Biblical Arguments for Eternal Marriage?

Dur­ing the course of cor­re­spond­ing with a beloved but dis­tant Mor­mon friend, I encoun­tered some of the first Bib­li­cal argu­ments for eter­nal mar­riage I had ever heard. Usu­ally the argu­ments have to do with want­ing to stay mar­ried after death or with strictly Mor­mon scrip­tures, and not with the Bible itself. Now, I don’t think the doc­trine of eter­nal mar­riage is a deal breaker by any means — it’s entirely periph­eral to the fun­da­men­tal incom­pat­i­bil­i­ties between Mor­mon the­ol­ogy and Chris­tian­ity. But for what it’s worth, I just hap­pen to think it is false. Here is how one of his Bib­li­cal argu­ments goes (rewrit­ten in log­i­cal form):

1. All that God does is eternal.

2. God did Adam’s and Eve’s marriage.

3. There­fore, Adam’s and Eve’s mar­riage is eternal.

This is a log­i­cally valid argu­ment (its con­clu­sion fol­lows inescapably from its premises), but I dis­pute the truth of (1). The heav­ens and the earth are a clear coun­terex­am­ple: We know that God cre­ated the heav­ens and the earth (Gen. 1:1), and yet He intends to destroy them (1 Pet. 3:7). There­fore, since the heav­ens and the earth are not eter­nal, despite being done by God, not every­thing God does is eternal.

When I pointed this out to him, instead of under­cut­ting my coun­terex­am­ple, he sim­ply tried to rein­force (1) with Bib­li­cal evi­dence (which would, I think, if strong enough, rebut my coun­terex­am­ple and force me to rein­ter­pret 1 Pet. 3:7). The text he used was Eccl. 3:14.

The prob­lem I see with that, is that it is set in a con­text of cyn­i­cism, hyper­bole, and other noto­ri­ously hermeneu­ti­cally trickly rhetor­i­cal devices. For exam­ple, start­ing just five verses later, in verses 19 and on, the author, taken at face value, seems to teach that men die just like beasts and are not priv­i­leged with an after­life. Nat­u­rally he and I both agree that this face value read­ing is false. For surely, as we allow scrip­ture to inter­pret scrip­ture, we see else­where teach­ings all about the afterlife.

What of this verse, then? Well, when we inves­ti­gate the his­tor­i­cal con­text of Eccle­si­astes we dis­cover some inter­est­ing facts. Solomon, the pre­sumed author, mar­ried the daugh­ter of an Egypt­ian ruler (1 Kings 3:1). Egypt­ian reli­gion was obsessed with the notion of a phys­i­cal after­life wherein a per­son could lug all of his earthly pos­ses­sions, wives, etc. on into eter­nity. The Egyp­tians never took seri­ously the final­ity and sever­ity of phys­i­cal death. Con­trast­ingly, the Bible teaches that death ought to be viewed as the “last enemy”, in which earthly pos­ses­sions count for noth­ing (1 Cor. 15:26, Matt. 6:19–20), and that it takes a spe­cial act of God through Jesus to over­come it (1 Cor. 15:55–57). There­fore it is far more likely that Solomon was rebut­ting the Egypt­ian notion of the after­life than that he was deny­ing the Jew­ish hope in res­ur­rec­tion dur­ing the Age to Come. But to under­stand this requires the use of a lit­tle elbow grease. A sur­face level read­ing of a verse from Eccle­si­astes with­out respect to con­text should be imme­di­ately approached with caution.

So what does this verse actu­ally mean then, if not that lit­er­ally every­thing God does is eter­nal? Well, in con­text, Eccle­si­astes com­prises exam­ple after exam­ple of the failed attempts by humans to gain insight into value and into the future. The point the author seems to be mak­ing here is that we must be faith­ful and con­tent with what lit­tle knowl­edge and respon­si­bil­ity we have, await­ing with patience the unfold­ing of God’s eter­nal plan. Jesus sums it up when He says that while the heav­ens and the earth will pass away, His word will not. In other words, what God intends to be eter­nal, His eter­nal plan, will be eter­nal. But that doesn’t mean God never does things for a finite period of time, or cre­ates tools or instruc­tions designed to be use­ful for a finite time (the Mosaic law comes to mind as a great exam­ple — surely the Mor­mon church doesn’t advo­cate the ston­ing of adul­ter­esses or the strict dietary laws of the Mosaic law?).

The next Bib­li­cal argu­ment for eter­nal mar­riage he gave goes like this:

4. God left Job with twice what Job had orig­i­nally had.

5. God left Job with ten children.

6. Job orig­i­nally had ten children.

7. There­fore, Job con­tin­ued to have his orig­i­nal ten chil­dren despite their deaths.

This, as you can see, is not even directly related to mar­riage, but to fam­ily. The doc­trine of eter­nal mar­riage is actu­ally a corol­lary to a deeper-​​running and more devel­oped doc­trine con­cern­ing the “seal­ing” of famil­ial rela­tion­ships — par­ent to child, hus­band to wife, etc.

And this too, is a log­i­cally valid argu­ment. And inter­est­ingly enough, I accept its con­clu­sion. The chil­dren who died were not oblit­er­ated, unlike Job’s mate­r­ial pos­ses­sions. But the mere facts that A) the chil­dren con­tin­ued to exist after their deaths, and that B) they will always each have the prop­erty of hav­ing been born to Job, do not imply that they were ever “sealed” to him in the strict Mor­mon sense.

I will admit this: if the Mor­mon doc­trine con­cern­ing seal­ing is true, then the Job nar­ra­tive may arguably be called a glimpse of it. But I don’t think that the Job nar­ra­tive on its own can be taken as con­sti­tut­ing any pos­i­tive evi­dence in favor of the doctrine.

