Sei sulla pagina 1di 49

!

The most philosophically of all the sciences:


"#$% &'(()$ #*+ (,-./0#% 0'.1'%'2-

HeIge Kragh
Cenlre for Science Sludies, Inslilule of Ihysics and Aslronomy
Aarhus Universily, 8000 Aarhus, Denmark
I-maiI: heIge.kragh+ivs.au.dk

34.5$#05 IrobIems of scienlific cosmoIogy onIy rareIy occur in lhe vorks of
KarI Ioer, lhe famous Auslrian-rilish hiIosoher. NeverlheIess, il vas a
sub|ecl lhal inleresled him and vhich he occasionaIIy commenled on. Whal is
more imorlanl, his generaI cIaim of faIsifiabiIily as a crilerion lhal
demarcales science from non-science has Iayed a significanl roIe in eriods of
lhe deveIomenl of modern hysicaI cosmoIogy. The aer examines lhe
hisloricaI conlexls of lhe inleraclion belveen cosmoIogy and Ioerian
philosophy of science. Apart from covering Poppers inspiration from Einstein
and his vievs on queslions of cosmoIogy, il focuses on lhe imacl of his
lhoughls in lvo eriods of conlroversy of modern cosmoIogy, lhe one reIaled
lo lhe sleady slale lheory and lhe olher lo lhe recenl muIliverse roosaI. Il
lurns oul lhal lhe imacl has been considerabIe, and conlinues lo be so, bul
aIso lhal lhe versions of Ioerian melhodoIogy discussed by cosmoIogisls
are somelimes far from vhal Ioer acluaIIy lhoughl and vrole.

6 7*5$'+805/'*
While Karl Poppers philosophy of science has only few followers among
modern hiIosohers, il is easiIy lhe viev of science vilh lhe biggesl imacl
on raclicing scienlisls. According lo Ieler Medavar, NobeI Iaureale and
eminenl hysioIogisl, Ioer vas lhe grealesl aulhorily ever on lhe scienlific
method. He praised the great strength of Karl Poppers conception of the
#

scienlific process, a main reason for the praise being that is realistic il
gives a relly fair iclure of vhat goes on in real life laboratories.
1
Iilher
exIicilIy or (more oflen) imIicilIy, many scienlisls subscribe lo some
version of simIified Ioerianism, usuaIIy by adoling lhe demarcalion
crilerion lhal a scienlific lheory musl be faIsifiabIe. This crilerion has Iayed
an imorlanl roIe in science ever since lhe ubIicalion of !"# %&'() &* +)(#,-(*()
.(/)&0#12 in 1959. One of lhe sciences vhere lhis conlinues lo be lhe case is
cosmology, which Popper once called the most philosophically of all the
sciences.
2
Yel his ovn commenls on cosmoIogicaI research vere fev and
scallered, in marked conlrasl lo lhe osilive recelion of his ideas in lhe
communily of hysicisls, aslronomers, and cosmoIogisls.
In lhis aer I deaI in some delaiI vilh lhe inleraclion belveen
Poppers philosophy of science and developments in physical cosmology
afler WorId War II. This is a loic lhal has received IillIe allenlion in eilher
lhe hisloricaI or hiIosohicaI Iileralure bul deserves a cIoser examinalion.
3

Aarl from examining vhal IillIe can be found in Poppers writings, I
incIude in Seclion 5 some nev informalion aboul his Iale oinion of
cosmoIogicaI modeIs, as he described il in a hilherlo unknovn Ieller shorlIy
before his dealh in 1994. The vay in vhich Ioerian melhodoIogy has
infIuenced lhe deveIomenl of cosmoIogy is anaIyzed by Iooking in
arlicuIar al lvo imorlanl conlroversies, lhe one (lhe sleady slale debale)
daling from lhe 1950s and lhe olher (lhe muIliverse debale) from lhe firsl
decade of lhe lvenly-firsl cenlury. Il is hardIy surrising lhal lhe version of
Ioerianism aearing in scienlific cosmoIogy, as in science in generaI,

1
Medavar (1990, . 100).
2
Ioer (1994, . 59).
3
Ior lhe case of lhe sleady slale conlroversy, see Kragh (1996, . 244-246), and for
the later impact of Poppers ideas, Sovacool (2005).
$

differs quile subslanliaIIy from vhal Ioer himseIf beIieved and vrole.
This sludy has lvo uroses: il aims al adding lo lhe Ioer schoIarshi by
looking at Poppers ideas of cosmology; and it aims at adding lo lhe hislory
of modern cosmology by investigating how Poppers philosophical ideas
infIuenced lhe deveIomenl. Thus, il is Iaced al lhe crossroad belveen lhe
hislories of science and hiIosohy.

9 &'(()$ #*+ :/*.5)/*
In }anuary 1921 eighleen-year oId Ioer Iislened lo a ubIic Ieclure Iinslein
gave in Viennas crovded Concerl HaII, bul lhe reIalivily lheory lhal Iinslein
talked about was quite beyond my understanding, as he later recaIIed. Yel,
shorlIy lhereafler he vas inlroduced lo Einsteins theory and what he called
his marvelIous idea of a nev cosmoIogy a finile bul unbounded
universe.
4
AIlhough he did nol deveIo an inleresl in lhis arlicuIar loic,
he was greatly inspired by the general theory of relativity and Einsteins
allilude lo observalionaI lesls. Iinslein insisled lhal if lhere vere )&,)34/(0#
evidence againsl a lheory, il had lo be abandoned. Referring lo lhe rediclion
of a gravitational redshift, he wrote: If lhe disIacemenl of seclraI Iines
lovards lhe red by lhe gravilalionaI olenliaI does nol exisl, lhen lhe generaI
theory of relativity will be untenable.
5
This vas an allilude lhal imressed
Ioer and vhich he arovingIy quoled in his aulobiograhy.
6


4
Ioer (1974a, . 28-29). The friend who introduced him to Einsteins general
lheory of reIalivily vas a sludenl of malhemalics by lhe name Max IIslein.
5
Iinslein (1920, . 132). Iinslein (2004, . 304) said lhe same in a Ieller lo Iddinglon
of 15 December 1919. See also Hentschel (1992), where Einsteins attitude to
emiricaI lesling is anaIyzed and rovided vilh documenlary evidence. As oinled
out by Holton (1998, p. 8), the earlier editions and printings of Einsteins 56#1 7(#
%

Ioer firsl ubIished his Ialer so famous demarcalion crilerion in a
Ieller of 1933 lo 819#,,-,(/, lhe |ournaI for anaIylic hiIosohy founded a fev
years earIier by Hans Reichenbach and RudoIf Carna. Il reads: Statements,
or syslems of slalemenls, convey informalion aboul lhe emiricaI vorId onIy
if lhey are caabIe of cIashing vilh exerience, or more reciseIy, onIy if lhey
can be subjected to tests which :('"- result in their refutation.
7
A fev
Iines Ialer he reformuIaled lhe demarcalion crilerion by arahrasing a
formulation by Einstein in an address of 1933: In so far as a scientific
slalemenl seaks aboul reaIily, il musl be faIsifiabIe, and in so far as il is nol
faIsifiabIe, il does nol seak aboul reality.
8
A very simiIar formuIalion
appeared in Poppers exlensive manuscril, .(# 6#(7#, ;14,7<1&63#:# 7#1
819#,,-,(/-"#&1(#, vhich vas onIy ubIished in 1979.
9

Al severaI occasions Ioer described key eIemenls in his
hiIosohy of science as cruciaIIy reIying on, or even derived from,
Iinsleinian hysics and melhodoIogy. In a C radio rogramme he
exressed his indebledness lo Iinslein as foIIovs: Einsteins influence on
my lhinking has been immense. I mighl even say lhal vhal I have done is
mainIy lo make exIicil cerlain oinls vhich are imIicil in lhe vork of
Einstein. The Einsteinian revolution has influenced my own views deeply:

/<#=(#33# 4,7 >33'#:#(,# ?#3>-(0(-@-/-"#&1(# (1917-1919) did nol incIude eilher lhis
senlence or somelhing lo lhe same effecl.
6
Ioer (1974a, . 29).
7
Ioer (1932-1933), as lransIaled in Ioer (1959, . 312-314). Quolalion on .
313.
8
Ioer (1959, . 313). In his address of 1921 on Geometry and Experience,
Iinslein (1962, . 119) said: In so far as lhe slalemenls of malhemalics refer lo
reaIily, lhey are uncerlain, and in so far lhal lhey are cerlain, lhey do nol refer lo
reality.
9
An IngIish lransIalion has onIy recenlIy aeared: Ioer (2008).
&

I feeI I vouId never had arrived al lhem vilhoul him.
10
This is aIso vhal he
said in his aulobiograhy:

What impressed me most was Einsteins own clear statement that he
vouId regard his lheory as unlenabIe if il shouId faiI in cerlain lesls.
Thus I arrived, by the end of 1919, at the conclusion that the
scienlific allilude vas lhe crilicaI allilude, vhich did nol Iook for
verificalions bul for cruciaI lesls, lesls vhich couId 1#*4-# lhe lheory
lesled, lhough lhey couId never eslabIish il.
11


Hovever, his relroseclive accounls of lhe genesis of his ideas on induclion
and demarcalion shouId nol be acceled uncrilicaIIy. These ideas may nol
have had lheir rools in Iinsleinian hysics versus sychoanaIysis and
Marxism, but rather in Poppers occupation vilh edagogy and sychoIogy.
Moreover, he mosl IikeIy arrived al lhe ideas onIy in 1929 or 1930, aboul a
decade after he learned about Einsteins attitude to critical tests.
12
On lhe
other hand, there is no reason to doubt Einsteins crucial influence on
Poppers development of his phiIosohy. This infIuence is iIIuslraled by lhe
firsl edilion of !"# %&'() &* +)(#,-(*() .(/)&0#12, vhere Iinslein is lhe second-
mosl ciled aulhor, vhelher scienlisl or hiIosoher. OnIy lhe German

10
Whilrov (1967, . 23-25).
11
Ioer (1974a, . 29). AIso in Ioer (1963, . 39) did he dale lhe faIsifiabiIily
crilerion lo lhe vinler of 1919-1920. At the time it seemed almost trivial to him,
and so il look lhirleen years before he ubIished il.
12
See ler Hark (2002) and aIso Iggers Hansen (2006) from vhich il aears lhal sliII
in 1927 Ioer considered Iavs of nalure lo be verifiabIe and based on induclive
reasoning. OnIy in .(# 6#(7#, ;14,7<1&63#:# did he reach lhe concIusion lhal Iavs of
nalure musl be faIsifiabIe and nol verifiabIe.
'

Iogician and hiIosoher RudoIf Carna, lhe Ieading figure in lhe Vienna
osilivisl circIe, received more references.
As lo Iinslein, he recognized in Ioer a kindred siril. Having read
%&'(9 7#1 A&1/)"4,' shorlIy afler il aeared in lhe aulumn of 1934, he vrole
lo Ioer: Your book has pleased me very much in many vays: re|eclion of
the inductive method from an epistemological standpoint. Also falsifiability
as the crucial element of a theory (of reality) . You have further defended
your positions really well and astutely.
13
He even offered lo bring Iopers
book lo lhe allenlion of his coIIeagues in hysics and asked Ioer of hov
he couId besl heI him.
The cIose affinily belveen lhe vievs of Iinslein and Ioer, and
lherefore lhe reason vhy Iinslein greeled %&'(9 7#1 A&1/)"4,' vilh such
symalhy, is erhas besl iIIuslraled by an essay Iinslein vrole in Iale 1919
in lhe B#13(,#1 !>'#63>--. Here he seIIed oul his conviclion lhal lhe crealive
scienlisl musl aIvays slarl vilh a hyolhesis, vhich may be nolhing bul a
reconceived viev of inluilion. The scienlisl lhen deduces cerlain
consequences from his hyolhesis, and lhese can be confronled vilh nalure:
A lheory can lhus be recognized as erroneous |4,1()"-('j if lhere is a
IogicaI error in ils deduclions, or as incorrecl |4,=4-1#**#,7j if a facl is
nol in agreemenl vilh ils consequences. ul lhe lrulh of a lheory can
never be roven. Ior one never knovs lhal even in lhe fulure no
exerience viII be encounlered vhich conlradicls ils consequences,

13
Iinslein lo Ioer, 15 }une 1935, as quoled in Van Dongen (2010, . 43).
(

and sliII olher syslems of lhoughl are aIvays conceivabIe vhich are
caabIe of |oining logelher lhe same given facls.
14

Also Einsteins mature philosophy of science, as he formulated it in a letter
of 1952 lo his friend Maurice SoIovine, vas in broad agreemenl vilh
Poppers views.
15

Irom 1937 lo 1945 Ioer slayed in Chrislchurch, Nev ZeaIand,
vhere he laughl hiIosohy al lhe Universily of Canlerbury. He
subsequenlIy moved lo IngIand lo lake u a osilion al lhe London SchooI
of Iconomics and IoIilicaI Science, and in 1949 he vas aoinled rofessor
of Iogic and scienlific melhod al lhe Universily of London. Iinslein had since
1933 been al lhe Inslilule for Advanced Sludy in Irincelon. Il vas onIy in
1950, vhen inviled lo give a aer in Irincelon on indelerminism in
quanlum hysics, lhal Ioer came lo meel lhe hysicisl he admired so
much. On lhis occasion he discussed, bolh vilh Iinslein and Kurl GdeI, a
cosmoIogicaI modeI lhal GdeI had recenlIy roosed on lhe basis of
Einsteins field equations and in which there was no cosmic lime.
16
Gdels
slrange universe vas rolaling and slalionary hence vilh no redshifls and
aIIoved for lime lraveIs inlo bolh lhe asl and lhe fulure. UnderslandabIy,
Iinslein re|ecled lhe modeI as nolhing bul lhe brainchiId of a cIever
malhemalician. Whal mallers here is lhal Ioer seems lo have been veII

