I've heard various accounts of the origin of what is known as
scientific method. Although it began emerging earlier, it gained a
significant boost in the hands of The Royal Society, in London, during The
Restoration (when the later Stewarts were invited back to the throne, after
King Charles I had lost his head over the balance of power within the English
state). This happened in the aftermath of the civil war (and The Protectorate,
which put the English off republicanism) and I
have seen it argued (not
unpersuasively) that the founders of The Royal Society's approach to doing
science was motivated by the desire to avoid conflicts. That provoked me to
want to elaborate on the cultural machinery from which science seems, to me,
to have sprung: the system of justice. In the case of The Royal Society, that
sytem is the
common law one used in Britain and various other
countries, which mostly speak the same language
as me, so I'll describe this as the
anglic system of justice.
The principal mode of trial in the anglic system of justice is the
trial whose salient features are its witness process, the common law
process of precedent and each juror's rôle in deciding (in principle
independently) both the credibility of the evidence and the true conclusion to
be drawn from it. It is notable that the early
(that's what early scientists, notably the founders of The Royal Society,
called themselves) were educated alongside, and generally belonged to the same
social strata as, the judges and magistrates of their day: thus these
processes were an integral part of the cultural milieu in which they
lived. Since natural philosophy tended to concern itself with
truths rather than the particular truths of some case in dispute, it
transmuted that process to suit its ends: but I find science's consensus
process intimately similar in spirit to the anglic (or
system of justice.
The anglic process of trial by jury has evolved substantially over the last millennium, but some rough truths pervade it. [To what extent other countries' systems of justice shared the same features, I know not: I single out the anglic system only because I know a little of it, without meaning to suggest it was unique in these respects.] Crucially, it is given to the jurors to decide issues of fact, based on evidence presented before them. The judge's rôle concerned only matters of law, notably what relief, if any, either party is entitled to in light of the facts, as established: and, in this, he was bound by the precedent, though statute had priority. In passing judgement, the judge explains his reasons: this educates as to what the law says; makes clear how his judgement follows precedent; and justifies, where it does not, his departures from precedent. The whole process is conducted, with rare exceptions, in public: thus can justice not only be done but be seen to be done – the public may be unable to affect the outcome, but the need to sustain the public's faith in the system of justice constrains the judge to apply rules fairly.
Where natural philosophers wished to persuade their peers of new ideas,
and could not do so by traditional philosophy's means – rhetoric
reference to authority (that is, to earlier writers) – they
turned to the other standard of proof (or, if you will, method of resolving
conflicts) available in their culture: the trial by jury. While one can
construe the spectators at a public demonstration (and other scientists
reproducing a result by repeating the published experimental protocol) as
witnesses, so as to draw the parallel in terms of their being available to
call as witnesses in an imagined trial, I am more inclined to cast the matter
in terms of each natural scientist as judge in the court of his own mind, with
the spectators serving as jury.
An idea proposed is a plaintiff, seeking as relief that it be recognised as true; the experimenter is its advocate. An idea attacked is defendant; the experimenter is a prosecutor seeking its condemnation. The experimental aparatus is the witness, the procedure of the experiment its interrogation and the experiment's results its answers. The observers (and those who duplicate the experiment elsewhere), as jury, decide the matters of fact established by the experiment; natural philosophers to whom these are reported judge whether these established facts entitle the proposed idea to be recognised as true. In deciding, they give their reasons: do the established facts bear on the question in hand; if so, do they suffice to decide it; if so, which way do they decide it; and why ?
In similar vein, consider the anglic justice system's use
precedent – this parallels the logician's precept of
consistency. As part of the
being seen to be done aspect of justice, if
a court allows (or rejects) some mode of argument or form of evidence in one
case then it (along, generally, with its peers) is bound to allow (or reject)
the same mode in other cases, unless clear grounds can be shown to distinguish
the latter from the earlier case. Failure to follow precedent exposes a court
to suspicions of capriciousness – the public may fairly object that at
least one of the cases, being handled differently than the other, was not
In effect, a court over-turning a precedent repudiates its own earlier
judgement: consequently, courts tend to do so only where the alternative would
be manifestly unjust. Both when over-turning precedents and when there is any
possibility that some will suppose their judgements to be at odds with
precedents, judges have thus taken great care to give, alongside the
conclusions of their deliberations, careful explanations of how they arrived
at their conclusions. Over the centuries, courts in the
jurisdictions (i.e. those which respect precedent) have – most
conservatively, and generally with the greatest deliberation – thereby
evolved a system of procedures which, when followed, command a high degree of
respect for their ability to lead to just outcomes.
