A charge Cheung makes against the idea that God is not the author of sin is that such makes him to be not sovereign. In his book The Author of Sin [Vincent Cheung, The Author of Sin (Boston, MA: Vincent Cheung, 2005)], Cheung makes the following statements:
I affirm the meaningfulness of so-called "second causes" only in the sense that these are the means by which God executes his immutable decrees; however, these second causes are not themselves self-existent, self-determined, self-caused, or self-powered. Rather, all so-called "secondary causes" are themselves immediately caused and controlled by God, and the objects on which these secondary causes supposedly act upon react in ways that are also immediately caused and controlled by God. (p. 20)
...
Now, "soft" determinism is used in contrast with "hard" determinism. Using these terms, the popular Reformed/Calvinistic position, which is compatibilism, would be called "soft" determinism, whereas my position would be called "hard" determinism.
The former is "softer" in quality and/or in quantity regarding the level and/or amount of control (determinism) that God exercises over his creation, whereas "hard" determinism is absolute, affirming that God exercises complete (in level or quality) and comprehensive (in amount or quantity) control over everything.
But this means that "soft" determinism is really partial determinism – that is, partial (not full) either in quality or in quantity, or both. And if what God does not absolutely determine can still actually happen, then this means that there is another (one or more) determining power in the universe. When we are speaking of God's relation to man, attributing only partial determinism to God necessarily implies attributing partial determinism to man also. So this becomes a version of dualism.
In other words, one who believes that God absolutely determines everything is a full determinist, since he believes that God fully determines everything, in terms of both quality and quantity, and in terms of both the level (extent) and the amount of control exercised. To believe anything less than this is not full; therefore, it is partial.
Also, since "soft" determinism really means partial determinism, this also necessarily means that it is partial indeterminism (that is, partial non-determinism). Granted, since Calvinists usually (claim to) affirm greater determining power to God than man, this indeterminism is a very "soft" indeterminism, but it is still partial indeterminism.
It becomes just a matter of emphasis as to which term one wishes to use. So the term "soft" determinism is at least a little misleading, making its adherents look better than they really should. To some, it has the effect of sounding "softer," kinder, and less extreme. But if we don't let the language deceive us, we see that it is really partial determinism, weak determinism, incomplete determinism, or "soft" indeterminism. And, at least by implication, dualism. (p. 25. Bola and emphasis original)
...
Those who see that it is impossible to disassociate God from the origination and continuation of evil still attempt to distance God from evil by suggesting that God merely "permits" evil, and that he does not cause it. However, since the Bible itself states that God actively decrees and causes everything, and that nothing can exist or happen apart from his will and power, it makes no sense to say that he merely permits something – nothing happens by God's mere permission. In fact, when it comes to ontology, "God's permission" is an unintelligible term. (p. 70)
According to Cheung therefore, to embrace the idea of contingencies and second causes, except as mere means, is to believe in "partial indeterminism." It invokes metaphysical distanciation of God from the event, introduces uncertainty into the equation, and thus lead to a compromise of God's sovereignty. God must exercise "complete and comprehensive control over everything," or He is not fully sovereign at all.
The problem with Cheung is that he is involved in categorical confusion, and reasons one-dimensionally. Firstly, Cheung confuses sovereignty with direct control. If agent G causes action x to agent A, and agent A causes action y to cause event B, why is agent G not as sovereign over event B as over agent A? The syllogism pairs 'If p, then q'; 'if q then r' would yield 'if p, then r' as a logical deduction. So upon what basis does Cheung claim that non-direct control implies a loss in sovereignty?
Cheung claims that language such as "permission" and the idea that secondary agents are actually really agents introduces indeterminacy into the equation. In other words, indirect control means that the indirect agents have some measure of determinacy. However, such is to confuse agency with sovereignty. Just looking at the basis syllogism pairs for instance, does the fact that agent A could choose otherwise factor into the fact that 'if p, then r'? No, the syllogism pairs (if correct) function independently apart from the state of agent A in doing q to get event r. Even if A does not want to get to event r, yet if the syllogism holds true, his doing q would necessitate the happening of event r. In other words, intention has nothing whatsoever to do with sovereignty. God and the secondary agents could both be free in their respective spheres (divine and creaturely freedom), and yet God is sovereign over all things.