14
Einstein, Induktion und Deduktion in der Physik, B#13(,#1 !>'#63>--, 25
December 1919, rerinled in Iinslein (2002, . 219). IngIish lransIalion in Adam
(2000), who discusses the significance of Einsteins essay in relation to Poppers
vievs. Hovever, Ioer vas unavare of lhe essay and onIy came lo knov aboul il
in 1983, vhen lhe hysicisl and Iinslein schoIar }ohn SlacheI aroached him on lhe
maller.
15
Iinslein lo SoIovine, 7 May 1952, as anaIyzed in HoIlon (1998, . 28-56).
16
For Gdels model universe and sources relevant to it, see Rindler (2009).
)

acquainted with Gdels model universe and its deficiencies, indicating that
he vas abIe lo foIIov lhe lechnicaI Iileralure in malhemalicaI cosmoIogy.
17

Il is debalabIe vhelher Iinslein, in his scienlific raclice, can be
characlerized as a Ioerian. Yel lhe affinily belveen his vievs and lhose of
Ioer vas nol Iimiled lo lhe rheloricaI IeveI, as may be iIIuslraled by his
resonse lo lhe lime-scaIe difficuIly lhal Iagued cosmoIogy for more lhan
lvenly years. In mosl reIalivislic modeIs of lhe exanding universe lhe age !
is of lhe order of lhe HubbIe lime, or lhe inverse of lhe HubbIe conslanl C.
The robIem vas lhal lhe age, as derived in lhis vay, vas nol onIy shorler
lhan lhe age of lhe slars bul aIso of lhe reIiabIy delermined age of lhe Iarlh.
Ior examIe, according lo lhe videIy favoured Iinslein-de Siller modeI, lhe
age is given by ! 2/3C, vhich resuIled in ! 1.2 biIIion years. There vere
vays lo avoid lhe aradox, bul Iinslein insisled lhal lhe anomaIy shouId be
laken seriousIy and nol be exIained avay. The cosmologic theory here
presented, he wrote in 1945, referring to the relativistic big bang theory,
would be 7(/<1&0#7 if il vere found lo conlradicl any such resuIls. In lhis
case I can see no reasonable solution.
18

Einsteins attitude was in the spirit of Popperian falsificationism,
such as vas his allilude in 1919 vilh resecl lo lhe gravilalionaI bending of
starlight. The if should be noticed, however, for it indicates that Einslein
vas nol, afler aII, convinced lhal lhe cosmoIogicaIIy derived age of lhe
universe vas an inconlroverlibIe facl. ul lhis quaIificalion, loo, agrees vilh

17
See Ioer (1974a, . 172).
18
Iinslein (1956, . 132), emhasis added. On lhe lime-scaIe difficuIly, see Kragh
(1996, . 73-79, 271-275) and rush (2001). The anomaIy onIy disaeared in lhe
1950s, vhen il vas reaIized lhal lhe HubbIe conslanl is much smaIIer lhan originaIIy
assumed. y 1960 lhe age of lhe Iinslein-de Siller universe vas laken lo be al Ieasl 7
biIIion years.
*

lhe viev of Ioer according lo vhom a scienlisl shouId nol inslanlIy
abandon a lheory lhal Ieads lo disagreemenls vilh dala (see Seclion 6).

; Physical cosmology in Poppers work
As ve shaII see in lhe foIIoving seclions, Poppers philosophy of science has
Iayed, and conlinues lo Iay, an imorlanl roIe in melhodoIogicaI debales
concerning cosmoIogy. On lhe olher hand, one Iooks in vain in his main
vorks for discussions of lhe science of lhe universe. Quanlum mechanics,
seciaI reIalivily lheory, lhermodynamics, and slalislicaI hysics aear
rominenlIy, and vilhin lhese branches of foundalionaI hysics Ioers
conlribulions are sliII recognized as imorlanl.
19
ul he vas nearIy siIenl
aboul cosmoIogy in lhe meaning of sace-lime lheories of lhe universe al
Iarge. The siIence is nol comIele, lhough, for if one Iooks cIoseIy al his books
and aers, queslions of cosmoIogy do lurn u, if onIy briefIy and scallered.
efore ve Iook al lhese, il needs lo be oinled oul lhal vhiIe lhe lerm
cosmology appears repeatedly in Poppers writings, in far mosl of lhe
cases il does nol refer lo scienlific cosmoIogy, vhelher of lhe hysicaI,
observalionaI, or mathematical kind. All science is cosmology, he wrote in
lhe reface lo lhe IngIish edilion of %&'(9 7#1 A&1/)"4,' lhal aeared in 1959
as !"# %&'() &* +)(#,-(*() .(/)&0#12.
20
Hovever, he used lhe lerm in a very
broad sense, stating that the problem of cosmology was the problem of
underslanding lhe vorId incIuding ourseIves, and our knovIedge, as arl
of the world. This was a sense not very different from, for example, the one
adoled by AIfred Norlh Whilehead, vhose main vorkD E1&)#// >,7 ?#>3(-2

19
See lhe survey in }ammer (1991).
20
Ioer (1959, . 15). The senlence aIso aeared in some of his Ialer vorks, e.g.,
Ioer (1982, . 1).
!+

from 1929, carried lhe sublilIe F, 8//>2 (, G&/:&3&'2. When Ioer deaIl
vilh lhe hislory of cosmoIogicaI lhoughl, il vas nol vilh deveIomenls in
lhe lvenlielh cenlury, bul vilh lhe Iresocralics, IIalo, Coernicus, and
Kanl.
21
Iarmenides and Kanl vere his favourile cosmoIogisls. In 1982 he said
that his ideas about the quantum world had been inspired by the problems
of physical cosmology, yet he used the term in a meaning entirely differenl
from lhe one adoled by aslronomers and hysicisls. TeIIingIy, his inleresl in
vhal he caIIed hysicaI cosmoIogy bul vhal olhers mighl erhas caII
melahysics or naluraI hiIosohy led him to biology, to the human
mind, and lo lhe roducls of the human mind.
22

In sile of his siIence aboul scienlific cosmoIogy in %&'(9 7#1
A&1/)"4,', Popper was aware of the developments initiated by Einsteins
seminaI aer of 1917 in vhich Iinslein had aIied his generaI lheory of
reIalivily lo lhe enlire universe. He seems lo have been broadIy avare of
vhal haened in cosmoIogy, bul nol lo have foIIoved lhe deveIomenl
cIoseIy. Shortly before his death in 1994, he described himself as an ardent
admirer of Friedmanns suggestion, a reference to the Russian physicist
AIexander Iriedmann vho in 1922 had shovn lhal lhere are exanding and
other dynamical solutions to Einsteins cosmological field equations.
According to Popper, Friedmanns cosmology made the explosion (nov
called the Big Bang) a simple explanation of all that was then known
empirically about the Universe.
23
Moreover, he considered il a

21
See, for examIe, the essay Kants critique and cosmology, reprinted in Popper
(1963, . 175-183). The essay vas firsl ubIished in !"# %(/-#,#1 51 (1954), 291-292,
303.
22
Ioer (1982, . 31).
23
K. Popper to H. Kragh, 10 June 1994. The letter, presumably one of Poppers last
(he died on 17 Selember 1994), vas a reIy lo some queslions I had asked him in
rearalion of a book on lhe hislory of modern cosmoIogy (Kragh 1996). Il is nov
!!

simplification of Einsteins original theory because it made the cosmological
constant an unnecessary assumption.
OnIy al one occasion, a IillIe knovn aer of 1940 lhal has never
been rerinled, did Ioer enler lhe discussion in cosmoIogy. Al lhal lime
lhe Iinslein-Iriedmann lheory vas chaIIenged by a very differenl syslem of
cosmoIogicaI hysics roosed by I. Arlhur MiIne, a briIIianl and
unorlhodox rilish aslrohysicisl. It was part of Milnes system that physical
evenls couId be described in differenl lime scaIes, Ieading lo differenl bul
hysicaIIy equivaIenl descrilions. Ior examIe, according lo one lime scaIe -
lhe universe vouId exand, vhiIe according lo anolher lime scaIe ,
IogarilhmicaIIy reIaled lo lhe firsl (- - Iog ), il vouId remain in a slalic
slale.
24
Poppers purpose with his letter in H>-41# vas lo discuss lhe generaIIy
acceled exansion of lhe universe in reIalion lo lheories lhal exIained
Edwin Hubbles redshift data on the assumption of special mechanisms
oeraling in a slalic universe. According lo one aIlernalive lhe seed of Iighl
decreased vilh lime, vhiIe anolher aIlernalive (lhe cIass of tired light
hyolheses) assumed lhal Iighl graduaIIy Iosl energy during ils |ourney
lhrough emly sace from lhe nebuIae lo lhe Iarlh.
CIearIy insired by MiIne, Ioer examined lvo aIlernalives lo lhe
reIalivislic lheory of cosmic exansion, arguing lhal lhey agreed in regard lo
observabIe effecls, lhal is, lhey Ied lo gaIaclic redshifls of lhe kind redicled

deosiled al lhe KarI Ioer Library, KIagenfurl Universily, Auslria. The main
conlenl of lhe Ieller is quoled in a nole forlhcoming in lhe I&41,>3 *&1 -"# C(/-&12 &*
F/-1&,&:2. Friedmanns ideas of a universe evolving in time (which included both
= 0 and 0) became generally known and accepted only after 1930, so Poppers
admiralion resumabIy refers lo lhe 1930s. Ior lhe earIy cosmoIogies of Iinslein, de
Siller, Iriedmann and olhers, see, for examIe, Norlh (1965).
24
On Milnes system, known as the theory of kinematic relativity, see Lepeltier
(2006) and Kragh (2011, . 101-108). Ior conlemorary assessmenls from a
hiIosohicaI oinl of viev, see Cohen (1950) and Grnbaum (1952).
!#

by the expansion theory. The three theories are logically equivalent, he
wrote, and lherefore do nol describe aIlernalive *>)-/, bul lhe same facls in
aIlernalive 3>,'4>'#/. But does the universe 1#>332 exand` Or does lhe
seed of Iighl inslead decrease` Or is il ralher lhe case lhal lhe frequency of
holons changes vilh lhe dislance lhey lraveI` According lo Ioer, lhe
question was not more legitimate than, when prices of goods fall
throughout the economic system, to ask whether in reality the value of
money has increased or the value of the goods has decreased.
25

NeverlheIess, he considered Milnes alternative a simpler and therefore more
attractive explanation of the observed redshifts. Poppers intervention in the
debale concerning lhe exanding universe did nol arouse much allenlion. Il
seems lo have received onIy one resonse, from lhe IngIish aslrohysicisl
and hiIosoher of science Herberl DingIe.
26

Al lvo olher occasions did Ioer lake u loics of cosmoIogicaI
reIevance, aIlhough in none of lhe cases in vays lhal direclIy reIaled lo
research going on in scienlific cosmoIogy. In 1953 lhe nevIy eslabIished
B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# announced a rize essay on lhe
IogicaI and scienlific slalus of lhe concel of lhe lemoraI origin and age of

25
Popper (1940, p. 70). According to Poppers letter to the author (note 23), he was
al lhe lime inleresled in non-DoIer exIanalions of lhe redshifls and Ialer
ubIished a remark in favour of redshifls increasing vilh dislance because of
coIIisions vilh arlicIes in inlergaIaclic sace. I have been unabIe lo lrace lhis
remark, vhich is nol menlioned in any of lhe Ioer bibIiograhies, and susecl
that Poppers memory failed him.
26
Dingle (1940), who criticized Poppers argument, accusing it of being an examIe
of a very dangerous tendency in modern physics, , namely, a retreat from
experience into the world of pure logic. Dingle was an outspoken critic of Milnes
lheory. He Ialer became friendIy vilh Ioer, vilh vhom he corresonded vilh
regard lo lhe vaIidily of lhe seciaI lheory of reIalivily. See Hayes (2010).
!$

lhe universe.
27
This vas of course a cenlraI queslion in lhe conlroversy
belveen lhe sleady slale lheory and lhe big bang lheory lhal vas raging al
lhe lime. The lvenly-six essays vere |udged by a commillee incIuding
Ioer, lhe olher members being HaroId }effreys (a malhemalician and
geohysicisl), Irilz Ianelh (a chemisl), and LanceIol Whyle (a hysicisl and
science vriler). HaIf of lhe firsl rize in lhe comelilion vas avarded lhe
American hiIosoher MichaeI Scriven, vho in his essay denied lhal il couId
be decided on a scienlific basis vhelher lhe age of lhe universe is finile or
infinile.
28
Ioer may have agreed.
The essay of lhe olher rize vinner, lhe hysicaI chemisl }. T. Davies
of Kings College, London, is of interest because it demonstrates the
influence of Poppers ideas. In his discussion of how to choose belveen rivaI
lheories of lhe universe, Davies slressed lhal lhe uIlimale lesl vas lo
compare deductions with new observations. He referred twice to Poppers
%&'(9 7#1 A&1/)"4,', vhich he evidenlIy had sludied carefuIIy: Unless we
can conceive of a crilicaI observalion vhich couId dislinguish belveen rivaI
theories, the latter are either unscientific or are stating the same thing in
differenl guises. As Ioer has emhasised, lhe crilerion of a scienlific
lheory is lhal il musl be ossibIe for an observalionaI check lo be devised
(hovever imraclicaI vilh exisling lechniques), by vhich il mighl be
disproved.
29

