Professional Documents
Culture Documents
God, Time and Knowledge - William Hasker
God, Time and Knowledge - William Hasker
UE Pinon py af oihoming anon
‘si
“The Historical Matix la
forbidden to call the Dominicans Calvinists, while che Dominicans
vere told that they must not call the Jensits Pelagians." Tes at
least ironic, and perhaps significant, chat the theory of middle
Knowledge, which was then under suspicion because it conceded
too much to fre will, should now be viewed with concern by 2
number of philosophers because t comes too close to detemin-
copa, Onan Sues 34 am ola in
‘iam ly Sex om ts corr sl pong
“foram exten evesew fat comoeny ee Froow' odacion 2
atin, Oy De Few[2]
Middle Knowledge
nning cis st of problems with the issue of middle knowh.
elge may seem an eccentric choice. As we have seen, the theory of
‘middle knowledge wa the last of the msjor postions on these
iaaves to make its appearance. Ie has never been lowe to a majority
‘pinion, and even today it commands only a relatively small group
of adherens. Furthermore, 2 dzece stack on middle knowledge
may be redundant. The theory of middle knowledge, in all its
historical forms, presuppores the compatiblity of divine fore-
knowledge and laman Freedom, 20 suecessfil argument for i=
compaiblism, iFone ean be mounted, would ender superfluous &
separate refutation of middle knowledge
All this etre, yet middle knowledge as much to recommend
asa stating point For one ching, che complexities ofthis dheory
lend i sn almost iresistibe philosophical fascination, With regard
to edundancy, it is cleae that in controversy as heated as this ne
it is a mort and not a defect to assemble multiple independent
arguments forthe same conchision; a single argument, even if
apparently unimpeachable, i too likely to be treated with suspic
‘ion when it contravenes strong inclinations. Furthermore, the
theory of middle knowledge, though dependent on the eth of
compauilism, also serves to strengthen compatitlsm in a lest
to ways. For one thing, i offers an account of how God is able vo
Middle Knowldge le
know future free actions, and the account of thi given by middle
knowledges fee of some ofthe difficulties for instance, etoae-
tive causation—that plague other accounts of the matter. Mote
important, middle knowledge provides the key toa unigusly pow=
‘erful conception of the operation of divine providence, almost
certainly the strongsst view of providence thats possible short of
‘complete theological determinism. In contrast with this, ic wil be
argued in the next chapter that foreknowledge without middle
knowledge—simple foreknowledge—docs not offer the beneis
for the doctrine of providence that ite adherents have sought £9
derive fom it In view ofthis iteould be argued that a good many
theists who are not explicit adherens of middle knowledge neve
cetheless hold to 2 conception of divine providence that implicitly
‘commits them to this theory.) And if this isso, refutation of
middle knowledge substantially weakens the doctrine of fore
Knowledge by removing one oft principal motivations. Finally,
it should be noted that a structural possibilty exists (ot to my
knowledge exploited by anyone as. yet) of combining middle
Iknoviledge with the theory of divin imeessness, hus gaining the
benefits of mide knowledge forthe doctrine of providence with-
fut incurring the difficulties of foreknowledge. In view of these
considerations, 2 refutation of middie knowledge which is inde~
pendent of the foreknowledge ise can hardy be considered s-
perfuoas.
(Our procedure inthis chapeer willbe a follows: Fist shall
‘consider the doctrine as it emerged inthe sixteothecentiry con=
‘woversy and mention briefly some ofthe arguments pro and con
that were put forward at that time. The, we shall turn to the
modern form of the doctrine as it has been revived—or rather,
reinvented*—by Alvin Plantings, Finally, a setes of objections to
this theory will be considered, leading wp to one that I consider
decisive
"nig point so asd Ringer, "Middle Rood and Csi Ctin|
‘Though Relig Ster 2 (os trea.
“Wlaninga developed hs ew independ wn Amory Kenny who
mel ou 9 bi the snny hewn en nd te cea they
‘le Knowledge Se Plana “Soe Pro”InjnsE omt si et
‘i Inwage, ede, Abe Pag, rfl vals (Dore Rl, 98)1 God, Tine, and Knowledge
“The Classical Theory
“The theory of middle knowledge bold that, foreach possible
fice ecature that might exis, and for exch possible situation in
cehuch coca exeate might make a fee choice here i aera,
howe to God prior to and independent of any decision on God's
fre concerning what definite choice that creature would freely
rirte if placed in that situation. In effect, middle knowledge ex-
Tends the doctrine of divine foreknowledge to include knowledge
‘he outcome of choices that might ave een made bur in Fac
“Gn casual consideration, midale knowledge may appear 10 be
simply an obvious implication of divine ompiscince: Wf God
noes everything, how could he fil to know shi? And by the
wae token, it may seem relatively innocuous. Both impressions
however a mistaken. Middle knowledge isnot 2 staightor~
svar implication of omniscence, because it snot evident that he
Tras postulated by this theory exist to be known. In oxdinary
foreknowledge i may be argued, what God knows isthe agents
rena devs todo one thing or anodher. But with repaed t0 2
Situation thse never in ft arises, no decision is ever made, and.
hone exists for God go know. And if the decision in question
apposed to be a fie decison, then al of the circumstances of the
mBPGinclaing the agents character and prior inclinations) are
SShsstent with any of dhe possible choices that might be made
Locking the agents aca! making of the choice, then, there is
rothing that disambiguate the situation and makes i¢ true that
Tome one ofthe options ithe one that wuld b selected. This line
‘tasgumentindieates the single most important objection thatthe
proponent of middle knowledge must seck to answer.
Thc he very sume feature that makes mide knowledge prob=
lematie (wz, that God can know the outcome of choices that are
hover actualy made) also makes it extraordinarily use for theo
Toeleal purposes. Consider the fllowing counteract: “It A
eure in circumstances, she would do X." According 9 middle
nowledge, God knows the truth of this whether or not A ever
ually & placed in circumstances C; indeed, God knows this
“Mhucher or not A even exists 30 that bis knowledge about this
Mid Knowledge ts
ily need of ny f Ca dion soa eon
fd providence. But this, of eure, make sch knowledge de
foe God 10 38 in ding whether orn to cate Aya Be
docs eet er, whether not plac ein Greumtances C. As
Moline sy
Godin ice Bw bys knee shige
Could: rhe ol ce hwo sd itl may oer
srr fl gen compe compen od poe
ie igh ocr thins dy wat Cel
IN ce swe proton lc onder Sow
‘Stnpenon kttoiscwn ie wit wool make et ey
Sth Saesdsuch an smoun of vie ence gen such and
{ek cpp, pts ot center nd ht
SSVI Seay desing alte tne he ay 1 do te
Sept he hn pyri spn nd i
‘phen (er: Ovnd Unvay Pp Pears2] God, Time, and Knowledge
exactly those outcomes that in fact oecur—though, ¢0 be sure,
Some of them may not be the outcomes he would most prefer. The
clement of ik is ently eliminated
"Ar ove have already seen, the chief ificulty thatthe proponent
‘of middle knovledge must confront isthe contention that the
truths God is alleged to know, commonty called “counterictals
Of fredom,” do not exist to be known. Most ofthe arguments for
Countefactals of freedom seem to depend on general consiera+
tions of philosophical plausibility, but in the medieval controversy
there were also arguments based on Seriprute. A favorit text for
this purposes found in [Samuel 23, which recouns an incident n
the troubled relationship of David with King Saul Davi, eut~
rently in occupation ofthe city of Keil, consults Yahweh by
fncars of the ephod about the rumors that Saul intends to attack
the cy:
wi Saul come down, a thy srvan as heal? © Lown the God
offal {beseech the, ll hy servants” And the LD si, "He
‘Silcome down." Then sid David, "Will he men of Kelas
ender me and my en i he Kan of Sol?" And the Low si
"they wl sarender you.” (Samuel 311-12, RSV)
‘The advocates of middle knowledge took this passage 25 evi
dence that God knew the following two propositions to be eu
(0). 1eDavidstayod in Keilah, Saul would besiege the sty
{G)_ te David stayed in Keiah and Sal Besiged the city, the
tmen of Keilh would sureender David to Sa
‘But (given the assumption tha Saul and the men of Keil would
fact ely in performing the speed accion), these two propos
‘ons ate counterfatoals of eedom, and the incident aa whole is
{ deamatie demonstation ofthe existence and practical efficacy of
‘middle knowledge,
‘Buc this argument is hardly compelling. As Anthony Kenny
or my dscns page ly ely on RL ML Ads, “Mile
Kee ae BeRe Prete of Eek" Aon Paonia Quy (97
Tope Sesto Ray The Gn he Pps, fp. 0-83
Midate Knowledge i
points out, the ephod seems to have been a yee-no device hardly
Possessing the subdety required to distinguish between various
possible condtionals that might have been asserced in answer to
David's questions. Kenny. indeed, suggests that we may under
stand material conditions here,* but that scems hardly likey,
since on that constrial both conditonals would be erue simpy in
virtue ofthe fact chat eeir antecedents ae false, Mach more pais
ible candidates are given by Robert Adams
(@) David stayed in Keilah, Saul woul potty besiege the
iy
(#) IF Dav saved in Keilah and Sal bsiged the city, the
men of Keil would probly surrender Davi to Saul
‘As Adams poins out, “(3 and () are enough for David vo at on,
ithe i prudent, but they will not satisfy the partisans of middle
knowledge."© The prospects for a scriptral proof of middle
knowledge, cherefore, da not scem promising
‘But of course, the argument just given shows only that the
responses to David’s questions need not be taken 3 sserting cone
terfcwals of fredom, not that they cannot be so understood. And
there ate not lacking situations in everyday fe in which t sens
plausible that we ae taking counterfactual offeeedom to be toe.