My friend is sharp, and he admits in his email that the Bible only offers “glimpses” of the doc­trine of eter­nal mar­riage. I hope he will even­tu­ally per­mit the use of sim­i­lar log­i­cal and hermeneu­ti­cal argu­ments in future cor­re­spon­dence con­cern­ing more fun­da­men­tal questions.

The Calvinist Reading of Gen. 50:20

Dur­ing the course of nearly every dia­logue with my Calvin­ist broth­ers in Christ (whom I love but with whom I occa­sion­ally dis­agree), Gen­e­sis 50:20 is brought up as a flag­ship exam­ple of the allegedly Bib­li­cal doc­trine of Prov­i­den­tial Con­cur­rence. Reformed the­olo­gian Wayne Gru­dem defines this doc­trine as “God coop­er­ates with cre­ated things in every action, direct­ing their dis­tinc­tive prop­er­ties to cause them to act as they do” (Gru­dem, 1994, p. 317).

Grudem’s meta­physics are never quite clear, but the phrase “cause them to act” is clear enough. Even though Grudem’s sys­tem­atic the­ol­ogy gives a cer­tain amount of lip ser­vice to the exis­tence of some­thing like a “will”, Gru­dem is ulti­mately a determinist.

And that would locate him smack dab in the main­stream of Reformed the­ol­ogy, accord­ing to Calvin­ist the­olo­gian and Pre­sup­po­si­tional apol­o­gist Gor­don Had­don Clark. Clark warns us not to assume that the denial of free will is “hyper–Calvin­ism”. He exposits the West­min­ster Con­fes­sion of Faith (the “high water­mark” of Protes­tantism), the Shorter Cat­e­chism, and like state­ments based on the cen­tury or so of dis­cus­sion lead­ing up to them. And his find­ings? All Reformed state­ments of faith and proper Reformed the­olo­gians are full-​​blown deter­min­ists (Clark, 1961, pp. 28–32).

Any talk of man’s “free agency”, “nat­ural lib­erty”, of “Com­pat­i­bal­ism”, or God never being the “author” of sin, to be found in the main­stream of the Reformed tra­di­tion is to be under­stood entirely deter­min­is­ti­cally, accord­ing to Clark. That is to say that the Reformed doc­trine of con­cur­rence holds man’s will is causally deter­mined in every detail by God, though some would say that because it is not deter­mined by phys­i­cal means, it can be called “free”. In this way God’s prov­i­dence is said to “estab­lish” (cause and con­trol) man’s will (and is there­fore “com­pat­i­ble” with it). Sim­i­larly God is said not to be the “author” of sin, in that while He causes oth­ers to sin, He some­how never directly sins Himself.

Call it what you will.

This is not to accuse my Calvin­ist broth­ers of the “Fatal­ism” label they are so quick to throw off when called “Deter­min­ists”. I’m not accus­ing them of believ­ing in any force in the uni­verse out­side of God. I under­stand that they believe that it is God Who is doing the determining.

And that’s exactly the prob­lem. It makes God the cause of evil (even if not the “author” of it under the man­gled, Reformed under­stand­ing of “author”). But being the cause of evil is bad enough, as it under­mines the omnibenev­o­lence of God and oth­er­wise runs roughshod over the Bib­li­cal pas­sages teach­ing God’s love and man’s free­dom and responsibility.

So the rel­e­vant part about my Calvin­ist brother’s asser­tion that Gen. 50:20 teaches Prov­i­den­tial Con­cur­rence is that it would mean that evil actions are not only “will­ingly” per­formed by their sub­jects, but also wholly, suf­fi­ciently, and causally deter­mined directly by God.

Upon inves­ti­ga­tion I have been pleased to find lex­i­cal, gram­mat­i­cal, and con­tex­tual rea­sons to lov­ingly but firmly disagree.

Here’s the verse in question:

As for you, you meant evil against me, but God meant it for good, to bring it about that many peo­ple should be kept alive, as they are today.

–Gen­e­sis 50:20 (ESV)

My posi­tion is that it is more plau­si­ble than not that the author of this verse did not intend it to teach Prov­i­den­tial Con­cur­rence or to con­sti­tute an exam­ple of it.

It seems the ini­tial ques­tion is what the word “meant” meant to the speaker in the nar­ra­tive, and to the author who recorded it. Here’s why I don’t think it meant “causally determined”:

1. Ancient Hebrews almost never discussed metaphysics.

Even the thor­oughly Calvin­ist Pres­by­ter­ian B. B. Warfield reluc­tantly admits, in the con­text God’s “irre­sistible prov­i­den­tial gov­ern­ment of the world” that “its meta­physics never come to explicit dis­cus­sion” in the Hebrew scrip­tures (Warfield, 1909).

2. Perhaps as a consequence of (1), ancient Hebrew almost never addresses metaphysics; it’s nearly always observational and pictorial.

The Tyn­dale Bible Dic­tio­nary explains that “Hebrew… con­cen­trates on obser­va­tion more than reflec­tion. That is, things are gen­er­ally observed accord­ing to their appear­ance as phe­nom­ena, not ana­lyzed as to their inward being or essence. Effects are observed but not traced through a series of causes… Hebrew is a pic­to­r­ial lan­guage…” (Elwell & Com­fort, 2001).

3. The meaning of the underlying Hebrew word חֲשָׁבָ (“Chashab” — Strong’s H2803), is nothing close to “metaphysically cause”.

Rather it’s “to think, account”, “to plan, devise, mean”, “to charge, impute, reckon”, “to esteem, value, regard”, or “to invent.” (Strong, 1996).

4. The use of the underlying Hebrew verb in nearly every other context in the Hebrew scriptures is quite obviously not intended to indicate metaphysical causation.