If science vere overIess lo decide belveen an elernaI and a finile-
age universe, erhas lhere couId be given hiIosohicaI and IogicaI
argumenls lhal ruIed oul an infinile asl` This is an oId queslion, much

27
See B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# 4 (1953): 92, 5 (1954): 179, and H>-41#
175 (1955): 68-69.
28
Scriven (1954). See aIso rush (2001, . 168).
29
Davies (1954, . 199).
!%

discussed aIready in lhe middIe ages, and in 1978 lhe aslronomer GeraId
Whilrov argued from reasons of Iogic lhal an infinileIy oId universe is
indeed imossibIe.
30
Ioer reIied lhal such a riori reasoning vas invaIid
and lhal lhe queslion mighl veII be scienlificaIIy meaningIess. Referring lo
Milnes two time scales, by means of which a universe without a beginning
couId be lransformed into one with a beginning, he wondered whether
lhere is an onloIogicaI difference corresonding lo lhe difference belveen a
lime co-ordinale reaching inlo an infinile asl and a lime co-ordinale vilh a
beginning.
31
His reasoning vas lhus of lhe same kind as in his aer of
1940. Of course, by lhe Iale 1970s lhe big bang modeI had oblained a
aradigmalic slalus and lhe Iarge ma|orily of cosmoIogisls vere convinced
lhal lhe universe had exisled for onIy a finile number of years. In his reIy lo
Whilrov, Ioer did nol refer lo lhe hysicaI argumenls for a finile-age
universe, bul he imIicilIy denied lhal lhese rigorousIy roved a big bang
universe of finile age. As far as he vas concerned, lhe universe mighl have
exisled for an infinily of lime.
Much inleresled in queslions of lhe direclion of lime and lhe
irreversibiIily of hysicaI rocesses, in lhe eriod 1956-1967 Ioer vrole
severaI aers on lhese and reIaled sub|ecls in H>-41# and lhe B1(-(/" I&41,>3
*&1 -"# E"(3&/&<"2 &* +)(#,)#. The robIem of lhe so-caIIed arrov of lime is a
cIassicaI one in naluraI hiIosohy, and il is no Iess imorlanl in
foundalionaI hysics. Does lhe second Iav of lhermodynamics rovide an
arrov of lime` Or do lhe Iavs of eIeclrodynamics erhas rovide one` The

30
Whilrov (1978).
31
Ioer (1978). See aIso Craig (1979), vho resonded lo lhe argumenls of
Whilrov and Ioer and reIaled lhem lo conlemorary deveIomenls in hysicaI
cosmology. Craig suggested that Poppers arguments were unconvincing and that
modern cosmoIogy and aslrohysics confirmed lhal lhe age of lhe universe vas of
lhe order len biIIion years.
!&

queslion enlered cosmoIogy in lhe 1960s, vhen severaI hysicisls and
cosmoIogisls discussed lhe exansion of lhe universe as a cosmoIogicaI
arrov of lime and hov il reIaled lo lhe one based on eIeclromagnelic signaIs.
Some advocales of lhe sleady slale rogramme, incIuding Ired HoyIe, }ayanl
NarIikar, and Dennis Sciama, argued lhal onIy lhe sleady slale lheory vas
abIe lo exIain vhy ve Iive in a vorId vilh a shar difference belveen
signaIs coming from lhe asl and from lhe fulure.
32

Al aboul lhe same lime lhal lhe arrov of lime robIem vas
discussed among cosmoIogisls, Ioer examined il indeendenlIy and from
a more generaI erseclive. In a aer of 1965 he argued lhal lhe enlroic
lheory of lime, according lo vhich lhe grovlh of enlroy as given by
lhermodynamics defines lhe direclion of lime, vas unfounded and in any
case of no cosmoIogicaI significance.
33
Radialion rocesses, he suggesled,
vere lhe onIy irreversibIe rocesses lhal vere reIevanl on a cosmic scaIe. Al a
Ialer occasion he exIained lhal vhen he insisled lhal lime has an arrov, il
was to emphasize that the universe has a history: The present age of the
universe is a perfectly good term in cosmology. In other words, the past,
resenl, and fulure are erfeclIy good lerms in cosmoIogy and astronomy.
34

Poppers approach was primarily philosophical, and he did not relate his
vork on lhe arrov of lime lo lhe conlemorary debale in lhe smaII
communily of cosmoIogisls.
In a re|oinder of 1967 lo crilicaI ob|eclions raised by a German
hiIosoher, W. cheI, Ioer exanded his discussion of irreversibiIily lo

32
See Kragh (1996, . 368-372) and Sanchez-Ron (1990).
33
Popper (1965). For a critical discussion of Poppers view, see Esfeld (2006).
34
Ioer (1974b, . 1143). The comIex of robIems Ioer deaIl vilh conlinues lo
allracl scienlific and hiIosohicaI inleresl. See, for examIe, lhe conlribulions in
+-47(#/ (, C(/-&12 >,7 E"(3&/&<"2 &* J&7#1, E"2/()/ 37: 3 (2006), a seciaI issue on lhe
arrovs of lime.
!'

cover some of lhe reIalivislic modeIs of lhe universe, incIuding lhe cycIic
modeI lhal Iinslein had roosed in 1931.
35
Ior lhis arlicuIar case, vhich
assumes a osilive curvalure of sace, he admilled lhal lhe enlroy vouId
increase conlinuaIIy. The 1967 nole demonslrales lhal Ioer had a cIear
gras of lhe essence of reIalivislic cosmoIogy and aIso indicales lhal he did
nol beIieve in a universe of the cyclic type. In the case of = 0, a osilive
sace curvalure necessilales an average mass densily grealer lhan a cerlain
vaIue knovn as lhe crilicaI densily and given by cril 3C
2
/, vhere C is
Hubbles expansion parameter and Iinsteins gravitational constant related
to Newtons constant ; by = 8;/)
2
. Ioer aarenlIy lhoughl lhal lhe
condilion > cril vas an unleslabIe hyolhesis, or ralher lhal Iinslein
thought so. Relying on Einsteins !"# J#>,(,' &* ?#3>-(0(-2, Ioer vrole:
He [Einstein] regarded lhe con|eclure > 3C
2
/ as unleslabIe, and lhe
oosile con|eclure as leslabIe or refulabIe, vhich mighl make il referabIe
to some.
36

As Iinslein oinled oul, lhe gravilaling mass densily is IikeIy lo
consisl of lvo comonenls, one due lo nebuIae and olher radialing maller r
and anolher lo non-radialing or dark maller d. The lolaI mass densily and
lherefore lhe curvalure of sace viII deend on r + d or d/r. If r < cril,
ve may sliII have > cril, a osilive curvalure. Iinslein shoved lhal lhe
Iover bound for viII deend on d/r, a quanlily vhich can be eslimaled
from lhe seclroscoicaIIy delermined veIocilies of lhe gaIaxies in a gIobuIar
cIusler. On the other hand, he wrote, I cannol lhink of any reasonabIy
reIiabIe delerminalion of an uer bound for . Ioer seems lo have read
Iinslein as saying lhal a cIosed universe cannol be faIsified, for lhe onIy vay

35
Ioer (1967). cheI (1967).
36
Ioer (1967). Iinslein (1956, . 131-132). Notice Poppers identification of
testable and refutable.
!(

viII be lo shov lhal cril, and even if il is shovn lhal r cril il can aIvays
be argued lhal r + d > cril.
Al aboul lhe lime Ioer vrole his 1967 nole lo H>-41#, lhere had
occurred somelhing Iike a aradigm shifl in cosmoIogy, vilh lhe emergence
of lhe slandard hol big bang lheory. Ioer did nol commenl exIicilIy on
lhe nev hysicaI cosmoIogy, bul in a fev cases he indicaled his viev. In a
plenary lecture on The Theory of the Ob|eclive Mind delivered in 1968 to
lhe 14lh InlernalionaI Congress of IhiIosohy in Vienna, he briefIy deaIl
vilh cosmoIogicaI modeIs the most interesting kind of aII hyolheses, he
caIIed lhem. Some of lhese, he said, can of course be lesled, and some have
been even sufficienlIy recise for refulalion. ul olhers, and very inleresling
ones, seem lo be unleslabIe, and may remain so.
37
He may have lhoughl of
lhe sleady slale lheory and lhe nev big bang lheory, resecliveIy, bul did nol
eIaborale. In a more general way Popper referred to the infinitely
improbable success of modern cosmology in the 1974 festschrift issued in
lhe %(61>12 &* %(0(,' E"(3&/&<"#1/ series. He said:
Our lheories leII us lhal lhe vorId is aImosl comIeleIy emly, and
lhal emly sace is fiIIed vilh chaolic radialion. And aImosl aII
Iaces vhich are nol emly are occuied eilher by chaolic dusl, or by
gases, or by very hol slars aII in condilions vhich seem lo make lhe
aIicalion of any hysicaI melhod of acquiring knovIedge
impossible. Modern cosmology teaches us that to generalize from

37
Ioer (1972, . 186). Iirsl ubIished 1968 in German, in lhe roceedings of lhe
Vienna inlernalionaI congress.
!)

observalions laken, for lhe mosl arl, in our incredibIy idiosyncralic
region of lhe universe vouId aImosl aIvays be quile invaIid.
38

Irom vhal IillIe Ioer vrole aboul cosmoIogy one gels lhe feeIing lhal he
did nol Iike lhe slandard big bang lheory. Al a conference in 1972 he
admilled lhal lhe theory was at presenl lhe mosl videIy acceled lheory of
lhe origin of the universe, but in his view it was nonetheless highly
recarious. The cosmoIogicaI origin of lhe conlenl of heIium in lhe universe
he called a speculation. Popper indicaled some symalhy for a recenl
roosaI lhal did nol resuose eilher a big bang or an exanding
universe.
39
More generaIIy he considered cosmoIogy and cosmogony
though immensely fascinating parts of physics lo be somevhal immalure
sciences, which, though they are becoming better leslabIe, are sliII aImosl
borderline cases of physical science.
40

Poppers most explicit public comments on modern cosmology
aeared in a Ieclure given in 1982 al lhe Iuroean AIbach Iorum in
AIbach, Auslria. CosmoIogy the most philosophically imorlanl of aII
lhe sciences, he reealed had undergone a revoIulionary deveIomenl
during lhe Iasl fev decades. He raised lhe by lhen defuncl ondi-GoId-
Hoyle theory as a very fine and promising theory, not because it was true
bul because il vas leslabIe and had in facl been faIsified. As a resuIl of
measurements based on methods of radio astronomy, it seems to have been
refuted in favour of the (older) big bang theory of expansion. Popper did

38
Ioer (1974b, . 1027). Since Ioer vrole lhis, lhe universe has lurned even
more exolic and lhe success of cosmoIogy even more imrobabIe. The universe is no
Ionger aImosl comIeleIy emly, bul fiIIed vilh dark energy and dark maller.
39
He referred lo Iecker, Roberls and Vigier (1972), vho roosed ineIaslic holon-
holon inleraclion as an exIanalion of gaIaclic redshifls.
40
Ioer (1974c, . 267-268), based on a Ieclure given al a conference on
reduclionism in bioIogy in eIIagio, IlaIy, 9-16 Selember 1972.
!*

nol menlion lhe cosmic microvave background radialion or olher evidence
(such as lhe measured amounl of heIium in lhe universe) lhal had Iaid lhe
sleady slale lheory in lhe grave. Wilhoul seaking oul againsl lhe big bang
theory, he remarked that we seem to be almost as helpless in the field of
cosmoIogy in lhe face of some of lhese revoIulionary resuIls as ve are in
politics when faced with the task of making peace.
41

Given that radio astronomy represented a most exciting and
revolutionary episode in the history of cosmology, Popper found it
erlinenl lo suggesl lhal lhe discovery of lhe melhod did ,&- originale from
lhe lechnicaI invenlion of an inslrumenl, lhe radio leIescoe. No, il vas lhe
(7#> of using radio vaves emilled by sleIIar bodies lhal vas imorlanl,
because I believe that the history of science is essenliaIIy a hislory of
ideas.
42
The remark underIines hov IillIe arecialion Ioer had for
discoveries based on advances in exerimenlaI and observalionaI lechniques.
He vas inleresled in lhe Iogic of scienlific discovery, bul discoveries
lhemseIves vere foreign lo his syslem of hiIosohy. This syslem orlrayed
science as a series of crilicaI discussions, nol of reaI discoveries made in lhe
Iaboralory or in lhe aslronomicaI observalory. As lhe AuslraIian hiIosoher
David Slore quied, in Popper any actual discovery would be as out of
place as a hippopotamus in a philosophy class.
43

Moreover, according lo Ioer exerimenls and observalions are
vhoIIy subordinaled ideas and lheories. Science slarls vilh robIems ralher

41
Ioer (1994, . 58-60). Iven before lhe discovery of lhe cosmic microvave
background in 1965, lhe sleady slale lheory vas conlradicled by surveys of radio
sources made by Marlin RyIe and olhers, vhich evidenlIy imressed Ioer. Ior
lhis melhod and ils roIe in lhe refulalion of lhe sleady slale lheory, see Kragh (1996,
. 305-317, 323-331).
42
Ioer (1994, . 59).
43
Slove (1982, . 13).
#+

lhan vilh observalions. Whal he caIIed his eislemoIogicaI lheory of
exerimenl vas lhis: The theoretician puts certain definite questions to the
exerimenler, and lhe Ialler, by his exerimenls, lries lo eIicil a decisive
ansver lo lhese queslions, and lo no olhers. Thus it is he [the theoretician]
who shows the experimenter the way.
44
Ierhas il vas lhis allilude lhal
caused him lo ignore lhe discovery of lhe cosmic microvave background
and ils cruciaI roIe in lhe revivaI of lhe big bang universe. The NobeI Irize-
revarded discovery of Arno Ienzias and Roberl WiIson vas serendiilous
and nol molivaled by lheory al aII.
45
Il vas nol a discovery lhal filled inlo
Poppers epistemological theory of experiment.

