1. What do scientists do?

 Inductionism, Kuhn, Lakatos and Popper's paradigm.
 Science, technology, gadgets and public health.
 Reproducibility, models and mechanisms and peer
 review. Publish or perish.

Inductionism
------------
 It used to be believed that inductionism was the method of science;
 certainly it was the traditional belief of non-scientists that this
 was the way that scientists proceeded. Facts were collected,
 regularities observed, and scientific laws were deduced from the facts
 so collected. A common example given was the joint work of Tycho
 Brahe and Kepler, the former collecting many observations of the
 position of the Planets and the latter deducing therefrom the famous
 Kepler's laws which enable the motion to be described by mathematical
 equations. Similarly Darwin observed the structure and behaviour of
 many animals and birds, finally enunciating the theory of evolution, a
 principle which, while not expressible mathematically, satisfactorily
 explained for him the diversity of life. At its most basic level
 induction says that if we observe the Sun to rise every 24hrs
 sufficiently often, then we are entitled to "induce" a law that it
 will always so rise.

 The history of inductionism is long and studded with famous names. The
 inductionist claims that the way forward in knowledge begins with the
 collection of facts, collection without pre-suppositions about what
 stucture may be later revealed, or even relevant. The first overt
 proposals for the extension of knowledge by the use of induction seem
 to have been made by Francis Bacon (1561-1626). He studied law at
 Cambridge and held a succession of posts under Elizabeth I and James I
 eventually becoming Lord Chancellor in 1618.

 In 1605 he published a book called "The Advancement of Learning" in
 which he argued against mysticism and tradition as the enemies of
 progress and argued in favour of experimental science as a means of
 improving the human condition. This was followed in 1620 by the "New
 Organon". The title was a reference to the work of Aristotle in which
 the latter demonstrated the rules of logic or deduction. Bacon was not,
 however a scientist or experimenter himself; his attempt to preserve a
 chicken by stuffing it with snow ended in a chill, bronchitis and
 death, and this is the only recorded experiment we have of his! Bacon's
 influence on the real experimenters of his age, Gilbert and Harvey, was
 negligible Harvey was in fact Bacon's own doctor! However Bacon's court
 position made experimentation fashionable and acceptable to the wealthy
 classes.

 In Bacon's "New Atlantis", published in 1627, he describes a fictional
 research institute, "Salomons House" where the idea of a scientific
 co-operative or research project appears. The declared objective is
 "the knowledge of causes and the secret motions of things; and the
 enlarging of the bounds of human empire, to the effecting of all things
 possible". There seems to be little doubt that this fiction had an
 impact on the founders of the Royal Society which had an enormous
 impact on English and indeed world science.


Scientific Revolutions.
----------------------
 Thomas Kuhn is a modern philosopher of science,
 Kuhn begins by saying that for most historians of science, "science"
 is the collection facts, theories and methods collected in texts
 current at a particular time and that the people who add to that corpus
 of knowledge are scientists. The history of science then
 becomes the chronicle of the piecemeal accretion of knowledge and
 techniques and an explanation of the errors and mistakes made in so
 doing. And Kuhn takes the view that a careful reading of original
 sources shows that ancient science, for example Greek dynamics, is no
 less "scientific" or idiosyncratic than the views held today. Kuhn
 defines as "normal" scientific activities as those of scientists who
 inhabit a mental world conditioned by their education and tradition.
 He claims that normal science "suppresses" new discoveries because
 they are "subversive" and uses a terminology that treats science as a
 form of politics. Only when scientists can no longer "evade"
 inconvenient facts does a scientific "revolution" take place and the
 new knowledge become absorbed into "normal" science. Scientific
 revolutions for Kuhn are the familiar changes recognized by
 historians, those associated with Copernicus, Newton, Lavoisier and
 Einstein.

