Faith and truth in science - was - RE: virus: Scientists and Philosophers

carlw (
Tue, 9 Feb 1999 21:19:45 -0600

FAITH, n. Belief without evidence in what is told by one who speaks without knowledge, of things without parallel. (The Devil's Dictionary Ambrose Bierce)

Speaking as a scientist, and as an engineer, and having delivered
"Philosophy of Science" courses at a post-graduate level (albeit in an
Engineering Faculty), I find the discussion here vastly stranger than it needs to be. Faith, belief, trust and related concepts have no place in good science. In contrary to some of the things stated here, we don't need to
"believe" that the entire universe has certain characteristics including
"constant-space linearity" to perform good science, in fact we have
reason to suppose it does not uniformly have this characteristic. What we do have, is a hypothesis that the universe is consistent within the ambit in which we perform our research, and we can hypothesize that the results of our experiments are applicable to the more general universe except under fairly unusual circumstances.

<humour warning> A good scientist knows that the universe is out to get him <humour ends>, and if he makes any untested assumptions, then it is probable that something in that untested assumption will come back to bite him (just ask Pons and Fleischmann about that).

Good scientists (and engineers) are very aware of the "conflict" between the utilitarean knowable and the theoretically elegant hypothesis. But in fact, from a theoretical perspective, the discoveries of modern physics do not oblige one to embrace any particular philosophical position, whether it be mystical organicism, dialectical materialism, or anything else. In the last resort, all such interpretations can be rejected by an astringent
"positivism". This being so, Ockham would suggest that we discard them as
foundations until such time as some blend of "positivism" and its offshoots fails to achieve a rational world view. In which case it will need to be discarded until a more useful tool becomes available. We can demonstrate this through a simple thought experiment.

There is nothing fundamentally wrong from postulating that the universe is a strange, unknown and unknowable environment except in our immediate vicinity. We can then analyse the environment in our immediate vicinity and develop a rational system to describe our immediate environment. We can then make the inductive step that all of the universe works the same way as our localised model. As and when we discover phenomena which confute our hypothesis, we simply modify our model of the localised universe to bring it into alignment with this new information. As anyone with a smattering of exposure to science will recognise, this is the very basis of the scientific method. As anyone with a slight exposure to the philosophy of science will recognise, this is the basis of the philosophy of science. As anyone with common sense will recognise, this does not take "faith". There is no presumption that all of the universe can be described in this fashion, but in our experience to date, nothing in the substantive universe has proven unamenable to this approach.

Examining an example, modern physics posits the existence of "quarks" and
"electrons". We speak of them as "things" as if they have an existence. In
fact, from a theoretical perspective this is not necessary. A scientific theory may be regarded as a formal structure, in which theorems are derived from a limited number of axioms, and in which some of these theorems are interpreted by so-called "correspondence rules" as statements about things which it is possible to measure or observe. In this way, one theoretical structure may co-ordinate a whole range of empirical laws. However, the terms in the axioms are not in general "directly interpreted" by being linked to the reports of observations or the results of measuring operations. "Quarks" and "electrons" on this account are theoretical terms which do not correspond to any directly observed entities. The nearest one gets to "direct observation" is with phenomena like the tracks made in a bubble chamber, but here of course, it is strings of bubbles which are being observed, and not the "particles" which supposedly produce them. In other types of apparatus, the direct observation may be the results of examination of a field via a scanning tunneling electron microscope, or a sequence of sparks, or the movements of a pointer, or the readings of a counter. So do the theoretical terms refer to any kind of "entity" ar all?

The positivistic interpretation of the question "Do electrons exist?" is
"Does electron theory make correct predictions?" Thus charecteristically
positivism sees no difference between an "instrumentalism" which argues that talk about "electrons" is just a fiction for co-ordinating the results of observation, and a "realism" which declares that "electrons" really exist behind the observations and independantly of our theories. So far as positivism is concerned, both "interpretations" agree that the theory is successful, and that is all that can be said. No need for "faith", no need for "trust", and no need for "belief". <humour warning> In fact, a firm expectation that instruments lie, variables won't and constants aren't is a prerequisit to experimental success. <humour ends>

We all posses a great deal of practical knowledge about the physical and social worlds and not even the most surreal post-modernist would be able to survive without it. Much of this knowledge is enshrined in recipes for doing things, set in frameworks of largely unarticulated assumptions, which may or may not be consistent with one another. People learn to "get by" without necessarily developing sophisticated theories. You can buy and sell without being able to expound a theory of money, and you can use a television without being able to say anything about the nature of the electron. This body of knowledge is both flexible and pretty robust. Because it is not co-ordinated in precise and explicit theory we tend to hold it in low esteem, even though everything else we do depends on it.