Usu­ally it’s some­thing like “regard” or “scheme” (for starters, this verb in the same mor­pho­log­i­cal state as it is in Gen. 50:20 can be found in 1 Sam. 18:25, 14:13, Esth. 8:3, 9:24–25, Job 35:2, Ps. 21:12, 14:3, 5, Isa. 33:8, 53:3–4, Jer. 11:19, 18:8, 48:2, 49:20, 30, 50:45, Lam. 2:8, and Amos 6:5).

5. The stem of the underlying Hebrew verb does not generally communicate causation.

This verb is found in the “Qal” form, that is, its unmod­i­fied root. While Qal verbs can com­mu­ni­cate actions…

6. …there is a stem in Hebrew entirely dedicated to indicating active causation (“Hiphil”).

There­fore the author could have specif­i­cally indi­cated cau­sa­tion if he had wanted to. For exam­ple, in Joshua 1:6b the author writes the word for “inherit” (נָחַל, “nachal”) with its Hiphil stem (תַּנְחִיל). Ren­dered in Eng­lish, this phrase there­fore reads “you shall cause this peo­ple to inherit the land that I swore to their fathers to give them.” (ESV).

Note that not even verbs with the Hiphil stem nec­es­sar­ily indi­cate meta­phys­i­cal cau­sa­tion, as God is not likely say­ing that Joshua will be in direct, “Prov­i­den­tially Con­cur­rent”, or “Com­pat­i­bal­ist” con­trol of every action and inten­tion of the peo­ple in order to lead them into the Promised Land (also cf. Bullinger, 1898, p. 821 for list of idiomatic active verb usages in Bib­li­cal Hebrew; I learned of Bullinger via Dave Miller & Kyle Butt, who call on Bullinger’s list for insight into God’s hard­en­ing of Pharoah’s heart–the other flag­ship exam­ple offered up in favor of the Calvin­is­tic for­mu­la­tion of divine prov­i­dence; I learned of the Miller & Butt arti­cle via Bran­don Ridley).

Even still, the author of Gen. 50:20 didn’t even use the causative mor­pho­log­i­cal device that was avail­able to him; he specif­i­cally chose instead a non-​​causative stem.

7. The “meant” translation of Darby, ASV, AV/​KJV, ESV, NASB95, NKJV, RSV and any others is metaphysically ambiguous.

If the trans­la­tors had thought that this verb meant “cause” in any meta­phys­i­cal sense, there are plenty of ways they could have indi­cated so in trans­la­tion (like, for exam­ple, “cause”).

The other trans­la­tions only worsen the Calvinist’s case:

• “devised” (Young’s Lit­eral Translation)

• “intended” (NET, NIV, TNIV, NLTNRSV)

• “planned” (God’ Word, NIrV)

• “turned” (New Cen­tury Version)

• “used” (The Message)

8. Many commentaries agree with me.

To take just one exam­ple, Pashall and Hobbs write that Joseph did not say to his broth­ers “…that God caused them to think evil against him, for they were respon­si­ble for their own thoughts. In his wis­dom and power, how­ever, God used their evil pur­poses to achieve his will.” (1972, p. 52).

On top of the above 8 points, I am still left with the hermeneu­ti­cal ques­tion about whether the speaker in the nar­ra­tive (Joseph) is intended to by the author of the nar­ra­tive (Moses) be a taken as meta­phys­i­cal author­ity to begin with. Else­where in the Hebrew scrip­tures we find our heros and fore­fa­thers say­ing and doing out­right bone-​​headed things. It’s not always clear which por­tions of the nar­ra­tives found in scrip­ture are meant to be didactic.

Now, you’re telling me that this is one of your flag­ship verses? This is among the clear­est, strongest evi­dence for your cen­trally impor­tant doc­trine that you can come up with?

I am no Hebrew scholar and really I am no scholar at all (at present). I con­sider this some­thing of a place­holder for future study. Even still, these seem like impor­tant points, and I would need them explained by any­one argu­ing for the Calvin­ist read­ing of Gen­e­sis 50:20.

What of the alter­na­tives to the Reformed doc­trine of Con­cur­rence? Among the many is the Clas­si­cal Armin­ian doc­trine of Con­cur­rence or the more philo­soph­i­cally stated “Molin­ism”, which at present looks far more coher­ent with the Bib­li­cal data to me.

Possible Straw Man in Kalam Rhetoric

1 Every­thing that begins to exist has a cause.

Craig says that “…if things really could come into being uncaused out of noth­ing, then it becomes inex­plic­a­ble why just any­thing or every­thing does not come into exis­tence uncaused from noth­ing. Why do bicy­cles and Beethoven and root beer not pop into being from noth­ing?” (p. 186, Craig & More­land, 2009).

But is it really all of that to deny that every­thing that begins to exist has a cause? Per­haps an objec­tor could hold that nearly every­thing that begins to exist has a cause. Per­haps there is exactly one thing that began to exist, but with­out cause. Per­haps the uni­verse as a whole is the counterexample.

Let me try to put this more for­mally. To deny (1) is to assert ~(1), or:

(It is not the case that) (every­thing that begins to exist has a cause.)

Which is equiv­a­lent to affirm­ing that there is at least one thing that began to exist with­out a cause. How­ever Craig seems to accuse every­one who holds ~(1) of holding:

(2) Every­thing that begins to exist could come into exis­tence with­out a cause.

But ~(1) ≉ (2).

Unless Craig can cough up an argu­ment for why accept­ing that there is at least one thing that begins to exist with­out a cause log­i­cally leads to the propo­si­tion that any­thing could come into being with­out cause, I think he com­mits a straw man fal­lacy here.

