Kuhnian Scientific Revolutions

There can be few publications on the philosophy of science that are as famous as "The Structure of Scientific Revolutions" by Thomas Kuhn. Frank Payares has written helpful synopses of it (first – second). Chalmers ("What is this thing called science?") devotes a chapter to it.

Kuhn's approach (to my reading) emphasised more the historical-factual-social aspects rather than the cause of revolutions. His core conclusion was that science has, in the past, evolved in a "normal science – crisis – revolution – retooling – normal science" cycle. The historical extrapolation is that this will continue to cycle. Now this is quite contrary to the theme of a recent book by John Horgan, "The end of science". In some ways Kuhn left many people dissatisfied and sceptical because he seemed to imply that science was "non-objective". However, to make a point effectively it is useful to play devil's advocate and caricaturise things to the point that the essential structures become obvious. Swapping paradigms (if not examined more deeply) may seem more like a religious conversion than a logical necessity. Chalmers explores all this in detail in the last chapter of his book and comes up, eventually, with his own concept where "unrepresentative realism" is a target in science rather than a search for the "truth".

The problem with all this philosophy stuff on science is that it is all seems airy fairy until you have recognised that there is a problem. Furthermore, when you live through a problem and see how a revolution could resolve the problems, you begin to realise that the inertia of conviction that afflicts many of the top cognoscenti carries them on, like a fleet of super-tankers, long after they should have examined their course and seen the advantages of turning about.

Like all things, until you feel some desire to know about something (because it offers you some advantage/power) you won't dedicate the time to study it. Until you assimilate it you won't understand it; until you understand it you won't appreciate its importance; and until you appreciate its importance you won't want to assimilate it.

Kuhn's account, emphasising as it did the historical-factual-social aspects, left the causes more dispersed in his essay (Barber 1961 addresses the causes well). I have become interested in the reasons why revolutions occur and that begs an emphasis on proportions (which are an important aspect of understanding - see below). These points are:

Models – at base science is more about "understanding" than "knowing". "pernosc(i)ence" is the real goal (ie, to thoroughly understand). Now, at base, the only way we can understand is by converting everyday experiences into a data stream of neuronal firings where the brain uses some algorithm to represent (and model) its environment. The closest we can get to any "truth" is to model it in this way. Now the modelling that scientists trust most are mathematical models. Mathematics tends to iron out errors incurred by inexact modelling and mathematicians are notoriously cruel to presumptions. Kuhn stated that:

"Philosophers of science have repeatedly demonstrated that more than one theoretical construction can always be placed upon a given collection of data."

These theoretical constructions (or models) are like analogies. Some are sophisticated mathematical analogies others are more mundane everyday analogies (see "At its simplest" at this site for one of these). These models may be parochially adequate (Newton), more universally applicable (Einstein) or they are alternative explicatory analogies (eg, light may behave as particles or as waves). Often, a more encompassing model (say quantum theory) will help to incorporate earlier – more na├»ve – models (particles vs waves). Whatever, they are all models and subject to the same problem – they are simplifications, conceived in human brains (using neurons, action potentials, synapses and networks) to help us better understand and manipulate our world. They are, at best, only approximations.

Assumptions – these are the very devil in science. Science tries to exorcise them. The ones that cause most trouble are the hidden assumptions – Bacon's idols. They are hard to recognise in prospect but stand out like beacons on a clear night once they have been rumbled. Assumptions are the (probable) reason that science and religion have always appeared to be at loggerheads. This ought to be more of a problem for the religious believer than the scientist (but, strangely, this may not be so). The religious believer will often bring along assumptions or dictates that (s)he sees as inviolate. The scientist should know that no concept, however apparently unassailable, can be considered inviolate and that (s)he frequently makes assumptions that (though probably valid) might eventually prove to be an impediment to true understanding. A religious fervour for a particular belief in science is an anachronism. Such belief drags an unwary discipline to the precipice of a Kuhnian revolution.