Plantinga, for example, says he believes that “If Bob Adanie were
to offer to take me climbing a Tahquite Rock the next ime come
to California, I would gladly (and freely) accept"? And Adams
notes hat “there doesnot normaly seem wo be any uncertainty at
alabou wha butcher for example, woul fae done had
asked hi to sell me a pound of ground beet, akhough we suppose
te would have ad et wil m the mates"
So the discussion of examples sems to end ina sand-of. iil,
the proponent of middle knowledge needs to addres the question
mentioned eatie: How is it possible for counterfacuals of fee=
ean, The Gl of he Pooper.
‘Ada. ie Rute pe
re aber Ann Tomb a ne, Aea God, Tim, and Knowledge
dom to he me? What isthe truth maker for dese propositions? At
{his point the advocate of midale knowledge is presented with an
Sfersaive opportunity, but one that iis imperative for her £0
fesit The opportunity is simply to claim that counteriatuas of|
freedom are tru in vine of the character and psychological tendencies
ifthe agents named in them, The attractiveness of this x evident
that in neatly all of the eases where we are dspoved to acepe such
ounterfactoals a tue, the epistemic grounds for ovr acceptance
‘Would be found precisely in our knowledge of such psychological
faew—Saul bescging Keil, Adam's compliant butcher, and
Plantinga climbing Tsbuite Rock ae all ces in point. But the
sweaknes ofthe suggestion becomes apparent when the following
(question 1 asked: Are the psychological facts aboot the agent,
gether with a description ofthe situation, plus relevant psycho=
Joga! Isws, supposed to ental that she agent would respond 35
indieated? I the answer is ye, then the counterfactual may be me
Dat ie isnot + counterfactual of fedom: the agent snot then free in
the relevant (libertarian) sense.” In the other and the answer is
fo, then how can those psychological fits provide good grounds
for the assertion that che agent definitely weald (as opposed, say, £0
very probably would) respond in that way?
‘rota the bes line for the proponent of mide knowledge 0
take heres the one suggested by Suarez: When a counterfactual of
Feeedom is true, itis simply animate fice about the fe agent in
(question that, i placed inthe indicated circumstances, she would
ice as the counterfactual states; this fict requires no analysis oF
metaphysical grounding in terms of farther, noncowmtefactal
Miter of afairs (Or, i€ the agent in question does not actully
‘hist eas face about a particular een that, fe were instantiated
icant fn et ne
sentence esses an a ur
Gein ces vf ee
ee eat ttn pa
Teoma i Etcetera
‘ctype an dan of fede. Bu 2c of hh so a ra ie
ERESCEE sw ence soa
Middle Knowledge Las
and its instantiation were placed in such circumstances, the in-
Stantston would acts stated.) Adams, commenting on this, ys,
‘do not think I have any conception -. of the sort of
property that Suiver aserbes to possible agen with respect 0
thet act under posible conditions. Nor do I think that Ihave ny
‘other primitive understanding of wha ie would be forthe relevant
subjunctive conditional to be rae." Nevertheless, he admits that
‘Sudter's view on this sof the “least clearly unsatisfactory type”
because “Its very dificult refute someone who claims to have a
primitive understanding which I seem noe to have."
‘The Modern Theory
‘The modern theory of middle knowledge" differs from he
classical version in virte ofthe application to dhe counterfatuals|
‘of freedom ofthe powerful possible-worlds semantics or couter
factual developed by Robert Stalnaker, David Lewis, and John L
Pollock. '# The ental des ofthis semantic i that a counterfactual
is tue if some possible world in which the antecedent and the
consequent are Both true is more similar co the actual world than
Sny in which the antecedent fs true and the consequent fle.
“Thos (a) above is correctly analyzed 38
(9) The actual world fe more similar to some possible world
in which David stays in Kea and Sal besicges the cry
"Adams, “Mile Kae” pe
"ie bic rare forthe cr theory of mie knoe Alin Pe
sing Te Nee of ety (Ono Onrd avery Pr 14), 9.
TRS fC Roa Sa
a Lapa Phy (Oxford: Bnd, tf Dad Less, Col (C-
ge Harerd Unverty Pre srs Jb ack, Sine Rai
If Dordc DR, 190 snl b moe, howeter that some cote
‘ary sens of mie owtege have tacos abt ta Smt,
el i camp rt ine enya
‘Smuts app teatro ecm) oreo nyse mpl cond
‘Sa tne cna ndererminc pena commanto)
"Stok sacs se the ent nor a a compare sty
bors hats posible world only change fo he nt wai 0
[Mismatch oil ue fee Sse Revoingp.
‘eag) alld’ ajumenrsee be sree, bo tbe ferene betwee
‘So mane iron sien for poppe eal ona ©
‘Slay the more aa ering1 God, Time, and Knowledge
{han to any pore world in which David stays in Kh
tn Saul does noe besiege the city!*
[Ac this point ie will be well to gota bit clearer about the exact
positions both of the advacste and of the opponent of middle
Knowledge, Fitst ofall it may be noted thatthe term “eounterfac
tual,” though customary and convenient, i not strictly accurate as
| designation ofthe propositions in question. tn some cases (name
Ty, those whore antecedents God decides to actualize) both the
antecedent and the consequent ofthe conditions will bene, and
fo not couverfactoal at all A better term, therefore, would bey 38
‘Adams suggests, “deliberative conditonals.” Having said that
however, we shall continue to refer to them as “counterfactuals of
freedom.”
‘But just what kind of conditionals ate these? Both Lewis and
Polloce distinguish “would” condtionals from “might” condi-
tional; the “might” conditional coresponding to (1) would be
(am) IF David stayed in Ki, Saul might bessge the cry.
‘But Pollock goes further and distinguishes thee diferent kinds of
“woul” conditions these distinctions ae no explicily made by
Lewis. There ate “simple subjuncives™ these are the eonlition-
sik most frequently, and most naturally, expressed by English sen-
fences of the form “If were the ease that , it would be the case
thar Q” Second, these are "even if” conditinals, of the form
“Even iit were che cate chat P, ie woud stil) be the case that Q.
"These are the condtionals Nelson Goodman calls “semi-ictuals”
they are asterted when their consequents are believed o be tue,
whereas their antecedents may or may aot be tue, and thee force
fs to deny thatthe truth ofthe antecedent would bring about the
falsity of the consequent, Finally, there are “necessiation cond
tions"; according to Pollack, “the notion of accssttion that |
fam trying to analyze here is that of the truth of one statement
“binging it about” that another statement i tue," 50 an appro-
aa "Mile Kae" p12
‘ste Pony Sui Rola bp, 2, Four Kind of Condon,
TE pp a-ae Pollock sys Peps th tery econ ppp
Midale Knowledge ia
priate formula might be “Rs being the case tht P would bring it
about that 0."
"Now, into which of these categories do we place the counterie-
twals of freedom? Evidently they cannot be "might" conditionals
“Even if” conditional ae tre only in (some of) those possible
worlds in which thei consequents are tue, but the truth of the
counterfactual of freedom mst be knovn to God quite indepen-
‘ently of whether or not their consequents are trae in the actal
‘world, Pollock shows chats simple subjunctive is equivalent tothe
Alisjunction of a necersitation conditional and an “even i condi
tional. If chen, we were to equate counteefaculs of feeedom with
simple subjunctive, it would follow that im those eases whete the
necessitation conditional is false the counterfactual of freedom
‘would be equivalent to an “even i conditional, which we have
sen tobe impossible. Soithe counterfactual of feedom ate to be
found among the varieties discussed by Pollock, they must be
necesitation condtionals. As be says, “All counterfactual cond
tonals express necesitation.""7
"Now that we have clarified the nature ofthe counterfsctals of
freedom, how exactly shall we charactrie the view taken of such
counteractals by the opponents of middle knowledge? There
seem to be three alternatives: One may deny that such propositions
txst at all; one may concede thet existence but deny tht they
possess trath-vahues or one may hold thatallsuch propositions are
false. The denial that there are svch propositions as counterfactual
of freedom does not seem to have much to recommend i, 38
Plankings s8ys, he may conceivably be wrong in believing that s€
‘Adams were to invite him eo dim Tahguite Rock ke would a
cept, but it would be passing strange to deny that thee is such 3
proposition a the one he claims to believe, think, infact, tht thie
view may best be understood as arising from an exigency: if one
thinks (asthe second view holds) chat there is no way to assign
“efor on Fhe in mind re, But hat ee unable nd be ee
ig all does vr deny ay ng Enea ocaton that comary
15.32. neem ony nap he ne res coh
‘Sealy coped by fom fertile te
‘Gin anne ta 2)
i aea) God, Time, and Knowledge
rth-values to counterfactuals of feedom, and ifame is also con=
‘Vince that every proposition must be either true or false, then one
js ily forced to deny that there are sch proposiions—eht is,
fone is forced to deny that the relevant Sentences express any propo~
‘Sons ae all
“The second view, according to which counterfctuals of fiee-
dom lack truth-vaes, probably arses from the refletion that
theres no way to assign the tuth-vales because (where the con-
Sequent expresses fee choice to be made in hypothetical cicum=
Stances) there i in principle no way of knowing whether the com
Sequent would be tre ifthe antecedent were eae, This, however,
overlooks the possibility that we might beable ro know whether
the counterfactual sere without knowing this,
‘But how i this possible? The general relationship between
counterfacuals and Hoestaran fice wills something that sll needs
to be worked out. (Indeed, itis eeally the central theme of the
present discussion.) Butan extremely plausible view to take isthe
Following: A situation in which an agent makes a ibertaran fice
choice wit respect to doing or not doing something is situation
Sn which che agent might do tae thing but also might refeain from
doing it, Suppose that A, if she found herself in citewmstances C,
would freely decide whether or not to do X. Then both of the
following counterfctls wll be true:
(6) IEA were in C, she might do X:
{G) IFA were in Gabe ight reftsin from doing X
But if this i 90 then there is no tue counterfactual of freedom
with respec to A's doing X in C: For (6 is inconsistent with
(@) IFA were in C, she wonld retain fom doing X.