Just so, if a natural philosopher refused to be persuaded of some view by its proponent's experiments, particularly where the same experiment persuaded its other witnesses, his peers might require him to explain the grounds of his obstinacy. If, in other disputes, he allowed himself to be persuaded despite the same grounds being applicable, his peers might justly impute some prejudice on his part. Thus, in effect, the community – by demanding at least self-consistency in modes of reasoning – mirrored the precedent process in a way that constrained each natural philosopher to reason consistently.
Such a demand of self-consistency is ultimately bound to drive out
assorted particular modes of argument. If you reject one claim on the basis
that the man propounding it is a scoundrel, yet accept some other claim, I
shall show you a scoundrel who propounds it and ask why you still accept
it. If you answer that there are others who agree with the latter and are not
scoundrels, I shall find you someone who is not a scoundrel yet accepts the
former. By similar means, a whole host of modes of
reasoning have been
discredited. Discarding those which have not been discredited is so great an
impediment to proving anything at all that this residue continues to be
accepted, if only because those who reject them are thereby too limited in
their ability to make any contribution the rest shall listen to.
It may be charged that the residue of modes of reasoning is culturally determined – it comprises those modes of reasoning that seemed self-evident to our forebears, that we have not actively debunked. While there is some fairness to this charge, the modes of reasoning in question have been refined with some care in the course of debunking others; and have survived scrutiny from a diversity of perspectives, by folk of several cultures. In any case, I bluntly assert that – absent some alternative sufficient to build up a cogent body of self-compatible working hypotheses with expressive and predictive power comparable to that presently in use by modern science – science shall continue to be done using what its practitioners have found to work.
All the same, in the aftermath of Gödel's work, the reductio ad absurdum process (roughly: if accepting some proposition and reasoning from it, along with what we all agreed on before, leads to an absurd conclusion, such as a contradiction, then the given proposition must be false) has come in for a good deal of well-deserved scrutiny. The constructivists seek to re-build mathematics without reductio (and some other axioms they mistrust – notably the transfinite form of the axiom of choice); but until they can build a foundation adequate to (at least) describe tensor fields on smooth manifolds the physicists shall continue to work with what they have, that works so well, for all that its powers of introspection are now seen to be somewhat compromosed. Physics does not rely on those powers, so it is not troubled.
The filter noted above – by which modes of reasoning are tested, and
a residue kept – has its analogue in the range of acceptable
skepticism. Those too unwilling to be persuaded of any new thing inescapably
have, as a result, little to contribute; hence they are ignored. Those
willing to be persuaded on too slight grounds are bound to diverge and differ:
when some meet one proposition before another at odds with it, they are
persuaded, but their peers who meet the propositions in the other order are
persuaded of the other. Those a little more skeptical will draw, from the
ensuing disputes among them, the conclusion that they were not skeptical
enough: and, in practice, exclude them from the community of discourse save in
so far as they are willing to, at least, acknowledge that others have good
cause to not be persuaded. Thus emerges a cultural norm of
skepticism tempered by a willingness to at least consider new ideas.
That balance of skepticism has an innate conservatism. It has always been safer to tacitly accept beliefs the ambient culture takes for granted, at least until contrary evidence became unequivocal. It could be fatal (or, at least, a career-limiting step) to contradict the church, or to undermine the religio-philosophical mumbo-jumbo used to justify the power of those in power. At the same time, aside from the common personal desire to avoid being made to look foolish, the reputation of natural philosophy as a whole stood to suffer if any of its followers were rendered laughable by their errors. Thus caution was strongly favoured, and that meant conservatism.
While that conservatism does imply some cultural bias, its main effect has been to slow the process of over-turning old beliefs – indeed, most of the beliefs it obliged early natural philosophers to at least tacitly accept have now been over-turned. Along the way, new prejudices have intruded themselves and we would be imprudent to suppose science is, or ever shall be, wholly free of such follies: none the less, science has a better track record than anything else when it comes to allowing its cherished beliefs to be over-turned when the evidence available contradicts them. To charges of cultural bias, science can at least respond that it is the least culturally biassed system of beliefs available. Indeed, the extent to which western culture and science internalise kindred beliefs and values today owes more to the culture bending to follow science as it broke free of its past than to science retaining the prejudices European culture once foisted on it.
Over many generations, then, the scientific community has thus evolved modes of reasoning conducive to fairness and a norm of moderate skepticism which, though conservative in character, leaves room for old ideas to be refined and improved upon, subject to a usually just mode of assessment of what constitutes an improvement.Written by Eddy.