Galileo's Telescopic Observations of Venus and Mars

Author(s): Alan Chalmers

Source: The British Journal for the Philosophy of Science , Jun., 1985, Vol. 36, No. 2
(Jun., 1985), pp. 175-184
Published by: Oxford University Press on behalf of The British Society for the
Philosophy of Science

Stable URL: https://www.jstor.org/stable/687033

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

and Oxford University Press are collaborating with JSTOR to digitize, preserve and extend
access to The British Journal for the Philosophy of Science

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
 Brit. J. Phil. Sci. 36 (1985), 175-191 Printed in Great Britain 175

Discussions
GALILEO'S TELESCOPIC OBSERVATIONS OF VENUS AND MARS

According to Paul Feyerabend, Galileo's use of telescopic evidence to

support the Copernican theory is not in conformity with any contemporary
versions of scientific method. In his view, as it is presented in Against
Method [1975], use of the telescope did not lead to increased factual support
for the Copernican theory in line with positivist methodology nor does it
lend support to the falsificationist idea that the triumph of Copernicanism
'consisted in the replacement of refuted theories by more general conjec-
tures which explained the refuted instances, made new predictions, and
were corroborated by the observations carried out to test these new
predictions' (op. cit., p. 143). Rather, according to Feyerabend, both the
Copernican theory and the reliability of telescopic observations in the realm
of astronomy are refuted by experience, and Galileo exploited the harmony
between these two refuted ideas to gain support for each of them. He thereby
furthered the Copernican cause 'by ad hoc hypotheses and clever techniques
of persuasion' (op. cit., p. 143). A key part of Feyerabend's case involves the
apparent sizes of the planets Venus and Mars as viewed from Earth. As
Galileo had stressed in Day 3 of his Dialogue [1967], the variations in their
sizes as gauged through a telescope are in conformity with the Copernican
theory whereas the corresponding variations as gauged with the naked eye
clash with that theory. Galileo introduced the phenomenon of irradiation to
argue for the preferability of the telescopic to the naked eye data. According
to Feyerabend, this hypothesis of Galileo received support mainly from its
agreement with the Copernican point of view and was therefore largely ad
hoc.

In the following discussion I argue against Feyerabend's claim that

Galileo's telescopic observations were defended by him in an ad hoc way.
Whilst I do not wish to challenge Feyerabend's general claim that the
transition from the Ptolomaic to the Copernican theory cannot be ad-
equately accounted for by appeal to contemporary accounts of scientific
method, I do wish to challenge the way in which he employs Galileo's use of
the telescope to support his case. The replacement of naked eye data by
telescopic data did constitute a transformation of the 'observational base'
and of epistemological standards, but Feyerabend has made a poor job of
demonstrating this.

According to the Copernican theory the distance of both Mars and Venus
from the Earth should vary appreciably during the course of the journeys of

