Definitions

Given (P), the concept of gratuitous evil is central to arguments from evil for the non-existence of God. Gratuitous evil is most simply understood as unjustified evil, an instance of suffering which God could have prevented without forfeiting some greater good or permitting some evil as bad or worse. Alan R. Rhoda has attempted an uncontroversial definition of gratuitous evil (Rhoda (Reference Rhoda2010), 281), which William Hasker has challenged (Hasker (Reference Hasker2010), 303). The controversy, argues Hasker, centres on two difficulties. First, uncertainty as to what an omniscient being actually knows and, second, the problem of arriving at a definition that is not demotivating with regard to assisting one's fellow human being. Both problems are relevant to one of the concerns discussed below and will be addressed at that time.

There is a common factor in the various definitions and concepts of gratuitous evil discussed by Rhoda and Hasker. Essential to determining if an instance of suffering is gratuitous or not are the intrinsic moral values of the consequences of the event, as well as the consequences of the best possible alternative that could have been actualized by God at the time. The argument presented here does not depend upon any particular definition discussed by Rhoda and Hasker, only upon the acceptance that the net moral value of all the relevant consequences are an essential component in what determines an event to be gratuitous or not. To clarify, one cannot discuss whether an event is gratuitous or not, without knowing something about its consequences.

In determining if an event is gratuitous, a few concepts and definitions are used here that arise from a method proposed by Kirk Durston (Durston (Reference Durston2000), 65).

Primary event: a single event that cannot be broken down any further into constituent events.

A car accident in which two people were injured, for example, is composed of at least two primary events, the immediate suffering of one of the victims and the immediate suffering of the other victim. There will be two interacting chains of consequences that unfold from these two primary events.

Intrinsic moral value: the moral value of a discrete event considered only by itself (in isolation from all its consequences and other events).

A discrete event may be a primary event initiating an exponentially increasing, interacting chain of consequences, or it may be a particular, discrete consequence of the primary event. Each consequence has its own intrinsic moral value that we could quantify as ranging from –10 to + 10 (or any other range of values a person deems fit just so long as all events are evaluated on the same scale).

Overall moral value: the sum of the intrinsic moral value of the primary event and the intrinsic moral values of each consequence to the end of history or so far as is logically possible for an omniscient mind to know.

The method originally presented by Durston sums the intrinsic moral values of all consequences to the end of actual and possible histories, which assumes that omniscience includes middle knowledge, that is, knowledge of the truth-value of all possible counterfactuals of freedom. The role of middle knowledge in omniscience is controversial (Hasker (Reference Hasker1986), 545), a fact that Rhoda and Hasker have taken into account as mentioned earlier. If omniscience is defined as knowing all that is logically possible to know (Mann (Reference Mann1975), 151), then the definition of overall moral value supplied above takes into account the controversy surrounding what God actually knows and the possibility that middle knowledge is not included in the attribute of omniscience. With this in mind, let

A = the overall moral value of the actual event and,

B = the overall moral value of the best alternative that could have been actualized by God

Net moral value: the difference between the overall moral values of the actual primary event

(A) and the best alternative that God could have actualized (B).

Given all of the above, we can define gratuitous evil as follows:

Gratuitous evil: An event is a gratuitous evil, or results in gratuitous evil if, and only if, the net moral value of that event is negative.

Since the argument to be presented concerns entire societies, the overall moral value of a society (actual or possible) at time t is defined as:

A_soc = the sum of the overall moral values of all the primary events occurring at time t in a society,

and the overall moral value of the best alternative to that society that God could have actualized is:

B_soc = the sum of the overall moral values of all the primary events of the best alternative that God could actualize at time t.

The net moral value of an entire society at time t would simply be the difference between A_soc and B_soc at t. A society results in gratuitous evil, therefore, when its net moral value is negative.

The argument

Recall (P): the existence of an all-powerful, all-knowing, and perfectly good God is incompatible with gratuitous evil. It follows from (P) that if the net moral value of a society at some time t in its history is positive (A_soc − B_soc) ⩾ 0, then God must permit that society to continue. To use an example from the destruction of the Canaanite society, in Genesis 15 God tells Abraham some 400 years prior to the conquest of Canaan that he and his descendants cannot possess that land for another 400 years because ‘the sin of the Amorites has not yet reached its full measure’, or (A_soc − B_soc) ⩾ 0 for Amorite society at that point in history.

(P) also entails that if, at some time t in the history of a society, the net moral value of that society becomes negative (A_soc − B_soc) < 0, then God must bring about an end to the society at that point. The argument can be summarized as follows:

1. 1. P.