Likewise, (9) s inconsistent with
6) IEA were in C, she would do X.
I propositions like (6) and (7) properly characterize a situation of
libertarian fee choice, thon all counterfactual of freedom are fale.
Middle Knowledge Lo
Inthe ensuing discussion, this dhe position which we shal s-
sme the opponent of middle knowledge © be asserting and de-
fending
(Objections to Middle Knowledge
[Now that the opposition between proponents and opponents of
riddle knowledge hasbeen delicate, how ean we make progress
fon resolving the issue? As noted, proponents seem willing 9 rest
their ease on general considerations of plausibby, perhaps but-
{uessed by allusions tothe alleged theological necessity of the doc-
trine. Opponents can do the same, ofcourse, and many do, but it
the discussion is to be advanced, more substancial arguments are
needed. And, in fact, such argument are available. In this section
thee brief arguments against the theory willbe spelled ont, and in
the next somewhat more detailed argument will be developed.
The first objection to be considered is one We have already al-
luded co: What, if anything, sth ground of the truth ofthe coun=
terfacuals of freedom? leis important to see cat the question here
Js metaphysical, not epistemological. The question isnot, How
can we low that a counterfacial of freedom is erae? Kemay be hat
‘we cannot know this, except pethaps in a very few cases, and
although itis claimes that Gad knows them, iis not clear that the
fiend of connterfactuls (or any other thst for that mater is
requied to explain how eis that God knows what he knows. The
‘question, rather, i What mabe the counterfictuls true what |
the ground of their ruth? As Adams says, “I do not understand
‘what it would be for [counterierals of eedom] to be tru.”"*
In replying to this Plantings finds this notion ofa requzement
tae there be something that "grounds" the wath ofa proposition
to be obscure. But insofar as the requirement oes hold, he chinks
the counterfactual of feedom are no worse off with respect 1 it
than are other propositions whose credentials ae unimpeachable
‘Suppose, then, sha yesterday I fely performed some aon
Wha was ors tat grounded or Founded my doing sot wasnt
Adan, “Mile Kole” p16wo] God, Time, and Knowledge
cued todo by anything ce; nothing relevant ails that 1 did so
‘So what gronde he truth of the proposition ngueton? Perhaps
yon wl sy that whit grounds tah jst hatin at Tid A,
‘Bc hist much fan answer: and at any rte the same kind of
tnwer avalbein the cae of Carle. For wht grounds the ith
‘ofthe counteract, we may sy in eatin face Carley is uch
‘hati he bid been offered 2835200 bbe, he would have fly
‘This answer of Plantings’ appears to be an endorsement ofthe
view already attributed to Sudeex: When a counterfactual of fee=
dom is true, ie simply an olkimate fice about the fice agent in
‘question that, if placed in the indeatedcicumstaness, she would
fact as the counterfctul sates; this fact requires no analysis oF
“grounding” in term of fureher, noncounterfctal states of af
fais I seems to me, hostever, thac there is something seriously
‘rrong about this answer In order to bring this out, I Want eo ery
sind formulate a cerain intuition—an intuition that, I believe, n=
derles Adams's objection even though Adams docs not explicly
Formulate it.The intuition this: In order For a (contingen) cond
tiogal state of afr to obtain, is obtaining must be grounded in
Some categorical state of airs, More colloquially, taths about
what would be theca...” must be grounded in truths about
‘hat in fat the ene. This rirement seems clay to be satis
Fd for the more familiar types of conditinals. The truth of 2
rateral conditional is grounded either in the wath of is conse-
‘quent, oF the fality of iss antecedent, or both?” More inter
‘tingly, the tcuth of causal conditional, and oftheir associated
‘counterfactual, are grounded inthe natures, cass] powers, ner
fet tendencies, and the like, of che natural entities described in|
them. The lack af anything like this a a basis forthe counteric-
nap to Robrt M, Adon” p74 Panga ee alae to am camp
Goon Pe Nemo Ney pe 7
‘Sout Panne ew ely pi by he epee on
papel Tie Nae Ney
Bet dy
“Samsung da wheres nates cal powers may eos |
coeliac ei, oe mao
Middle Knowledge to
talk of freedom seems to me to be 2 serious problem for the
‘theory
Perhaps tis worthwhile to repeat here that the grounding cannot
be found n the character, prychologiealtendencien, and the ike of
the agent. This point i, in effect, conceded by the defenders of
riddle knowledge; they recognive that such paychological facts ate
insufficient asa basis for the counterfctsls, And yet there isthe
following point: I visually every case where we sem have plausible
examples of me counefcuae of fedom, the plow is grounded
‘precy in such psychological fact as thee, (Again we reall Sal be=
Seging Kelah, Plantingn climbing Tahquity Rock, and Adams's
Dutcher selling him 2 pound of hamburger.) And this, 1 think,
‘ought to make us very suspicions of those examples. the basis for
the plausibility ofthe examples i i all eases found in something
that has no tendency to show that the examples ae correct—no
tendency, that is, to show thatthe propositions in question reilly
tre true counterfactuale of feedom=*—then the examples lose all
force as suppor forthe theory. And without the examples, dhereis|
very litle in sight that even looks lke supporting evidence. =
‘he comin inl rounds nore wate of is camp a
‘ey cme owes ats ene oy eg,
trem pst aur ail owe “ts it he pt ot he
leer ot wanda! posse word soa or tance ona For
"uly amd atthe aries mone pombe wes veh eh
Semana conspire quan of tua he onc
{iotshoppnte eon wh he codon qos ae tnt Tate
‘Eoninpent ih tndad pe were soman ar sant conde
‘mals respons eter, then Molise omy tr popes |
{ewatee cpie of aang ral with topes to cond! ee
{ogenfodscon to as be Molin Ox Ds Foe ne Pat Va he
Crm) an Alfed | Fredo [sc NV: Conall Une Pr
‘ota 2)
"Nac, fever, ht ich papcolgil fc might very well prowde
rounding fr cdma ach hy aneing hone pen coed
‘Spee genes wo pray et
"Siang ye Surly there se many noe id many creates ch at
ats ht wea ae io of henl God, Time, and Knowledge
‘The second dificuly tobe considered—one, so far a | know,
not noticed in the literature to date—eoncerns the modal stars of
‘ounterfactuals of freedom. To do the job required of them, these
Counterfacaals ust be logically contingent—but I shall argue
that, based on the assumptions of che theory of counterfactual
logit, certain crucial counterfactsals should be regarded rather 3¢
nocessiry triths indeed they are er at all26
“The examples of counterfactuals considered so far (8. “IC
David stayed in Keilah, Saul would besiege the city") are in 2
certuin way notoriously incomplete. The antecedent specifies a
Single craial fact but leaves unstated many other facts about the
situation which would undoubtedly be relevant to Sauls deci
sion-—facts about Sauls character and state of ming, but also fats
bout the strengel and readiness of Sauls own military forces,
about other threats to the kingdom, and so on. Now, it cannot
Seriously be supposed that the counterfactuals God considers in
deciding about his own activity in cettion and providence are
incomplete in this vay. Surely, the antecedents ofthe conltionals|
The considers must include everthing that might conceivably be
relevant co Sat’ deciding one way othe other. In order to have
Some grasp om this sor of counterfactual, suggest that we chink in
terms of inii-spmentcurterfacals, in which the antecedent spec~
ies a omplee inital segment of porble aor yp toa given point
in ime, and the consequent an event that may oF not take place at
that time, (OF course, the antecedent will include any relevant
Causal laws that have held up unl chat cime in that possible
‘Selden ofeach coer of feedon” ("plo Rob Mt
Rate The ame to as However, vse The su ef ch
‘Smet but Ga stn rvs Go aon eon tn
[Stn way ut honane, fr tet pare hae ne such con ttn
‘Ser choc they mip be ed upon fo make=or, nko they do ave tem,
‘Telnet a to grown wed sl coor
“Sonata enn salySld tate exec of dam ae cone
‘aint inte or oh ie rr, feefonatan Kran Te Ps of
(ot aowing Ge Wook uM 8), pp. ate). Buen se
{te tn No nial coon, pon a wha cane
"ty arde hy tee the notion ofan tl seme” of 3 pombe
wold wea te dstson btw hd ean softs he sr
|
Middle Knowledge tn
world.) [Fnow we symbolize such counteeFactuals using a capital
lester fllowed by an asterisk to stand fr he antecedent, then the
iniilsegment countertacual corresponding t0 (2) would be:
(40) AT Saul besiege Kea,
where “AM represents a proposition specifying the entice inital
segment of the possible World envisaged by God as the one in
‘which Saul makes his decision, The contrary counterfactual then
‘would be
(22) AY Sl doesnot besiege Kel
If, 25 we have been assuming all along, (t) s tue, then (1) also
willbe true, and (1) fle,
‘The interesting question, however, is whether (1) i 3 contingent
cor a mec tuth, Clestly, the theory of middle knowledge =
‘quires that tbe contingent: if on the contrary itis necessary, hen
Soul's decision is onsiled by a complete statement of antecedent
conditions and his action isnot fee. 10) in fc, i tobe evaluated
inte same way 8 any other counterfactual proposition: To assert
(fo) is in efece to assere that some world in which “A” is tue and
Saul besieges Kesh is more similar othe actual world than any in
which “A°" is tue and Stel docs not do this. But, we may ask if
(1o) is contingent, then under what possible circumstances would
itbe false? The answer is that (10) might be false ifthe actual world
‘were different chan i is; wha is crucial isthe similarity of envis-
aged possible worlds o the actal world, and so ifthe actual would
were a diferent world (in ways we need nor atempe to specify)
than the one which sin et acta, i might cura out tac the world
specified in (r2) would be more similar to that world than isthe
‘world specified in (10), in which case (11) would be ere and (10)
fake
‘But this, T want to say, violates the fundamental idea that under-
lies the possble-worlds semantics for counterfictuls. For why
exactly i ie that counterfactual are to be evalted in terms of
‘comparative similarity of posble worlds wo dhe acwal word? The
answer to this is crucially related to the incompleteness, notedul God, Time, and Knowledge
above, whith attaches tothe antecedents ofthe counterfacuas we