< =,) 0'*5$'>)$.- '>)$ 5,) .5)#+- .5#5) 1'+)%
The sleady slale lheory of lhe universe aroused a greal deaI of hiIosohicaI
interest, in part because of the theorys controversial claim of continual
crealion of maller and more generaIIy because of ils aeaI lo hiIosohy
and melhods of science. Ior examIe, in lheir aer of 1948 ondi and GoId
argued lhal lhe nev sleady slale lheory vas referabIe from a
melhodoIogicaI oinl of viev, as il vas simIer, more direcl, and more
rediclive lhan lhe cosmoIogicaI lheories based on generaI reIalivily. The
Ialler cIass of theories, they said, was utterly unsatisfactory since it covered
a vhoIe seclrum of lheories lhal couId onIy be confronled vilh lhe
observed universe if suIied vilh more or Iess arbilrary assumlions and
parameters: In general relativity a very wide range of modeIs is avaiIabIe
and lhe comarisons |belveen lheory and observalionj mereIy alleml lo
find vhich of lhese modeIs fils lhe facls besl. The number of free aramelers

44
Ioer (1959, . 107).
45
On lhis discovery and ils consequences, see Kragh (1996, . 343-355).
#!

is so much Iarger lhan lhe number of observalionaI oinls lhal a fil cerlainIy
exists and not even all the parameters can be fixed.
46
ReIalivislic cosmoIogy
soreIy Iacked lhe deduclive characler of lhe sleady slale lheory, vhich
uniqueIy Ied lo a number of rediclions, such as lhe mean densily of maller,
lhe curvalure of sace, and lhe average age of gaIaxies. According lo ondi
and GoId (bul nol HoyIe), lhe rediclions vere essenliaIIy based on lhe so-
called perfect cosmological principle, the postulate that there is neither a
riviIeged Iace nor a riviIeged lime in lhe universe.
Whelher in lhe ondi-GoId or lhe HoyIe version, lhe sleady slale
lheory vas crilicaIIy discussed by many hiIosohers and hiIosohicaIIy
minded aslronomers and hysicisls. To lhe firsl calegory beIonged AdoIf
Grnbaum, Mario unge, MiIlon Munilz, Norvood RusseII Hanson, and
Rom Harre, and lo lhe Ialler DingIe, Whilrov, WiIIiam McCrea, and WiIIiam
Davidson. Much of lhe melhodoIogicaI discussion in lhe 1950s focused on
lhe crileria on vhich lo |udge lhe scienlific nalure of lhe sleady slale lheory,
or of cosmoIogy in generaI.
47
It was in this context that Poppers demarcation
crilerion came lo Iay a roIe in lhe cosmoIogicaI debale. The Ieading
Ioerian cosmoIogisl in lhe eriod vas undoubledIy Hermann ondi, al
lhe lime rofessor of aIied malhemalics at Kings College, London. WhiIe
ondi look Ioer very seriousIy indeed, lhe more unhiIosohicaI HoyIe
did nol.
48
His osl-1960 modificalions of lhe sleady slale lheory, deveIoed

46
ondi and GoId (1948, . 269 and . 262). On lhe hiIosohicaI foundalion of lhe
sleady slale lheory, see aIso aIashov (1994).
47
Ior lhis discussion and reIevanl references lo lhe Iileralure, see Kragh (1996, .
224-256).
48
To my knovIedge, HoyIe onIy referred once lo Ioer, and lhen in conneclion
vilh lhe anlhroic rinciIe and nol lhe sleady slale cosmoIogy. Al a conference of
1989, he defended the anthropic principle if our existence leads to a potentially
falsifiable prediction in the sense of Popper. See Hoyle (1993, p. 85).
##

in coIIaboralion vilh NarIikar, vere mosl un-Ioerian, vhich vas a ma|or
reason vhy ondi vouId have nolhing lo do vilh lhem.
Although Poppers philosophical views only became generally
knovn vilh lhe ubIicalion in 1959 of %&'() &* +)(#,-(*() .(/)&0#12, among
hiIosohers and some scienlisls lhey had allracled allenlion al an earIier
dale. The surprisingly successful %&'(9 7#1 A&1/)"4,' vas favourabIy
received and revieved aIso in IngIish |ournaIs, and as a consequence Ioer
vas inviled lo IngIand in 1935 lo give Ieclures.
49
ondi may have become
acquainled vilh Ioers views lhrough lhe German originaI. HimseIf an
Auslrian immigranl he came lo IngIand in 1937 and became a rilish
sub|ecl in 1946 he may have read %&'(9 7#1 A&1/)"4,' in lhe Iale 1940s.
50
Al
any rale, he had adoled Ioerian slandards of science and used lhem in
lhe cosmoIogicaI debale even before 1959. Thus, in lhe B1(-(/" I&41,>3 *&1 -"#
E"(3&/&<"2 &* +)(#,)# of Iebruary 1954 ondi and Whilrov engaged in a
ubIic discussion aboul lhe scienlific slalus of hysicaI cosmoIogy. In lhis
memorabIe exchange of ideas, ondi defended lhe olimislic viev lhal
cosmoIogy had aIready become a science on ar vilh olher sciences, vhiIe
Whilrov, slressing lhe unique domain of cosmoIogy, argued lhal il vas nol
lruIy scienlific and robabIy never vouId be so. Il vouId remain, he lhoughl,
a borderIand sub|ecl belveen science and hiIosohy. ondi, on lhe olher
hand, suggesled lhal lhe haIImark of science vas faIsifiabiIily of lheories and

49
Ioer (1974a, . 86-88). AIlhough videIy read, lhe book onIy soId in a fev
hundred coies. On lhe imacl of lhe book and lhe reaclions il occasioned, see
Hacohen (2002).
50
According lo a Ieller from H. ondi lo H. Kragh, }anuary 1995. In his obiluary of
Popper, Bondi (1994) wrote that His seminal work in the philosophy of science
became avaiIabIe in IngIish onIy in 1959, bul ils rinciaI ideas vere aIready veII
arecialed in lhe IngIish-speaking countries by that time. At an earlier occasion,
Poppers eightieth birthday, Bondi (1982) recalled that it was thirty years ago,
vhen I vas firsl so deeply and lastingly influenced by Poppers analysis.
#$

that on this criterion cosmology was indeed a science. Every advocale of
any |cosmoIogicaIj lheory viII aIvays be found lo slress eseciaIIy lhe
suosedIy exceIIenl agreemenl belveen lhe forecasls of his lheory and lhe
sparse observational results, he admitted. And yet,
The accelance of lhe ossibiIily of exerimenlaI and observalionaI
disroof of any lheory is as universaI and undisuled in cosmoIogy as
in any olher science, and, lhough lhe ossibiIily of IogicaI disroof is
nol denied in cosmoIogy, il is nol denied in any olher science eilher. y
lhis lesl, lhe cardinaI lesl of any science, modern cosmoIogy musl be
regarded as a science. I consider universaI accelance of lhe
ossibiIily of exerimenlaI disroof of any cIaim an absoIule lesl of
vhal conslilules a science.
51

AIlhough nol menlioning Ioer by name, ondi vas cIearIy defending a
main melhodoIogicaI oinl in Ioerian hiIosohy. Whilrov, vho vas aIso
well acquainted with Poppers views, did not disagree, although he warned
lhal faIsifiabiIily shouId nol be considered a finaI and absoIule crilerion: The
imorlanl roIe of disroof in science, vhich has been so cogenlIy argued by
K. R. Ioer, is inlimaleIy reIaled lo lhe seIf-correcling lendency of science
and this, in my view, is another aspect of the pursuit of unanimity.
52
A fev
years Ialer, in an arlicIe on hiIosohicaI robIems of cosmoIogy, ondi
reealed lhal lhe scienlific slalus of a lheory vas given by ils abiIily lo
roduce rediclions lhal couId be emiricaIIy disroved. This, he argued,

51
Whilrov and ondi (1954, . 279 and . 282). Ior lhe ondi-Whilrov discussion,
see aIso Kragh (1996, . 233-237).
52
Whilrov and Ioer (1954, . 280).
#%

eslabIished lhe scienlific slalus of lhe erfecl cosmoIogicaI rinciIe and
hence of lhe sleady slale lheory buiIding on il.
53

olh in lhe debale vilh Whilrov and al some Ialer occasions, ondi
seems to have conceived Poppers methods as prescriptions to be
imIemenled in concrele scienlific vork, vhich may nol be vhal Ioer had
in mind. Whilrov ul more emhasis on lhe asymmelry belveen roof and
disroof being ureIy IogicaI, vhereas in scienlific raclice a concIusive
disroof or faIsificalion can never be roduced. Indeed, lhis vas cIearIy
recognized by Ioer.
54
In raclice il vas aIso recognized by ondi and olher
sleady slale rolagonisls, vho in concrele cases, such as vhen confronled
with Ryles data from radio astronomy, denied that falsifying evidence
amounled lo acluaI disroof of lhe lheory. They did nol aIvays behave in
accordance vilh vhal lhey erceived lo be Ioerian slandards.
55

Lalesl by 1960, ondi had become a convinced and enlhusiaslic
advocate of Poppers philosophy, such as is evident from a glowing review
essay he vrole of !"# %&'() &* +)(#,-(*() .(/)&0#12 logelher vilh his coIIeague
at Kings College, the mathemalicaI hysicisl CIive KiImisler. The lvo
physicists were full of praise of Poppers splendid book, which they found
was so well written that it was almost desirable bed-time reading. More
importantly, its main appeal to scientists was that it rings true. According
to the two reviewers, Popper speaks as a working scientist to the working
scientist in a language that time and again comes straight out of ones

53
ondi (1957). In a oslscril of 1963 lo lhe second edilion ubIished in 1966,
ondi referred lo lhe cosmoIogicaI significance of some recenl deveIomenls in
nuclear astrophysics. Expressing himself in Popperian language, he said that the
sleady-slale lheory has effectively passed a severe test.
54
Ioer (1959, . 50). See aIso Seclion 5.
55
Kragh (1996, . 327-328).
#&

heart.
56
ondi and KiImisler did nol miss lhe oorlunily lo oinl oul lhe
reIevance of lhe Ioerian faIsifiabiIily crilerion lo lhe conlemorary
silualion in cosmoIogy. According to Popper, Once a hypothesis has been
roosed and lesled, and has roved ils mellIe, il may nol be aIIoved lo
drop out without good reason, which quotation Bondi and KiImisler used
as an argumenl for lhe sleady slale lheory.
57
A good reason might be the
reIacemenl of lhe hyolhesis by anolher, beller leslabIe one, and no such
aIlernalive exisled. Seaking of lhe sleady slale lheory: For here the correct
argumenl has aIvays been lhal lhe sleady slale modeI vas lhe one lhal couId
be disroved mosl easiIy by observalion. Therefore, il shouId lake
recedence over olher Iess disrovabIe ones unliI il has been disroved.
Tvenly years before lhe inlroduclion of lhis hyolhesis, Ioer |in 1934j
formulated the conditions with the utmost accuracy.
Al aboul lhe same lime as ondi sludied !"# %&'() &* +)(#,-(*()
.(/)&0#12, he arlicialed in lvo C broadcasls on modern cosmoIogy
logelher vilh WiIIiam onnor, Raymond LyllIelon, and Whilrov. In lhe
ubIished version, he once again aid lribule lo Ioer as lhe reeminenl
hiIosoher of science, summarizing his viev as foIIovs:
y far lhe mosl successfuI anaIysis of scienlific melhod is due lo
Professor Karl Popper. The urose of a lheory is lo make forecasls
lhal can be checked againsl observalion and exerimenl. A scienlific
lheory is one lhal il is in rinciIe ossibIe lo disrove by emiricaI
means. Il is lhis suremacy of emiricaI disroof lhal dislinguishes
science from olher human aclivilies. We can never regard a lheory as

56
ondi and KiImisler (1959-1960, . 55).
57
Ibid. . 56. Ioer (1959, . 53). Norlh (1965, . 294-295), discusses lhe meaning
of Poppers quote and the use made of it by Bondi and Kilmister.
#'

roved, because aII ve can say is lhal, so far, lhere have been no
exerimenls conlradicling il. A scienlific lheory, lo be usefuI, musl be
leslabIe and vuInerabIe.
58

Irom lhe mid-1960s, afler lhe sleady slale lheory had faIIen in disfavour,
ondi increasingIy lurned lo sludies of gravilalionaI radialion and olher
robIems of generaI reIalivily. ul he vas nol yel quile ready lo abandon his
referred cosmoIogicaI modeI, vhich he conlinued lo defend for some lime
from a melhodoIogicaI oinl of viev ralher lhan as a viabIe scienlific modeI
of lhe universe. Wilhoul embracing lhe viclorious big bang lheory, he
admilled lhal lhe ballIe vas Iosl. The sleady slale modeI mighl be vrong, bul
il vas neverlheIess more scienlific lhan lhe reIalivislic big bang lheories:
The steady-slale lheory is far more leslabIe lhan any olher. According lo
Ioer and olher hiIosohers of science, lhis makes il cIearIy referabIe lo
aIlernalive lheories. The onIy oen queslion is vhelher il has indeed aIready
been disproved.
59
Bondis high appreciation of Poppers philosophy was
slrenglhened by his vork on sleady slale cosmoIogy, bul il did nol deend
on il. Iven afler having abandoned cosmoIogy as a research fieId, he
continued to praise Poppers system in the strongest possible words, such as:
There is no more to science than its method, and there is no more to its
method than Popper has said.
60
On the occasion of Poppers 90lh birlhday,