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
176 Alan Chalmers
the three of them about the Sun. When Mars and Venus are on the same side
of the Sun as the Earth they will be relatively close together, whilst when
they are on opposite sides of the Sun as the Earth they will be relatively
distant. In the case of Mars the distance from Earth varies by a factor of the
order of 8 and in the case of Venus by a factor of about 6. Consequently, the
diameters of the planets as viewed from Earth should vary by a similar
factor. However, when viewed with the naked eye, Mars appears to change
size by a factor not much greater than 2, whilst the apparent change in size of
Venus is negligible. It was for this reason that Galileo described Mars as
launching a 'ferocious attack' on the Copernican system and Venus as
presenting a "greater difficulty" ([19671, p. 334). When the two planets are
observed through a telescope the difficulty is removed. The observed
changes in size are in accordance with the predictions of the Copernican
theory.
If the situation is presented thus, it would appear that the Copernican
theory is vindicated provided a case can be made for the superiority of the
telescopic to the naked-eye data. Feyerabend ([1975], chapters 9 and io)
convincingly argues that the issue was by no means straightforward. Galileo
had no adequate theory of the telescope to appeal to, to support its use,1
whilst the theory that his contemporary, Kepler, had developed clashed
with psychological aspects of vision through lenses. As far as practice was
concerned, the reliability of the telescope when used terrestrially could be
demonstrated, but that success could not automatically be assumed to be
transferable to the quite different and unfamiliar realm of astronomical
phenomena. Indeed, Galileo himself witnessed some glaring inconsistencies
in telescopic observations of the celestial region. For example, the telescope
magnified the stars much less than it magnified the moon, whilst it removed
irradiation from the planets but to a much lesser extent from the stars.2 In
the light of the problematic nature of the telescopic data, then, it cannot be
simply assumed without further ado that the telescopic observations of the
changes in size of Mars and Venus offered genuine support for the
Copernican theory. Feyerabend argues that Galileo could give no adequate
independent support for the veracity of his telescopic observations so that
his support for Copernicus was ad hoc in that respect.
I will not pursue the issue of ad hocness further in this section because it is
time for me to indicate that the situation as it is presented above, which
follows the main lines of Feyerabend's own presentation and in many
respects Galileo's also, totally misrepresents the logic of the situation. There
was no question of the telescopic sightings of the apparent size of the planets
offering support for the Copernican theory, as compared to its rivals, the
1 Feyerabend's case can be strengthened here by reference to a passage in The Assayer in which
Galileo attempted to offer his theory of the telescope in response to a challenge. Quite apart
from a blatant non-sequitur the 'theory' is almost laughably inadequate. For the passage in
question see Drake ([19571, pp. 245-6).
2 Galileo himself noted these inconsistencies when reporting his first telescopic observations.
See Drake ([19571, PP. 46-7).

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
 Galileo's Telescopic Observations of Venus and Mars 177

Ptolemaic and Tychonian systems, because both of the latter predicted

precisely the same variations in size as predicted by Copernicus. Variations
in distance from Earth, leading to predicted changes in apparent size, arise
in the Ptolemaic system because the planets move closer, then further from
the Earth as they traverse the epicycles superimposed on the deferents,
which latter do define paths equidistant from Earth, whilst they occur in
Tycho Brahe's system for the same reason as they occur in the Copernican
system, since the two systems are geometrically equivalent. Derek J. de S.
Price [1959] has shown quite generally that this must be so once the
epicycles in the systems are adjusted so that they are compatible with the
observed angular positions of the planets and Sun. That the apparent sizes
of the planets had posed a problem for the major astronomical theories since
antiquity is acknowldged by Osiander in his introduction to Copernicus's
Revolutions of the Heavenly Spheres. It is also noted by Feyerabend ([19751,
p. Ioz) but the point is not given appropriate emphasis by him.
In the light of this clarification of the logic of the situation it is possible to
make a Feyerabend-style criticism of Galileo's presentation of his case in the
Dialogue. Galileo himself (Drake and O'Malley [1960], p. 184) had noted in
earlier writings that observations of Mars and Venus clashed with both the
Ptolemaic and Copernican theories. Nevertheless, when describing his
telescopic observations of the size of Mars and Venus in the Dialogue he
presented them as if they constituted confirmation of the Copernican
theory. According to Galileo ([19671, p. 335), the Copernican prediction
that Mars and Venus should appear to change size during the course of the
year turns out to be 'not only not contrary to the Copernican system, but to
absolutely favour it, and greatly', once the telescope is employed. In
presenting the case in this way, and suppressing the fact that the rival
astronomical theories equally benefit from the telescopic estimates of the
sizes of Mars and Venus, Galileo was engaging in what can justifiably be
called propaganda for the Copernican system.
A characterisation of the epistemological situation which differs from that
offered by Galileo and Feyerabend is as follows. For two thousand years
there had been a clash between astronomical theory and the observed
magnitudes of Mars and Venus. Replacement of naked eye data with
telescopic data removed this long-standing difficulty. This constituted an
argument in favour of the telescopic data. Of course, if we use the
compatibility between theory and the telescopic observations of variations
in size in this way to support the veracity of the latter, we cannot also do the
reverse, and use the telescopic observations of size to support the theory,
both because of the circularity involved and because, as we have seen, that
evidence does not serve to distinguish between the rival theories. If we use
the telescopic sightings of the size of Mars and Venus as evidence for the
veracity of the telescopic data in the astronomical domain, we are then in a
stronger position to appeal to other telescopic evidence, such as observation
of the phases of Venus, as genuine support for Copernicus.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
178 Alan Chalmers
II