"Biologists can be just as sensitive to heresy as theologians" - H.G. Wells.

Religions have frequently cried "heresy" – and professed Christians have often been in the vanguard. But, for them, the very person (real, embroidered or apocryphal) they claim to follow was a philosopher-heretic. Heresy is good for all disciplines – it provokes open debate and exposes the weak links. It is the fear of having to face up to these weak links that makes people emotional (ie, it exposes their "animal").

Power – at base it is all about power – not the "truth". The truth is a bankrupt resource in science. An extreme example of this is the following. Say the truth was that god created life in just the biblical way that fundamental Christians believe. Would it provide any power in manipulation of genes, understanding the reason why new world monkeys are different from the old? (and a host of other observations). No – their definition of God is that he was powerful enough to zap it there, at will ("here's one I prepared earlier"!) in the way stated by Genesis. OK, it's possible (if, to my mind, vanishingly improbable) but, were it the "truth", God has either gone to extreme lengths to make us believe it happened by gradual evolution (to test our faith?) or he used a mental algorithm to conjure up all these life forms that now makes us believe it happened by gradual evolution. Regardless, the powerful assumption (the one that lets us predict and manipulate the system) is to accept the apparent process of gradual evolution. I know this is an extreme argument but I think it is valid. Getting hooked on the truth puts blinkers over our eyes when it comes to searching for predictive/manipulative power. Let's go for the power and leave the truth to fend for itself. My bet is that the power can be trusted to deliver us close to the "truth" anyway. The truth is like a mirage – we think we can see it but as we are getting close the horizons change, new vistas appear and the truth recedes. Note that, if faith healing had been 100% effective in treating illness, we would never have bothered to develop medical science; so it's the power that makes a science grow.

Major scientific advances tend to take us from parochial to more universal perceptions. The shift from "the truth" to "power" means that we can continue to use more parochial models where the errors incurred are negligible and their application is simpler. Newtonian physics is fine for navigating satellites around the solar system and pretty much useless in bubble chambers.

Power again – tooling up for serious research produces enormous inertia (a sort of kinetic potential or, more simply, power). The system is up and running; much investment has been placed in it. A community is not likely, on a whim, to re-tool and head off in another direction. It will take something substantial to precipitate such a change of direction. We are fully aware of how powerful the "lemming" instinct is – people will look around and behave as if

"one hundred thousand lemmings can't be wrong".

(My apologies to lemmings who don't seem to be so daft – this "fact" about their behaviour seems to be, largely, "an old wive's tale".)

What is more, a community of interacting minds – I guess – has a consciousness much like a community of neurons. As this emergent property "thinks" more slowly (months to generations contrasted with the microseconds to minutes of a single brain) we are instantaneously blinded to this level of consciousness (much as a single neuron probably has no clear appreciation of the brain's consciousness). Unless we look hard for it and over a long interval (eg, over generations) we will not appreciate its existence. Science advances more through such communal consciousness that through individual consciousness. The contribution of an individual who "wants to rock the boat" is probably more akin to an epileptic focus than to the step by step progression of social thinking. We will gain more communal power if we explore ways of harnessing, recognising, facilitating and communicating with this high-order, social consciousness.

Power – yet again – but this time it is not the power of models and manipulating our environment. This time, it is the more sinister social powers of position, importance and influence. These are pernicious and are frequently at the root of " (mis)conceptual skyscrapers". It is this power that so often leads to the cry of "heresy". It is this power that allows a flawed edifice to grow to gargantuan proportions. It is this power that is largely responsible for the severity of the eventual conceptual earthquake that will devastate the respective community when it strikes. Now, I suspect, this will be true of biological science. Should the relevant cognoscenti have had more insight, and been better able to discriminate their animal from their intellectual natures, then they would have been able to pull down the bricks earlier and rebuild them without major upheaval. The longer a revolutionary concept is (r)ejected, the bigger the eventual collapse becomes. Leaving pioneers alone to slowly piece it together builds up a debt of the very power they crave, stacking it up for the pioneers' future credit. I have a suspicion that awarding credit to an individual who is dead carries a much smaller threat to influential-power than awarding it to an individual who is still alive.