Nat­u­rally I agree with (1), but I am begin­ning to think the rhetoric he offers up in sup­port of it might be in need of some reform.

On the “Infinitely Remote Beginning” II

While I think Pop­per, Mackie, and Cobb’s ver­sions of the “Assumed Infi­nitely Remote Begin­ning Point” objec­tion com­mit the straw man fal­lacy against Kant and by exten­sion G. J. Whitrow and Craig, I stum­bled across another ver­sion of the objec­tion for­mu­lated by Quentin Smith against Whitrow, P. M. Huby, and Craig that I think suc­ceeds [10].

Actu­ally, I don’t believe that Smith’s objec­tion holds against Kant’s argu­ment or any of the well-​​formed argu­ments against the for­mu­la­tion of an infi­nite set by suc­ces­sive addi­tion. Rather I think that Smith points out more selec­tively where Craig, et al. make sup­ple­men­tal argu­ments, or where they rephrase their orig­i­nal argu­ments, in ways that do com­mit the error in question.

Here is the per­pe­trat­ing pas­sage in Whitrow:

Con­se­quently, if the chain of events form­ing the past of E is in finite, there must have occurred events that are sep­a­rated from E by an infi­nite num­ber of inter­me­di­ate events [11].

(I can­not locate where in Huby’s paper the per­pe­trat­ing pas­sage is alleged to be [12].) And Craig’s:

If the chain of events prior to E is infi­nite, then there must be an event O that is sep­a­rated from E by an infi­nite num­ber of inter­me­di­ate events [13].

On the “Infinitely Remote Beginning”

In 1781, a 57-​​year-​​old Pruss­ian man by the name of Immanuel Kant pub­lished a lit­tle work enti­tled “Kri­tik der reinen Ver­nunft” or “Cri­tique of Pure Rea­son”. In it he dis­cusses four antin­o­mies. Each com­prises a the­sis and an antithe­sis, for both of which Kant argues. The first antin­omy con­cerns space and time.

The the­sis of Kant’s first antin­omy is that space and time are finite. He begins by argu­ing for the fini­tude of the past. If the past were infi­nite, an infi­nite num­ber of moments would have elapsed prior to any given moment. But an infi­nite series can never be com­pleted through suc­ces­sive syn­the­sis (time ticks by one moment at a time, and you’ll never form an infi­nite col­lec­tion of moments that way). So the pass­ing of every moment proves that its own tem­po­ral past is finite. In argu­ing this way, Kant antic­i­pates William Lane Craig’s sec­ond argu­ment for the sec­ond premise of his Kalam [6].

I have come across sev­eral philoso­phers who have each inde­pen­dently raised the same objec­tion to Kant’s argu­ment or some ver­sion of it [7]. The objec­tion says that Kant pre­sup­poses that there is a begin­ning point in time, but one that is infi­nitely far away. In this post I first try to imag­ine what it would look like to posit a first moment in time that occurred infi­nitely long ago. Then I explore why posit­ing a first moment in time like this might inval­i­date Kant’s argu­ment. Finally I exam­ine why one of Kant’s crit­ics says his the­sis does in fact com­mit this error, and I find that the crit­i­cism attacks only a straw-​​man ver­sion of Kant’s the­sis by ignor­ing the tensed nature of time.

An easy way to think about time is as a rope with tick marks on it. Each tick mark is labeled with an inte­ger, and the present moment is labeled “0”. So the moment prior to the present one is “-1″, and the next moment after the present will be “+1″. Of course one dif­fer­ence is that a rope is sta­tic and doesn’t move, so we can look at the full length of one all at once (or if it is a long rope, we can at least look at more than one “moment” on the rope at a time). Whereas moments ticks by one at a time; the past moments are gone and we can­not directly expe­ri­ence them, and future moments have not yet come to pass. So let’s imag­ine that we are hold­ing the rope and mov­ing our grip on it from one tick mark to the next dur­ing each moment in time that elapses in order to sym­bol­ize to our­selves what the pass­ing of time is like. (At present we are grip­ping “0”.)

Kant’s the­sis is that this num­ber line-​​rope can­not extend infi­nitely back­wards. Now, what would it look like for Kant to be assum­ing that there is a begin­ning point in time, but one that is infi­nitely far away? It would be like imag­in­ing that the rope has a last neg­a­tive num­ber on it, mark­ing one of the ends of the length of rope, but that this end is infi­nitely far away from 0.

Why would think­ing this way about the rope cause prob­lems? For starters, if Kant is try­ing to prove that the rope has a finite num­ber of neg­a­tive inte­gers, it would seem question-​​begging to assume from the get-​​go that the rope has an end on that side of it. Nat­u­rally, if the rope has an end, then it would be pos­si­ble to reach that end by count­ing moments back­ward to it. This is com­mon sense — to be finite is to be bounded, and to be infi­nite is to be unbounded. Assum­ing a bound prior to argu­ing for fini­tude is question-​​begging.

Sim­i­larly, if there is a begin­ning point in time, it would be impos­si­ble for such a point to be infi­nitely far away. To put this more math­e­mat­i­cally: for every pair of col­in­ear points x and y, the dis­tance between x and y is finite. So if the begin­ning point of time is a point and the present moment is also a point, then the span between the begin­ning point of time and the present moment is finite.

But beg­ging the ques­tion isn’t the only prob­lem. Kant’s argu­ment is meant to be some­thing of a reduc­tio ad absur­dum of the notion of an infi­nite past, that is it assumes for the sake of argu­ment that the past is infi­nite and then shows how such a view would lead to absur­dity in order to dis­credit it. But if Kant is assum­ing that the past has a start­ing point, and such is tan­ta­mount to assum­ing the past is finite, and he is assum­ing that the past is infi­nite, which is tan­ta­mount to assum­ing that the past has no start­ing point, then he is tak­ing the very two con­tra­dic­tory propo­si­tions between which he is try­ing to adju­di­cate, as premises in his argu­ment. And that’s just wrong.