Compatibility of telescopic observations of variations in size of Mars an

Venus with all the major astronomical theories of his day was not the on
reason available to Galileo for preferring those observations to the co
responding naked eye estimates. He appealed to the phenomenon of
irradiation as providing independent grounds for the preference. Galileo's
hypothesis was that the eye 'introduces a hindrance of its own' when it views
small, bright, distant light sources. Because of this such objects appear
'festooned with adventitious and alien rays' ([19671, p. 333). Thus, if stars
'are viewed by means of unaided natural vision, they present themselves to
us not as of their simple (and, so to speak, their physical) size but as
irradiated by a certain fulgor and as fringed with sparkling rays' (Drake
[19571, p. 46). In the case of the planets, irradiation is removed by the
telescope. According to Feyerabend ([1975], pp. 138-9), 'Galileo's hypo-
thesis received support mainly from its agreement with the Copernican
point of view and was, therefore, largely ad hoc' in spite of the fact that he
rendered it plausible by attempting to show how irradiation can be removed
by means other than the telescope.
In this section I will argue that Feyerabend's charge of ad hocness cannot
be justified in this context. If this is to be done adequately it is important to
be clear about the sense or senses in which 'ad hoc' is being used, something
that Feyerabend does not do. For this purpose it will be convenient to
consider the various senses of 'ad hoc' distinguished by I. Lakatos and his
followers. I shall argue that Galileo's hypothesis concerning irradiation was
not ad hoc in any of those senses.
Following E. Zahar's improved rendering of Lakatos's distinctions
([1976], pp. 216-17), we can distinguish three senses of 'ad hoc'. An
hypothesis is ad hoc, with respect to some problem situation if there are no
testable consequences of the hypothesis that are not already part of that
problem situation.' If it is not to be ad hocl an hypothesis must be
independently testable.2 An hypothesis will be ad hoc, with respect to the
problem situation involving the Copernican theory if there are no con-
sequences of it that are not already consequences of the Copernican theory
and its background. An hypothesis is ad hoc2 if its independently testable
consequences have not survived tests, either because those tests have not
been carried out or, more seriously, because the results of such tests are
negative. The characterisation of a third sense of 'ad hoc', 'ad hoc'3, is not so
straightforward. Lakatos initially characterised it in temporal terms. An
hypothesis is ad hoc3 with respect to some evidence if the evidence precedes
its formulation. Lakatos came to accept Zahar's alternative formulation,
according to which an hypothesis is ad hoc3 with respect to some evidence if