But open minded-ness does bring its own problems. It's all very well being open minded but how many new ideas can we consider without halting the research process? We can't stop and reconsider fundamentals every time some crackpot comes along thinking they have a good answer for all the problems. However, that is no excuse for NOT highlighting the anomalies and searching out the stranded tautologies. An important list to hang in every lab or theorist's office is "These are the known anomalies that the current paradigm fails to resolve; if you notice any stranded tautologies, please list them below".

Proportions – have you seen those pictures of the motor and sensory areas of the brain and how much space is dedicated to the lips, tongue and hands? The result is a distorted picture of a human who has enormously expanded lips, tongue and finger tips. Now, the development of science (the parochial way in which it accumulates its component facts) leads to the same sort of non-proportionality. Just about every idea previously developed in science will probably retain its parochial value but this value can often be lost if it is extrapolated to the whole. It is "getting things out of proportion" that leads to the real crises in science. This is pertinent to the immunology community. For example, "life without lymphocytes might be very awkward but, without phagocytes, it is impossible". Now, you wouldn't pick up this proportionality if you opened up a 1995 textbook of immunology. There are many more examples of non-proportionality in this subject as it stands today. The reason is that the accumulation of "facts" in a science occurs in a non-proportional way. It follows whim, ease, fashion, precedent and the prospect of prizes. We tend to forget that this is a serious Achilles heel in the research process.

So, at root, it is their own animal heritage (impediment) and flawed reasoning that allows scientists to continue paddling up blind creeks long after a dispassionately applied logic should have told them to turn about. This and the inertial property of various "powers" is what makes Kuhnian revolutions inevitable.

It is a salutary principle to believe only in what you trust and to be suspicious of all views paraded as immutable truths. Before lending credence to any purported intellectual content in absolute views, examine carefully what primitive animal responses are polarising the holder's perspectives. Power and authority are frequent culprits and insight will often be dim. The views could eventually prove to have justification but you are likely to be trumped in a power game if you go straight to examining the intellectual content.

It is of considerable interest that a true, "big" Kuhnian revolution has not occurred in science since Kuhn wrote his book. For this reason, doubt persists that Kuhn's view was valid. I predict that this disbelief will soon evaporate.

This page, up to the last paragraph, was last updated on 18/11/11. (Neurone changed to neuron.)

Further notes.

The vast majority of scientific research is differentiative. For that reason, calling it "re"search (from rechercher – to seek or look for again) is arguably inappropriate unless there is an degree of re-interpretation. This is an important point because a perspective has emerged (particularly in bio-medicine) that sees all but empirical (differentiative) studies as inferior science – even non-science (allusion to non-sense intended) or valueless science. Most conventional studies are a search for greater and greater detail using more and more sophisticated equipment. Historians are more likely to do "true" research ("seeking again") when they are studying a subject. Re-interpretation is a major part of their quest even though it is founded on much detailed differentiation as the groundwork for this re-interpretation. Integration (which involves a lot of re-interpretation) is far less apparent in bio–medical journals. Re-interpretation is most dramatically evidenced in a Kuhnian paradigm shift. Few people, in doing their research project, aim for integration across vast landscapes of previously unlinked (or loosely linked) subject matters.

My experience is that differentiative studies are easily identified within their title. Equally, articles containing some attempt at re-interpretation or integration are usually identifiable within their title. One give away is the number of terms like IgAs, TLRs, nfkappaBs, p53s, BAXs, T-regs, T-h2s, statistical significance and so on . . . and so on . . . that appear (especially) in the title or in the summary. Don't misunderstand! These studies are the bedrock of authoritative science but they become mere collections of observations in the absence of any interpretation by integrative science. Indeed, however hard the observer tries, observations are nearly always viewed and interpreted using the language and imagery of extant, accepted assumptions.