tse in everyday discourse. We simply donot have che resources 10
specify in the antecedents of our counterGcwal everything that
might be relevant to the occurence ofthe consequent, and even
‘when we ate clear in our own minds what the cicumstanees
should be, we often do not take the trouble to state them. The
hotion of similarity to dhe actual world, chen, removes what would
otherwise be the ambiguity of our counterfactuals by specifying
how the unstated conditions are so be understood: We are to think
of the actual world as being modified as litle as possible so 28 to
fccommodate the counterfactual antecedent. Thus, David Lewis
states tha che point of his “system of spheres representing m=
parative similarity of worlds” s "to rle out of consideration many
fof the various ways the antecedent could hold, especialy the more
bizare ways." He also says
‘A coumtericul § C+ is rca wor if and only 9 olde a
ertin Qoworas but certainly noe all worlds mae. "If Aa
re dn el hey would pple oer” trae or Ese, she ese
nay be) at our words quite without regard those posible words
Ivhore kangaroos walk around on catches, and sy peight eae
Say, Those worl ar oo fa awsy fom ours Whats mean by
the countrfice ht, ting Being pty dha they are—che
Seavey of erutches for kangaroos being prey mich a acualy
{the kangaroos inability to ne ertces being prey ench a
cally and so nif kangaroos had noe they would tople
So the point of the notion of comparative similarity berween
possible words sto place finite on che works tha ate eleva for
the evaluation of 2 given counterfactual. But of course, (00) is
leeady maximally Lmited in this way, i already includes every
‘thing abous te envisaged world up unt the ime when Saul makes
hs decision. With raed to niial-segmentcounterfcrls, then
comparative similarity has no work ef o do. Ask yourself this
(question: In evaluating (10), why should it make + difference
Pee Company 9
Mile Knowledge Las
whether the actual worlds asitis, or isa world differen in various
‘ways from cis one? Afterall, A” were actual, then neiier"oue™
cial world nor that other one would be acwal—so why should
the truth of (10) depend in any way on which of those worlds is
actual as things now stand? This contrast sharply with thesia
tion as regards Lewis's kangaroos: If, for instance, we lived in a
‘world in which a large and active Animal Friendship League was
assiduously providing prosthecic devices for “handicapped” ani
mals, then we would "il in” dhese conditions a6 se evaluate his
counterfactual and would very likely jodge ito be fle. But with
inigasegment counterfacuale chere ix just no room for this to
happen: there are no spaces lef tobe fled in,
The situation, then, as fllows: The theory of middle know
‘edge is obliged to hold that some iniia-segment counterfactual
fare logically contingent. But in order to-do this, the theory most
apply to hese counterfactual the notion of comparative similarity
tothe actual world, and { have argued that this notion has no
legitimate application here—which isto s4y, the notion i mip
pled. The correct conchasion to be drawn from counterfctl
Togie, then is that if inital-segment counterfacuas are tre tll
they ate tsue in all worlds and thos are neesrly true. Bat this
conclusion i fatal to middle knowledge.
‘There is another, closely rated point, one that connects his
second argument withthe fist one given. Plantinga admits chat
"We can't look to similarity, among possible worlds, as explaining
counterfacuality, or 35 founding ot grounding it. (Indeed, any
founding of grounding inthe neighbothood goes in the opposite
Alirection.)" This means that (in some cases 2¢ leat) of 890
‘worlds W and W’, one is more similar tothe actual world preily
Because shares countericals with the act world—i tot the
case that, because one of thse worlds is more similar tothe accal
‘world in other respects, certzin counterfesals are tue, But this, 2 1
have argued above, violates the reason for introducing the com-
paraive-similaiy notion inthe fre place—that reason being, 26
‘explained by Lewis, to secure that counterfatals are evalated in
‘worlds sufficiently simular tothe actual world i noncuntrital
ely Rober M. Adam” 7hx) God, Time, and Knowledge
respects. How can Plantinga justify elying on the principles of
‘Counterfactual logic when atthe same time he undercuts the ration-
tle for accepting those same principles?
“The third (and final) objection ofthis group is one that was
discovered independently by Robert Adams and Anthony Ken-
hy.2" This dificuly arias 36 we bring together the account given
‘ofthe eruth-conditions for counterfactual and the use God is said
to make of them. As Kenny sas, “Ifieis tobe posible for God to
know which world he i actaaizng, then his middle knowledge
must be logically prior to his decision to actualize; whereas if mide
dle knowledge isto have an object, the acwalization must already
n place."
Let's spell this out bit moe, We will suppose, contrary to the
argument the preceding section, that some inital-segmentcoun=
terfactoals (nantly, chose whose consequents involve Geely
‘chosen actions) are contingently tru. Their eth, according to the
thcory, depends on the similarity of various possible worlds tothe
actual world, and this ie depends on which world isthe actual
world. But, which word is acral depends, im part atleast, on
God's decision about what to creat: Iis only by deciding to create
that God scles which word sacral, and therefore which coun-
terfactaals are tue. So rather than the counterfictals providing
_guiance for God's decision aboue what to create, the fact is that
their truth is determined only as 8 conssquence of that very dec-
lantnga's answer to this spelled out in his reply to Adams In
order forthe tr ofthe countrfactals to be “available” to God
ashe makes his creative decisions, i need not be already sted in
very respest which world is the actual world. What needs to be
sealed, in order forthe truth of «given counterfactual tobe deter-
‘minate and knowable, is only thatthe actual world isa member of
the set of words in which tha particular counterfctal is tue
Now, wy shouldn't his be the case, even prior to God's decision
about which parcular world to actuaize? Why shouléa’e it be the
have
Se Ram “Mite Rooweie“ Prt Key Th Ga i Pi
iAP cede Ppt, p2
Mide Knowiedge tar
‘ase nother words, tht the same counterfamals of eden are te i
al the worlds God could aaualizé® Why shoulda’ the truth ofthe
‘ounterfactasls of freedom be “cumerfumaly independent of the
‘arious courses of ation God could have taken”? If his 8 $0,
then the tath ofthe counterfactual seted prior to God's dec
som about which world co stale, andthe Adsme-Kenny objec
tion collapses.”
so enble the theory to wont Hest nn nn
Ie ear ari) [HC rtd Ada apd here woo be
‘oa hn ers a wn Gt
tae eaves) sabe. or apps) wa have ben ef Gn
Td rete oie eur We cn lige God een fallows
‘0d than moral ei sod mould be bor fo have more mot good Ban
Sele Bienen (9 ol ty
‘oer hat) ware col pete aller hs exon
ii waon br cing Asm a Ere Chap fo Reb Rate
30)
11h mgt teint oom elongation
pothesis, (8) would be tru if God created fee ceatres bu ae ihe did 0%
ect eg So hun Si ee
pth ronpect of tol god) Foes Cl, oti bck on doc ress
‘wal Set xen, cm rly ym ld deed or cee
[Sim an Eve fri had ceed thm the ol ae eo woe more
‘han oral go.” But hi consis he apposten af se xp wich
Sate wero fe cs) ee She eae
Sie ee mora ma
shot ert eect bec seme pablo th fr ea
‘ord wh hte ave feces sl would beer ke + wei ie
Itost fee creatrely decom ae bad ones C Middle Knowledge p14). Pan
Ep, however kamovedby ter we hve aeady not he sn oe
site err re ot re amra1 God, Time, and Knowledge
le must be acknowledged that this reply slices as a formal
answer to the objection, but I think it leaves us with a freer,
major problem, How are wet expla ee alleged fact that the same
‘countefactuals of ffeedom are tre in all the worlds God could
sctuslize? These counterfacruals, according to the theory, ae not
necessary truths. Their truth, furthermore, is not due t0 God's
decison; on the contrary, they constitute an absolute inion which
worlds God is able to actalize. For example: There ae possible
Worlds, plenty of them, in which eis ere that, if God had created
[Adam and placed him in Eden just as he did in the acwal world,
‘Adam would frcely have refrained from sinning. (We wil symbok-
ine this inia-aegment counterfactual as "E" > Adam refrains
from sinning.”) Now, why didn't God actuaize one of tho
"worlds in preference to this one? The answers, that infact the trae
counterfactoal, the one tre in ll the worlds God could acuaize,
{is "E> Adam sins." Buc why is this counterfacval true? Not
because of God's decision, and not because af any noncounterfac~
tual truths about the centres God has created. We wil se inthe
next section that a very feof these counterfacuas ate sid to be
true in vreue of che fee choices made by created beings, but even
if this answer proves tenable, it ean account only fora tiny propor
tion of the whole, So me are confronted with ths vast array of
countefactals— probably, thousands or even millions for each
acral or possible Fee ereature—almost ll of which simply aye me
‘without any explanation whatever ofthis fat being given, Is this
not 3 deeply puzzling, even balling state of airs?
The three objections in this scion have been developed inde
pendently, yet on close inepecton they reveal a common theme.
‘The frst objection complained about the lack of 2 ground forthe
teuth of counterfactuals in nonbypothetcal, noncounterfactul r=
ality, The reply is that no such ground is needed. The sceond
‘objection points eu that if this sso, chen te rationale isc from
tinder the principles of counterfactual logic on which the theory
relies thus making such reliance dubious at best. In the third ob=
Jetion, dhe groundlessnes of counterfactual reappears a higher
level, not concerned this time merely with individual counterfa
tuals but rather with the whole vast aay of them, all allegedly
te in al the worlds God could have actualized, and tue wichout
¥
Middle Knowledge bo
there being any ground for this ether inthe nature and actions of
God, or in the natures of eeated beings, oF (except for a tiny
fraction) inthe choices of created fee agents, Without doubt, see
are here confronted with something deeply miysterious—but this
the mystery of God's creation, or simply the mysteriousness of
misguided philosophical theory?