Scientific activity, in contrast, is explicitly theory guided, but this is not to deny that it is underpinned by the same kind of practical recipes. A physicist will learn how to wire up a circuit, how to use an oscilloscope, how to bend an electron beam by a specific amount. Theory itself may be construed simply as an instrument of prediction and control, and in one sense, to have such knowledge is indeed to "know what the world is like." A strongly positivistic interpretation of theory generates an
"instrumentalist" account of science, and implies that attempts to integrate
the results of science into frameworks of wider significance are scientifically and literally meaningless. Thus positivism can function as a professional ideology, appropriate for defending the territory of a technical puzzle-solving community, which is confident in its own expertise and contemptuous of amateur attempts to meddle in its practices. Now to say that positivism,in one or another of its gises, cna function as a professional ideology is not to refute it. As I am using the word here, an
"ideology" is a system of "beliefs" about people, society and the world
which serves the interests of some group or other. Whether the "beliefs" are true or not is a seperate matter. IMO the trouble with positivism is not that it is an ideology but that it reduces science to a rudderless cargo of techniques, and while this is by no means a disproof of positivism, it certainly limits its appeal.

Science's inherited images of itself conflict with arid instrumentalism. It has a pantheon of heroes, populated with the good and the great, all discoverers of some aspect of "The Truth". These figures serve as mileposts, signposts and guardian angels on the route to "Man's Unending Quest For Knowledge." These heroes, their quest and the truths they have found, have all come to play a role as cultural symbols. Thus science is seen not just as a means to other ends, however socially useful they may be, but as an end in itself. Science is pictured searching for the "Key to the Universe," hidden somewhere just beyond the frontier of current theory. Thus the advances made in fundamental physics by their very existence proffer a justification for the social milieux within which they were produced.

Scientific progress may be variously claimed as a justification for a "free market in ideas" or for "scientific materialism" or for "tough-minded positivism", but everyone (almost) agrees in seeing it as a pinnacle of human achievement, integrated into a scheme of social goals and values. And such considerations may provide a strong motive to those who commit themselves to a scientific career. Those who have interpretted twentieth-century physics have often been motivated by the desire to promote some value-laden world view (or meme).

Now "values" can enter science in a number of different ways. So far as scientific practice is concerned, value-commitments such as "telling the truth" are essential to it. It is also clear that ethical considerations may prohibit certain kinds of investigation; such knowledge is, as it were,
"taboo." It is also evident that the choice of a problem for research
depends explicitly or implicitly on value-judgements. But while such decisions may affect the "neutrality" of science by making it the servant of particular interests, they do not affect its "objectivity." Indeed, the value commitments intrinsic to scientific activity are designed to safeguard science's objectivity. Sometimes, it is true, scientists distort the evidence, but if this is done deliberately then they risk expulsion from the scientific community: fraud is a "mortal sin." However, as I have argued, all evidence is mediated by theory-laden descriptions and to that extent its acceptance must be provisional.

Theoretical commitments are inevitable and mean that the kind of objectivity sought by empiricism is unattainable. Of itself, however, this does not mean that science cannot be impartial and critical and thus "objective" in a different sense. Still the evidence shows that outside influences on science can go deeper and affect not only the interpretation of the significance of a theory, but the way in which it is presented and even the criteria which govern whether it is acceptable.