 Because a new theory impinges on the prior work of those specialists
 working in the same field, says Kuhn, it will be resisted by them.
 Normal scientists, in Kuhn's view, work totally within the "paradigm"
 to which they give their allegiance, exemplified by him as a seminal
 work such as Newtonian Dynamics, and normal scientists do not disagree
 over the fundamental laws, theories, application or instrumentation
 associated with their paradigm. Kuhn discusses theories of light to
 show that different paradigms were used in different historical
 periods. Before Newton there was no paradigm and no common set of
 beliefs and the choice of experiments or phenomena to describe was
 un-constrained. This, for Kuhn, is "less than scientific". This
 pre-paradigmic activity, where authors were more concerned with
 converting others to their views than working within an agreed basis,
 is seen by Kuhn as radically different from post-Newtonian work on
 optics.

 In the absence of a paradigm ALL facts may seem equally relevant and
 it is difficult to be other than random in gathering facts. Thus the
 inductive method is doomed to failure because the absence of a
 paradigm, any paradigm, causes uncritical fact collection. Once
 electrical researchers had characterised electricity as a fluid,
 creating a paradigm, someone was bound to try and bottle it. The
 result was the Leyden jar, or capacitor which, Kuhn says, could hardly
 have been discovered without the framework provided by the paradigm.
 The further result of the existence of a paradigm is that the basic
 facts etc. are now agreed and need no longer be described for the
 general worker in the field, they can be encapsulated in the textbooks
 for students, and leave the field free for research on extending the
 paradigm among the professional practitioners of the paradigm, the
 normal scientists. This research will be published in Journal or
 reports and will be inaccessible to those who have not undergone the
 apprenticeship of studying the texts. The normal scientist will only
 write volumes of autobiography or text-books. It takes a
 revolutionary to write a book with a new paradigm addressed to a
 receptive readership.

 A new paradigm achieves its revolutionary success by its power to
 explain phenomena more convincingly but also by containing implicitly
 a program on consolidation and "normal" scientific work. Most
 scientists throughout their career are concerned with "mopping-up"
 operations according to Kuhn. Normal experimental science has three
 foci.

 Firstly the facts which illustrate the particular nature of the
 paradigm can be determined more accurately and in a larger variety of
 situations. For example, in astronomy, the positions and magnitudes
 of the stars, the periods of planets. Complex experimental equipment
 is developed to assist in this program.

 Secondly there is the determination of those facts which can be
 checked against the predictions of the paradigm theory. Kuhn states
 that there are usually fe such predictions, and again they usually
 require large investments in time and equipment. As an example he
 quotes the predictions of Einstein's theory of relativity.

 Thirdly there are experiments to resolve remaining problems and
 ambiguities left unresolved by the paradigm. Again Kuhn refers to
 astronomy pointing out that Cavendish's work on the value of
 the universal gravitational constant which is a resolution of material
 in Newton's Principia

 Normal theoretical science can be divided into the same type of
 activities. Firstly the use of the theoretical paradigm to predict the
 occurrence of phenomena, for example astronomical ephemerides; this is
 usually seen as a technical problem, not suitable for advanced work.
 Then there are extensions of the theory to further cases and
 reformulations of the theory with different enphases. Kuhn believes
 that normal science is is not concerned with novelty, but with puzzle
 solving. Failure to reach the expected result is considered a
 failure. Kuhn believes that only soluble problems are attempted;
 problems which will require skill certainly, but which with technique
 can be soved within the terms of the paradigm. "Difficult" problems
 will not be attempted because they invite failure.

 It seems to me that Kuhn's work is an attempt to describe how people
 work in science by examining their published work and not by asking
 them, a very academic exercise. Increasingly experimental scientists
 are sceptical of the work of their theoretical cousins and do not, it
 seems to me, elevate their paradigms to such a predominant position.
 In particular one on the failings of Kuhn's work is that it is based
 very largely on physics, where theory is predominantly taught. The
 situation in biology is, I think, different.