Cheung of course has a counter in the idea that "freedom" is defined from God, and thus creatures are not free, as defined by freedom from God. Contingencies of second causes are not free from God and thus not free (Cheung, 23). Thus, Cheung divorces creaturely intentions from the equation altogether. But such is a redefinition of "freedom" as well as a strange definition of "direct control." Upon what basis does Cheung glosses over human intentions, and claim that, since God is sovereign, therefore "direct control" implies "ultimate control"? It's as if Cheung claims that God's ultimate control over all things means that the intention of the secondary agent can be overloooked, when the issue under discussion is whether God's control is direct, not whether God's control is ultimate. How can one claim that God's control is direct by writing the creaturely intention out of the equation, and then claim that proving God's ultimate control implies proving God's direct control? One can only do that by illegitimately claiming that denying direct causation implies denying ultimate causation, a claim which is logically fallacious (as seen in our syllogism pairs) and having a unbiblical idea of what sovereignty implies, which brings us to our second point.
Cheung's second error is that he reasons one-dimensonally. In theological terms, he confuses the archetypal/ectypal distinction. But even worse than that is his flattening of all categories into a one-dimensional strait-jacket. Even secular philosophers may think in more than one level, whereas for Cheung everything must be reduced to one level. Thus, for Cheung, sovereignty implies direct control, and he has no place whatever for more sophisticated understanding of sovereignty. In other words, by his definition, if a theory denies direct control, they necessarily must deny sovereignty since that is all his strait-jacket thinking can conceive of.
Cheung's thinking of sovereignty looks like this:
Fig. 1
The problem here of course is that the whole process is one-dimensional. There is in this case metaphysical distanciation because God is removed from B greater than from A; (x + y) is certainly greater than x.
If one however sees a difference in the manner of causation of God and the agent A, then one could hold on to a two-dimensional view of causality, as follows:
Fig. 2
Fig. 3
In figures 2 and 3, the distance between God and event B is a vector of x.y, as opposed to the simple scaler equation of (x + y). As seen in figure 3, the vector can function in such a way that the distance between God and event B is exactly the same distance as that from God and agent A (in an isoceles triange), but such is not necessarily the case.
For a Christian however, figure 4 betters fits the biblical data of what we know about God, as follows:
Fig. 4
In figure 4, God is at the center of the sphere. God is sui generis and is totally unlike the rest of creation, which are on the surface of the sphere. God is at once totally immanent and totally transcendent, at once close to His creation (since it's a distance away), but far from it (since it's a distance away). God thus causes agent A through action x to do action y resulting in event B, but all things (agents, events) are at the same distance away from the center by the radius of the sphere. God is thus just as near and just as far away from secondary, tertiary and further removed events. Just as the distance from the center to any point in the surface of the sphere is the same, so likewise God is just as sovereign and in control of the effects as the agents, without being directly causing them.
So God is sovereign over all things. All things are under God's control, directly or indirectly, and there is no distanciation involved by positing the contingencies of secondary causes. God's freedom is freedom indeed, as there is an infinite number of lines possible to be drawn from the center to the surface of the sphere. Creaturely freedom is also real (by which we do not mean freedom from God, a ridiculous redefinition of "freedom"), for there are an infinite number of points chosen for A to connect. A could connect to B, C, D, E, F and so on.
God is sovereign, and Man is free. That is not a contradiction, when one understands that creaturely freedom functions on a different plane from God's freedom. Cheung's errors in the whole issue is that of confusing categories, flattening the categories into a rationalistic straitjacket, and in so doing impinges the character of God by either making God the culpable agent of evil, or by divorcing God's will from His nature.