Review of Edwards’ Arguments Against LFW

, posted by Godismyjudge

This article defines Libertarian Free Will (LFW) and then reviews Jonathan Edwards’ arguments against LFW based both on causation and divine foreknowledge.


LFW is the idea that man is able to choose otherwise than he will choose. It’s contrasted with Compatiblism Free Will (CFW), the idea that free will and determinism are compatible. These are alternative views of the will; both can’t be true about a persons’ will at the same time.

The descriptions “libertarian” and “free” distinguish LFW from CFW, but are otherwise redundant. For those holding to LFW, the will is always at liberty, and is always free, else it’s not a will. Arminius put it: “the will cannot be forced”.
The bible says people we have wills and make choices. Since LFW and CFW are alternative views and LFW is reducible to biblical phrase, “the will”, the question is which is right: LFW or CFW? If LFW is coherent and CFW is incoherent, LFW is biblical and CFW isn’t. Jonathan Edwards realized that the converse is true as well, and that’s why the bulk of “the Freedom of the Will” attempts to demonstrate the incoherence of LFW. I argued that Edwards’ view is inconsistent here, and plan on addressing Edwards’ arguments latter. But since LFW is about to undergo a detailed inspection by Edwards, I thought it might be helpful if we first described LFW. Contrast is helpful, so let’s start with what LFW is not.
What LFW is Not
LFW is not:
1) The ability to choose between good and evil – the fall disabled man from doing good, but fallen man is able to choose among evil options
2) The ability to avoid the consequences of our choices
3) The ability of our body to accomplish anything we choose – choices are mental resolutions which start the bodies action, whether or not the action is completed. If someone pulls a gun on you and says your money or your life and you attempt to escape, you made a choice to escape whether you get away or not.
4) The ability to choose both alternatives simultaneously – you can’t have your cake and eat it too
5) The ability to change the past – causation works forward in time – there’s no going back
6) The ability to create ex nihilo – We require God’s providential concurrence with every phase of choice
7) The ability to falsify God’s foreknowledge – We can, but will not do the opposite of what God foreknows. Foreknowledge and the future event are logically related, so if we suppose we will do the opposite, we must also suppose God foreknew that event.
What LFW Is
Choice is the action of the will selecting one alternative over another. The will is the faculty within man’s soul that’s able to choose either alternative. Being able to choose either option implies both options are possible, which implies neither option is necessary. When we say the will is free, it’s free from necessity, and it’s free to choose either alternative. With respect to time, the ability of the will precedes the will’s choosing. At one moment we have a choice, the next moment we actually choose.


Brief Outline of Edwards’ Arguments in Part II of Freedom of the Will
Edwards attacks LFW in two broad categories: causation and divine foreknowledge. Under causation, Edwards argues that LFW either leads to an infinite regression of causes or is an action without a cause. Edwards then argues that actions without causes are absurd because: 1) they would violate the common sense idea that nothing ever comes to pass without a cause, 2) then we wouldn’t be able to reason from cause to effect, 3) all proof of God’s existence is taken away, and 4) actions produced by a causeless cause would be both random and irrational, and therefore not a basis of moral accountability.

Infinite Regression of Causes or Causeless
Edwards first argues under LFW choices can’t have causes. Asserting they do “brings us directly to a contradiction: for it supposes an act of the Will preceding the first act in the whole train, determining the rest; or a free act of the Will, before the first free act of the Will.” If our choices have causes, either whatever causes our choices was causeless or it had a cause. And this argument repeats till we either reach something causeless or we continue with an infinite regression of causes.

My Response
The way Edwards defines causes our choices don’t have causes, but the way I define causes, they do. I don’t think Edwards defines cause correctly or consistently. In part II.III Edwards defines cause as : I sometimes use the word Cause, in this inquiry, to signify any antecedent, either natural or moral, positive or negative, on which an Event, either a thing, or the manner and circumstance of a thing, so depends, that it is the ground and reason, either in whole, or in part, why it is, rather than not; or why it is as it is, rather than otherwise. In this sense our choices have causes. But in part II.VIII Edwards says: If there are some events which are not necessarily connected with their causes, then it will follow, that there are some things which come to pass without any cause and also, if the effect be not necessarily connected with the cause… this instance is a proof, in fact, that the influence of the cause is not sufficient to produce the effect. For if it had been sufficient, it would have done it.

On the one hand Edwards defines cause broadly as anything on which effects have complete or partial dependence, but on the other hand Edwards defines cause narrowly as having a necessary connection with the effect and being sufficient to produce the effect. That’s the problem. Under the first definition, our choices have causes, under the second they don’t.

