ p.223
Again, no evidence is offered and yet this is clearly a remark about real-world practice and again this is the domain where empirical support for his assertion would be appropriate. Feyerabend finishes by commenting on the structure of a free society in which different traditions have equal access to ‘education and other positions of power’.
In this chapter, Feyerabend starts to draw some conclusions. He asserts that there are problems with rationality and that there are limits on all methods. Again, I don’t think these points have been sufficiently justified in preceding chapters but this general conclusion seems reasonable. However he is not explicit about the methodologies that are being considered as alternatives to rationalism. He returns to rationalism and suggests that he is recommending an interplay between rationalism and practice so that each is guided by the other. He advocates against the systematisation of the treatment of theories as this will essentially create a conservative culture that will react against new theories. However, Feyerabend considers this topic in a very small chapter and is not sufficiently bold in his exploration. Feyerabend has not given consideration to the use of augmented science – that is the use of artificial intelligence solutions to support scientific inquiry and he also misses the possibility of a systematisation that is progressive, developing through appropriate processes which include outcome-based assessment of theory-generating and theory-developing methodologies. His other main suggestion in this chapter is to approach a given research area with an open-minded and novel approach.
In this chapter Feyerabend asks several questions about science. In this chapter, at the very end of the book, he asks ‘what is science’ and answers that it is difficult to define given the multiple branches of science that exist. It is curious that this point is made at the very end of the book and that the case of Galileo has been used to generalise to all of science. This statement supports my argument that a philosophy of science should be developed which examines individual branches of science and bases the inquiry on a firm foundation of knowledge. Feyerabend is again dismissive of medicine and again without any justifications. He suggests that there are fashions and schools in medicine and psychology. He asserts that medicine is not a science and that sickness and health are culture-dependent concepts. There is no justification for the assertion that medicine is not a science and he is presumably unaware of the sciences that underlie medicine some examples of which include anatomy, physiology, pharmacology, pathology, biochemistry, histology, neuroscience and genetics. Although many illnesses are culture-dependent, not all are. For instance, the course of Tuberculosis has a characteristic course which if untreated would be seen in different cultures. In other illnesses, normative data is important in the diagnostic criteria and here culture may play a role but this is to shape the expression of the illness, other people’s responses and health-seeking behaviour to name just a few examples. This does not negate the underlying illness however and it is curious that Feyerabend has invoked this observation. The only possible interpretation that would make sense is if Feyerabend thought that an illness was entirely a social construct and that if the culture were different the underying illness would disappear. While on technical grounds it can be argued that an illness could be denied within a culture this does not make it go away. It is difficult to know if this possibility is what is being suggested by Feyerabend as again he does not qualify the statement further. The existence of different schools and ‘fashions’ can result from a multiplicity of effective treatments for which there are skilled practitioners and Feyerabend’s point is unclear. He goes on to suggest science is used by society in a way that fits with the values of the culture.
In this chapter, Feyerabend goes on to discuss some of the events that influenced the views that he discusses in this book.
Firstly there are a number of difficulties that I have with ‘Against Method’.
1. Feyerabend is drawing conclusions about science. If the book was written with the intent of having an effect on the scientific community then his conclusions should be assessed in the same way as any scientific theory. Thus if the scientific community or affiliated communities are to act on Feyerabend’s conclusions, then those same conclusions should be validated through appropriate studies else there is the potential for resources to be wasted.
2. The conclusion from 1. can be generalised to the entire field of the philosophy of science if this field informs the practice of science. Thus in the philosophy of science there should be at least two domains of conclusions
a. Those that are entirely theoretical. These would consist of theories or hypotheses about how science works. In a similar vein to the hierarchical evidence base used in review articles in the Cochrane database, they should be assigned a value corresponding to a judgement of their validity.
b. Those that are tested empirically. Again there are different gradations of evidence and specificity of evidence which could be utilised.
3. A clearly defined field within science could be developed which incorporates the methods of philosophical inquiry and which would be concerned with theory generation, hypothesis generation, theory refinement and analysis of methodology and related areas. This discipline would be represented within each branch and sub-branch and would be conducted by scientists with training in this area as well as by philosophers recruited into this area.
4. I found the book could benefit from a clearer structure. While short statements begin each chapter, the subsequent assumptions and chains in the argument were difficult to find as they were interspersed with discussion and insights. I have observed a similar style being used in psychotherapy papers that I have reviewed but it makes it difficult to analyse the material and to get to the point quickly.