1 For Lakatos's original distinctions see his [1968].

2 This notion of independent testability is discussed by K. R. Popper in his [1969], pp.
191-203.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
 Galileo's Telescopic Observations of Venus and Mars 179
the evidence plays some heuristic role in the formulation of the hypothesis.
For the purposes of this discussion a precise formulation of ad hoc3 is not
necessary because it will be shown quite decisively that Galileo's irradiation
hypothesis was not ad hoc3 in anything like the senses that have been
attributed to that notion.
Since Galileo's hypothesis involves the claim that irradiation arises as a
consequence of the brightness, smallness and distance of the sighted source,
it can be tested by modifying those three factors in a variety of ways, many of
which do not involve use of the telescope. A number of ways are explicitly
mentioned by Galileo. The brightness of stars and planets can be reduced by
viewing them through a cloud, a black veil, coloured glass, a tube, a gap
between the fingers or a pinhole in a card (Drake [19571, P- 46). In the case of
planets, the irradiation is removed by these techniques, so that they 'show
their globes perfectly round and definitely bounded' whereas, in the case of
stars, the irradiation is never completely removed so that they are 'never
seen to be bounded by a circular periphery, but have rather the aspect of
blazes whose rays vibrate about them and scintillate a great deal' (op. cit., p.
47). In the context of Galileo's hypothesis, perhaps it should be concluded
from this that any estimates of the magnitudes of stars, whether they be
viewed through a telescope or otherwise, should be treated with caution if
not scepticism.1
As far as the dependence of irradiation on the apparent size of observed
light sources is concerned, Galileo's hypothesis is borne out by the fact that
the Moon and the Sun are not subject to irradiation ([19671, p. 338). This
aspect of Galileo's hypothesis, as well as the associated dependence of
irradiation on the distance of the source can be subject to a direct terrestrial
test. A lighted torch can be viewed from near or far, and at day or night.
When viewed at a distance at night, when it is bright compared with its
surroundings, it appears larger than its true size. When viewed in the day, or
close at hand the apparent size is in conformity with the real size of the torch.
Galileo appeals to this consideration to argue that his predecessors,
including Tycho and Clavius, should have proceeded with more caution
when estimating the size of stars. 'I will not believe', wrote Galileo ([1967],
p. 361), 'that they thought that the true disc of a torch was as it appears in
profound darkness, rather than as it is when perceived in lighted surround-
ings; for our lights seen from afar at night look large, but from near at hand
their true flames are seen to be small and circumscribed'. The dependence of
irradiation on the brightness of a source relative to its surroundings is
further confirmed by the appearance of stars at twilight, which appear much

We now know that the observed 'size' of a star is produced entirely by the interaction of light
from the star with the intervening medium and the eye and has no bearing on its physical size.
Incidentally, Galileo's hypothesis concerning irradiation gives him at least the beginnings of
an explanation of the 'inconsistency' between the magnification of planets and stars by the
telescope. Since the stars are more distant, the irradiation is more pronounced, perhaps too
pronounced to be entirely removed by the telescope. As we shall see, Galileo offered a direct
test of the dependence of irradiation on the distance of the source.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
180 Alan Chalmers

smaller then than at night, and of Venus when observed in broad d

which appears 'so small that it takes sharp eyesight to see it, though i
following night it appears like a great torch' ([1967], p. 361).
This latter effect provides an approximate way of testing the co
bility of the Copernican (and other) theories with the observed siz
Venus which does not involve an appeal to telescopic evidence. The
be made with the naked eye provided observations are restricted to da
There are two reasons why this test will be difficult and not enti
satisfactory. The first is that under these conditions, Venus appears so
as to make accurate estimates of its apparent size difficult. Secondly, i
possible to carry out this test when Venus is close to its maximum
minimum apparent sizes because at those times it appears very clos
Sun. Consequently, it cannot be observed in the daytime because o
glare of the Sun, but only after the Sun sets, when Venus is close to th
and at its largest, or before it rises when it is furthest from the Earth a
smallest. Nevertheless, according to Galileo at least, although the chang
size of Venus can only be precisely observed with the telescope, t
'quite perceptible to the naked eye' (Drake [1957], p. I31).
This abundance of independent tests, then, is sufficient to refu
claim along the lines that Galileo's irradiation hypothesis was ad ho
respect to the Copernican theory. Further, since there is no reason to
that Galileo did not carry out the tests that he mentions, and, indeed,
he could hardly have known of some of the effects mentioned had
witnessed them, it cannot be said that Galileo's hypothesis was ad hoc2
Some straightforward facts about the emergence of the irradiat
hypothesis in Galileo's writings are sufficient to remove any suspicion
the hypothesis was ad hoc3. The hypothesis first emerged in Gal
Siderius Nuncius of i6io, which contains an account of his first te
observations of the heavens, made in the period December 1609 to Febr
i6Io. It is important for my purpose to note that at the time of w
Galileo had not, nor could he have, observed Venus through the tel
because, during that three-month period, it was too close to the Su
seen (Drake [I1978], p. 158). Consequently, his irradiation hyp
preceded his telescopic sightings of the variations in size of Venus and
not have been influenced by that phenomenon. There is no eviden
Galileo made sightings of Mars in that period either, and even if he di
three month period was too short for the variations in size to be n
takes Mars of the order of a year to pass from that of its maximum to
its minimum size. When Galileo first formulated his irradiation hypot
appealing to naked eye observations of Venus and to many of th
mentioned above, as well as to telescopic sightings of the stars, he cou
have been influenced in any way by the variations in size of Mars and
as they appear through the telescope. His claims about the veracity
telescopic observations in this respect were not ad hoc3 with respect t
Copernican theory.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
 Galileo's Telescopic Observations of Venus and Mars 181

