It seems to me, upon reflection, that there may be a problem with presuming both that God has middle knowledge, and that we have free will in the libertarian sense. If we do have free will in the libertarian sense then our decisions when acting in a free capacity are literally not determined by anything other than ourselves (we act as a special kind of ‘first mover’, though not unmoved ourselves). However, the Molinist thesis defended by Alvin Plantinga, and William Lane Craig (perhaps the two most able defenders of Molina today) requires that God know counterfactuals such as what I would freely do in X circumstances. For instance, Craig might suggest that if the Canaanites had been allowed to flourish they would have caused much more death and devastation in the world, and thus God acted in a morally prudent way when he decided to kill the Canaanites off (or have them killed off). Craig has also suggested that this view allows one to maintain that those people who lived life never hearing about the Gospel were in those circumstances precisely because God knew the counterfactual to be true that if they were to have heard the Gospel they would have rejected it. This position seems crudely proposed by Craig, however the real problem is that maintaining this molinist view and libertarian free will seems to mean entertaining incompatible positions.
However, in rejecting molinism on these grounds, I would also be rejecting the idea of God’s middle knowledge (which is just his knowledge of all counterfactuals). This compounds the problem, however, since counterfactuals are typically taken to be coherent modal concepts, and yet my objection here would be a case against counterfactual conditionals (at least counterfactual conditionals which are of the form C⊃B where C stands for the whole situational circumstance of some free agent acting in a morally free capacity, and B stands for some choice corresponding to the choice of the morally free agent acting in a morally free capacity in that situational circumstance).
The principle philosophical problem with counterfactuals usually advanced is that there is nothing which grounds them. As Wikipedia reads:
“The grounding objection is at present the most debated objection to Molinism, and often considered the strongest.”
For one such argument, see Pruss here. Wikipedia continues:
“Alvin Plantinga responds to the grounding objection by saying “It seems to me much clearer that some counterfactuals of freedom are at least possibly true than that the truth of propositions must, in general, be grounded in this way.” William Lane Craig follows up on this by pointing out the burden of proof the grounding objector bears. The grounding objection “asserts that there are no true counterfactuals about how creatures would freely act under any given set of circumstances. This assertion is no mere ostensibly undercutting defeater of Molinism, but a putatively rebutting defeater. It makes a bold and positive assertion and therefore requires warrant in excess of that which attends the Molinist assumption that there are true counterfactuals about creaturely free actions.” Latter, Craig points out “Anti–Molinists have not even begun the task of showing that counterfactuals of creaturely freedom are members of the set of propositions or statements which require truth–makers if they are to be true.”
In response to Plantinga, I would simply say that I do not share his intuition. It seems to me much more intuitive to say that all truths are grounded than to say that some counterfactuals are possibly true. This intuition makes Craig’s follow up even harder for me to swallow, as it seems self-evident that true propositions require truth-makers.
“The Molinist position, exemplified by Craig in the preceding paragraph, is that a choice can be free, and yet the way in which an agent will make that choice can be known apart from observation of the actualized choice itself (and even apart from the actualization of the choice entirely)”
In order to understand this discussion in depth it will be useful, at this point, to turn to the Catholic Encyclopedia on Molinism:
Objects or events viewed merely as possible, God is said to apprehend by simple intelligence (simplex intelligentia). Events which will happen He knows by vision (scientia visionis). Intermediate between these are conditionally future events–things which would occur were certain conditions fulfilled. God’s knowledge of this class of contingencies they term scientia media. For instance Christ affirmed that, if certain miracles had been wrought in Tyre and Sidon, the inhabitants would have been converted. The condition was not realized, yet the statement of Christ must have been true. About all such conditional contingencies propositions may be framed which are either true or false–and Infinite Intelligence must know all truth. The conditions in many cases will not be realized, so God must know them apart from any decrees determining their realization. He knows them therefore, this school holds, in seipsis, in themselves as conditionally future events. This knowledge is the scientia media, “middle knowledge”, intermediate between vision of the actual future and simple understanding of the merely possible. Acting now in the light of this scientia media with respect to human volitions, God freely decides according to His own wisdom whether He shall supply the requisite conditions, including His co-operation in the action, or abstain from so doing, and thus render possible or prevent the realization of the event. In other words, the infinite intelligence of God sees clearly what would happen in any conceivable circumstances.
~New Advent Catholic Encyclopedia, Free Will
Now, this scientia media (middle knowledge), I had always thought was unnecessary if only one could argue that God’s knowledge of counterfactuals comes precisely from quantifying over all logically possible worlds to find what description of counterfactuals best fits the widest range of logically possible worlds. For instance, if God knew that in some set of circumstances C I would choose B in place of A in more logically possible worlds than those in which I would choose A in place of B, then God can have at least probabilistic counterfactual knowledge that ‘if C then B’. Maybe this kind of quasi-middle knowledge is sufficient for God to make prudential decisions, such as to have some infant die before she/he becomes a mass murderer (or very likely becomes a mass murderer) – or more significantly for Christian theology: that the Canaanites’ survival would (likely) have been the undoing of so many other peoples, including perhaps the Jewish people. Wikipedia, notes an objection:
“So instead of truths of the following sort: “God knows that in circumstance C creature X will freely do A” God knows truths of this sort: “God knows that in circumstances C creature X would probably do A.” Yet, as Edward Wierenga has pointed out, probable counterfactuals are also contingent truths and fall victim to the same grounding objection.
However, I’m not sure that they do fall to this objection. The probabilistic knowledge of counterfactuals is grounded in God’s knowing, or being aware of, all logically possible worlds. That seems sufficient for God to know the likelihood of anything or everything, even if there are infinitely many logically possible worlds. Logically possible worlds themselves are grounded in the contingent nature of the actual world.
The problem left over would be how to explain Jesus’ statement cited above in the Catholic Encyclopedia. How would Jesus know that if he had done the same miracles in another place/time, then his audience would have been more receptive to his message? Certainly it seems that such a statement must be true, but in virtue of what can it be true? Perhaps Jesus has the knowledge that the set of all logically possible worlds in which many, most or all of the people in Tyre and Sidon would have believed in his message is just significantly larger than the set in which they would not have.
Another thought is that, on my view, God would know some counterfactuals to be true (some counterfactuals would be grounded) but not the molinist-type counterfactuals. Thus God might now that if I were to choose B, then C would occur in a determined sense, where C is not a choice but some event which is in part a consequence of B. However, explaining Jesus’ statement in these terms seems incredulous, since the ‘C’ in question was the act of conversion, which is necessarily an act of the heart, and freely chosen. One could perhaps suggest that Jesus was speaking about a sort of external conversion rather than a real internal one, but then it seems that the relevance of his statement is diminished too significantly.
Or perhaps counterfactuals can be grounded in some way in the world at which they are true – for instance if in some world W it is the case that, as things are now, if things now are analogous to the way they are in another world W* then in all worlds relevantly similar to W* (or fitting the description of W*) some counterfactual X is true. On this view, perhaps the world in which the counterfactual proposition is uttered in some way grounds the truth of that same counterfactual (so that in other logically possible worlds the very same counterfactual would be false). The difficulty here is that I can’t quite see how to demonstrate this.
It is interesting to note, finally, that Jesus’ statement that there are people in Tyre and Sidon who would have believed had they been presented with the same miracles he presented to others undermines the proposal that people who live without hearing the Gospel are all people who would not have accepted it if they had heard it.