Choices have indeterministic causes, not sufficient causes. Indeterministic causation is a tricky subject, but there are a few things we can say about it. First off, indeterministic causes do not predetermine the effects. The effect may or may not happen given the presence of the indeterministic cause. So that’s what indeterministic causes aren’t, but what are they?Indeterministic causes:

  1. are necessary to produce effects – without the indeterministic cause the effect cannot happen
  2. produce desires within us. We desire more than one thing: I want to eat the cake and I want to lose weight. An indeterministic cause can give us the desire for something, without actually forcing the choice
  3. eliminate the “no choice” option – an indeterministic cause can force us to choose between two paths, as opposed to waiting at the Y intersection
  4. At times, a special subcategory of indeterministic causation can be sufficient to produce a choice, without opposition. That is to say, given the presence of the indeterministic cause and without our will specifically opposing that indeterministic cause, the choice will be produced.

#4 is rather abstract, so perhaps an example or two will help. One example of this is prevenient grace. If we do nothing, prevenient grace causes us to choose God. If we resist, we block prevenient grace. Think of two sumo wrestlers. If one stops pushing, the other causes him to go flying out of the ring. But if he continues pushing, he holds his ground. In the same way, this subcategory of indeterministic causation is sufficient without resistance and insufficient with resistance.

Another example, for guys, is looking at a beautiful woman. Christ says if we look for the purpose of lusting after her, we are committing adultery. So it’s look and lust. But don’t think that if a beautiful woman catches your eye you can continue looking without lusting. She’s not just beautiful, she’s attractive. If you don’t quickly choose to look away, you will be caused to lust after her.

Of course, indeterministic causes don’t predetermine outcomes. Our choices do not have sufficient causes. The agent is the source of his actions and his act of choosing is the first cause in a sufficient causation (rather than indeterminisic causation) sense. Edwards argues that everything has a sufficient cause, but we disagree.

It’s true that most of what we observe in nature has a sufficient cause. For example, the strike of the cue ball is sufficient to sink the eight ball. This works in the physical world, although perhaps exceptions exist at the quantum level. But remember, the will isn’t physical, it’s part of your immaterial soul. If a mad scientist were to dissect me, he wouldn’t find my will. The will can’t be moved with respect to location, because it’s not spatially extended. So LFW is not really an exception to physical causation.

The short answer to Edwards argument is that his equivocation of the word “cause” leads to a false dichotomy. This response also refutes Edwards’ points #1 & 2.

All proof of God’s existence is taken away
In part II.III Edwards argues: If this grand principle of common sense [nothing ever comes to pass without a cause] be taken away, all arguing from effects to causes ceaseth, and so all knowledge of any existence, besides what we have by the most direct and immediate intuition, particularly all our proof of the being of God, ceases: we argue His being from our own being, and the being of other things, which we are sensible once were not, but have begun to be; and from the being of the world, with all its constituent parts, and the manner of their existence; all which we see plainly are not necessary in their own nature, and so not self-existent, and therefore must have a Cause. But if things, not in themselves necessary, may begin to be without a Cause, all this arguing is vain.

My Response
In the first place, the cosmological argument isn’t the only argument for the existence of God. In my opinion, Anselm’s ontological argument is stronger. In the second place, Edwards’ idea that everything that came to pass has a cause actually disproves the biblical account of God. Consider God’s first act, creation, as recorded in Genesis 1:1. Did creation have a preceding sufficient cause? If it did, it wasn’t in the beginning. The cosmological argument relies on agent causation, which falsifies Edwards’ premise that everything that comes to pass must have a sufficient cause. In the third place, Edwards’ equivocation over the word “cause” reveals that every action of creation does in fact have a cause. And in the fourth place, God’s providential concurrence reconciles the cosmological argument with LFW.

Actions produced by a causeless cause would be random and irrational, and therefore not a basis of moral accountability
In part II.XIII Edwards argues that if our choices don’t have a cause, then reason isn’t a cause of our choices. If reason isn’t the cause of our choices, then we are choosing irrationally.

My Response
If we choose to act without considering the consequences, we are choosing irrationally. If we consider an action and our reason tells us that action is altogether bad, our reason eliminated that action as a possible option for us, and we can’t choose it. If our reason either identifies some aspect of an action as good, or is uncertain as to the outcome, that action remains an option to us. In which case our reason tells us that multiple options have good aspects, or at least they could have good aspects. In which case, choosing either option would be rational.

Let’s take the example of a milk shake and a diet. Our reason projects the milk shake will taste good and also that not having the milk shake help with weight loss. They both have some good aspects, so they remain options. If we choose either, we are acting in accordance with a plan, and our actions are rational.