We are now in a position to sum up some aspects of the situation with

respect to Galileo's case for the Copernican theory based on evidence
revealed by the telescope. Galileo had a number of independent reasons for
accepting the veracity of his estimates of the variations in size of Mars and
Venus as observed through the telescope. Firstly, they were consistent with
all the major astronomical theories of his day. Secondly, they could be
supported by appeal to the independently testable irradiation hypothesis.
Thirdly, they were supported to some extent by naked eye observations of
Venus in daylight. As we have seen, this state of affairs cannot be taken as
support for the Copernican theory over its rivals. I suggest, rather, that it be
interpreted as a case for the veracity of telescopic observations in the
astronomical realm. On this basis, we do obtain genuine evidence for the
Copernican theory when we turn to the fact that, when Galileo observed
Venus through the telescope he observed, not only its variations in size, but
also its phases. Having made a case for the veracity of the former, we are in a
strong position to accept the veracity of the latter. Not only did the changes
in the phases of Venus conform to Copernicus's prediction, but they were
correlated with the variations in apparent size in the way that the
Copernican theory required. Venus appeared very nearly as a full circle
when its diameter was at its smallest and only as a crescent when its diameter
was at its largest. This occurrence did constitute genuine support for the
Copernican theory over its Ptolemaic rival, although it did not serve to
distinguish it from Tycho's system, which is, incidentally, another point
that is suppressed by Galileo in the Dialogue. The evidence of the telescope
also told in favour of Copernicus in other ways. His observations of the
moons of Jupiter served to diffuse an Aristotelian objection to the motion of
the Earth, whilst his sightings of the imperfections of the Moon and Sun
served to undermine the celestial/terrestrial distinction. However, enough
has already been said to establish the untenability of Feyerabend's view that
Galileo exploited 'the harmony between two iriteresting but refuted ideas',
namely, the Copernican theory, and the reliability of the telescope, 'in order
to prevent the elimination of either' ([I975], p. 143).

III

Feyerabend has made a poor job of appealing to Galileo's use of the

telescope to support his case against method. Nevertheless, there is a case
against method to be made here.
The episode involving Galileo's introduction of the telescope into
astronomy certainly clashes with any extreme empiricist view that the senses
supply the raw material on which scientific knowledge is based in some
unproblematic and secure way, whether it be Aristotle's view that 'it falls to
experience to supply the principles of any subject'1 or that of a twentieth
1 Aristotle, Prior Analytics, I 30, 46a 17.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
182 Alan Chalmers