‘A Refutation of Middle Knowledge
We turn now to final, and slightly more complex, argument”
In his argument we shall ot, as previously, argue dell against
the counterfserals of fedom. Instead, we shall concede, provi-
Sonally, that there are ue counterfactuals of feedom and ask
bout them the question suggested in the lst section: Who or what
Js it GF anything) that brings abou? chat these propositions are
In order to give the discussion a touch of concreteness, imagine
the following situation: Elizabeth, a doctoral stadent in antheopol
‘ogy, i in the concluding phase of her course work and is begin-
fing to make plans for her disertation field eesearch. Her advisor
has been asked to make a recommendation fora foundation grant
to be awarded for observation of a recently discovered tribe in
New Guinea. This assignment offers exciting prospects for new
ic es min gh 0 the cis th fm he
{fe proponent of ml ledge maybe epee wi the na
[cif obvi fo the that compra sniinty wou Boa
‘Sodium uuson ine gene elem peel em
‘onan. estos that eng df een aes thet soe
rt tack ta elas» poo de fhe teary Pxopbe
‘Sisnce, Tate plloehs ie smile] wh ep nan oe
{oe mpion of comer enol hve he psn cond we
eS o snysyemate soar fh seman
SS wal Fake “A Ruan of Mile Rowe,
“Site cone of aging shin employed ere ad cbewkere toughowt the
{he concept ofa oymmtrl reson of dependence of what bros sont
‘nthe ston or ever tht bangs about te spe ngustion ra eB
Nae 30 (8:wl God, Time, and Knowledge
iscoveries ut would also involve considerable hardship and per-
Sonal risk, The advisor asks himself whether Elizabeth would
Choose to undertake this study, or whether she would prefer 10
Continue with ner preset plans eo study a relatively placid group
ff South Sea islanders. He wonders, in other words, which ofthe
following two coonterfictals of feedom is tre:
(12) IP Bbcabeh were offered de grant, she would accep (in
symbols, O— 4),
(13) IeElabeth weve ofered the gran, she would not accep it
(o= a).
Now, Elzabeths advisor may find himself unable to decide
which counterfactual strc, oF he may rach the wrong conchi-
son about this, But according co the theory of mille knowledge,
fone of the two counterfacoals i true, and God, if no one else,
knows which one. For the sake of our discussion, se willasume it
is (12) thats rue rather than (13; we shall assume, moreover that
Elizabeth is in fact offered the grant and she aceeps itll this,
however, is merely preparstory to raising the question already
suggested: Who or what ie that brings st about chat this proposition
Inthe previous section we have considered the restons why it
ennot be God who brings it about that counteracuas of freedom
te true; we shall not rehearse those reasons here, The answer to
this question thats infact given by the ends of middle knowl
Cele i that iis the agen named inthe counterfictual who brings it
Shout thatthe counterfactual is re. More precisely, isthe agent
‘who brings thie about those posible words in which the antecedent
te Ie this claim, then, that will be the principal subject of
discussion throughout this section.
"Howr might it be possible forthe agent to bring it about that a
given counterfactual of fredom is tue? Ie would seem that the
fnly possible way fr the agen to do this isto perform the action
Specified inthe consequent ofthe counterfactual under the condi-
ris ew {have head satin cso by Png kw af 0
swt source nny cet sabe waa a we labo Be
‘Eesing te come fo le Kavoedg hi sm mr ae
Middle Knowledge La
ons stated in the antecedent. That is to say: Inthe case of =
genuinely fee action, the only way 10 insure the acton’s being
dane is to do it [believe the proponents of middle knowledge
accep this, which s why they claim that the agent brings about che
truth ofthe counterfactual only those possible world in shih he
scent iste. I nother words an scepted principle that
(14) 1s in an agents power to bring it about data given
‘counteracualofficedom is tue, ony its uth would be
broughe about by the agen’ performing the action spec-
ifed in the consequent ofthe conditional under che condi-
tions specifi inthe antecedent
Batis it posible forthe agent to bring about the erath of
counterfactual of treedom in ths way? What is equired feist be
the ease that a particular evene brings i about tit a proposition is
true? It seems initially plausible that
(45) ICE brings it about that “Q" sera, then “Q” would be
‘ue if Eeccured and would be fei E didnot occur (CE
occurs) + Q and ~(E vecurs) + =)
sso aoe aa
Sesto aber ano
ict tba ye
Siac ar etunennchec gente21 God, Time, and Knowledge
rolled a7 or an 8.) ln each ese the problem arses because the event
in question sa token of type of event such thatthe occurrence of
any event of that eype (omeone's knocking on the door, your
tolling» or better on yous las ball) would bring about the ruth
‘ofthe proposition in question, With this in mind, we evise (15) 35
Tellows
(6) ICE brings it sou hae “Q" terse, then Bis token ofan
ceventype T sch that (ome taken of T occurs) + Qand
Some token of T occurs) > ~Q. and E's the fist token
of T which occur
If then we add the simplifying assumpzon tha if E were not to
‘cur, no other roken of T would occur, we ge (15) a8 a special
‘ase, When, on the other hand, we have an event and 3 proposition
‘such tha the conditions specified in (rs) and (6) ae not satisfied,
‘re wll say thatthe euth ofthe proposition i independent of the
event in question
‘Applying this to our example, what we need to know i whether
Elizabeth brings about the ruth of the counterfactual of freedom
*"O-+ A” by accepting the grant, or whether es tuth i independent
‘of er action, in the sense just specified. In order to determine this,
‘we need to know whether the following propositions are true:
(7) Ezabeth were eo aceep the grat, ic would be eruc that
O- Alie, A= (0-04),
(08) WElzaech were not 1 accep he grant, i would e te
that O— A (ne, “A> (0 A}
“There ean be no question about the truth of (17): if both “O” and
“A” are tue inthe actual world, the counterfactal willbe rue. Ie
might seem equally obvious chat (18) i lb: If Elizabeth does not
sccept the grant, how can ie be ere that, if offered it, she would
fsccept i? This, however, is a mistake. If (28) seems tous to be
‘obvibuiy false, we are probably misreading (18) a8
(49) IP Bhzabeth were wo reece the grant, it would be tue that
OA ie, (08 ~A) > (0 AY)
rT
Midle Knowhdge Las
‘This is indeed obviously fs, but i is not the same as (1); the
antecedent of (18) says, mot thit Elizabeth ejects the offer, ut
merely that she doesnot accept i. It is consistent both wth her
rejecting the offer, and with the ofer's never having been made I
she rejects it, then "O-—> A” must be fale, but ifn offers made,
"OA" wal sil be rue, So now we have to evaluat the cou
verfactal
(20) IPBlzabech doesnot accepe the oe i wil be Beaute she
rejected (re. A> (O&A)
(21) Elizabeth does noc accep the afer, i wil be besa the
fer was not made (Le, A=» (0 & ~A).
1 (20) i tr, (18) wl be false, but (21) is tue, soi (48,
How shall we decide his question? According to our semantics
for counterfaesals, the question sbout (20) and (2t) comes down
to this: fea worldin which Elizabeth reeived the ofer and rejected
it more oF les similar eo the actoal world (in which the offer was
accepted) than a world in which the offer wat neither made nor
accepted?
‘One's first thought might be that the world specified in (2),
which difers from the actual world with respect to Elizabeth's
cceptance ofthe off, more similar tothe actual world than the
‘world speciied in (21), which difers with respect both to the
making of the offer and to its acceptance. I 50, however, then
one's irs thought (ss So often the ease in matters connterfatl)
‘would have overlooked important considerations, To see why.
consider the fllowing example: Ihave been hard at work making
2 poste announcing an upcoming event, and just as the postr is
nearly completed I knock over my ink bot, spilling ink on the
Poster and forcing me to stare all ver agnn, As Ido this, pause
From cursing my chimsiness long enough to wonder what it would
have been like not to have had my poster rained inthis way. Two
possibiiies occur to me: I might have tefiained from knocking
{Tata maton) rh) poet my emp ne
‘ould sell ec hive sy Scone sate, Nope empha betad
Sem ai gly omen se popesa“) God, Time, and Knowledge
‘over the ink bottle inthe first place, or, I might have knocked i
‘over just I didn the actual wood, but instead of pling any nk,
the botle spontaneously righted tsefand come wo rest again ins
‘original portion. I then wonder which of these somarios would
have occurred if had not gotten che ink spilled on my poste. 1am
wondesing, in other words, which ofthe Following countertic-
tual i rae
(22) Ifmojnk had boon spied on my pote, ie would have Been
because did nor knock over my ink tle (=~S —» (= K&
=5).
(23). Ifnoink had been spilled on my poster, it would have Been
beste I knocked over my ik bowls but no ink spilled
Cs a9),
| purale over this for afew moments, but my question is quickly
answered along the same lines already suggested for (20) and (21),
‘The (22)-world would have differed from the actual world with
respect both to the bottl’s being knocked over and the ink’ spill
Jing, whereas the (23)-world diflers from itonlyin the Iter re=
spect. So the (23)-world is more simile tothe actual world than
‘he (23)-world iis 23) chat eee and not (23), and I elze tae
ry poster bad not been ruined, the reason for this would have
been, not that I was careful about my ink borte, bur chat afer 1
knocked over the bowl it miraculously sighted self wichour spill
ing any ink. And tht makes me fs ile better abot my eim=
(OF course this is absurd, bu why ist absurd? What exactly is
‘wrong with the reasoning that led me to conclude that (23) ere
{atherthan (23)? The anewer seems to be tix: Inthe actual world
‘certain counterfactual are true, among them
(24) IFT were to knock my inksborle ia sueh-and-uch a way,
‘he bot would fll ver and spi nk on my poste,
“This counterfictul is true in the actual world (as events have
shown), and i also tre inthe (23)-world, but notin the (23)-
¥
Middle Knowledge Las
‘orld, And in weighing the comparative similarity to che actual
‘world of the (22)-world and the (23)-world, the truth in the
(@2)-world of che counterfactual (24) counts fa more heavily than
the slightly grester similarity of the (23)-world with respec co
‘eal content. So a6 we thought all along, itis (22) that is true
rather than (23),
‘Buc of course exactly similar considerations apply in the case of|
(20) and (2). In ehe actual word, itis cue that
(42) Ue Elinabeth wore oferod the grant she would accepe i.