On the other hand, if one doesn’t pre­sup­pose that there was a first moment in time, accord­ing to this objec­tion, one can see that the past may stretch back­ward infi­nitely. This would be like our num­ber line-​​rope hav­ing no end, no largest neg­a­tive inte­ger, or if you pre­fer no begin­ning. In this case, though any two points on the rope will only be a finite dis­tance away from one another, the points will go on into the neg­a­tive num­bers for­ever with­out stopping.

So why think that Kant pre­sup­poses a begin­ning point, before he even gets started try­ing to rea­son to the fini­tude of the past? Charles Cobb says that to say that “no infi­nite series can be com­pleted… is false unless we sup­pose that the infi­nite series has a begin­ning” (empha­sis mine) [8]. A few lines later he repeats that “to say that a series is not infi­nite because it has a last term is to say that it has a first term” (empha­sis mine again).

What does he mean by this? If a series we know to be infi­nite “has a first term”, he says “and each term has a def­i­nite suc­ces­sor, there can be no last term; for exam­ple, the infi­nite series 0, 1, 2, 3… begin­ning with zero, can have no last term” (since we know it to be infi­nite). But if we turn it around and con­sider “the [infi­nite] series …3, 2, 1, 0″ we can see that it has a last term, but is nev­er­the­less infi­nite. “There­fore”, he says, “to say that a series is not infi­nite because it has a last term is to [assume] that it has a first term” (empha­sis mine). He gives the exam­ple “3, 2, 1, 0, which has a last term zero, [and] is not infi­nite because it has a first term, 3″.

What are we to say to this? Is Kant’s posi­tion really that any series that has a last term must be finite? Would he deny that the set of all neg­a­tive inte­gers is infi­nite? I am inclined to say no. The dif­fer­ence between a series of sta­t­i­cally exist­ing mem­bers that has a last term, such as the series of neg­a­tive inte­gers, and the series of moments in time is that moments in time elapse [9]. Kant’s the­sis is that a series of elaps­ing mem­bers that has a last term can­not be finite — Cobb, et al. com­pletely ignore Kant’s “suc­ces­sive syn­the­sis” clause and thus mis­un­der­stand his argu­ment. Kant doesn’t presup­pose that a series with a last term must have had a first term, he demon­strates that an elaps­ing series with a last term must have had a first term, and thus must be past-​​finite.

Why Frankfurt’s Examples Aren’t Counter

Long ago, while under­go­ing the most pro­found par­a­digm shift I’ve expe­ri­enced to date, I ran across what I have since come to under­stand are noto­ri­ous coun­terex­am­ples to the view that moral respon­si­bil­ity is grounded in the abil­ity for an agent to choose between alter­na­tives. Harry G. Frank­furt, who also wrote a hilar­i­ous work called “On Bull­shit”, pro­posed these so-​​called coun­terex­am­ples to the view he called the “Prin­ci­ple of Alter­nate Pos­si­bil­i­ties” in a paper in the year 1969 [5]. The below block quotes are taken from that paper with the excep­tion of the quote from a related Wikipedia article.

Frank­furt­ian coun­terex­am­ples have run in and out of my mind since then and I recently got the inspi­ra­tion to read Frankfurt’s orig­i­nal paper and set down in writ­ing my basic crit­i­cism of it. If I were to ever get some­thing like this into a pub­lish­able for­mat, it would require research into how these thought exper­i­ments have been treated in the lit­er­a­ture since ’69, but I think that get­ting out my per­sonal reac­tion to the orig­i­nal paper is suit­able blog­ging material.

Wikipedia does a fair job of describ­ing Frank­furt­ian coun­terex­am­ples to the Prin­ci­ple of Alter­nate Pos­si­bil­i­ties and their significance:

The prin­ci­ple of alter­na­tive pos­si­bil­i­ties forms part of an influ­en­tial argu­ment for the incom­pat­i­bil­ity of respon­si­bil­ity and causal deter­min­ism, as detailed below:

(1) PAP: An agent is respon­si­ble for an action only if said agent is free.
(2) An agent is free only if causal deter­min­ism is false.
(3) There­fore, an agent is respon­si­ble for an action only if causal deter­min­ism is false.

Tra­di­tion­ally, com­pat­i­bilists (defend­ers of the com­pat­i­bil­ity of moral respon­si­bil­ity and deter­min­ism, like Alfred Ayer, Wal­ter Ter­ence Stace[,] and Daniel C. Den­nett) try to reject premise two, argu­ing that, prop­erly under­stood, free will is not incom­pat­i­ble with determinism…

From the PAP def­i­n­i­tion “a per­son is morally respon­si­ble for what he has done only if he could have done oth­er­wise”, Frank­furt infers that a per­son is not morally respon­si­ble for what he has done if he could not have done oth­er­wise – a point with which he takes issue: our the­o­ret­i­cal abil­ity to do oth­er­wise, he says, does not nec­es­sar­ily make it pos­si­ble for us to do otherwise.

Frankfurt’s exam­ples are sig­nif­i­cant because they sug­gest an alter­na­tive way to defend com­pat­i­bil­ism, in par­tic­u­lar by reject­ing the first premise of the argu­ment. Accord­ing to this view, respon­si­bil­ity is com­pat­i­ble with deter­min­ism because respon­si­bil­ity does not require the free­dom to do otherwise.

Frankfurt’s exam­ples involve agents who are intu­itively respon­si­ble for their behav­iour even though they lack the free­dom to act otherwise.