58
ondi (1960, . 12). WhiIe onnor, a lheorelicaI hysicisl al Queen IIizabelh
CoIIege, London, oosed lhe sleady slale modeI, lhe Cambridge aslronomer
LyllIelon vas in favour of il. Al lhe lime he coIIaboraled vilh ondi and GoId on a
new electrical version of the steady state theory. On this cosmological hypothesis,
vuInerabIe lo exerimenlaI disroof and indeed quickIy disroved, see Kragh
(1997).
59
ondi (1966, . 32).
60
As reorled by lhe hiIosoher ryan Magee in his biograhy of Ioer. See
Magee (1973, . 2), vhich gives no source for lhe quolalion.
#(

ondi spoke in a similar laudatory language, stating that Poppers view of
empirical disproof has profoundly influenced me and many others.
61

IinaIIy, in an obiluary in H>-41#: Although many scientists have little
interest in the philosophy of science, to me his lhoughls came as a fIash of
brilliant light.
62

AIlhough Ioerian crileria of science Iayed a considerabIe roIe
during lhe cosmoIogicaI conlroversy, and vere highIighled by lhe sleady
slale roonenls in arlicuIar, lhey vere rareIy an issue of disule. y and
Iarge, crileria of a Ioerian kind vere acceled aIso by many cosmoIogisls
favouring an evoIving universe governed by lhe Iavs of generaI reIalivily.
One of lhem vas George McVillie, a rilish-American aslronomer slrongIy
oosed lo lhe sleady slale lheory and olher lheories he susecled vere
based on a riori rinciIes. He described lhe hiIosohicaI foundalion of
lhe ondi-Gold theory as Karl Poppers dictum that a scientific theory can
never be roved lo be lrue bul, inslead, lhal cerlain lheories can be roved lo
be false by an appeal to observation. While he considered the dictum to be a
probably unimpeachable doctrine, he parodied Bondis use of it. If one
followed Bondis vulgar version of Poppers philosophy, we should be
|uslified in invenling a lheory of gravilalion vhich vouId rove lhal lhe
orbil of every Ianel vas necessariIy a circIe. The lheory vouId be mosl
vulnerable to observation and could, indeed, be immediately shot down.
63

AIso onnor referred lo Ioerian slandards in his defense of
reIalivislic cosmoIogy, mainlaining lhal il vas in facl emiricaIIy refulabIe, if
admilledIy nol as easiIy or cruciaIIy as the steady state alternative. Indeed,

61
ondi (1992).
62
ondi (1994).
63
McVillie (1961, . 1231). On McVillie and his emiricisl allilude lo cosmoIogy, see
Sanchez-Ron (2005).
#)

if lhe sleady-slale lheory is righl, lhen reIalivislic cosmoIogy is vrong, and,
in my opinion, we have to scrap general relativity as well, he wrote. But he
did nol see il as a queslion bearing on Poppers criterion: I make this point
because every scienlific lheory musl be caabIe of 7(/<1&&*, olhervise il says
nothing.
64


5 Poppers late views of cosmological models
As menlioned in Seclion 3, Ioer vas lo some exlenl symalhelic lo lhe
sleady slale lheory, of vhich he vas veII informed. Concerning his allilude
lo lhis lheory, in his Ieller lo me of 1994 he vrole as foIIovs:
65

I discussed lhe conlroversiaI osilion vilh ondi in considerabIe delaiI,
and so I vas quile veII informed, bul I did nol read lhose aers of his.
IncidenlaIIy, I Iiked his lheory, bul nol lhe so-called cosmological
principle and even less its (temporal) extension. (Because I dislike
making of our 3>)9 of knovIedge a rinciIe of 9,&K(,' /&:#-"(,'.) I
lherefore vas ,&- shocked when Ryle killed the theory, for he only
killed this temporal extension of a principle that had nothing to do
vilh lhe 6>/() (7#> of lhe ondi-GoId lheory: lhal maller mighl be
crealed oul of some fieId and so kee lhe universe going more or Iess
steady. For since mass-crealion vas a random rocess anyvay, lhere
was no need to assume that it was steady &0#1 comparatively small
parts of the universe (even if we assume that it was steady over
large parts vhalever large may mean): the propensity of creating

64
onnor (1964, . 158).
65
See nole 23. Ioer divided his Ieller in five arls, denoled (1) lo (5). I have nol
incIuded lhis numbering in lhe lranscrilion. The Ieller is nov deosiled al lhe KarI
Ioer CoIIeclion al lhe Universily Library in KIagenfurl, Auslria.
#*

a rolon (or eIeclron, or some olher arlicIe) may deend on (lhe
gravilalionaI` Or lhe eIeclromagnelic`) fieIds Iinked vilh lhe nearesl
gaIaxies exIaining gaIaclic (`) evoIulion and offering an ansver lo
RyIes arguments.
WhiIe Ioer lhus vas symalhelic lo a version of sleady slale lheory, if nol
lo lhe originaI ondi-GoId lheory based on lhe erfecl cosmoIogicaI
rinciIe, he came lo disIike and re|ecl lhe viclorious big bang lheory. Aboul
lhis he vrole:
I nol onIy Iiked lhe idea of a :&1# &1 3#// steady mass creation :4)"
6#--#1 lhan lhe big bang for obvious reasons: I mean lhe inexIicabiIily
of a beginning of lime. ul aIso because big bang lheory became raidIy
more and more comIicaled. And my resenl viev is lhal lhe number
of auxiIiary hyolheses is simIy inloIerabIe: according lo my lheory of
science, -"(/ (/ ,&- /)(#,)#. Il is (1) inlroducing a nev auxiIiary hyolhesis
every lime lhe lheory is refuled, and (2), il is :4-4>3 /4<<&1- of
cosmoIogicaI -"#&12 and arlicIe -"#&12 bul crilicism, and crilicaI
exerimenls ( allemled refulalions) are ignored oul of hand. And nol
onIy is il ,&- slressed by lhe uhoIders of lhe lheory lhal il is aII
secuIalion vilhoul lesls, bul il is resenled as if lhe lheory vere a
roven *>)-. This is horrid, imermissibIe, againsl scienlific elhics.
By contrast, Einsteins General Relativity is a marvelous theory.
ul il may be suerseded, as aII lheories may, and I acluaIIy susecl
lhal il may be suerseded aIready. ul lhese are big robIems, and I
cannol nov go i|njlo discussing lhem.
Ioer summed u his allilude lo cosmoIogy as foIIovs:
$+

I once K>/ an enthusiastic admirer of (Friedmanns) Big Bang. I am ,&K
a disgusted opponent. As to the steady state theory, it is insufficiently
developed, and Ryles criticism insufficiently discussed. And the
cosmological principles were, I fear, dogmas that should not have
been roosed.
It is noteworthy that Poppers main ob|eclion lo lhe cIassicaI sleady slale
lheory vas lhe erfecl cosmoIogicaI rinciIe, vhich he considered
unscienlific and dogmalic. He vanled lo reIace il vilh maller crealion.
ondi and GoId, on lhe olher hand, slressed lhe scienlific nalure of lhe
faIsifiabIe erfecl cosmoIogicaI rinciIe and regarded il as rimary reIalive
lo lhe conlinuaI crealion of maller. According lo lhem, maller crealion vas a
direcl consequence of lhe erfecl cosmoIogicaI rinciIe vhen combined
vilh lhe observalionaIIy eslabIished exansion of lhe universe. Wilhoul lhis
rinciIe, cosmoIogy couId nol be lruIy scienlific.
66
Il is aIso vorlh lo observe
that much of Poppers criticism of the big bang theory agreed with the
ob|eclions raised by HoyIe and olhers al aboul lhe same lime. This crilicism
Ied lo lhe so-caIIed quasi sleady slale cosmoIogy (QSSC), vhich Ioer, had
he knovn aboul il, mighl have found a more salisfaclory lheory of lhe
universe.
67


? =,) /*@%8)*0) '@ &'(()$/#* (,/%'.'(,-
Poppers falsificationist philosohy has been infIuenliaI in a broad range of
sciences, from bolany lo lheorelicaI cosmoIogy. According lo David Slove, a

66
If it does not hold, cosmology is no longer a science. Bondi and Gold 1948, p.
255.
67
A fuII descrilion of lhe QSSC aIlernalive is given in HoyIe, urbidge, and
NarIikar 2000. See aIso Kragh (2011, . 133-137).
$!

crilic of Ioer, if you scratch a scientist of middle age or older, you are
aImosl cerlain lo meel vilh a hiIosohy of science vhich consisls of haIf-
remembered scraps of Popperism.
68
The infIuence of faIsificalionisl
hiIosohy L 3> Ioer varies of course from one science lo anolher, and il
may be arlicuIarIy slrong in lhe aslronomicaI sciences. According lo a sludy
by en|amin Sovacool, astronomers and cosmologists often invoke Poppers
ideas as a guide for conslrucling and evaIualing lheories, aIlhough lhey
rareIy reveaI a deeer famiIiarily vilh lhese ideas.
69
In addilion lo vhal has
aIready been menlioned, a fev more examIes viII indicale lhe enduring
infIuence of Ioerian melhodoIogy al Ieasl on lhe rheloricaI IeveI. In a
reviev arlicIe on lhe comaralive merils of slandard big bang cosmoIogy
and QSSC lheory, NarIikar and his coaulhor T. Iadmanabhan emhasize lhe
cruciaI roIe of lesls in dislinguishing belveen cosmoIogicaI modeIs:
Whal lesl can be erformed lhal couId in rinciIe disrove lhis
|slandard or QSSCj cosmoIogy` This queslion is in lhe siril of KarI
Poppers view of a scientific theory, that it should be disprovabIe.
Thus if such a lesl is erformed and ils resuIls disagree vilh lhe
rediclion of lhe lheory, lhe lheory is considered disroved. If lhe
lheory seeks survivaI by adding an exlra aramelric dimension, lhal
is againsl lhe siril of lhis queslion. On lhe olher hand, if lhe
rediclion is borne oul, our confidence in lhe lheory may be
enhanced, bul lhe lheory sliII cannol be considered roven.
70

y lhese slandards NarIikar and Iadmanabhan consider slandard reIalivislic
cosmoIogy lo be a scienlific lheory because there are tests that are decisive

68
Slove (1999, . 8).
69
SovacooI (2005).
70
NarIikar and Iadmanabhan (2001, . 241).
$#

in disproving or strongly discrediting the theory. As anolher examIe, in a
reviev of lhe slale of cosmoIogy al lhe miIIennium, lhe American
aslrohysicisl MichaeI Turner advocaled lhe nev slandard modeI of lhe
universe, incIuding lhe cosmoIogicaI conslanl and dark maller, by aeaIing
lo ils melhodoIogicaI virlues: However, with its unidentified dark matter
and myslerious dark energy, il is currenlIy very much oul on a Iimb.
According to Karl Popper thats what strong theories do! Inflation + cold
dark matter is bold and testable.
71

Slehen Havking, lhe ceIebraled cosmoIogisl and lheorelicaI
hysicisl, rales highIy lhe over of malhemalicaI eIegance and IogicaI
consislency in lheory conslruclion. WhiIe lhese are lhe cruciaI sources in lhe
conlexl of discovery, in lhe conlexl of |uslificalion lhe lheory needs lo be
confronled vilh nalure. In his besl-seIIing F B1(#* C(/-&12 &* !(:#, Havking
vrole:
Any hysicaI lheory is aIvays rovisionaI, in lhe sense lhal il is onIy
a hyolhesis: you can never rove il. ... On lhe olher hand, you can
disrove a lheory by finding even a singIe observalion lhal disagrees
vilh lhe rediclions of lhe lheory. As hiIosoher of science KarI
Ioer has emhasized, a good lheory is characlerized by lhe facl
lhal il makes a number of rediclions lhal couId in rinciIe be
disroved or faIsified by observalion. Iach lime nev exerimenls are
observed lo agree vilh lhe rediclions lhe lheory survives, and our
confidence in il is increased, bul if ever a nev observalion is found lo
disagree, ve have lo abandon or modify lhe lheory.
72


71
Turner (2001, . 656).
72
Havking (1989, . 11).
$$

Havking reaIized lhal lhis is a normalive rescrilion ralher lhan a
descrilion of reaI research raclices, and lhal consequenlIy il has lo be
quaIified: Or ralher, lhal is vhal is suosed lo haen. In raclice, eoIe
are very reIuclanl lo give u a lheory in vhich lhey have invesled a Iol of
lime and efforl. They usuaIIy slarl by queslioning lhe accuracy of lhe
observalions. If lhal faiIs, lhey lry lo modify lhe lheory in an ad hoc
manner.
73
WhiIe lhe firsl of Hawkings quotations is a arahrase of
Ioers philosophy, the second one may aear lo deviale from Ioerian
slandards.
ul il does nol, for Ioer did nol cIaim lhal lhese slandards
refIecled lhe acluaI raclices of scienlisls. Iirsl of aII, he never heId lhal
faIsifiabiIily is a sufficienl condilion for a lheory being scienlific, bul onIy
lhal il is a necessary condilion. He knev veII lhal lhe demarcalion crilerion
cannol be very shar, bul musl be assigned degrees, a loic lo vhich he
devoled a vhoIe chaler in !"# %&'() &* +)(#,-(*() .(/)&0#12.
74
AIlhough
somevhal ambiguous vilh regard lo lhe reIalionshi belveen his
melhodoIogicaI ruIes and scienlific raclice, he admilled lhal slricl
faIsifiabiIily does nol beIong lo lhe reaI vorId of science:
In oinl of facl, no concIusive disroof of a lheory can ever be
roduced, for il is aIvays ossibIe lo say lhal lhe exerimenlaI
resuIls are nol reIiabIe, or lhal lhe discreancies vhich are asserled lo
exisl belveen lhe exerimenlaI resuIls and lhe lheory are onIy
aarenl and lhal lhey viII disaear vilh lhe advance of our
understanding. If you insist on strict proof (or strict disproof) in

