Neurath - On The Foundations of The History of Optics PDF
Neurath - On The Foundations of The History of Optics PDF
Neurath - On The Foundations of The History of Optics PDF
sopbical question. Those who wish to give more weight to the imagery
of hypotheses (as I believe one must in some cases), may without con-
tradiction add this to the analysis.
Let us illustrate this from the history of optics between the early 17th
and early 19th centuries: in that period a series of outstanding thinkers
discussed a fairly narrowly defined field of phenomena which was yet
rich enough to evoke complicated hypotheses. Acoustics early reached
a fairly complete form, and could be developed without great controver-
sies. Electricity, on the contrary, shows a bewildering range of phenomena
and systems of hypotheses. Often the cognate is severed and the disparate
linked. The hypotheses are so different that it is hard to compare them.
Moreover, able experimenters and astute mathematicians were so evenly
ranged behind either theory that the battle swayed for a long time. Beyond
the early 19th century we can no longer pursue a simple history of optics,
because it soon became linked with findings from electricity, magnetism
and the theory of heat. We can no longer encompass the intricacies. Only
recent decades, or better, years, are slowly restoring a feeling that
hypotheses and theories will soon so far settle down that on looking
back we may discern their origins. For to find in a tangled skein of
hypotheses the origins of what will come later is to nurse forebodings
of the hypotheses of the future, an enterprise beyond the daring of any
oftoday's experts.
Accounts in the history of physics are rarely started from a complete
dissection of the several theories into elementary propositions. Hence the
lack of an adequate classification of different views that takes note of
their constituent parts. Writers usually describe schools of thought
by one outstanding characteristic without consulting, as they should,
the totality of characteristics. The case of Descartes' optics shows what
this may lead to: Whewell reckons him among emission theorists, while
Littrow (Whewell's translator into German), in a comment, calls him a
forerunner of the wave theory. How can this be? Descartes' account
contains elements of both theories. Some of his remarks show that he
began from the assumption that a luminous body emits corpuscles, or at
least that luminous corpuscles are ejected from its neighbourhood and
travel towards the eye. The sensation of light is caused not by the original
particles reaching the eye, but through their pushing against other
particles which transmit a pressure. If we stress the displacement of the
104 EMPIRICISM AND SOCIOLOGY
abc
ab c*
ab* c
a b* c*
a* b c
a* b c*
a* b* c
a* b* c*
If, for classification, we use only a, there are four a and four non-a
theories. But usually we describe one group of views by a and the other
by c. Clearly, this cannot yield a complete division. In fact we mostly
find 80me only of the possible outlooks, which leaves the set to be classified
less transparent, as for instance
ab c*
a b* c*
abc
a* b c
a* b* c
Writers then usually call the first two pure a-theories, the last two pure
FOUNDA TIONS OF THE HISTORY OF OPTICS 105
Huyghens, too, light resembles sound, but he does not consider sound to
be periodic. The 'Huyghens ray' is not periodic, but may be polarized.
From his discussion we may at best infer that light consists of spherically
propagated single impulses. He thought of light rays roughly as many
did of X-rays before Laue and others revealed their periodicity through
interference. Even before this, X-rays were considered as a kind of wave
radiation, if we go by the constant velocity of propagation in vacuo
rather than by periodicity. Huyghens' principle is important: all points
of the sphere become centres of new spheres expanding uniformly in all
directions and together forming a new spherical surface. Thus besides
the direction of propagation he introduces a second direction of motion,
just as Malebranche had already vaguely done. If we knew nothing
about Huyghens save that he found his principle, we might well assume
he had introduced it to account for diffraction, since this can easily
be deduced. We know that this was not his motive, he did not even know
of Grimaldi's experiments when he wrote his book on light. Since his
hypothesis leads to diffraction, he has to explain its absence in experiment
from the assumption that diffraction produces luminous effects too weak
to be seen, just as today we explain that light travels in straight lines.
Interference he touches not at all, nor could he deal with it by means
of his hypotheses. But he does treat of spheroidal waves, double refraction
and polarization of light rays.
In Newton (1642-1727) much is vague and vacillating. That, partly,
makes him great. Like Grimaldi he senses the vast range of optical
phenomena. Although he strives, by means of farflung abstractions, to
escape from the wealth of things seen, elsewhere he does them justice.
He even touches on the possibility that the movement of the aether might
be adduced for the theory of light and uses it in trying to explain the
colours of thin layers. Thus in Newton there are many more elementary
notions and hypotheses than in Huyghens. Some are worked out but
none is so central as with Huyghens. On his own showing Newton hardly
values speculations on the nature of light. But from his Optics (1704)
which followed an earlier Treatise on Light (1672), and from the Principia
(1686) we infer how deeply he was preoccupied with, and influenced
by, this problem. In the Principia, the vital statements are in the section
on the motion of very small bodies impelled by centripetal forces directed
to the various parts of a large body; likewise in the section on the propaga-
108 EMPIRICISM AND SOCIOLOGY
tion of motion in liquids. There he points out, against the view that light is
wave motion, that the entire wall behind a slit should have to be illu-
minated. He does indeed know diffraction, but fails to link it with the
wave character of light, regarding it rather as due to the material of the
slit. This outlook is not an isolated auxiliary hypothesis for he uses
similar notions for reflection. Even a well polished mirror should, accord-
ing to him, show strong dispersion; what prevented it was that the mirror's
material set up around the body a uniform layer of force which was the
real cause of reflection. His great number of elementary notions afforded
many possibilities for later physicists to work out. Newton's ray is
periodic and can be polarized. Periodicity first made possible a theory
of periodic waves and interference. Newton attributed periodicity to
his particles of light in order to explain the colour of thin layers. As
to when periodicity arises, his view vacillates; according to his Optics,
Book Two, proposition twelve, not until transit through the refracting
surface. The periodically recurring arrangement of particles, the surges
of easy reflection or easy transition already involve the notion of interval
that figures again in the wave theory. Later thinkers, like Brewster (1781-
1868) see the particle of light as having two poles, with now one and
now the other leading in the direction of motion, because of rotation
about an axis normal to it. If the positive end collides, there is transit,
if the negative, reflection. The difficulties arising from this hypothesis
led to many auxiliary hypotheses, many of which are discussed at length
by Biot (1774-1862). The hypothesis of surges involves periodicity but
not interference. In Proposition twelve, Newton gives a hypothesis for
those who need one: it recognizes that vibrations of the illuminated
material influence the light ray, but not that interference occurs. Period-
icity does not entail interference. Nothing, at first, is said about what
happens when a positive and a negative end hit the material together.