2. 2. A society results in gratuitous evil if its net moral value is negative.

3. 3. If P and (2), then God must terminate a society at the point in its history when its net moral value becomes negative.

4. 4. Therefore, God must terminate a society at the point in its history when its net moral value becomes negative.

This concept can be illustrated as shown in Figure 1. In contemplating the curve in Figure 1, it is important to understand the distinction between the net moral value (A_soc − B_soc) of a hypothetical society, the overall moral value A_soc of a society, and the intrinsic moral value of a society at some time t. For example, at t _term in Figure 1, the intrinsic moral value of the society at that point might actually be positive and A_soc might also be positive at that point in time, which entails that there is a B_soc that is higher in overall moral value than A_soc (i.e. there is another option God could actualize at t _term that is better than permitting the actual society to continue). Of course, it may also be the case that the intrinsic value of the hypothetical society at t _term is already quite negative, but (A_soc − B_soc) = 0. In other words, one cannot infer the intrinsic moral value of a society at any time t, or the overall moral value A_soc at time t by simply looking at the curve in Figure 1.

Fig. 1 Net moral value vs time Each point on the descending curve represents the net moral value (A_soc – B_soc) of a hypothetical society versus time t. This is not to be confused with either the intrinsic moral value or the overall moral value (A_soc) of the society at any point in time. If those points were plotted, the resulting curves would be quite different from the (A_soc – B_soc) curve shown.

The argument presented here raises various worries and problems, four of which are addressed below.

Problem one: what about good people in an evil society?

Let us define a completely depraved society as a society where all individuals freely choose to do wrongly all the time, and a completely perfect society as a society where all individuals freely choose to do rightly all the time. It is reasonable to grant that the vast majority of societies fall somewhere on a continuum between the two extremes. In other words, the vast majority of societies are composed of a set of individuals who form that society, a subset of which contains people whose goodness exceeds their badness and another subset of which contains people whose badness exceeds their goodness. Along with this, recall that the intrinsic moral value of an event or person may be negative but their net moral value may still be positive when all the consequences to the end of actual and possible histories are considered. Similarly, there may be individuals with a positive intrinsic moral value but who have a negative net moral value when all future consequences are considered. With this in mind, a society that reaches the crossover point into gratuitous evil may still contain a subset of good people, or people who have a positive net moral value. If a society must be destroyed to prevent gratuitous evil but, in doing so, people who have a positive net moral value are destroyed, is not one gratuitous evil replaced with another?

The above worry arises out of the assumption that if God must terminate a society, then all people in that society must be terminated. It does not logically follow, however, that the termination of a society entails the termination or death of every individual within that society. One option is simply to disperse the individuals of that society broadly enough such that the net moral value of the larger dispersion becomes positive. A danger with this, however, is that if the subset of evil people within the society is large and evil enough, complete dispersion may end up fanning the flames of evil over a larger area. Another option is for a filtering process to occur whereby the good are dispersed with only the evil remaining behind for destruction. To use an example of dispersal from the account of the conquest of Canaan mentioned earlier, God is quoted as saying, ‘I will drive them out little by little until you . . . take possession of the land’ (Exodus 23:30–31).

If a filtering process is used, however, whereby the good are dispersed leaving behind only the evil, then the concentration of people with a negative net moral value will be much higher, resulting in a post-filtered society with an even more negative (A_soc − B_soc). For the remaining members for whom (A_soc − B_soc) < 0, given (P), God must terminate those individuals unless there is a way to bring their net moral value into a positive position.

To summarize, the termination of a society does not entail the termination or death of all its constituent members.

Problem two: can we know if a society is gratuitous?

Can we justify the conclusion that a particular society has declined to the point where it has become a gratuitous evil? The underlying worry here is whether we can destroy a society on the grounds that, in our view, it has become a gratuitous evil.

Durston has argued that the consequential complexity of history is such that we are in no position at all to know if an event or, in this case, a society is gratuitous. The amount of relevant information we have is minuscule in comparison to what we would need to know, to infer inductively if an event is gratuitous or not (Durston (Reference Durston2000), 65). I will only very briefly summarize his argument here as follows.

Each event in history unleashes an exponentially increasing number of interacting consequences to the end of history. To evaluate whether a primary event is gratuitous or not, we must have a reasonable idea of the net moral value (A–B) of that event. Recall that (A) is the overall moral value of the event with all its consequences to the end of history (or as far as omniscience permits). If we let (E) represent the intrinsic moral value of the primary event and C₁, C₂, C₃, . . . C_end represent the intrinsic moral values of each of the consequences of that event to the end of history, then we can represent (A) as follows.