While read­ing his paper, I was sur­prised by how much intu­itive appeal and pop­u­lar­ity among philoso­phers that Frank­furt con­cedes up front to Incompatibilism:

A dom­i­nant role in nearly all recent inquiries into the free-​​will prob­lem has been played by a prin­ci­ple which I shall call “the prin­ci­ple of alter­nate pos­si­bil­i­ties.” This prin­ci­ple states that a per­son is morally respon­si­ble for what he has done only if he could have done oth­er­wise… no one, how­ever, seems inclined to deny or even to ques­tion that the prin­ci­ple of alter­nate pos­si­bil­i­ties… is true. It has gen­er­ally seemed so over­whelm­ingly plau­si­ble that some philoso­phers have even char­ac­ter­ized it as an a pri­ori truth. Peo­ple whose accounts of free will or of moral respon­si­bil­ity are rad­i­cally at odds evi­dently find in it a firm and con­ve­nient com­mon ground upon which they can prof­itably take their oppos­ing stands.

In the mid­dle of his intro­duc­tion, he even goes so far as to con­cede that

…a per­son who has been coerced to do some­thing did not do it freely and is not morally respon­si­ble for hav­ing done it.

This made me curi­ous as to how exactly he could reject PAP. I found myself sur­prised yet again at the end of the paper, when he merely pro­poses an amend­ment to the prin­ci­ple, rather than an out­right rejec­tion of it alto­gether. In some ways this con­di­tioned me for a more char­i­ta­ble read­ing of Frank­furt than I was antic­i­pat­ing, but in other ways it laid the ground­work for a stronger argu­ment on Frankfurt’s part. The less you are try­ing to prove, the stronger your case for it can be. And if his amended prin­ci­ple plays to the Com­pat­i­bilist, then his paper has the chance of being more dif­fi­cult to cri­tique than I had per­haps ini­tially expected.

At any rate, after the up-​​front pleas­antries, Frank­furt pulls no punches in ham­mer­ing us with his thesis:

But the prin­ci­ple of alter­nate pos­si­bil­i­ties is false.

He goes on:

A per­son may well be morally respon­si­ble for what he has done even though he could not have done otherwise.

…there may be cir­cum­stances that con­sti­tute suf­fi­cient con­di­tions for a cer­tain action to be per­formed by some­one and that there­fore make it impos­si­ble for the per­son to do oth­er­wise, but that do not actu­ally impel the per­son to act or in any way pro­duce his action. A per­son may do some­thing in cir­cum­stances that leave him no alter­na­tive to doing it, with­out these cir­cum­stances actu­ally mov­ing him or lead­ing him to do it-​​without them play­ing any role, indeed, in bring­ing it about that he does what he does.

Here is how he sets it up:

…Jones decides for rea­sons of his own to do some­thing, then some­one threat­ens him with a very harsh penalty (so harsh that any rea­son­able per­son would sub­mit to the threat) unless he does pre­cisely that, and Jones does it. Will we hold Jones morally respon­si­ble for what he has done? I think this will depend on the roles we think were played, in lead­ing him to act, by his orig­i­nal deci­sion and by the threat.

After this he tin­kers with some vari­ables to describe two dis­tinct but sim­i­lar sit­u­a­tions, which finally lead up to his main counterexample:

…Now con­sider a third pos­si­bil­ity. Jones was nei­ther stam­peded by the threat nor indif­fer­ent to it. The threat impressed him, as it would impress any rea­son­able man, and he would have sub­mit­ted to it whole­heart­edly if he had not already made a deci­sion that coin­cided with the one demanded of him. In fact, how­ever, he per­formed the action in ques­tion on the basis of the deci­sion he had made before the threat was issued. When he acted, he was not actu­ally moti­vated by the threat but solely by the con­sid­er­a­tions that had orig­i­nally com­mended the action to him. It was not the threat that led him to act, though it would have done so if he had not already pro­vided him­self with a suf­fi­cient motive for per­form­ing the action in question.

…I think we would be jus­ti­fied in regard­ing his moral respon­si­bil­ity for what he did as unaf­fected by the threat even though, since he would in any case have sub­mit­ted to the threat, he could not have avoided doing what he did.

…When we excuse a per­son who has been coerced, we do not excuse him because he was unable to do oth­er­wise. Even though a per­son is sub­ject to a coer­cive force that pre­cludes his per­form­ing any action but one, he may nonethe­less bear full moral respon­si­bil­ity for per­form­ing that action.

…For the irre­sistibil­ity of the threat to which Jones is sub­jected might well be taken to mean that he can­not but per­form the action he per­forms. And yet the threat, since Jones per­forms the action with­out regard to it, does not reduce his moral respon­si­bil­ity for what he does.

The first issue that strikes me about Frankfurt’s exam­ple is the sharp divi­sion it makes between cog­ni­tive states and behav­ior. Jones is intu­itively respon­si­ble for his actions pre­cisely because he orig­i­nally freely decided upon them (he may not have been able to behave dif­fer­ently, but he was able to will dif­fer­ently). And this is all the Prin­ci­ple of Alter­nate Pos­si­bil­i­ties addresses to begin with. The details that Frank­furt adds all occur after the free choice was made and if you will remem­ber, he even admits that

…a per­son who has been coerced to do some­thing did not do it freely and is not morally respon­si­ble for hav­ing done it.

So surely if Jones’ cog­ni­tive states were them­selves causally deter­mined we wouldn’t intuit that he is morally respon­si­ble for them. So the “Com­pat­i­bilist” world that Frank­furt paints is one which is deter­min­is­tic only in regards to exter­nal behav­iors, but in which agents some­how enjoy free­dom of the will in terms of their own cog­ni­tive states and are thus morally respon­si­ble for their actions only when their choices line up with them (which could even be believed to hold for every action, as I take Leibniz’s Occa­sion­al­ism to assert). But such a world is hardly deter­min­is­tic, as there are things in it which are not causally deter­mined (cog­ni­tive states).