This conclusion may seem close to heresy; have we not been taught that physics and mathematics give us a knowledge of a kind which is absolutly hard, secure and objective? Morever, if we are to believe science's heroic legends, all atttempts to mould it in the service of some ideology have led to disaster, and it can hardly be denied that the process of scientific development is, to a large extent, driven by problems and goals set internally by the scientific community's own endeavours, rather than laid upon it as external obligations. But this does not mean that the scientific community is wholly cut off from the rest of society. In the case of mathematics, its special status as the supreme exemplar of objectivity in knowledge is a little curious. The truths and proofs of mathematics do not depend on the evidence of the senses, and in consequence some have postulated that they depend on "transcendant objects", accessible only to Reason. But, as in the case of moral and political authority, invoking a transcendendent realm to underpin a set of social institutions and practices may simply be a way of disguising the fact that they are grounded on a social consensus. Mathematics, after all, is a human invention.

In physics, the situation is obviously different: after all, there is a
"real" physical world for physics to be about. Nevertheless, a similar
misrepresentation can arise, for one can come to think of one's theories and concepts as themselves possessing the characteristics of the physical world which they purport to describe and explain. Thus one is tempted to speak of knowledge as "rock-hard", "solid" and "real", and to think that when something is "obvious" that it is the facts which have spoken rather than ourselves. When we speak, however, we draw upon the cultural resources of a language which reflects particular interests and ways of seeing the world.

What then of the idea of "Absolute Truth"? Obviously, except in the dreams of some post-modernist religionists, there is no transcendent realm of concepts and theories which rest forever in perfect correspondence with the states of affairs to be found in the world, and quite independent of all human conventions. Concepts are "social institutions" forged by a language using community, not "things" whether transcendent or otherwise. "Truth" we may grant is a relation of correspondence between what we say and the world, but it follows that a "truth" has both an objective and a conventional pole: it depends both upon the implicit rules governing particular concepts and upon what the world is like.

There could be (and are) many systems of concepts capable of being used to describe the world "correctly" according to their own implicit criteria. Greengrocers classify strawberries, raspberries, blackberries and loganberries together as "berries". Students of elementary botany are taught to say that they are wrong, and that berries are fruit like bananas, cucumbers, tamatoes and - as luck would have it - gooseberries. They learn to say that a strawberry is "really" a swollen receptacle covered with achenes, and that raspberries and blackberries are "really" clusters of drupes. But this criticism of greengrocers is a piece of gratuitous academic imperialism. Greengrocers and their customers have different interests from those of botanists: they are concerned with taste, appearance and whether you eat them with cream: it matters not at all for the practice of
"greengrocing" that some vegetables are tubers or rhizomes rather than
roots. There is no sense in asking which system is more "correct".

This does not mean that all systems of concepts are equally good. Evidently some systems of concepts are vastly superior to others, relative to certain kinds of pursuit. Science is one kind of pursuit, or perhaps one should say, a family of more or less related pursuits. However, to say that the goal of physical science is an "understanding" of the "laws of nature" is not particularly helpful. It suggests "tuning in" to theories "laid up in heaven", but it does not offer any criteria which might guide scientific practice. The instrumentalists insist that the goal of science is
"prediction and control" (which fits some fields of science more happily
than others), and this provides a means for comparing the relative fruitfulness of different theories. A falkse physical theory, employing
"mistaken" concepts, can then be defined as one which fails to generate
successful predictions, and we judge the concepts to be mistaken because of the failure of the theories within which they are embedded.

As we can see in the history of science, the criteria by which a theory is judged acceptable may undergo changes. Of particular importance here are what we may term the "regulative principles" of a science, and in the "big" scientific revolutions it has been such principles which have been overthrown. Thus the mechanical philosophies of the seventeenth-century expunged sensory qualities and purposes from the vocabulary of physics, replacing them by matter in motion and causal action by contact. The success of Newton's theory of gravitation, however, required another change of viewpoint: good theories need not embody a plausible "mechanical" picture, but they must contain a mathematical formulation of the laws governing the forces acting in a system. Nineteenth-century field theory implied that
"action-at-a-distance" theories were not really intelligible after all, and
reinstated action by continuous contact. The aether theories offered to explain all in terms of picturable mechanisms once again, but when special relativity triumphed it gave priority to "invariance" over "mechanism". And when quantum mechanics was born, to Einstein's horror, it required abandonment of the age-old ambition of calculating with certainty every detail of the behavior of any system and reintroiduced the concept of
"spooky or strange action at a distance". In each case, the transition
involved a change in the ideals of scientific explanation: across such discontinuities scientists themselves may stand in mutual incomprehension. The "convert" needs to accept not only new evidence, but a new way of looking at things.