century positivist that 'science is a system of statements based on dir

experience' so that there is 'a sharp distinction ... between the raw materia
of scientific investigation and its organisation' (Carnap [198 1], p. 152). As
result of the introduction of the telescope into astronomy, claims based o
direct experience via the sense of sight, and which, indeed, Osiander ([197
p. xvi) had described as borne out 'by the experience of every age', w
discarded as false, and were replaced by new ones. There is no secur
observational base given to us by way of our senses. What are considered t
be the observable facts change as our knowledge progresses.
Quite apart from posing problems for positivism and other extreme for
of empiricism, the use of the telescope can be appealed to in order t
challenge the claim that there can be any universal, ahistorical account of
scientific method which purports to offer criteria for guiding the work o
scientists or for judging the hypotheses that they produce. Galileo de
onstrated in practice, or at least went a long way towards demonstrating i
practice, the superiority of telescopic evidence over the evidence of t
unaided senses in some circumstances.1 This practical discovery
demonstration eventually led to the overthrow of the criterion inherent i
much Aristotelian orthodoxy, namely, 'that the senses and experien
should be our guide in philosophising', as an Aristotelian adversary o
Galileo put it.2 More recently, M. Clavelin ([19741, p. 384) has remar
that 'the chief maxim of Peripatetic physics was never to oppose the
evidence of the senses' and S. Gaukroger ([1978], p. 92) writes of 'a
fundamental and exclusive reliance on sense-perception in Aristotle's
work'. In using the telescope Galileo departed from this Aristotelian
methodology. He was not guided by it and his work constituted a challenge
to the standards that it embodied.
A point that Feyerabend omits concerns the extent to which Galileo's use
of the telescope was a threat to the teleological aspect of medieval versions of
the Aristotelian orthodoxy concerning sense perception.3 From that teleo-
logical point of view it makes no sense to claim that tht human senses might
err when doing what they are meant to be doing under normal circum-
stances. This would mean that they are generally inadequate for the purpose
they are designed for. This point might be countered by the remark that
viewing celestial objects does not constitute normal circumstances as far as
the sense of sight is concerned, so that replacing naked eye by telescopic
observations of the heavens was no basic threat to the teleological aspect of
the Aristotelian theory. Nevertheless, that it did constitute a threat seems to
have been the view of those Aristotelian contemporaries of Galileo who

This is in keeping with Feyerabend's remarks on Reason and Practice in his [1978], pp.
16-3I.
2 This was the 'criterion of science itself' according to Scipio Ciaramonti (1565-1652), the
author of Anti Tycho quoted by Galileo ([19671, p. 248).
3 On this aspect of Aristotle's view see, for example, I. Block [1961] and S. Gaukroger ([1981],
PP. 75-89).

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
 Galileo's Telescopic Observations of Venus and Mars 183

responded to the latter's use of the telescope by observing that 'if God had
meant man to use such a contrivance in acquiring knowledge, He would
have endowed men with telescopic eyes' (Kuhn [1959], p. 226) and it seems
to have been in Galileo's mind when he wrote that 'nature has not granted so
perfect [a sense of sight] to men that they can succeed in discerning such
distinctions [as the variations in size of Mars and Venus and the phases of the
latter]', whereas, in the telescope, 'it has pleased God to concede to human
ingenuity an invention so wonderful as to have the power of increasing
vision.. .' ([1967], p. 335).
By using the telescope, then, Galileo undermined in practice standards
implicit in Aristotelian methodology. In a similar way, a century later, the
generally held view that action at a distance is an absurdity was undermined,
in practice, through the success of Newton's theory involving gravitational
attraction. It is implausible, to say the least, that such moves could be
anticipated and legislated for in advance by methodologists. We can expect
methodologies to alter in the light of new discoveries, including practical
discoveries, and for this reason the notion of a universal, ahistorical account
of method that can serve as a standard, not only for present but also for
future knowledge, is an absurdity. In so far as the episode in science
involving Galileo's use of the telescope is an example of standards changing
in the light of practical discoveries, it provides the opportunity for
constructing a case against a universal method. I suggest that the argument
of this article indicates that Feyerabend has not taken full advantage of that
opportunity, and spoils his case by exaggerated, unsubstantiated and
unsupportable claims.