5. Feyerabend does not consistently divide science up into its constituent parts
a. The scientific process
(i) The ideal scientific process
(ii) The actual scientific process constituting social and anthropological aspects
b. The body of scientific knowledge
c. The image of science
By using these concepts interchangeably and at times confusing them with wider culture, there are opportunities for developing
6. In the introduction to the Chinese version and in later references to China, Feyerabend describes the effects of a mono-culture on a different civilisation when referring to science
a. He confuses science with culture
b. A socio-political trend is used to support his argument for a multiplicity of methods, a trend which has been reversed at the time of writing
c. He selects from the culture those aspects which fit with his arguments but neglects significant scientific developments in Chinese culture independent of Western science which supports the notion that science or the rudiments of science represent a universal human activity
7. Feyerabend states that his purpose in writing the book is not intellectual but humanitarian. This follows from his belief in 5. that science is a pursuit which suppresses alternative ‘traditions’. Nevertheless as noted in 5. in this argument he is confusing culture with science. Indeed he later writes that the way in which science is used is decided on by the culture. If this holds then it suggests that his book is written on the basis of false premises – a confusion of science with the perceived use to which it is put to by society.
8. Feyerabend refers to the ‘hardest sciences’ and implicitly refers to a hierachy within science which he makes no attempt to qualify
9. Despite asserting that the book is written on humanitarian grounds, Feyerabend repeatedly writes dismissive comments about medicine. Such is the brevity of the statements and their nature that I have referred to them as pseudophilosophy in that the statement masquerade as philosophy (by repeatedly appearing in a book about the philosophy of science) but are not justified by either reason or evidence.
10. Feyerabend criticises rational argument as a political manoeuvre and yet rational arguments constitute a significant portion of his book in his efforts to persuade the reader that ‘anything goes’
11. Feyerabend’s support of a languageless element in science circumvents the burden of proof that is required in the shared communication that is part of the scientific process
12. ‘Anything goes’ is a vague phrase. While it infers tolerant values, in concrete terms it does not sufficiently inform actions at a group or organisational level. While on an individual level different approaches will suit different people’s needs, at the organisational level, the absence of a structured approach will create the possibility of duplication and other inefficiencies. The absence of a direction in the form of strategies or goals makes evaluation difficult.
13. Feyerabend chooses Galileo as the subject of his inquiries. This choice has potential drawbacks
a. There are no living witnesses to the events described
b. He (and therefore we) are dependent on a translation of an Italian text
c. We are dependent on a translation of text(s) from the 16th century
d. Feyerabend selects the material to use in his arguments introducing possible confirmation bias
e. Feyerabend undermines the credibility of Galileo’s writings at several points although these serve as the central point for his arguments
f. Feyerabend does not appropriately consider the effect of cultural sensitivities on Galileo’s writings
g. Galileo’s achievements are judged on the basis of his selected writings and yet much of his thinking may have remained undocumented
h. The focus on a single period and area of activity cannot be representative of all of science
14. Feyerabend uses Galileo’s drawings of the lunar landscape as evidence of faulty equipment and yet Galileo himself apparently expressed a wish to return to draw the lunar landscapes to take into consideration confounding factors
15. Feyerabend negates the considerable number of telescopic observations that Galileo is purported to have made in establishing an empirical basis for his works and also in the development of an intuition
16. There is no focus on the considerable period of time over which the possibility that the earth moved had been considered. Indeed it had been considered by the ancient Greeks and was therefore one of the competing theories to the established view which Galileo came to challenge.
17. Feyerabend has argued in a dichotomous manner that the theory by Copernicus was in trouble because it had been challenged. It would not be unreasonable to suppose that every significant theory has been challenged and it does not necessarily mean that it is in trouble.
18. Again in a dichotomous manner, Feyerabend argues that falsification would destroy science. In practice, the effect of conflicting facts on a theory is a matter of judgement and this would be expected to occur in a probabilistic manner.
19. The suggestion that the physiology of the ancient Greeks differed from our own is highly speculative and the example of Egyptian canonical art supports the argument that art can be used for a specific purpose by society and that can impact on how art is expressed. This favours the impact of culture of biology in this particular case. Feyerabend has again chosen a point in the distant past which is less than ideal for closer inspection.
20. A number of Feyerabend’s arguments are not supported by empirical evidence when it is clear that this would strengthen his speculation
21. Much of Feyerabend’s discussion crosses into interpretations of ancient history which are contested elsewhere and which moves the focus away from a philosophical inquiry into science
In summary, I thought that Feyerabend tackles the subject of science in an overly simplistic manner, focusing in particular on one area of science several hundred years ago. The choice of subject material means that the variable quality of evidence being used in his arguments distracts from the main discussion of a philosophy of science. Thus challenges to a number of his arguments involve a closer look at historical events and this reveals a great deal of uncertainty in a number of areas. I found his arguments against falsifiability and rationalism unconvincing. I thought that a number of his arguments used absolutes whereas science in the real world operates within the bounds of probability and pragmatics. The use of probability and a pragmatic approach to science has produced useful results and mirrors expectations people have of their interactions with the external world. In other words, we get simple things right most of the time (walking, talking, eating and so on), simple things wrong some of the time (e.g falls) and complex things are a lot more variable.