<Memetic Flag> A "worldview" presents both a picture of the physical world and an account of human values in a co-ordinated fashion. It is sufficient for people to believe that there are connections between moral and physical concepts for changes in scientific theories to be taken to have wider significance. As we have noted, many people hold that our whole conception of a moral order would founder if it wwere to be shown that the behaviour of human beings were mechanically determined. Newton and the followers of the
"corpuscular philosophy" insisted that matter was "passive" and was capable
of generating neither order nor motion of itself. Thus not only did their physics enable them to invoke the "divine intelligence" in accounting for natural order, thus underpinning the values which sustained their social order, but it gave them an anology for the "proper" governance of the land under the aegis of the civil authorities and the established church. The late-nineteenth-century aether theorists saw connections of a different kind between their worldview and their theories. Of course, you may argue that such connections are "extraneous", and that the "real content" is given by the equations and the experiments. But this too is an interpretation, and one which is particularly adapted to the professional scientist, intent on getting results for the journals. <End Memetics>

Science has the goal of bringing its knowledge under a small unified set of postulates. Thus it differs from "common-sense knowledge" in being systematic, and from the systems of the metaphysicians (such as those found on the memetics list :-) ) in that its unifying postulates can, albeit with difficulty, be brought under empirical scrutiny. It is this striving for a logically consistent unified set of postulates, which can be judged for their predictive success or otherwise, which leads us to say that science is an organised "search for the Truth" (capitalization intended).

This way of organizing knowledge focuses attention on those parts which may seem to be the most speculative, since they are of the greatest generality. The highest level of axioms is regarded as the foundation: it is here that work is said to be most "fundamental", where the deepest "secrets" are being
"unlocked". This way of speaking, however, is misleading. A house collapses
if you undermine its foundations; not so our knowledge of the physical world. Even if special relativity were refuted, radio, television and nuclear weapons would still continue to work. You would not hesitate, philosophically, before turning on the light switch. Paradoxically, the
"foundations" of physics are insecure points at the summit of its
theorizing. Most of the accumulated knowledge of the scientific community lies not up Olympus, but in its repository of technical knowhow. And every good scientist (and engineer) is fully aware, even if it is usually at a subconcious level, of the lurking instability in the basement. This is what allows a scientist to view the overthrow of an entire theoretical basis for his work if not with equanamity, at least with resignation. This is also the reason why "faith" in a system is not only not required, it is a positive hindrance to any scientist interested in pursuing original work.

Even granted the importance of of successful predictions, there is an ineradicable plasticity in the interpretation of physical theory. A tough positivism will seek that interpretation of theory which is most
"economical" in dealing with experience (thank you William of Ockham). Those
whose worldviews embody some account of the nature of things will find room amongst the conventions deployed in physics to structure an interpretation in conformity with their metaphysical preferences. Who knows whether one of these interpretations is "right"? The growth of knowledge of the world is shown in our increased practical competence, but all of our theories are entwined with conventional elements which reduce their testability. To make these conventions explicit is to reveal the extent to which our theories can tell us nothing for certain about the world.

Faith in such uncertainty is not only foolish. It is misplaced. And the
"Truths" which seem so self-evident and perpetual to the metaphysicists are
perceived as being transitory and ephermal by the physicist. Thrive on chaos, the world is not only stranger than we can imagine, but the strangeness metamorphises every time we approach it, as perception of
"truth" self-modifies our worldview.


-----Original Message-----
From: []On Behalf Of David McFadzean
Sent: Tuesday, February 09, 1999 2:58 PM To:
Subject: RE: virus: Scientists and Philosophers

At 02:36 PM 2/9/99 -0500, sodom wrote:
>The Faith thing comes up all the time. I swear that sometimes I am the only >person here who does NOT see Faith as a human requirement.

Some friendly advice from someone who survived the Great Virus Faith Wars: when you see "faith", read "trust". It may not be a perfect translation, but it will almost certainly help communication (assuming that's the goal).

David McFadzean       
Memetic Engineer      
Church of Virus