There­fore it seems that not only do Frankfurt’s exam­ples fail to actu­ally counter the Prin­ci­ple of Alter­nate Pos­si­bil­i­ties, but they also fail to faith­fully rep­re­sent a full-​​blown Com­pat­i­bilist metaphysic.

Post Script
Inci­den­tally, I think the alter­nate pos­si­bil­i­ties in Frankfurt’s work cor­re­spond to what I else­where call “real” pos­si­bil­i­ties: those alter­nate meta­phys­i­cal pos­si­bil­i­ties that dis­tin­guish worlds which share phys­i­cal prop­er­ties and a full his­tory con­gru­ent with our own up to a given point, and between which free agents can choose. I take Frank­furt to be say­ing that there can be moral respon­si­bil­ity in worlds lack­ing real behav­ioral possibilities.

FactChecking Mike D on Kalam

Self-​​proclaimed apos­tate and per­sonal trainer Mike D cri­tiques the argu­ments Dr. William Lane Craig used in a debate with noted jour­nal­ist Christo­pher Hitchens not long ago. How­ever Mike D’s blog reflects igno­rance of the depth of Craig’s schol­arly research in sev­eral areas and a gen­eral atti­tude of dog­ma­tism. Today I am inter­ested in cor­rect­ing this par­tic­u­lar post’s pos­ture toward Craig’s Kalam Cos­mo­log­i­cal Argument.

Craig’s for­mu­la­tion of the cos­mo­log­i­cal argu­ment goes like this:

1. Every­thing that begins to exist has a cause.
2. The uni­verse began to exist.
3. There­fore the uni­verse has a cause.

Mike D’s first objec­tion is a rebut­ting defeater of the sec­ond premise, his sec­ond objec­tion is sim­ply a mis­un­der­stand­ing of the con­clu­sion, and his third objec­tion is a very unclear rebut­ting defeater of the first premise. He never dis­putes the valid­ity of the logic.

He first argues that the uni­verse could exist in “a state in which time [func­tions] non-​​linearly as another dimen­sion of space, negat­ing… prob­lems with infi­nite regres­sions of time”. Some­thing like Mike D’s posi­tion was orig­i­nally advanced by John McTag­gart Ellis McTag­gart (yes, that’s his real name) in his famous 1908 arti­cle in Mind, “The Unre­al­ity of Time”, as “B-​​series” time. This view of time would see any begin­ning of the uni­verse like the begin­ning of a yard stick, whose length is tense­less. Most laypeo­ple how­ever, are unknow­ing believ­ers in what McTag­gart called “A-​​series” time, and pas­sively inter­pret their expe­ri­ence of tensed time as real.

Craig is aware that his view pre­sup­poses A-​​series time. In fact, he has done exten­sive schol­arly work, much of which is avail­able online, in under­cut­ting the B-​​series view and advanc­ing an A-​​series view (includ­ing a full book-​​length treat­ment of each). The qual­ity of his research is reflected by the fact that he was invited to pre­side over the Phi­los­o­phy of Time Soci­ety from 1999–2006. To sim­ply point out the well-​​known fact that there is a view of time which would pre­clude Craig’s argu­ment, and then declare Craig wrong with­out any fur­ther com­ment on the mat­ter, is both igno­rant and dog­matic [1][2].

Mr. D’s sec­ond “objec­tion” is sim­ply a mis­un­der­stand­ing of what it is that the argu­ment is try­ing to prove. Accord­ing to Mike D, the “sec­ond prob­lem… is that [Craig’s] argu­ment fails to answer why, even if a first cause is nec­es­sary, it must be ‘God’” [3]. Here Mr. D mis­un­der­stands the claim made by the argu­ment, which is only that the uni­verse has a cause, not that the cause is God. Inci­den­tally, Craig is always pre­pared with other argu­ments on hand that build on his kalam by attempt­ing to recover from it cer­tain attrib­utes that must be true of a uni­ver­sal cause [4].

Why didn’t Craig go into finer tech­ni­cal detail about these issues? Because Hitchens has no for­mal train­ing in either sci­ence or phi­los­o­phy, and demon­strated a pro­found igno­rance of the tech­ni­cal issues that have bear­ing on the dis­cus­sion. In fact, Craig is often frus­trated at just how shal­low his debates are, because his inter­locu­tors rarely bring up any meaty objec­tions at all (as was most cer­tainly the case with Hitchens, whom Craig “spanked… like a fool­ish child”, accord­ing to the infa­mous “Com­mon Sense Athe­ism” post on the debate).

Mr. D’s final objec­tion is unclear, but I think that it may be under­stood as a rebut­ting defeater of the first premise (though parts of it seem to assume he is right about the sec­ond premise being false). In it he ini­tially asserts that causal­ity requires space and lin­ear time. Yet he pro­vides no argu­ment what­so­ever for his meta­phys­i­cal dogma. Are we to take it on faith?

Next, while his first objec­tion defended a view of time which is non-​​linear, this final objec­tion claims that ‘lin­ear’ time is needed for causal­ity to occur. Are we to take it then that Mr. D denies the real­ity of causal­ity entirely?

After this he says that if the uni­verse were caused, we would all actu­ally be igno­rant of how causal­ity works out­side of the uni­verse. But if that’s the case, how does ‘he’ know that causal­ity requires things pre­cluded by Craig’s view?