“This counterfactual tue inthe actual world, and also in the (1)=
world, bu notin the (20)-woeld. Andina compacson of the later
two worlds, the rath of the counterfactual (12) outweighs the
slight difereace with respect to similarity in factual content, 30
thatthe (21)-world is indeed more similar to the actual world than
the (20)-world, and ii (21) that iste rather than (20).
Te might be uggested thatthe reason (24) is decisive with respect
to the decision between (22) and (23) is that (2) i backed by laws
‘of nature; counterfactual of freedom such a5 (12) donot have sch
backing and are therefore not decisive with regard tothe choice
(for example) berween (20) and (21). Bu this rally will not do
For one thing, Plantingahimsel's prey clearly committed to the
view that, in deciding the comparative similarity of possible
Wrorlds, counterfactuals outweigh diflerences in matters of fict
whether or no they are backed by laws of nature.”? And there are
reasons that make every dificult eo justify weighting counterfa
tale of freedom less hei than les of matre. First, dete is che
contention, noted in the lst section, that the same counterfacuals
fof freedom are tre in all the worlds Gad ean actualize and com
seiute absolute limitations on God's power to bring about states of
afsirs (Lawes of mature, clearly, do nat limi God's power inthis
‘way: he could have created a world in which different laws ob-
tained)
The proponent of middle knowledge, however, may object to
“take thi ese ups of be argument given in The Nabe of Ns,1 God, Time, and Knowledge
this piece of reasoning. He may point out that, although God can-
‘not control which counterfactuls of freedom are tru, the human
legs in question—the agents named in the countertactals—do
Ihave conteol over this, since itis they who, by making the choices
that they do bring about that those counterfactual are tue. Now.
fof course, whether oF not the agent brings about the wuth ofthe
‘counterfactual s the very point ise inthe present discussion. In
view ofthis, one might tend to consider it question-begging (0
introduce this poin on cite side at this stage ofthe argument:
Bat the proponent of middle knowledge may fc this snfsr to
him. The claim thatthe agent brings about che erat of counteric-
twals of Freedom is, he points out, an integral pare of his postion,
‘ne that he should be permitted to appeal to dnt and less iis
‘efited by his opponent. Suppote we concede this point and
agree to evaluate the immediate point in queston—the question,
that i, whether ie (20) or (21) chat is rue—in the light of the
claim thatthe agent decides which ofthe coumterfcrals about her
factions are tue. How will this affect the outcome of the discus
‘A natural view to take would seem to be tht this point made by
‘he proponent of middle knowledge tends to balance off, and thas
torneuttalize, the last point made i the previous paragraph, There
it was pointed out that God has control over which laws of nature
obtain, but not over which counterfaetal of freedom are true
“The eeoinder i that human beings have control over some coun
terfacteals of freedom, but not over natural laws. 38 woul em
tobe the eas, these considerations weight about equally on ether
‘in acincs me of bersing the gun (in my “Reto of Mie
Kowaige) by ang tat tce Cala conor wh cum
fans af esiy fey ae ts see do we (Hikes Rear of
Mie Koel", 20. Bath be, 10 pts his Tope for
‘mong th remiss of my argument. What shoe tt Es
andr th po rahe proponent col heath he
seme feo) fm my Sac of whe coum fee
prot a en wre he Bren of pol named by he mt sd de
ere ho
Yr
Middle Knowledge le
side ofthe argument, the upshot would seem to be chat we cannot
decide, on the basis ofthese considerations alone, whether coun-
terfactuals of feedom are more fondamental than lws of nature,
‘oF vice versa Ifanything, what secs to be suggested i thatthe
‘90 are roughly ata party. I'we wish fora more definitive answer
toe question, we must look further.
[Now what is at issue is whether ic is counterfictals backed by
laws of ature or counterfietal of fcedomn that have countcrex-
samples in possible worlds “closer” tothe actual world. le rele
‘ant in this connection that we now know with viru cereinty
that the fundamental laws of nature are probabilistic rather than
strictly deterministic, chu, che countefacaals backed bythe laws
of nature (suchas [24) ain fict would probably conditions rather
than qrue necesstation conditonals. Surely, however, necessit-
tion conditional (such ay the counterfactual of feedom are supe
posed tobe) have to be weighted more heavily than “would-proba-
bly” conditional in determining the relative closeness of possible
worlds, There i also the important poin tat God, according t0
(Christian belief, can and does work miracles. If this is so, then
some counterfactual backed by las of mature have counterexam
pls inthe actual world sel, and therefore also in posible works 25
close to the actual world as you pleas. In view of all this, che
‘counterfactual of freedom seem t> be considerably more funda-
‘mental, with espet to explaining why dings ae they ae, than
the laws of nate fron, they are more fundamental than par=
ticular facts such a0 that Elizabeth is offered the grant."
‘Fenn commenting onthe en fh eure enn th ale “A
Refs of Me Knowledge is that te argument ea ave
cane of my um he "he proponents frie know et ool
id tah oh of eourcetal of csdom Gd he woe cone
{Shen word he asin of ete ated akon, On
PN {ot ik ny cone of do al
ava of my wn caning ety af sma herr om
‘cence ftom pmlc th eo shoul be cred by te propents of
‘ue Enowlige te para conn rely Fen eed et
{0 be sronger hn ware ts wold nox cre sor airy
‘Sit docrmning the eve sary of pore wri, comer
‘al of fdom surwegh ptr et sch hur Eke sole the“) God, Time, and Knowledge
But i (21) 6 true, then sos (18) and (nce [17] i also tr) it
follows thatthe truth of the counterfactual "OA" i indepen-
dene of whether or noe Elizabeth actually acceps the grant. (le
rot toe if she rj the gran, bur tat i another matter.) And it
ako follows (by [16) that Elizabeth's acceptance ofthe offer does
tot bring it sbou thatthe counterfactual "O—> A etre. And in
feneral, it snot rae thatthe cath ofa counterfictual of freedom s
brought about by the agent.
Fan hating ngen ed in ot 4 tow sams to commit him ool
Epa toy do, adn ne fi the ify be dosnt em spel
fram by no means convined ths he condi ave dean sto stony
bees spot gen athe ot pode cogent earn baling
thre cc or ya te
‘ns age wth dh they ae to pron Gee) ely eset
“Sm nie o Avan Png fr a cxemely rnin objection thi
{Tal ot werk ne tune thecal spp shy” Spon che
ena word tise tat) OA, ert int (3) O29, spon
‘here that Eesha ec Wechen
‘elven eb ese ep And mn nm ny
roth oa ty ce uy
Sciacca sb cell the pr bee she wat tor ofr
sion dat "“O"» “A iste, ad shen fet ret the grant ee koe the
{ch hae thangs were deen tat ay, Ehesbeth would ace the pan
toy emt be sos wb she awe usury whe ust a swe
edly seminal Change ould ke that al her eke oppor ri
‘oc nelle show so ht she dc oe scp ce he eae
‘ithe pening ot her ec willbe pospoed dt iwc ern te
Fopontonsnog shined te gr eto mod rooms
‘hough °O"> A be) Supposing ths tobe the cave, we then tk, IE
tren pth rat, woul be bane "OF" es ue ber an “O. a
ower lb af abe wero pith woul eas he meanest
Srv dlr shanty te th acl wel wih she races and nt
[SSuse commerical “0 = A world tet So all wn eat
Elcotcth rns abt oush of «seach buy
shih the orl eae dors
Yr
Middle Knowledge lo
Does the conclusion we have reached constiute a serious prob-
Jem for middle knovledge? Perhaps not. David Basinger has r=
cently argued thatthe proponent of middle knowledge need not
find should not hold tha the eruth ofthe counterfactual of freedom
js brought about by the agent” To be sue, the view chat che truth
fof these counterfietals is broughe about by God must also be
excluded, forthe reasons already discussed. Rather, these counter
factuals simply ave nue without their eruch having been brought
about either by God a by anyone cle. "Who responsible forthe
‘rath of [the counterfctals of fred] inthe acta world? The
answer is that no one i responsible."
This proposal, however, creates serious dificutes for mide
knowledge. On the proposed view, Fleabeth snot responsible for
the fice hat, ifshe were offered the grant, she would accept it (ie
for the tuth ofthe counterfactual "O~» A"). Nor, we may a=
sume, is she responsible for the truth ofthe antecedent —that i, for
the ft that she e offered the grant. Buti she is responsible for
rcither ofthese things, itis difficult see how she ean be responsi
ble for accepting the grant—a conclusion that is entirely «n=
sweleome to the proponents of middle knovledge
‘But there is another, even more fundamental, difficulty, We
have learned chat Elizabeth does not bring it about thatthe coun
terfacoal "O— A is tue, What effect, iF 2ny, does this have on
the question of what i in he power when the grant offer is made?