 Popper - Conjectures and Refutations
 ----------------------------------------
 One of the problems in talking about science with non-scientists is
 the use of a specialized vocabulary that has meaning independent of
 the speaker. This can be contrasted with other subjects of human
 endeavour, where what a word means often depends on who the speaker is.
 Nowhere is this more obvious than in the the use of the word "law".
 In elementary science courses we are taught about Ohm's Law, Newton's
 Law and so on, although very few of the scientists of the 20th Century
 have given rise to "Laws", and perhaps this is due to a shift in
 perception of what a scientific "Law" is. For the Greeks there were
 "Laws of Nature" which were thought to be rules laid down by the Gods
 about how the world should work. Thus it was as possible for these
 laws to be "broken" by other Gods and miracles to be worked. But a
 scientific "Law" is descriptive, not prescriptive, and, as such, can
 be superceded at will by a better description.

 The distinction was well understood by the Scottish philosopher Hume
 (1711 - 1776) who pointed out that the other characteristic of
 scientific laws was that while inductionism was seductive it was not
 logical. There was no necessary law that connected cause and effect.
 Just because the Sun rises every morning is no guarantee that it will
 rise tomorrow. What we think of as "causation" is merely succession
 in time.

 No one has offered a satisfactory solution to Hume's problem of
 inductionism. If science only advances by induction of laws from
 data, and induction itself is not firmly based in logic, then, so the
 argument goes, we are on very uncertain ground. Most practising
 scientists merely ignore this problem, some finding solace in
 religion, but it is a strange astronomer who worries about sunrise
 when he goes to bed.

 Karl Popper is almost unique among philosophers of science in that
 some scientists, at least, feel that what he has to say about science
 actually corresponds to what they do in their working life. Popper
 points out that although Hume's problem is still there, a scientific
 law is not, in some sense, symmetrical. What he means by this is that
 you can make many observations that confirm a scientific law but
 although these may make you more certain, it is only psychological,
 like a climber's psychological belay. When a climber cannot protect
 himself with a strong belay he may use two or more weak ones, although
 he "knows" that he is not protected, he may gain enough courage to
 make his next move. But one observation which goes counter to the
 law is sufficient to refute it. This many/one distinction between
 confirmatory observations and a single refutation is the keystone of
 Popper's theory.

 The typical example given is that of the birdwatcher and the swans.
 The birdwatcher sees a large white bird and calls it a swan. Many
 similar birds are seen on subsequent occasions, and the birdwatcher
 induces the scientific law - all swans are white. But one day she
 sees a large bird with all the same characteristics - but it is BLACK!
 This single observation is sufficient to disprove the law, despite all
 the other white swans that were seen.

 Further, Popper points out that for something to be "scientific", for
 it to be possible to describe something by a "scientific" law, the law
 must be "testable". That is to say, it must be possible to describe
 an experiment and an observation which would refute the law.
 Applying this criterion of demarcation there are many theories which,
 while descriptive, cannot be described as scientific. One of the
 most famous examples given is that of Freud's psychoanalytic theory.
 While saying nothing about "truth", Medawar, a distinguished scientist
 and Popperian, destroys any pretensions of Freud's work as science.
 It is just not possible to design and carry out an experiment which
 would refute Freud's theory of the unconscious mind.

 Lakatos - Research Programs
 ---------------------------
 Imre Lakatos suggests that in reality science may not proceed like the
 birdwatcher and the swans. He suggests that when the black swan is
 found a research program is initiated to discover the pollutants that
 coloured the swan. None are found, so a grant is obtained to
 investigate the genetic structure of black swans with a view to
 dicovering the cause of the mutation which caused the white swans
 offspring to become melanic. None is found so another research
 program is started to investigate whether a form of accelerated
 evolution is resulting in all swans becoming black, and so on.

 Science, technology, gadgets and public health.
 ----------------------------------------------
 We are increasingly dependent for our functioning on what I would call
 gadgetry, devices and products that derive directly from scientific
 advances this century. What seem to distinguish this mode of existence
 from that which we had previously is that we can no longer understand
 or fix the gadgets when they go wrong. In the days of the blacksmith it
 was easy to see how to make a horseshoe and the equipment to make one
 could itself be made by a man working alone. It is very different with
 a pneumatic tyre. Similarly a cats-whisker radio is easily made by 9
 year old, but a modern television set is often junked when it goes
 wrong, the number of people who can fix them is limited and their
 charges are high. When it costs 4 pounds an hour to clean a room, 60
 pounds an hour to clear a blocked drain, 100 pounds an hour for a
 lawyer and 1000 pounds an hour for a top male model we can see how
 society values the varying services. You can get quite a good scientist
 for 20 pounds an hour and an adult education lecturer for 13 pounds an
 hour. We can survive quite well without either, but without our gadgets
 we would be lost.