73
Havking (1994, . 36). A senlence lo lhe same effecl foIIoved lhe revious
quotation: At least, that is what is supposed to happen, but you can always
question the competence of the person who carried out the observation.
74
Ioer (1959, . 112-145).
$%

lhe emiricaI sciences, you viII never benefil from exerience, and
never Iearn from il hov vrong you are.
75

Again, in a section in the 1974 festschrift on Difficulties of the Demarcation
Proposal Popper made it clear that he did not assign any absolute value to
lhe crilerion of faIsifiabiIily and did nol consider il a 7#*(,(-(&, of science. He
recognized lhal lhe dislinclion belveen melahysics and science is oflen
blurred. What was a metaphysical idea yesterday can become a testable
theory tomorrow, such as happened with speculative atomism at the time of
DaIlon.
76
Iar from eIevaling faIsificalionism lo an invioIabIe rinciIe, he
suggesled lhal il is ilseIf faIIibIe and lhal il may be ralionaI lo kee even an
admilledIy vrong lheory aIive for some lime:
There is a Iegilimale Iace for dogmalism, lhough a very Iimiled
Iace. He vho gives u his lheory loo easiIy in lhe face of aarenl
refulalions viII never discover lhe ossibiIilies inherenl in his lheory.
!"#1# (/ 1&&: (, /)(#,)# *&1 7#6>-#: for allack and lherefore aIso for
defence. OnIy if ve lry lo defend lhem can ve Iearn aII lhe differenl
ossibiIilies inherenl in our lheories. As aIvays, science is con|eclure.
You have lo con|eclure vhen lo slo defending a favourile lheory,
and vhen lo lry a nev one.
77


75
Ibid., p. 50. In a note appended to the English edition, Popper remarked that I
have been conslanlIy misinlerreled as uhoIding a crilerion (and moreover one of
:#>,(,' ralher lhan of 7#:>1)>-(&,) based upon a doctrine of complete or
conclusive falsifiability. On the ambiguity between the prescriptive and
descriptive in Poppers writings, see Mulkay and Gilbert (1981).
76
Ioer (1974b, . 981). In Ioer (1959, . 38), he referred lo alomism as an
examIe iIIuslraling that scientific discovery is impossible without faith in ideas
which are of a purely speculative kind.
77
Ioer (1974b, . 984).
$&

The same theme appeared in Poppers autobiography, where he, referring to
his early studies, said: I also reaIized lhal ve musl nol excIude aII
immunizalions, nol even aII vhich inlroduce >7 "&) auxiliary hypotheses.
AII lhis shovs nol onIy lhal some degree of dogmalism is fruilfuI, even in
science, bul aIso lhal IogicaIIy seaking faIsifiabiIily, or leslabiIily, cannol be
regarded as a very sharp criterion.
78
The viev indicaled by lhese quoles is
indeed far from lhe slricl or nave faIsificalionism oflen discussed by
scienlisls eilher for or againsl Ioer.
The aIIegiance lo Ioerian slandards among cosmoIogisls may be
furlher iIIuslraled by a nev cycIic modeI of lhe infinile universe lhal vas
roosed in 2002 by IauI Sleinhardl and NeiI Turok as an aIlernalive lo lhe
ouIar infIalionary lheory. In lheir efforls lo romole lhe cycIic cosmoIogy
Sleinhardl and Turok emhasized ils faIsifiabiIily in lhe form of rediclions
of delaiIs in lhe cosmic microvave background. These delaiIs differ from
lhose redicled by lhe rivaI infIalionary lheory and are of such a kind lhal, if
they are not found, then this would support the inflationary picture and
definiliveIy rule out the cyclic model.
79
An examIe of a differenl kind is lhe
IsraeIi hysicisl en|amin GaI-Ors G&/:&3&'2D E"2/()/D >,7 E"(3&/&<"2, a
lexlbook lhal ambiliousIy aims al inlegraling lhe lhree sub|ecls. AIlhough
heavily influenced by Poppers philosophy, which it quotes extensively, in

78
Popper (1974a, p. 32). The reference to immunizations is to attempts to protect a
lheory againsl refulalion.
79
Sleinhardl (2004, . 469). WhiIe infIalion redicls a seclrum of rimordiaI
gravilalionaI vaves, according lo lhe cycIic modeI lhere shouId be no gravilalionaI
signalure. Ior a survey of lhe nev cycIic modeI, see Kragh (2011, . 202-208).
$'

lhis case il is nol lhe demarcalion crilerion vhich is in focus bul ralher
Poppers critical rationalism and general attitude to science and philosophy.
80

Il is nol difficuIl lo find more examIes of lhe infIuence of
Ioerianism on lhe vievs of modern cosmoIogisls lhan lhose aIready
menlioned, and a fev more viII be ciled in lhe foIIoving seclion. Yel il is
cIear from lhe Iileralure, bolh in cosmoIogy and in olher sciences, lhal
Ioer and lhe melhodoIogicaI ruIes associaled vilh his name moslIy Iay a
roIe in generaI discussions and onIy very rareIy in research aers. This vas
lhe case during lhe earIier cosmoIogicaI conlroversy, and il is sliII lhe case.
When cosmoIogisls refer lo Ioerian crileria of science, il is lyicaIIy in
ouIar books, in reviev arlicIes of a broad scoe, in ubIic Ieclures, or
somelimes in conference roceedings. A series of inlervievs vilh
biochemisls conducled by MichaeI MuIkay and NigeI GiIberl indicaled lhal
although Poppers philosophical message was well knovn among lhem, and
many of lhe scienlisls subscribed lo il, his hiIosohy of science had onIy
very Iimiled infIuence on lheir acluaI scienlific raclice.
81
The cosmoIogicaI
research Iileralure rovides evidence lhal lhe silualion in cosmoIogy is aboul
lhe same.
Few of the scientists commenting on the merits or faults of Poppers
hiIosohy of science have acluaIIy read him, bul reIy on vhal lhey have
been loId or haen lo knov. This resuIls in discussions lhal are oflen
simIislic and somelimes based on misunderslandings. Whal cosmoIogisls
(and olher scienlisls) discuss is mosl oflen nave faIsificalionism ralher lhan

80
GaI-Or (1981), with a foreword by Popper. Using the criterion of scientific
testability, refutability and falsifiability, Gal-Or concIudes lhal lhe IIanck Ienglh, as
given by /

10
-33
cm, is a myth (p. 242).
81
MuIkay and GiIberl (1981).
$(

lhe sohislicaled versions of aulhenlic Ioerianism.
82
In a Ieclure in Cracov
in 1973, lhe anlhroic cosmoIogicaI rinciIe vas inlroduced by 31-year oId
aslrohysicisl randon Carler. He soon reaIized lhal lhe rinciIe disagreed
with generally accepted standards of physics, such as Poppers emphasis on
rediclivily and faIsifiabiIily. Carler vas al lhe lime aware of Poppers ideas,
bul vilhoul having sludied lhem.
83
In a paper of 1983 he criticized the
doctrine that scientific theories are never verifiable but only falsifiable,
arguing that the Popperian doctrine implied that all existing theories are not
onIy faIsifiabIe, bul may safely be assumed in advance to be false.
84

Six years Ialer Carler amplified his attack on lhe videsread
misunderslanding lhal has Ied lo undiscriminaling insislence on lhe
requirement that a theory should satisfy the requirement of refutability. If
a consequence of a lheory is confirmed, he argued, il is lurned from a
hyolhesis inlo a facl, and lhen il ceases lo be refulabIe, and yel lhe slrenglh
of lhe lheory increases. Hovever, lhis misreresenls lhe Ioerian viev, for
according lo Ioer confirmalion mereIy corroborales a lheory, il does nol
change ils degree of faIsifiabiIily. In any case, Carler beIieved lhal, far from
being indisensabIe, such refutability is definitely less satisfactory
(scientifically) than an equal amount of (irrefutable) verified osldiclive
output provided the latter is deduced logically from the same amount of
independently hypothesised input information.
85
WhiIe Carler and a fev

82
HeIIer (2009, . 88-89).
83
In an e-mail to the author of 6 February 2010, Carter wrote: Simplified folklore
versions of Poppers falsifiability criterion (and Ockhams razor criterion) were
aIready famiIiar lo me from coffee labIe discussions vhen I vas a sludenl, bul I
did not get to know about Kuhns very apposite ideas until much later.
Reroduced vilh lhe ermission of . Carler.
84
Carler (1983, . 352).
85
Carter (1989, p. 194). For Carters ideas and the historical context of the
conlroversiaI anlhroic rinciIe, see Kragh (2011, . 217-254).
$)

olher roonenls of lhe anlhroic rinciIe ob|ecled lo lhe norms of
faIsificalionism (or vhal lhey look lo be lhe norms), anlagonisls considered
lhe confIicl belveen lhe norms and anlhroic reasoning as an indicalion lhal
lhe Ialler did nol beIong lo science. One of lhe anlagonisls, lhe American
hysicisl Heinz IageIs, cIaimed in 1985 lhal lhe anlhroic rinciIe vas
unleslabIe and immune to experimental falsification a sure sign lhal il is
not a scientific principle.
86


A B8%5/>)$.) #*+ &'(()$/#*/.1
Togelher vilh slring lheory and lhe infIalionary scenario, lhe anlhroic
hyolhesis is a key element in the modern hypothesis of the multiverse
lhal for a decade or so has been holIy debaled in lhe cosmoIogy communily.
According lo lhis hyolhesis or roosaI, our universe is onIy one inslance
among a huge ensembIe of universes and erhas lhe onIy one riviIeged by
lhe exislence of inleIIigenl and cognizabIe beings. The numerous olher
universes are causaIIy disconnecled from ours, hence unobservabIe in
rinciIe.
87
They may have lheir ovn Iavs of hysics, ovn conlenl of
eIemenlary arlicIes, and erhas even lheir ovn number of sace-lime
dimensions. WhiIe some hysicisls and cosmoIogisls are enlhusiaslic
advocales of lhe muIliverse, olhers lend lo dismiss il as an inleresling bul
unscienlific secuIalion.

86
IageIs (1990, . 177). Neilher IageIs in 1985 nor Carler in his ubIicalions of 1983
and 1989 menlioned Ioer by name.
87
There are differenl kinds of muIliverse modeIs, and in some of lhem lhe conslanls
of nalure and basic Iavs of hysics are lhe same in aII universes. On lhe muIliverse
idea and lhe conlroversy il has aroused, see Carr (2007) and Kragh (2011, . 255-
290).
$*

One robIem vilh lhe muIliverse hyolhesis is lhal lhe excessive
amounl of universes seems lo aIIov aImosl any hysicaI slale of affairs if
nol in our universe, lhen in some olher. This, logelher vilh lhe
unobservabiIily of lhe olher universes, makes lhe muIliverse robIemalic,
nol lo say unaccelabIe, from a Ioerian oinl of viev. According lo
Poppers philosophy, a scientific theory must be falsifiable and therefore set
constraints to the results of possible observations: Every good scientific
lheory is a rohibilion: il forbids cerlain things to happen, as he said in a
lecture of 1953. The more a theory forbids, the better it is.
88
Al Ieasl in some
versions, muIliverse cosmoIogy suffers from an exlreme Iack of
rohibiliveness.
In the ongoing controversy over the multiverse Poppers views play
a significanl roIe, somelimes exIicilIy and al olhers limes imIicilIy. Do
muIliverse lheories quaIify as scienlific according lo Ioerian slandards`
Consider Lee SmoIin, a Ieading lheorelicaI cosmoIogisl and crilic of lhe
muIliverse and anlhroic reasoning in hysics:
According lo Ioer, a lheory is faIsifiabIe if one can derive from il
unambiguous rediclions for raclicaI exerimenls, such lhal vere
conlrary resuIls seen al Ieasl one remise of lhe lheory vouId have
been proven not true. Confirmation of a prediction of a theory
does nol shov lhal lhe lheory is lrue, bul faIsificalion of a rediclion
can shov il is faIse.
89


88
Ioer (1963, . 36).
89
SmoIin (2007, . 323-324). Emphasis added. Smolins Popperian view of science
is nol unIike lhe one earIier esoused by ondi. See aIso SmoIin (2008, . 369), vhere
he proudly declares himself a Popperazo.
%+

Smolin even goes as far as arguing that scientists have an #-"()>3 (:<#1>-(0#
lo consider onIy faIsifiabIe lheories as ossibIe exIanalions of naluraI
phenomena. AIso urlon Richler, a NobeI Iaureale in hysics, aeaIs lo
Karl Poppers definition of science in his attack on the anthropic
muIliverse. This suosed definilion that science is limited to models with
testable and falsifiable consequences ruIes oul muIliverse cosmoIogy and
reIegales il lo lhe domain of melahysics.
90