If we change the example to let’s say eating or not eating a slug. Let’s also ignore extreme cases, like starvation or winning a bet. Your reason isn’t going to tell you eating the slug is good, so after you’ve thought about it, it’s not an option for you to eat the slug. Could you not even think about it at all and haphazardly pop the slug in your mouth? I suppose, but that would be irrational. So again, as long as you have thought about your options, projected the consequences of your possible choices and your reason is telling you there is (or could be) some good aspects of the choice, both choices are rational.
God willing, we will address Edwards’ arguments on divine foreknowledge in a separate post.


In part 2 section 12, Edwards attempts three demonstrations of the incompatibility of LFW and God’s foreknowledge: 1) based on the connection between foreknowledge and the event, 2) based on the impossibility of knowing things without evidence and 3) based on knowing a contingent event with certainty.

The Connection between Foreknowledge and the Event
Edwards’ Argument:
P1: Things in the past are now necessary
P2: In the past, God infallibly foreknew our future choices
C1: therefore, God’s foreknowledge of our future choices is now necessary
P3: if something necessary is infallibly connected with something else, that something else is also necessary
P4: God’s necessary foreknowledge is infallibly connected with our future choices
C2: therefore, our future choices are necessary

My Response
P3 is equivical. “Connection” could mean either a logical connection or a causal one. Additionally, “is” can mean either a logical or a temporal moment.

If connection is understood as a causal connection, we grant P3, but C2 doesn’t follow, because foreknowledge is connected logically (not causally) with events. Edwards admits foreknowledge doesn’t cause events. Nor do events cause foreknowledge. God’s knowledge of events is immediate; there isn’t a moment in time that an event happens and God doesn’t know about it. This implies that God’s knowledge isn’t caused, because effects temporally follow causes. God’s knowldge is unique in this regard.

If “connection” is understood as a logical connection, we must distinguish logicial necessity into simple and compound. Something is simply logically necessary if the opposite contradicts itself. (i.e. square circles) Something is compoundly logically necessary if the opposite contradicts another truth. (i.e. X = 1 contradicts X =2) Undersood as simple logical necessity, we deny P3. The event isn’t necessary such that propositions about the event not happening contradict themself. Rather they contradict statements about God’s foreknowledge.

If the connection is understood as logical and the necessity is understood as compound necessisty, we must distinguish between logical and temporal moments. If “is” relates to a logical sequence, we grant P3. But the logical sequence is: the event -> true propositions about the event -> foreknowledge. That is to say, the event is the logical basis of truth of propositions about the event, which is the logical basis of foreknowledge. Understanding this, the conclusion is inverted. C2 should be: given the logical connection between the event and foreknowledge, at the logical moment of the event and the connection between the event and foreknowledge, foreknowledge is necessary. One can conclude foreknowledge is necessary based on the event, but not the other way around.

But if the sequence is understood temporally, we deny P3. Truths are omni-temporal. The proposition “on May 17th 2008 Dan is typing a post” is true at all points in time. Thus, temporally sequencing a logical connection is inappropriate.

Impossibility of Knowing Things without Evidence
Edwards argues that nothing can be know without evidence. Evidence consists of either 1) “self-evidence” or 2) “the necessity of it’s nature“. Edwards denies future contingents give self-evidence, because they have not “present existence“. Regarding option 2, Edwards also points out that if something is necessary, it isn’t contingent.

My Response
The future event is the evidence by which God foreknows the future event. Thus the event is “self-evident”, using Edwards terms. What is required to be self-evident isn’t present existence, but future existence. Otherwise God would know the event, not foreknow it. To require present existence as evidence collapses time. What God knows would no longer be the future, if it has present existence.

Even though we can say the future is the evidence supporting God’s knowledge of the future, we do not know how God knows the future. Some say it’s because God is outside of time, but I am not sure about that. But Calvinism doesn’t explain how God knows the future either (if God knows the future in Calvinism). For more, see this post, and my exchange with Steve.

Knowing Something Certainly which is Contingent
Edwards argues that God’s knowledge would be inconsistent with itself, if He knew something contingent with certainty. To Edwards, this is like saying: “he now knows a proposition to be of certain infallible truth, which he knows to be of contingent uncertain truth“.

My Response
Edwards conflates “can” and “will”. Saying an event can happen, isn’t the same as saying it will happen. Can relates to possibility, will relates to future occurrence. God knows, out of the many things that can happen, what will happen. His knowledge is certain, which means His knowledge is correct, not that the event is necessary. Only by switching “can” and “will” does Edwards get to his conclusion God knows something both certainly and uncertainly; a move that that cannot be made without relying on the two flawed arguments above.

For original posts with comments see here,here, and here.