I’m going to use an analogy now. It’s not a great analogy but I thought it might be useful in reframing the discussion. When a light (lightbulb) is switched on – most of the time it will come on, but at some point it won’t. However when we switch a light on, we expect it to come on and in most cases we will be right. This approach is incredibly useful in helping us to get on with our lives. The approach that Feyerabend takes in some of his arguments is much akin to focusing on the potential of the light to not come on and basing expectations on this and generalising the consequences to all other scenarios. Such an approach is useful in considering how to respond when the light doesn’t come on, but for most of the time it’s not too useful. To take the analogy further, Feyerabend might recommend gathering candles, lanterns and a variety of other methods for generating light in order to meet the eventuality of the light not working and even using these methods simultaneously. Thus in one room we might use a light bulb and in another, we might use a candle and so on. However Feyerabend has been sufficiently vague that in this analogy we do not know what alternatives we should be using, how costly it will be, where we should be using them, how often we should be using them. As for why we should be using them, all we know is that sometimes the light bulb might not work and therefore we need to use alternatives as well.
Taking this further, it is possible to see that the light bulb has become the dominant domestic light source in many nations. Should we say in a slightly light-hearted manner that the light bulb has displaced other light sources ‘without a single glance in their direction’ in an imperialistic march into the household. This of course would be ridiculous. If there is a demand for the light-bulb then the market will fill this demand soon enough. People have a choice between candles, light-bulbs (and their successors) and any number of alternatives. If they want to light their house with candles they can do so. The light-bulb though has still been dominant. It is convenient, effective and reliable. There is a place for alternatives when an atmosphere is to be created and these will be occasionally. The light bulb has been adopted by large groups of people in different communities. Choices have been made and the alternatives considered. In just the same way, scientists have presumably considered alternative careers or may immerse themselves in the arts when outside of their scientific endeavours. People have the choice to believe in alternatives to the scientific knowledge that is provided. To believe that science can be like a fallen empire however is another matter. Money has been successful and regardless of empires rising and falling, has remained. Making clothing, building shelters, creating music. These are just a few examples of activities which have persisted through empires across the world. So too has science. Science offers us the promise of a knowledge of the world around us, a knowledge that is tangible and useful to us. A knowledge that we can relate to. There may be many ways of getting to that knowledge but common to this is a thinking, exploring community of people across the world testing this and that. Seeing what works and what doesn’t. It is this network with a common aim, a plan of furthering knowledge about the world that partly underlies this success. Perhaps the underlying methodology is still advancing as it is being tested.
This network has different properties to the networks in other endeavours. Perhaps it is characterised by a willingness to readily embrace the fruits of science – technology – in the pursuit of its goals. Maybe in a hundred years this network will have properties far different to those of today. While this happens, people in other networks, from other traditions observe these networks with interest and their summated conclusions determine how science is perceived. In such paradigms, the most successful networks regard each other with mutual interest.
Although I have been critical of many of Feyerabend’s arguments as outlined above, I agree with some of the main values he espouses – that we should be respectful of other traditions. In the process of discussing these values however, Feyerabend puts them squarely up against science and here it is that I disagree fundamentally with him. In many ways, I found ‘Against Method’ to be a political statement although I have read it for the purpose of gaining a better understanding of the philosophy of science. In the end though, I was not persuaded by Feyerabend’s central argument that ‘Anything goes’. For me, Feyerabend would have been more successful had he demonstrated this principle with several examples from current practice. I would argue that a systematic methodology underlies an incremental progress. A multiplicity of methods satisfies the need for individual expression in scientific work but I would argue that this is far different from the structured approach that is required at the organisational or ‘network’ level to draw on the earlier analogy. Indeed it is the forging of global networks that is so important to the advance of scientific knowledge and the benefits for humanity.
1. Feyerabend Paul. Against Method. 3rd Edition. 1993 (2002 Reprint).
7. Swerdlow, N. and O. Neugebauer, 1984, Mathematical Astronomy in Copernicus’s De Revolutionibus, 2 vols., New York: Springer-Verlag
Intuition has been examined by the philosophers Immanuel Kant and Henri Bergson
You can follow ‘The Amazing World of Psychiatry’ Twitter by clicking on this link
You can listen to this post on Odiogo by clicking on this link (there may be a small delay between publishing of the blog article and the availability of the podcast).
You can follow the TAWOP Channel on YouTube by clicking on this link
If you have any comments, you can leave them below or alternatively e-mail firstname.lastname@example.org
The comments made here represent the opinions of the author and do not represent the profession or any body/organisation. The comments made here are not meant as a source of medical advice and those seeking medical advice are advised to consult with their own doctor. The author is not responsible for the contents of any external sites that are linked to in this blog.