Attempts to extend the hypothesis of surges to account for interference
as well came later. Indeed, Biot laid it down as a condition for visibility
that particles of opposite polarity must not hit the eye at the same time.
To explain why there was no effect on a photographic plate, special
hypotheses were adopted concerning the chemical nature of light. Thus,
it was one of the main founders of emission theory who introduced
periodicity, which is so vital in modern light theory. But Newton is aware,
too, that light may be polarized at right angles to the direction of propaga-
FOUNDATIONS OF THE HISTORY OF OPTICS 109
tion. He is the first to use the term 'polarization' and compares the effect
of feldspar crystals on light with magnetic influences. He regarded
polarization as proof of emission theory, though later it was used as a
prop of wave theory. It could be explained only if particles could
be magnetized. The emission hypothesis thus recognized a double
polarity of corpuscles, one in the direction of propagation (surges) and
the other normal to it, when the ray has passed through feldspar. Thence-
forth periodicity and polarizability are backed now by emission the-
orists, now by wave theorists. Because, on the whole, wave theory won,
many even among the scientifically cultured have been led to imagine
that through its greater wealth it outpaced emission theory, which they
regard as rather primitive. The unfair contrast between complete and
incomplete theories partly springs from the unfortunate dichotomy of
emission and wave. How differently the history of physics would have
figured in human thought if physicists had been divided into periodics
and non-periodics. This too would have been faulty, as must be any
classification based on a single characteristic. Because Newton linked
the periodicity of light rays with the colour of thin layers, emission
theory was long superior to wave theory, if we may reckon Huyghens
~nd his followers as wave theorists, although periodicity had not been
adopted in their doctrine. It was Malus (1775-1811), one of the main
opponents of wave theory, who advanced the theory of polarization in
unforeseen ways by extending magnetic analogies. Had he lived to see
today's discoveries, such as the Zeeman effect, they would merely have
spurred him on to stretch his magnetic theory of light still further.
Newton's opponents long failed to produce an equally valuable
hypothesis to explain optical phenomena. Euler (1707-1783), who was
a wave theorist, indeed knows periodicity; like Malebranche, he explains
the colours from the rate of vibration, but works without Huyghens'
principle. His main objection to emission theory was incidentally that
one could not grasp how emitted particles might penetrate solid bodies,
a criticism since superseded, given that alpha and beta particles penetrate
aluminum as corpuscular rays.
At the end of the 18th and the early 19th centuries, optical controversies
became very lively. Young (1773-1829) in England and Fresnel (1788-
1826) in France extended the wave theory, perfecting it by introducing
interference and also transverse instead of longitudinal oscillations, and
110 EMPIRICISM AND SOCIOLOGY
to corpuscular rays: Newton's rays are much closer to gamma rays with
their many properties, than to alpha or beta rays. For gamma rays have
periodicity, while alpha and beta rays have none. How far a hypothesis
meets all demands may depend on the efforts made on its behalf. Duhem
said that if physicists had offered a prize for an optical system based on
emission while agreeing with Foucault's finding that light travelled faster
in air than in water, such a theory would have been found. Often promis-
ing theories were dropped because the young are always eager to tamper
with the work of their elders. Thus the electric fluid hypothesis was
quickly abandoned, though Lodge could show much later that it was
more serviceable than had ever been suspected, as in his hydraulic
model of a Leyden jar discharge. Likewise for the doctrine of two elec-
tricities which today is on the rise again. Before fighting a hypothesis,
one ought to give it its most finished form, a task often too great for a
lifetime's work. But the historian of physics will have gained much if
at times he tries to rehearse the scope of an older theory. Oersted used
to advise his young chemists to translate old theories into the new ones
as best they might, in order to recognize their value or uselessness. He
showed, ingeniously, how many propositions in the theory of oxydation
already figured in phlogiston theory, which had, for instance, recognized
that breathing was a form of burning.
We shall not enquire whether the creative impulse suffers when several
theories are worked out at once, but the historian of physics cannot
avoid such comparative surveys. Today many would dismiss the views
of Duhem and Poincare sketched above as a whim of fashion: yet this
fashion is very old, it arises wherever several highly finished theories
live side by side. Equality is claimed either by the champions of defeated
theories who wish to save them, or by protagonists of theories that have
yet to make good, who wish to soften objections. Sometimes impartial
critics adopt such views. A hundred years ago, others beside Brewster
held that emission and wave theories could equally be perfected; witness
Herschel's (1792-1871) view that Newton's theory, if pursued as carefully
as Huyghens', might lead to just as viable an account of phenomena
hitherto held to lie beyond its scope, if only we were to extend Biot's
hypothesis about spinning particles of light: emission of particles at
equal intervals, and a corresponding motion of the luminous body
would then account for interference without an aether. Similarly,
112 EMPIRICISM AND SOCIOLOGY
REFERENCE