In order to investigate this, we need what I call power nto
prinple, principles that state that an agents possesing the power
to perform 3 certain kind of action ental thatthe agent also pos
estes the power to perform another kindof action. More will be
‘sid about power entailment principles in chapter 6 bat a principle
that will suffice for our present purposes i
(PEP) Its in As power to bring it about that P, and “P” entails
"Qi" and "Qi Be, then st isi A's power to bring ie
about that Q
‘2-Divne Omnis and Human Freda AM Knowle Pepe
sig” Fah and Piya 990
“ios1 God, Time, and Knowledge
Aide thought wil show this pinple ob corm 1°" etal
("weno be the cn that Panes tia oe hot IF
(ised tw, then te eanet pees had to's
bait bing abot at (Seething
‘injowr power tring about tat yous snr een
though you comply ck he power to gst te snc
ial) Buf "Qe tate, rn pone for you to Bring i
hour tha Punls i saopoable for you to rig sbou cit
Uappoach your hour wih he intense of ging Your
Scovbal ony todcover that you do not haves dtl Ulos
‘Sin my power fo ring taboo tha ou tae Sooke
by imtllng one myself or having onc itd—ie natn my
poser tang your door.)
How dow tpn ppy to the marin and? node to
proccd ew make he stumpon, which is matoned byt
Neary of idle Knowledge, tut of ny pai of counertictuas
Such (1) and (3) one oth ter yh pst the
‘Ejucton of tie two men tv, toe ral posse
rors, Gres his ssunpton, we veh lowing areca
ts
(25) Ut Blzabeth is offered the grant nd acepe it, tis tra hat
O-- A (O&A) > (0 A).
(26) I Elzabeth is offered the grat and eet i, ti ere hat
0 =4(0& =A) > (0 ~A),
‘That is to say, Elzabet’s acceptance or teection of the grant
tna the truth ofthe corresponding counterfactual of freedom.”
“ci anion Mali Die onc 5) te
ae
(a) PG tereore,P—>
sept at a) td (ane dered witht recut the mt
{hn moet text But (f) ay onion sone acme Lew nim a
wo com ery odo pk te complsy cele he no
esrngth after ut tse the ouierstl CP 8 "(Cam
tmp 39. Lews pgs semansn which (A) would not hl rh
en, Ei tery a) etd
Tete ter tlio hte © be oud in be dations made by
Middle Knowledge Ta
Now we ae ready to consider what isin Elizabeth's power
when the offers made. Fist ofall etn her power co ace the
grant? One would suppose that itis, since in Fact she seal doce
50. And (PEP) places no obstacle in the way ofthis conclusion. Her
accepting the offer entail the tuth ofthe counterfactual" A,"
bt thar counterfactual is infact eue, ad s0 the question of wheth-
crits in her power to bring abou is tuth does not ate,
But now let us ask, does she have it inher poster to rjc the
rant? Her rejecting the gran ental that the eonterfatu
A" be tue, but ths counterfactual i infact alse. So according
‘0 (PEP}—she can have the power to reject che grant only fii in
her power to bring it sbout tht this counterfactual i tru, IF she
does not have cis power, chen she lacks power to reject the gant.
‘And now the situation becomes serous, We have sen that i
would be in Elizabeth's power to bring it about that the counter=
factual "O > ~A" is ee only if the truth of this counterfactual
‘would be brought about by her rejecting the offer. But we have
ako seen that the tuth of 2 counterfactual of freedom is not
brought about in this way. I fllows that Elizabeth does not haveit
in her power to bring it about that O—> ~A, and lacking this, she
also—by (PEP)—lacks the power to reject che ole.
is time o summarize. In cis setion we are investigating the
‘question, Who ox what brings it about thatthe counterfactuals of |
freedom ate true? We frst considered the possibility that its the
agent named inthe counterfactual who does this—in terms of out
‘xample, hat Elizabeth by acepting the grant offer brings about
that OA. Ie turns out, however that thie counteract i true
independently of whether or not she accepts the offer: Ie would be
true ifshe were o accepe the offer, andi would also be true were
she not to accept the offer. To be sue, ie would not be tru if she
ere 10 ree the offer, bu this tens out not to be relevant i she
did nor accep the offer, this would be becuse the offer was never
‘made and not because i was made and she rejected it. Bat since the
Pelloc terwen various kind of jontive ional (A) a fr singe
‘bet spdabe for "een condone bara wlan so the coe
{Gea of eam se nessa soa for ee cena (8)
lai (See Pll, Suction chap 3 (8 them, nt te
Counties2) God, Time, and Knowledge
counterfactual is true independently of whether or not she accepts
the offer, it cannot be the ease that she brings about the rath ofthe
‘counterfactual by her acceptance ofthe grant
‘We then went on to consider what Elizabeth has it in her power
to-do when the grant is offered to het. Clearly, she has it in bet
power to accept the grant, and she demonstrates ths by doing so.
But does she also have che power to reject the grant? Of pasiculae
importance her is the fact that (given the ruth of the theory of
tml Knowledge) her ejection ofthe grane entails the auth of|
the counterfactual "O> ~A." But this counterfactual is not tue,
foi ean be in her power to reject the grant nly ifit is also her
power to bring about the truth ofthis counterfactual. But we have
[ready seen that this is impossible. She could have the power to
‘bring sbout the truth ofthe counterfactual "O —» ~A" only ifs
truth could be brought about by her rejection ofthe offer, but we
have sen thatthe agent cit in this way bring about the truth of
2 counterfactual of freedom, So i notin her power to reject the
rant
“The conclusion to be drawn from this is chat the concession
made eatlit—that some counterfctals of freedom ae rue—as
warranted, Ie turns out from our consideration ofthe case of
Flizabeth that insofar as such counterfctuals ae me, they are not
counterfactual of fvdom Ifthe counterfactual “O> A" stro, i
js notin Elizabeth's power t reject the oer, and shes not free in
the raquired sense. And, on the other hand, iasfat as an agent
frenuinely fie, there ae no true counterfactual stating what the
gent would defintely do under various possible circumstances.
‘And so the theory of middle knowledge is seen to be untenable:
‘Thee ane mote coerfutualsof reed
yr
[3]
Simple Foreknowledge
This chapter isin the mature of an interlude, The preceding
chapter dale with the possibilty of middle knowledge; the next
several chapters willbe occupied with the logisl possibilty of |
‘comprehensive divine foreknowledge in a world containing fee
choices. Bu theres another important question about forcknov-
fedge which has received mach less attention than it deserves—
namely, what would be the use of foreknowledge if i existed?
Why, in other words, is foreknowledge important? This isthe
topic for the present chapter.
‘One answiee to our question which comes readily to mind isthe
following: That God as foreknowledge of all things whatever is
aninescapable implication of divine omicente—and, since omnis
tence i recognized as an essential atibute of the theistic God,
what is at stake in the foreknowledge controversy i nothing lest
than theism itself. This answer, however, wll not withstand in-
vestigation. It is clear thr there are versions of the doctrine of
vine ommiscience which are compatible with God's having less
than complete knowledge of fare events: this would be the case,
for instance, sf certain propositions about the futute lack trith-
valees, and similar result follows if omniscience is defined by
saying that an omniscient being is one that knows everything that
itis logically possible o know. Now, theists who fel strongly4) God, Time, and Knowledge
about the necessity of foreknowledge may be les than enthusiastic
bout these sorts of definition of omniscience—bat if this isso, the
‘efnition of omtniscience is tse in part a rut ofthe importance
ascribed tothe foreknovledge ise; icannoc therefore, be used ro
‘explain that importance
"believe thatthe importance attached to foreknowledge finds its
explanation in cereain aspects of theism which are more “r=
Tipious” and closer to the actual experince and concerns of re
ligious persons, than are the somewhat aid topics with which
philosophers of religion usually concern themselves, Iria pat of
heist eigions suchas Jodaism and Christianity that certain per=
Sons, speaking in God's name and with hi authori, fell the
“atic But (fe may be asked) if Gd does not knew what i going to
happen, how ean he tell the prophets what to say about it? Further
‘more, theistic religions make extensive claims—elaims that are
xperientially important to reigioes persons—about God's pro-
‘idental guidance and contea of worldly affairs. And many smes|
these aime seem f0 “cach out i terms of the doctrine of fore=
knowledge: Since God aleady knows everything that going t0
hhappen—especally, perhaps, he knows of some severe testing I
am going to have ta face-—he hat already arranged things S0 28 (0
work out inthe best posible way, say. by developing my charac-
ter to meet the anticipated challenge, and perhaps by prearranging
‘other circumstances $038 to provide a solution for otherwise une
Surmountable diffuses" OF 3 pice with this, though perhaps
Tess common in practice, isthe belie that answer Jo prayer can be
understood along these same lines—that God, knowing before=
hand that one of his children will ack in faith fora certain blessing,
may prearrnge things long in advance (perhaps even 35 early 36
the inal stae ofthe entire universe) in such a way that, in the
natural course of ehings and without further dec vine interven-
tion, the desived answer wil el
Prayer, providence, and prophecy these are mates of intense
concer to many ordinary religious believes, chough they re less
common in the books af philosophers. But then, some of the
"Form extensive declopet of hse pe, se David Bangs, “Mile
‘grape nd Chesed Chan Tg Res She 3:0} 38
Lg
Simple Forcknowledge bss
philosophers are themselves religious believers, and if seems to
these philosophers that vital religious concerns are at stake inthe
foreknowledge controversy, they may see compaibiism asa post
tion tobe defended tothe lat ditch.