 List of gadgets.

 Convenience foods; freeze dried coffee,post ww2; frozen food 1950.

 Chip-based calculators 1972

 Lasers - Bar coding books/goods/carriages,Surveying/ranging/compact
 discs/welding/surgery/telecomms.

 Antibiotics.penicillin/streptomycin.

 Microwaves.ovens/air traffic control.

 Video.entertainment/security/inspection in hostile environments

 Contraceptive pill 1960 Carl Djerassi in California

 Smoke detector using radioactive source

 Time measurement by quartz crystal.

 Plastics contact lenses/containersMan made fabrics, nylon/terylene
 Teflon non-stick frypans/clothing/insulation heart valves/replacement
 joints.

 Shatter proof glass.

 Satellite telephone and video links optical fibre low interference
 cables

 Computers/copiers/facsimile transmission Holograms as security devices.

 Reproducibility, models and mechanisms and peer review.
 ------------------------------------------------------
 The most important fact about scientific knowledge is, it seems to me,
 its reproducibility, across time across space, and independent of the
 observer. This gives rise to as yet unsolved problems in respect of
 Quantum Theory, which seems to show that there are phenomena which in
 some way depend on the observer. Thankfully these phenomena are all
 at the microscopic level and have no effect in the universe of trees,
 houses and people which we normally inhabit. Speculations as to the
 meaning of Quantum Theory for everyday life can be left to
 philosophers and mystics but we must be aware that such theories are
 models of reality and are themselves subject to revision from time to
 time. Indeed, the practical scientist is only too well aware that
 mathematical constructs, such as Quantum theory or Newtonian dynamics
 are models which are subject to revision and sometimes radical
 revision in the light of observal, practical measurements. An
 astronomer has likened theories to great battleships, assembled by
 ingenious workmen, which steam around in the ocean of truth.
 Sometimes a battleship encounters a small bubble of observational
 fact, and to the surprise of all, the battleship capsizes and sinks,
 while the bubble sails serenely on.

 By peer review, by the repetition of experiments by others in
 different places and different times, the scientific community ensures
 that the reports of individuals are confirmed and embodied in
 scientific knowledge. Examples of efects that failed this acid test
 are the production of neutrons from Zeta, polywater and cold fusion.

 Publish or perish.
 ------------------
 One of the slightly disturbing developments of the last fifty years is
 the way in which scientific developments have become almost completely
 the prerogative of large corporations and universities. This has
 resulted in two opposing trends. For large corporations it implies
 that scientific knowledge is an asset and is subjected to legal
 protection by means of patents and copyrights. Documents are marked
 commercial in confidence and scientists and technicians have to sign
 away the right to publicise work in which they are involved. On the
 other hand in Universities it is necessary to publish work in order to
 achieve tenure or peer approval. This results in the publication of
 much that is of marginal importance and the inflation of routine
 results. There is an old joke about "Biologically active
 substances", these are defined as those substances, which, when
 injected into a laboratory rat, produce a learned paper. Derek de
 Solla Price, a Cambridge historian of science, showed 40 years ago
 that the volume of scientific papers was increasing exponentially.
 this results in another joke; the undergraduate scientist reads
 text-books, the graduate scientist reads papers, the established
 scientist reads abstracts, when promoted, the professor reads titles
 and on retirement the scientist reads detective stories.


References
----------
1. B.McGee
 Popper
 Fontana 1973

2. T.S.Kuhn
 The Structure of Scientific Revolutions
 University of Chicago Press 1962

3. Anthony Quinton
 Francis Bacon
 Oxford 1980

Last updated 23rd December 1998