A = E + C₁ + C₂ + C₃ + billions of unknown intrinsic values + C_end.

Similarly, the net moral value of the best alternative that God could actualize can be represented as follows,

B = E′ + C′₁ + C′₂ + C′₃ + billions of unknown intrinsic values + C′_end.

The intrinsic moral value for each one of the billions of consequences has some unknown value of unknown sign, positive or negative. Given that our knowledge of the consequences to the end of history is minuscule in comparison to the billions of unknowns, the most rational position appears to be agnosticism with regard to the net moral value of any event. A hypothetical society is composed of many primary events at any time t. If the net moral value of each of the constituent events is unknown, then the summation of those unknown values in order to calculate (A_soc − B_soc) at time t is also an unknown. We are thus in no position at all to know if a society has become a gratuitous evil.

A possible complication arises if we grant that there is controversy with regard to what an omniscient being knows, as discussed earlier. In response, it is reasonable to assume that even if we do not know how much an omniscient mind knows, it is very likely that a mind that knows all that is logically possible to know, knows vastly more than we do about the future consequences of any event. Furthermore, even if there is controversy over how much an omniscient mind knows, the consequential complexity of history keeps us just as much in the dark as ever, independent of the question of what omniscience includes. The most defensible position to take, therefore, is agnosticism with regard to the net moral value of any society, owing to the consequential complexity of history. It follows from this that humanity cannot justify the conclusion that any particular society has a negative net moral value.

Problem three: were destroyed societies evil?

There are societies that no longer exist and others that have been decimated in history. If the argument presented earlier is granted, does it follow that these societies were evil?

The first point to make by way of response is that even if (A_soc − B_soc) is negative, it does not follow that the intrinsic moral value of that society at the time it was destroyed was negative. From the previous section, we can see that (A_soc − B_soc) is the difference between two values, each of which has billions of variables of unknown value and sign, positive or negative. If all we know is that (A_soc − B_soc) is negative, we are in no position at all to infer the intrinsic moral value (E_soc) of the society at the time it was destroyed or decimated; it may actually have been positive. Nick Trakakis has pointed out that the consequential complexity of history also prevents us from knowing if a primary event that has a positive intrinsic value leads to a gratuitous evil (Trakakis (Reference Trakakis2003), 451). To clarify, for a single event, its intrinsic moral value (E) may actually be positive but the consequences that unfold from that event, when compared to the best alternative, may result in a negative (A − B). Similarly, it is possible that (E_soc) is positive but (A_soc − B_soc) is negative. In other words, it is possible that a society with a positive intrinsic moral value may actually result in gratuitous evil when the consequences of allowing that society to continue are considered. Conversely, a society that has a negative intrinsic moral value, at a particular time in history, may actually result in a positive net moral value when all its consequences are considered to the end of history (or the limits of what is logically possible to know). We are not, therefore, in a position to conclude that a society that was destroyed or decimated must have been an evil society. If God exists, and if we grant (P), then all we can conclude was that the consequences of letting that society continue would have either led to a greater evil, or prevented a greater good when compared to terminating that society at that time in history.

Problem four: moral indecision

Nick Trakakis also raises the problem of moral indecision. If the consequential complexity of history puts us in a position of having no idea if any event, good or evil, will result in gratuitous evil, then on what basis can we act? If we have no idea if a society will result in gratuitous evil, then could we be paralyzed into moral indecision when that society appears to need help?

In his reply to Trakakis, Durston has proposed a premise that seems to arise out of our moral intuitions and which is embraced by many judicial systems. He states it as follows :

The above premise (J) is applicable both to the problem of moral indecision presented here, as well as to the problem addressed by Rhoda, mentioned earlier. Recall that Rhoda's concern was that any definition of gratuitous evil should not produce moral inaction because of the assumption that if God exists, all actual events either prevent a greater evil or bring about a greater good, so we can sit back and let things happen.

The motivation for (J) arises not just from our moral intuitions, but out of Alvin Plantinga's free-will defence as well, specifically, the idea that the state of any possible world at time t that contains free agents, is not solely determined by God, but also by how well the free agents in that world freely cooperate with God. Plantinga writes (Plantinga (Reference Plantinga1979), 164–193):

It follows from Plantinga's idea that the more the free agents of a possible world will cooperate, the better the world God is able to actualize. Cooperation entails that we make the best moral decisions we can, given the knowledge we can reasonably be expected to acquire, resulting in Durston's premise (J).

If premise (J) is granted, then the problem of moral indecision and moral inaction has a possible solution despite our minuscule knowledge due to the consequential complexity of history.