Mr. D says it’s a “fal­lacy then to talk about a time ‘before’ the uni­verse”, but Craig does no such thing. His view is that “with­out” (not “before”) the uni­verse, God is time­less. And on this mat­ter too, Craig has writ­ten exten­sively at the schol­arly level, includ­ing a book-​​length treat­ment and dozens of arti­cles, many of which are avail­able online. Should Mr. D ever read any of them, he may remain in dis­agree­ment, but his post reflects only igno­rance of Craig’s research.

An hon­est piece describ­ing the thoughts of an arm­chair philoso­pher in process would be one thing, but Mr. D’s post is packed with dog­matic rhetoric and scathing dis­missals of a man whose peers regard him as eas­ily falling within the top 1 per­cent of prac­tic­ing philoso­phers in the West­ern world.

Girardeau’s Calvinism Reviewed: Introduction

While work­ing Powell’s City of Books over dur­ing one of my days on vaca­tion, I acquired a hand­some vol­ume of “Calvin­ism and Evan­gel­i­cal Armini­an­ism” by John Lafayette Girardeau, a remark­able man of French Huguenot and Scot­tish Pres­by­ter­ian descent, who pas­tored slaves and slave own­ers and taught at the orig­i­nal Colum­bia sem­i­nary in the South, and voiced the sole “nay” in the vote to seg­re­gate the South­ern Pres­by­ter­ian Church in 1874.

Some­one named James M. Bul­man, in the intro­duc­tion, cites “one of the Hodges” as admir­ing this book as “the most con­vinc­ing argu­ment for Calvin­ism to be seen any­where”. Mr. Bul­man agrees, cit­ing “lit­er­ary crafts­man­ship befit­ting French extrac­tion; and some­thing of the gen­uinely ora­tor­i­cal, pul­sat­ing with warmth of reli­gious devo­tion”, as well as the unique abil­ity to fully com­mu­ni­cate the strength of a the­o­log­i­cal sys­tem afforded by defend­ing one par­tic­u­lar, fully-​​orbed view (in this case Girardeau’s par­tic­u­lar brand of fed­er­al­ist sub­lap­sar­ian Calvin­ism) against its strongest con­tender (Wes­leyan, or “Evan­gel­i­cal” Arminianism).

Hav­ing once held to all five of the points upheld at the Synod of Dort, and still hold­ing to a unique respect for, and par­tial agree­ment with, the Reformed tra­di­tion, I have decided to attempt a char­i­ta­ble review and exact­ing cri­tique of Girardeau’s treat­ment of the objec­tions to Uncon­di­tional Elec­tion from God’s good­ness and from man’s moral respon­si­bil­ity. These treat­ments only make up a frac­tion of the book, but they are for­mi­da­ble in their own right and are the sec­tions I am most inter­ested in.

My Modality Series & Blogs Generally

My recent series on modal­ity (which ter­mi­nated in this post) was never meant to con­clu­sively prove the com­pos­si­bil­ity of fore­knowl­edge and lib­er­tar­ian free will. It was only to resolve the appar­ent con­tra­dic­tion cre­ated by the facts that fore­knowl­edge requires that there be no alter­nate pos­si­bil­i­ties, yet free will requires that there be alter­nate pos­si­bil­i­ties. (If there are alter­nate pos­si­ble futures, how can it be known which will occur? And if the future is known, how could mat­ters unfold any other way?) The res­o­lu­tion is seen by under­stand­ing that the type of alter­na­tives required to be lack­ing to ground fore­knowl­edge is dif­fer­ent than the type of alter­na­tives required to be present to ground free will. By show­ing that there are dif­fer­ent modes of pos­si­bil­ity I believe I nuanced the tra­di­tional Prin­ci­ple of Alter­nate Pos­si­bil­i­ties in order to show that, as far as this one par­tic­u­lar mat­ter is con­cerned, fore­knowl­edge and free will are not mutu­ally meta­phys­i­cally exclusive.

Being a blog and not a pub­lished work, my series did not posi­tion itself into the broader ongo­ing dia­logue in the pub­lished lit­er­a­ture by way of cita­tions, though it did make use of some con­ven­tional terms (in some cases inten­tion­ally flex­ing them). I think this is a good use for a blog: to set down some ideas whose rela­tion to the ideas of oth­ers is not yet fully under­stood, which are not yet fully devel­oped, which lack rhetor­i­cal pol­ish, or for other rea­sons are not close enough to pub­li­ca­tion to be kept under wraps and about which a thinker is open to pub­lic feedback.

Too late for an Independence Day post?

And God saw every­thing that he had made, and behold, it was very good.

-Gen­e­sis 1:31a

One func­tion of this blog is to doc­u­ment things I have inklings about in order to cre­ate place­hold­ers for future study. And this par­tic­u­lar post is to do just that.

Today I was read­ing in Gen­e­sis and the words “saw” and “was” popped out at me. After each cre­ation day, the author says God “saw” that what He had cre­ated “was” good. God didn’t cre­ate things that had no intrin­sic value and then sim­ply declare them good, like we arbi­trar­ily declare worth­less green bills valu­able. He cre­ated good things. And the best part about the nar­ra­tive is that when it gets to the cre­ation of mankind, it says that God saw every­thing that He had made, and it was very good. And surely, God Him­self is good! And thus beings fash­ioned after His like­ness would be as well!

Could an exeget­i­cal study of the Gen­e­sis cre­ation nar­ra­tive lend Bib­li­cal war­rant for a the­ory of inher­ent rights, as opposed to inputed rights?

We hold these truths to be self-​​evident, that all men are cre­ated equal, that they are endowed by their Cre­ator with cer­tain unalien­able Rights, that among these are Life, Lib­erty and the pur­suit of Happiness.

-Dec­la­ra­tion of Independence