AA this is entirely understandable, and it sheds light om the
intensity with which the foreknowledge issue is pursued. But 1
‘vant to say, nevertheles, that al of cis misguided-—dhat the
doctrine of divine foreknowledge, in its most widely held form, i
fof mo importance whatever forthe religiously significant concerns
about prayer, providence and prophecy. If am right about this,
‘hen compatiilist and incompatbilist alike en reuen to thee dis
fassion with lighter hearts and lowered pulse rates. They will
continue, and rightly s0, to pour thee best pilosophical efforts
into a fscinatng logical and conceptual problem. Bt they need
‘no fea that central religious interest willbe vitally affected by the
‘way the controversy comes ou
‘The Theory of Simple Foreknowledge
‘Whar as bean sid above doc nt apply tall versions ofthe
tate in forcnowlege. Ie doesnot apply to version of theism
‘which are straightorwardy deteministic® but neither dest ap-
Bly co the theory of middle Knowledge. Rather, i relies on the
Fefutaton of middle knowlege given inthe previous chaper—
nd on the oer hand, pilsopiers unconvinced by those 5
tents may welcome che present chapter ae demonstration of
how essential middle knowledge ely i. This chapter i con
cerned wih single orbnowlede, foreknowledge that embraces ll
tual fice dies, including thos tht ae yee be made, but nt
se Teh te pa ap Pel po ab te
Bay Sniee’ Tiel encpin sateen broad
{ets Sco loa si ol ple yw
LR ie er,so God, Time, and Knowledge
(es in middle knowledge) those choices that might have Ben made
Tt in face never ate
Tes possible tobe 3 bit more precise about the nature of simple
foreknowledge by asking the following question: Just how ii that
God is able co know futute events? Ie seems clear chat the notion of
4 mere accidental correlation between God's belies about the fir
ture and the actual events sehen they oceur is inadequate here
Evidently, one way God could know the facure is by knowing the
‘tual antecedent that wil lead tothe future evens in question; this
‘iew is open to soe determinists but not ro libertarians. Another
‘vay is given by the theory of middle knowledge: God, by know
ing both the conditions in which a choice would be made and the
relevant counterfactual of freedom, knows inflibly which choice
would be made under those circumstances. But if a has been
argved, there are no tue counterfacuals of freedom, this posi
ity also is climinated, Ie seems, the, thatthe only way left forthe
adherent of simple foreknowledge is 0 claim that God dvealy
Ibnows the acmal future event that God's belie about the mater in
{question i somchow brought abou hy the fur even self. Thus,
sidvocites of foreknowledge speak of God a having “direct vision”
(ofthe future a iin 3 crystal ball ora telescope.® This of course
involves something, very much lke reroactive causation; the 2d-
vocate of simple foreknowledge, if she i wise, will cheetilly
‘admit this and challenge her opponent to prove that retroactive
“ausiton is impossible. Ie may be tempting to suppose chat this
lke means thatthe fature mo in some way aleay exis other
Wise, hove could it have elects in the present (and past)? But
perhaps this temptation can be resisted. The inference will hold
only if we assume that causes must exist at the ime at which thei
cllccs ate produced, and proponents of foreknowledge and reeo-
Sctive cusition may decline to accept this, They may rather main
‘tin, pending proof to the contrary, tha he facure event can have
effects in the present eventhough doesnot ye eis
Fore cll compro, sce Stephen T Davi, Lap ad he Ne of
‘Gal Grand Rapa Nh Eada 6p rhe elope, sonata
Elvan The Proof Wal naps Debts Merl ep posh.
Shou! bed ht Edward fn ds osu oth wy mig
‘Eedoowice)
. J
Simple Foreknowledge Lo
‘What sort of understanding of providence can emerge from this
conception of foreknowledge? In order to make this question con-
‘rete, lee’ consider an example suggested by Basinger concerning
2 young woman deciding which of two mien to marty. If she is
‘devout, she wil wer ikely seek divine guidance on the matter: out
‘question is, what sort of guidance mighe be available? To begin
‘vith, God will have full and complete knowledge of the physical
and personality characteristics ofeach ofthe three people involve,
including hopes, ambitions, and commitments, latent fears and
cravings and the like. He knovs, inshore, 2 great deal more dhan
‘even the wisest and best informed marriage counselor could hope
to find out. All ofthis, however, comes from God's preset know!
‘eige, whereas our question concems what benefit can be derived
from God's knowledge ofthe fuare. What diference,ifany, does
this make to the sor of guidance God is able to ive?
Tn seeking to ansver this question, it may be helpful to remind
‘ourselves how i would be answered from the wandpoin of middle
knowledge. IfGod has middle knowledge, he as no need to base
his guidance onthe potential for happiness and success of the two
‘envisioned marriages. Rather, he knows in full detail exactly what
‘would happen tothe young woman if she mie either of these
men or ifshe refused both of them. The guidance, inthis es, is
‘aed on fll and complete knowledge
“The situation if God has simple foreknowledge, however, is
‘quite different. An immediate, and stiking, result, i that God
amet use his foreknowledge in guiding the young woman about
hher marriage decison, For the future situation which God forc-
knows is, of course + stution in which she aleady is marvied to
‘one of the ro men (or, perhaps, to neither)—and since the deci
Son's actually havi Ben made i presupposed by God's knowledge
of the future, he cannot possibly wse that knowledge in deciding
hhow to influewe that decision. And in general, iis clea that God's
oreknowledge cannot be used either to bring shout the gccurrence
‘of foreknown event or to preven uch an event from occurring.
For what God foreknovs isnot certain antecedents whic, unless
imerfered with in some way, will ead the occurrence of he event
rather, tthe evo tel hat is Foreknown as occuring, and ies
contradictory 0 suppose that an event is brown to occur but#1 God, Time, and Knowledge
then alo is prevented fiom accurting. In the logical order of depen
dence of events, one might say, by the "ime" God knows some
thing will happen, is “too late” ether to rig abou is happening
for to prevent it fot happening. God's guidance of the young.
‘woman with respect to her marriage must be independent of God's
Knowledge of her acrual future.
‘What may seem to be porsibe, however, i hiss God, because
Ihe frcknovs that a certain event wil ceue, may preaeange other
cos in the scuation in such a way as to produce the best overall
result, Such “prearrangement” is really what is involved in the
concrete applications of foreknowledge mentioned earlier In
prophecy, God prearanges the annourconent of the foreknown
{vent this announcement may serve any ofa number of purposes,
Including the accreditation of God's spokesman (when the proph-
coy ie filed) 26 walla influencing hearers of the prophecy to
conduct themselves in the ways God intend for them In answer
sng prayers, God can prearrnge rcumstances in such a way that
‘when the prayer is made (as he knows it will be), the answer
comes, ot asa result of any immediate intervention by God at chat
‘moment, but a a consequence ofthe circumstances that have been
prearranged months or millenia before. The general case of prov
‘dence differs from tis only in that the foreknown event for which
the circumstances are prearranged is not specified 35a prayer that
will t some future time be offered. And to return to out example,
although God cannot ase his foreknowledge of the young worn
an’s marriage decision to influence that decision itself would
sem that he may use it fo arrange other aspects inthe situation in
such a way a to improve the outcome of the choice she actually
rakes,
“To make thie still more concrete let ws consider an actl exam=
ple, Ins fimous Woeld War Il bade the Allied armies were en-
fired by the Germans at Dunkirk in June 1940. The military
balance heavily favored dhe Germans, ad the situation ofthe AIL
lied forces was desperate. However, unusually calm weather on
the English Channel and fog that inbited dive bombing by the
LLufewatfe made possible the evacuation by sea of most of the
Allied troops with far fewer losses than could have been expected
fee assume (8 many of es would) that the Allied cause was just
7
Simple Foreknowlodge Lo
and that God desired an Allied victory inthe war, we can interpret
these event along the following lines! God, having foreknowledge
ofthe encirclement, prearranged the causal factors that would de-
termine the weather at the time of the evacuation, thus making
possible the relatively favorable outcome instead ofthe anni
tion or mass surrender of che Allied forces. This example, I be-
live, gives 4 fair pieote of che way in which simple Forckaowh
‘edge can be incorporated into an understanding of providence.
‘The Uselessness of Simple Forcknowledge
1 now argue that this conception of divine providence is in-
coherent and that, in fact, simple foreknowledge is eel wees
for the doctrine of providence. Let us begin with a distinction
concerning the way God knows the future. The distinction have
in mind i beewen knowing proportions about the future and
Knowing the concrete evans of the fuare—that s the actual sac
time proces that will acer. The distinction may be lastated by
the story of King Czoesus, who upon consulting the Delphic Ora-
cle a tld, “Ifyou eros the Halys River, a great empire will be
destroyed.” He knew, let us assume, the tuth of this proposition,
but he was not aeqoained with the comet coeus that would make
the proposition erue—in particular, he did not know tha the that
empire that was to be destroyed ws his own,
"Now, it isnot tobe thought of tht, n knowing the fature, God
knows merely the truth of certsin propositions. All accounts of
God's knowledge stress that past and future events ae for God wo
le vivid or comple in ther epistic presence than those that are
presently occuring. (This sa lest one ofthe points made by the
{yal ball and telescope metaphors.) God, we may say, not only
knows thet 2 given proposition about the futuee is ere; be also
knows in the fillet detail exactly how itis trae, down to the
rinuest fact chat might romain forever inaccessible to even the
‘mort assiduous human investigator, We may, say, then, that sf
God has simple foreknowledge, he knows the conte events ofthe
Faure, and not merely propositions about che fotre
‘With this in mind, let us recur to the example ofthe young
woman's marrage. God, we may assume, knows the truth ofthe@) God, Time, and Knowledge
proposition “One year hence Susan will be martied to Kenneth”
Bathe not only knows tat this is tue, be knows in fll dail howe
its erue—he knows, n other words, the fll complex pattern of|
‘mutual interactions that wall consiute Susan and Kenneth's mat-
ried life together one year hence (Les call this pattern, the con-
‘rete event of Susan and Kenneth’ marrage, Mf.) Now, let ss
suppose that God, on the basis ofthis foreknowledge, ats inthe
present in some way s038 to improve the eventual hsppines ofthe
Inarsigge—for example, he infloences Kenneth to work on over=
coming his tendency toward impatience so as to become a more
sympathetic and understanding husband. Ics evident, given the