WhiIe SmoIin and some olher crilics dismiss lhe muIliverse on lhe
ground that it violates Poppers demarcation criterion for science, il seaks lo
lhe aulhorily of lhis crilerion lhal il is invoked aIso by some cosmoIogisls
favourabIy incIined lo lhe muIliverse hyolhesis. Ior examIe, Mario Livio
(vho is one such cosmoIogisl) emhasizes lhal a lheory lhal cannol be lesled
even in principle can hardly be counted as scientific. It goes against the
rinciIes of lhe scienlific melhod, and in arlicuIar il vioIales lhe basic
concel lhal every scienlific lheory shouId be faIsifiable, he says.
91
Anolher
sympathizer, the French cosmologist Aurlien Barrau, maintains that the
multiverse remains within the realm of Popperian science, although he adds
that falsifiability is just one criterion among many possible ones.
92
Again,
Max Tegmark, roosing a muIliverse lheory of lhe TOI (lheory of
everything) kind, discusses its relation to Poppers hardly controversial
faIsifiabiIily crilerion. Ralher lhan queslioning lhe crilerion, he argues lhal

90
Richler (2006). The misunderstanding that Poppers demarcation criterion is a
7#*(,(-(&, of science is videsread among scienlisls and science vrilers. Thus, an
article on multiverse cosmology carrying the subtitle Do we need to change the
definition of science? speaks of falsifiability as the defining characteristic of real
science, whereas Popper considered it a proposal scientists can decide to follow
(Mallhevs 2008).
91
Livio (2000, . 187).
92
arrau (2007).
%!

the TOE we have proposed makes a large number of slalislicaI rediclions,
and lherefore )>, eventually be ruled out.
93

Olher muIliverse hysicisls admil lhal lhe muIliverse is hardIy
reconciIabIe vilh Ioerian slandards of science, bul lhey lend lo consider
lhe lension a virlue ralher lhan a robIem. Leonard Susskind, one of lhe
falhers of lhe modern muIliverse hyolhesis and an ardenl advocale of il, has
no respect at all for Popper and the Popperazi who follow him. Here is
Susskinds opinion of the falsifiability criterion and what he calls
overzealous Popperism:
Throughoul my Iong exerience as a scienlisl I have heard
unfaIsifiabiIily hurIed al so many imorlanl ideas lhal I am incIined
lo lhink lhal no idea can have greal meril unIess il has dravn lhis
criticism. Good scientific methodology is not an abstract set of
ruIes diclaled by hiIosohers. ... IaIsificalion, in my oinion, is a
red herring, bul confirmalion is anolher slory.
94

Whal Susskind refers lo as confirmalion incIudes malhemalicaI consislency,
vhich he and olher slring hysicisls rale highIy as a melhod of lesling. In
lhis conlexl il shouId be oinled oul lhal Ioer (robabIy unknovn lo
Susskind) aIso incIuded inlernaI consislency as a vay of lesling a lheory. In
%&'() &* +)(#,-(*() .(/)&0#12 he dislinguished belveen four differenl Iines
aIong vhich a lheory can be lesled, and onIy one of lhem vas emiricaI

93
Tegmark (1998, . 1 and . 42). This is one of lhe fev research aers vilh secific
references to Poppers demarcation criterion.
94
Susskind (2006, . 193-195).
%#

lesling or faIsificalion. He considered seIf-consislency lo be a crilerion
formaIIy anaIogous lo faIsificalion.
95

More couId be said aboul lhe roIe of Ioerian slandards in lhe
modern debale of vhelher lhe universe is unique or a muIliverse consisling
of differenl universes. I have mereIy inlended lo iIIuslrale lhal lhe Iegacy of
Ioer is sliII very much aIive in lvenly-firsl cenlury cosmoIogy.

C D'*0%8./'*
Whereas Ioer deaIl exlensiveIy and crilicaIIy vilh robIems in slalislicaI
hysics and quanlum mechanics, he had disaoinlingIy IillIe lo say aboul
lhe nev kind of hysicaI cosmoIogy lhal emerged and malured during his
ovn career as a hiIosoher.
96
He foIIoved lhe deveIomenl from lhe
sideIine, bul vilhoul examining any of lhe rivaI cosmoIogies from lhe
erseclive of his hiIosohy of science. In regard of lhe roIe lhal his vievs
Iayed in lhe sleady slale conlroversy, and lhe aeaI lo Ioerian
slandards lhal ondi highIighled in arlicuIar, lhis is somevhal surrising.
From Poppers scattered remarks, supplemented with his testimony of 1994,
il seems safe lo concIude lhal he vas scelicaI vilh regard lo lhe scienlific
slalus of cosmoIogy. He had a cerlain reference for lhe faIsifiabIe sleady
slale lheory, vhereas he came lo see lhe successfuI big bang lheory as
essenliaIIy ad hoc and unscienlific. Al lhe end of his Iife he exressed in

95
Ioer (1959, . 32, 92, and 314). AIlhough dislinguishing sharIy belveen
faIsificalion and confirmalion, Ioer did nol necessariIy give riorily lo emiricaI
faIsificalion.
96
The absence in Poppers books and articles of discussions of modern cosmology is
onIy surrising in lhe Iighl of lhe roIe lhal his ideas Iayed among cosmoIogisls
during his ovn Iifelime. Mosl olher rominenl hiIosohers of science incIuding
Thomas Kuhn, Imre Lakalos, IauI Ieyerabend, and Larry Laudan aIso ignored lhe
nev deveIomenls in hysicaI cosmoIogy.
%$

slrong vords his disIike of lhe big bang lheory, eseciaIIy for melhodoIogicaI
reasons (Seclion 5). He characlerized lhe melhods on vhich lhe lheory buiIl
as intolerable, impermissible, and against scientific ethics.
From about 1955 and until the present Poppers ideas have been
highIy visibIe in lhe deveIomenl of cosmoIogy much more so lhan lhe
ideas of olher hiIosohers, incIuding Thomas Kuhn. Il aears lhal lhe
significance of his vievs has been IargeIy Iimiled lo eriods of conceluaI
uncerlainly or rivaIry belveen comeling modeIs. References lo Ioer,
somelimes exIicil and al olher limes imIicil, aear frequenlIy in lhe
ouIar Iileralure and generaI discussions, bul onIy very rareIy in lhe
research aers. Anolher fealure vorlh nolicing, and one vhich is nol
ecuIiar lo cosmoIogy, is lhal lhe version of Ioerianism lhal is discussed
in scienlific and semi-scienlific conlexls is aImosl aIvays a simIified foIkIore
version. This vas lhe case in lhe 1950s, and it is still the case. Poppers views
of vhal characlerizes science and hov il rogresses are oflen dislorled and
frequenlIy boiIed dovn lo lhe seducliveIy simIe formuIa lhal a non-
faIsifiabIe lheory is unscienlific er definilion. This is a formuIa lhal has IillIe
lo do vilh aulhenlic Ioerian hiIosohy.
Hov unique is lhe case here described` The vievs of hiIosohers
have Iayed a roIe in many olher sciences in lhe modern eriod, bul lhey
have usuaIIy been delached from lhe acluaI research raclices. Mosl oflen
lhey have been inlerrelalions of vhal has haened, relroseclive
evaIualions of recenl deveIomenls. In some fieIds of lhe bioIogicaI sciences
Ioerian slandards have been infIuenliaI, and il has been discussed vhal
lhese slandards reaIIy are and if bioIogicaI lheories shouId Iive u lo lhem.
97

Anolher case of ossibIe reIevance lo lhe one here discussed is lhe Iale

97
HeIfenbein and DeSaIIe (2005). RieeI (2008).
%%

tectonics revolution of the 1960s, where it was Kuhns philosophy rather
than Poppers that entered the discussion.
98
Hovever, lhis moslIy occurred
<&/- *>)-4:, in allemls lo eslabIish vhelher or nol lhe deveIomenl lhal Ied
lo lhe nev iclure of lhe mobiIe Iarlh foIIoved a allern of lhe kind argued
by Kuhn. The case is neverlheIess reIevanl because il shares vilh lhe one of
cosmoIogy lhe fealure lhal hiIosohies of science vere discussed by lhe
scienlisls lhemseIves and, lo some exlenl, became arl of lhe scienlific
discourse.
99


F)9,&K3#7':#,-/: I vouId Iike lo lhank lhe KarI Ioer Library al KIagenfurl
Universily for permission to quote from Poppers letter.

E)@)$)*0).
Adam, AvshaIom M. 2000. Farewell to certitude: Einsteins novelty on induction
and deduclion, faIIibiIism. I&41,>3 *&1 ;#,#1>3 E"(3&/&<"2 &* +)(#,)# 31: 19-37.
aIashov, Yuri. 1994. Uniformilarianism in cosmoIogy: ackground and
hiIosohicaI imIicalions of lhe sleady-slale lheory. +-47(#/ (, C(/-&12 >,7
E"(3&/&<"2 &* +)(#,)# 25: 933-958.
arrau, AureIien. 2007. Ihysics in lhe muIliverse. G#1, G&41(#1, 20 November.
ondi, Hermann. 1960. The sleady-slale lheory of lhe universe. In ?(0>3 !"#&1(#/ &*
G&/:&3&'2, eds. H. ondi el aI., 12-21. London: Oxford Universily Iress.
ondi, Hermann. 1966. Sleady-slale cosmoIogy. In F--( 7#3 G&,0#',& /433> G&/:&3&'(>,
eds. L. Rosino el aI., 31-36. IIorence: G. arbera.

98
HaIIam (1973). Laudan (1980).
99
The Canadian geohysicisl }ohn Tuzo WiIson, one of lhe ioneers of Iale
leclonics, vas infIuenced by Kuhns +-14)-41# &* +)(#,-(*() ?#0&34-(&,. As earIy as 1968
he referred secificaIIy to Kuhns brilliant analysis of scientific methods, arguing
that what had recently taken place in the earth sciences constituted a Wegenerian
revolution in Kuhns sense. See WiIson (1968, . 317) and aIso Cohen (1985, .
564-565). Il is nol hard lo see an anaIogy belveen WiIson and ondi, lhe firsl
advocaling Kuhn in lhe case of lhe earlh sciences and lhe Ialler advocaling Ioer
in lhe case of cosmoIogy and lhe hysicaI sciences generaIIy.
%&

ondi, Hermann (1982). Sir KarI Ioer on his eighlielh birlhday. A&4,7>-(&,/ &*
E"2/()/ 12: 821-823.
ondi, Hermann. 1992. The hiIosoher for science. H>-41# 358: 363.
ondi, Hermann. 1994. KarI Ioer (1902-1994). H>-41# 371: 478.
ondi, Hermann, and Thomas GoId. 1948. The sleady-slale lheory of lhe exanding
universe. J&,-"32 H&-()#/ &* -"# ?&2>3 F/-1&,&:()>3 +&)(#-2 108: 252-270.
ondi, Hermann, and CIive W. KiImisler. 1959-1960. The imacl of Logik der
Iorschung. B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# 10: 55-57.
onnor, WiIIiam. 1964. !"# J2/-#12 &* -"# 8M<>,7(,' N,(0#1/#. Nev York: MacmiIIan.
rush, Slehen G. 2001. Is lhe earlh loo oId` The imacl of geochronoIogy on
cosmoIogy, 1929-1952. In !"# F'# &* -"# 8>1-"O A1&: PQQP BG -& F. RQQR, eds. C.
L. I. Levis and S. }. KneII, 157-175. London: GeoIogicaI Sociely.
cheI, W. 1967. Inlroy and informalion in lhe universe. H>-41# 213: 319-320.
Carr, ernard, ed. 2007. N,(0#1/# &1 J43-(0#1/#S Cambridge: Cambridge Universily
Iress.
Carler, randon. 1983. The anlhroic rinciIe and ils imIicalions for bioIogicaI
evoIulion. E"(3&/&<"()>3 !1>,/>)-(&,/ &* -"# ?&2>3 +&)(#-2 F 310: 347-363.
Carler, randon. 1989. The anlhroic rinciIe: SeIf-seIeclion as an ad|uncl lo
naluraI seIeclion. In G&/:() E#1/<#)-(0#/, eds. S. K. isvas, D. MaIIik, and C. V.
Vishveshvara, 185-206. Cambridge: Cambridge Universily Iress.
Cohen, I. ernard. 1985. ?#0&34-(&, (, +)(#,)#. Cambridge, Mass.: Harvard Universily
Iress.
Cohen, Roberl S. 1950. IislemoIogy and cosmoIogy: I. A. Milnes theory of
reIalivily. ?#0(#K &* J#-><"2/()/ 3: 385-405.
Craig, WiIIiam L. 1979. Whilrov and Ioer on lhe imossibiIily of an infinile asl.
B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# 30: 165-170.
Davies, }. T. 1954. The age of lhe universe. B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)#
5: 191-202.
DingIe, Herberl. 1940. Red-shifls in nebuIar seclra and scienlific raclice. H>-41#
145: 224-225.
Iggers Hansen, TroeIs. 2006. Which came firsl, lhe robIem of induclion or lhe
robIem of demarcalion` In T>13 E&<<#1O F G#,-#,>12 F//#//:#,-U V&3U WO +)(#,)#,
eds. Ian }arvie, KarI MiIford, and David MiIIer, 67-82. AIdershol: Ashgale.
Iinslein, AIberl. 1920. ?#3>-(0(-2O !"# +<#)(>3 >,7 -"# ;#,#1>3 !"#&12. London: Melhuen
& Co.
Iinslein, AIberl. 1956. !"# J#>,(,' &* ?#3>-(0(-2. Irincelon: Irincelon Universily
Iress.
%'