ALAN CHALMERS

University of Sydne

REFERENCES

BLOCK, I. [I961]: 'Truth and Error in Aristotle's Theory o

Quarterly, ii, pp. I-9.
CARNAP, R. [1981]: 'The Unity of Science', in O. Hanfling (ed.
Positivism. Blackwell.
CLAVELIN, M. [1974]: The Natural Philosophy of Galileo. M.I.T. Press.
DRAKE, S. [1957]: Discoveries and Opinions of Galileo. Doubleday Anchor.
DRAKE, S. [1978]: Galileo at Work. University of Chicago Press.
DRAKE, S. and O'MALLEY, C. D. [1960]: The Controversy of the Comets of i618. University
Pennsylvania Press.
FEYERABEND, P. [1975]: Against Method. New Left Books.
FEYERABEND, P. [1978]: Science in a Free Society. New Left Books.
GALILEI, GALILEO [1967]: Dialogue Concerning the Two Chief World Systems. Universit
California Press.
GAUKROGER, S. [1978]: Explanatory Structures. Harvester.
GAUKROGER, S. [1981]: 'Aristotle on the Function of Sense Perception', Studies in History a
Philosophy of Science, 12, pp. 75-89.
KUHN, T. [1959]: The Copernican Revolution. Random House.
LAKATOS, I. [1968]: 'Criticism and the Methodology of Scientific Research Programme
Proceedings of the Aristotelian Society, 69, pp. 149-86.

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms
184 I. J. Good
OSIANDER [1978]: Forward to Copernicus, N.: On the Revolutions of the Heavenly Spheres, E.
Rosen (trans.) Macmillan.
POPPER, K. [1969]: Objective Knowledge. Oxford University Press.
PRICE, D. J. DE S. [1959]: 'A Critical Re-Estimation of the Mathematical Planetary Theory of
Ptolemy, Copernicus and Kepler' in M. Clagett (ed.): Critical Problems in the History of
Science. University of Wisconsin Press.
ZAHAR, E. [I1976]: 'Why Did Einstein's Programme Supersede Lorentz's?' in C. Howson (ed.),
Method and Appraisal in the Physical Sciences. Cambridge University Press.

A HISTORICAL COMMENT CONCERNING NOVEL CONFIRMATION

Campbell and Vinci ([19831, p. 325) give the following obj

'standard' Bayesian approach to the confirmation or corrob
hypothesis h. (I have changed their oblique stroke to a vert
this is the standard notation in modern books on statistics or mathematical
probability. Also I have written P for pr.) They say:

Suppose as before that P(e jh & b)= i but this time imagine that e is known
[contained in the background information] so that P(e I b) = i. It follows im-
mediately from Bayes' Theorem that P(h I e & b) = P(h I b). Hence known evidence
cannot offer any additional confirmation. But this implication is counter-intuitive,
since there are central examples, as we noted earlier, where already known evidence
has increased the confirmation of a hypothesis.

Later, the authors resolve this difficulty in their long paper. My main
purpose is to point out that the difficulty was briefly resolved by an
argument published in 1968, which was based on the concept of dynamic
probability.
To make the argument concrete we may plug in 'values' for e, h and b. Let
h denote Newton's inverse square law of gravitation, let b, denote the
detailed observations of the positions and motions of planets, let b2 denote
other background information, so that b = b, & b2, and let e denote the fact
that the orbits of the planets are (approximately) elliptical. Then h implies e,
so P(bl I h & b2)/P(bl I - h & b2) would rightly be judged to be large by most
astronomers, that is, b has already provided substantial evidence in favour of
h. This corroboration of h is not increased by stating the evidence a second
time, if it has already been taken into account. Sometimes the background
information supports a hypothesis h, but the support has been overlooked.
When the support is noticed for the first time the multisubjective
probability of h increases suddenly. An example was the support of
Newton's inverse square law of gravitation by the previously known
approximate elliptical orbits of the planets. Before Newton did the
mathematics, this support was not noticed. This point was made by Good
([1968], p. 129). The argument exemplified the concept of dynamic
probability, called there shifting, slippery, sliding, superficial, surface,
changeable, or evolving probability. The same example has been recently
used by Niiniluoto [1983] who refers to the shift in conclusions owing to the

This content downloaded from

143.117.16.36 on Thu, 04 Mar 2021 15:05:09 UTC
All use subject to https://about.jstor.org/terms