Iinslein, AIberl. 1962. J#(, X#3-6(37, ed. CarI SeeIig. Irankfurl am Main: UIIslein
VerIag.
Iinslein, AIberl. 2002. G&33#)-#7 E><#1/ &* F36#1- 8(,/-#(,, voI. 7, eds. MicheI }anssen el
aI. Irincelon: Irincelon Universily Iress.
Iinslein, AIberl. 2004. G&33#)-#7 E><#1/ &* F36#1- 8(,/-#(,, voI. 9, eds. Diana K.
uchvaId el aI. Irincelon: Irincelon Universily Iress.
IsfeId, MichaeI. 2006. Ioer on irreversibiIily and lhe arrov of lime. In T>13 E&<<#1O
F G#,-#,>12 F//#//:#,-U V&3U WWWO +)(#,)#, eds. Ian }arvie, KarI MiIford, and
David MiIIer, 57-70. AIdershol: Ashgale.
GaI-Or, en|amin. 1981U G&/:&3&'2D E"2/()/D >,7 E"(3&/&<"2. Nev York: Sringer-
VerIag.
Grnbaum, AdoIf. 1952. Some highIighls of modern cosmoIogy and cosmogony.
?#0(#K &* J#-><"2/()/ 5: 481-498.
Hacohen, MaIachi H. 2002. T>13 E&<<#1O !"# A&1:>-(0# Y#>1/ Z[QR\Z[P]. Cambridge:
Cambridge Universily Iress.
HaIIam, Anlhony. 1973. F ?#0&34-(&, (, -"# 8>1-" +)(#,)#/O A1&: G&,-(,#,->3 .1(*- -&
E3>-# !#)-&,()/. Oxford: CIarendon Iress.
Havking, Slehen. 1989. F B1(#* C(/-&12 &* !(:#O A1&: !"# B(' B>,' -& B3>)9 C&3#/.
Nev York: anlam ooks.
Havking, Slehen. 1994. B3>)9 C&3#/ >,7 B>62 N,(0#1/#/D >,7 ^-"#1 8//>2/. London:
anlam ooks.
Hayes, Ieler. 2010. Poppers response to Dingle on special relativity and lhe
robIem of lhe observer. +-47(#/ (, C(/-&12 >,7 E"(3&/&<"2 &* J&7#1, E"2/()/ 41:
354-361.
HeIfenbein, Kevin G., and Rob DeSaIIe. 2005. IaIsificalion and corroboralion: KarI
Ioppers influence on systematic. J&3#)43>1 E"23&'#,#-()/ >,7 80&34-(&, 35: 271-
280.
HeIIer, MichaeI. 2009. N3-(:>-# 8M<3>,>-(&,/ &* -"# N,(0#1/#. erIin: Sringer-VerIag.
HenlscheI, KIaus. 1992. Einsteins allilude lovards exerimenls: Tesling reIalivily
lheory 1907-1927. +-47(#/ (, C(/-&12 >,7 E"(3&/&<"2 &* +)(#,)# 23: 593-624.
HoIlon, GeraId. 1998. !"# F70>,)#:#,- &* +)(#,)#D >,7 (-/ B417#,/. Cambridge, Mass.:
Harvard Universily Iress.
HoyIe, Ired. 1993. The anlhroic and erfecl cosmoIogicaI rinciIes: SimiIarilies
and differences. In !"# F,-"1&<() E1(,)(<3#, eds. Irancesco erloIa and Umberlo
Curi, 85-89. Cambridge: Cambridge Universily Iress.
HoyIe, Ired, Geoffrey urbidge, and }ayanl NarIikar. 2000. F .(**#1#,- F<<1&>)" -&
G&/:&3&'2. Cambridge: Cambridge Universily Iress.
}ammer, Max. 1991. Sir KarI Ioer and his hiIosohy of hysics. A&4,7>-(&,/ &*
E"2/()/ 21: 1357-1368.
%(

Kragh, HeIge. 1996. G&/:&3&'2 >,7 G&,-1&0#1/2O !"# C(/-&1()>3 .#0#3&<:#,- &* !K&
!"#&1(#/ &* -"# N,(0#1/#. Irincelon: Irincelon Universily Iress.
Kragh, HeIge. 1997. The eIeclricaI universe: Grand cosmoIogicaI lheory versus
mundane exerimenls. E#1/<#)-(0#/ &, +)(#,)# 5: 199-231.
Kragh, HeIge. 2011. C('"#1 +<#)43>-(&,/O ;1>,7 !"#&1(#/ >,7 A>(3#7 ?#0&34-(&,/ (,
E"2/()/ >,7 G&/:&3&'2. Oxford: Oxford Universily Iress.
Laudan, RacheI. 1980. The recent revolution in geology and Kuhns lheory of
scienlific change. In E>1>7(':/ >,7 ?#0&34-(&,/O F<<1>(/>3/ >,7 F<<3()>-(&,/ &*
Thomas Kuhns E"(3&/&<"2 &* +)(#,)#, ed. Gary Gulling. Nolre Dame: Universily
of Nolre Dame Iress.
LeeIlier, Thomas. 2006. Edward Milnes infIuence on modern cosmoIogy. F,,>3/ &*
+)(#,)# 63: 471-481.
Livio, Mario. 2000. !"# F))#3#1>-(,' N,(0#1/#O W,*(,(-# 8M<>,/(&,D -"# G&/:&3&'()>3
G&,/->,-D >,7 -"# B#>4-2 &* -"# G&/:&/. Nev York: }ohn WiIey & Sons.
Magee, ryan. 1973. E&<<#1. Nev York: Viking Iress.
Mallhevs, Roberl. 2008. Some svans are grey. H#K +)(#,-(/- 198 (10 May): 44-47.
McVillie, George C. 1961. RalionaIism versus emiricism in cosmoIogy. +)(#,)# 133:
1231-1236.
Medavar, Ieler . 1990. !"# !"1#>- >,7 -"# ;3&12O ?#*3#)-(&,/ &, +)(#,)# >,7 +)(#,-(/-/.
Nev York: HarerCoIIins.
MuIkay, MichaeI, and G. NigeI GiIberl. 1981. Iulling hiIosohy lo vork: KarI
Poppers infIuence on scienlific raclice. E"(3&/&<"2 &* -"# +&)(>3 +)(#,)#/ 11: 389-
407.
NarIikar, }ayanl, and T. Iadmanabhan. 2001. Slandard cosmoIogy and aIlernalives:
A crilicaI araisaI. F,,4>3 ?#0(#K &* F/-1&,&:2 >,7 F/-1&<"2/()/ 39: 211-248.
Norlh, }ohn. 1965. !"# J#>/41# &* -"# N,(0#1/#O F C(/-&12 &* J&7#1, G&/:&3&'2.
London: Oxford Universily Iress.
IageIs, Heinz. 1990. A cozy cosmoIogy. In E"2/()>3 G&/:&3&'2 >,7 E"(3&/&<"2, ed.
}ohn LesIie, 174-180. Nev York: MacmiIIan.
Iecker, }. C., A. I. Roberls, and }. I. Vigier. 1972. Non-veIocily redshifls and holon-
holon inleraclions. H>-41# 237: 227-229.
Ioer, KarI R. 1932-1933. Iin Krilerium des emirischen Characlers lheorelischer
Sysleme. 819#,,-,(/ 3: 426-427.
Ioer, KarI R. 1940. Inlerrelalions of nebuIar red-shifls. H>-41# 145: 69-70, 701.
Ioer, KarI R. 1959. !"# %&'() &* +)(#,-(*() .(/)&0#12. Nev York: asic ooks.
Ioer, KarI R. 1963. G&,_#)-41#/ >,7 ?#*4->-(&,/. Nev York: RoulIedge.
Popper, Karl R. 1965. Times arrow and entropy. H>-41# 207: 233-234.
Ioer, KarI R. 1967. Times arrow and feeding on negenlroy. H>-41# 213: 320.
%)

Ioer, KarI R. 1972. ^6_#)-(0# T,&K3#7'#O F, 80&34-(&,>12 F<<1&>)". Oxford:
CIarendon Iress.
Ioer, KarI R. 1974a. Aulobiograhy. In !"# E"(3&/&<"2 &* T>13 E&<<#1, ed. IauI A.
SchiI, 3-181. La SaIIe, IL: Oen Courl IubIishing House.
Ioer, KarI R. 1974b. ReIies lo my crilics. In !"# E"(3&/&<"2 &* T>13 E&<<#1, ed. IauI
A. SchiI, 961-1200. La SaIIe, IL: Oen Courl IubIishing House.
Ioer, KarI R. 1974c. Scienlific reduclion and lhe essenliaI incomIeleness of aII
science. In +-47(#/ (, -"# E"(3&/&<"2 &* B(&3&'2, eds. Iransisco }. AyaIa and
Theodosius Dobzhansky, 259-284. erkeIey: Universily of CaIifornia Iress.
Ioer, KarI R. 1978. On lhe ossibiIily of an infinile asl: A reIy lo Whilrov.
B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# 29: 47-48.
Ioer, KarI R. 1982. `4>,-4: !"#&12 >,7 -"# +)"(/: (, E"2/()/, ed. W. W. arlIey.
London: Unvin Hyman.
Ioer, KarI R. 1994. W, +#>1)" &* > B#--#1 X&137O %#)-41#/ >,7 8//>2/ *1&: !"(1-2 Y#>1/.
London: RoulIedge.
Ioer, KarI R. 2008. !"# !K& A4,7>:#,->3 E1&63#:/ &* -"# !"#&12 &* T,&K3#7'#, ed.
TroeIs Iggers Hansen. London: RoulIedge.
Richler, urlon. 2006. Theory in arlicIe hysics: TheoIogicaI secuIalion versus
raclicaI knovIedge. E"2/()/ !&7>2 59 (Oclober): 8-9.
RieeI, OIivier. 2008. Ioer and syslemalics. +2/-#:>-() B(&3&'2 52: 259-271.
RindIer, WoIfgang. 2009. Gdel, Einstein, Mach, Gamow, and Lanczos: Gdels
remarkabIe excursion inlo cosmoIogy. F:#1()>, I&41,>3 &* E"2/()/ 77: 498-510.
Sanchez-Ron, }ose-ManueI. 1990. Sleady-slale cosmoIogy, lhe arrov of lime, and
Hoyle and Narlikars theories. In J&7#1, G&/:&3&'2 (, ?#-1&/<#)-, eds. runo
erlolli el aI., 233-246. Cambridge: Cambridge Universily Iress.
Sanchez-Ron, }ose-ManueI. 2005. George McVillie, lhe uncomromising emiricisl.
In !"# N,(0#1/# &* ;#,#1>3 ?#3>-(0(-2, eds. A. }. Kox and }ean Iisenslaedl, 189-
222. oslon: irkhuser.
Scriven, MichaeI. 1954. The age of lhe universe. B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &*
+)(#,)# 5: 181-190.
SmoIin, Lee. 2007. Scienlific aIlernalives lo lhe anlhroic rinciIe. In N,(0#1/# &1
J43-(0#1/#S, ed. ernard Carr, 323-366. Cambridge: Cambridge Universily
Iress.
SmoIin, Lee. 2008. !"# !1&463# K(-" E"2/()/. London: Ienguin ooks.
SovacooI, en|amin. 2005. IaIsificalion and demarcalion in aslronomy and
cosmoIogy. B433#-(, &* +)(#,)#D !#)",&3&'2 a +&)(#-2 25: 53-62.
Sleinhardl, IauI. 2004. The endIess universe: A brief inlroduclion. E1&)##7(,'/ &* -"#
F:#1()>, E"(3&/&<"()>3 +&)(#-2 148: 464-470.
%*

Slove, David. 1982. E&<<#1 >,7 F*-#1O A&41 J&7#1, W11>-(&,>3(/-/. Oxford: Iergamon
Iress.
Slove, David. 1999. F'>(,/- -"# W7&3/ &* -"# F'#. Nev runsvick, N}: Transaclion
IubIishers.
Susskind, Leonard. 2006. !"# G&/:() %>,7/)><#O +-1(,' -"#&12 >,7 -"# W334/(&, &*
W,-#33('#,- .#/(',. Nev York: LillIe, rovn and Comany.
Tegmark, Max. 1998. Is the theory of everything merely the ultimate ensemble
lheory` F,,>3/ &* E"2/()/ 270: 1-51.
Ter Hark, MicheI. 2002. elveen aulobiograhy and reality: Poppers inductive
years. +-47(#/ (, C(/-&12 >,7 E"(3&/&<"2 &* +)(#,)# 33 A: 79-104.
Turner, MichaeI S. 2001. A sober assessmenl of cosmoIogy al lhe nev miIIennium.
E463()>-(&,/ &* -"# F/-1&,&:()>3 +&)(#-2 &* -"# E>)(*() 113: 653-657.
Van Dongen, }eroen. 2010. Einsteins U,(*()>-(&,. Cambridge: Cambridge Universily
Iress.
Whilrov, GeraId }., ed., 1967. 8(,/-#(,O !"# J>, >,7 C(/ F)"(#0#:#,-. London: rilish
roadcasling Cororalion.
Whilrov, GeraId }. 1978. On lhe imossibiIily of an infinile asl. B1(-(/" I&41,>3 *&1
-"# E"(3&/&<"2 &* +)(#,)# 29: 39-45.
Whilrov, GeraId }., and Hermann ondi. 1954. Is hysicaI cosmoIogy a science`
B1(-(/" I&41,>3 *&1 -"# E"(3&/&<"2 &* +)(#,)# 4: 271-283.
WiIson, }. Tuzo. 1968. Slalic or mobiIe earlh: The currenl scienlific revoIulion.
E1&)##7(,'/ &* -"# F:#1()>, E"(3&/&<"()>3 +&)(#-2 112: 309-320.

Potrebbero piacerti anche