Wednesday, November 08, 2006

Zerah & Perez, Supervenience, and Systemic Evil

[Audio .mp3 20MB 22Mins] When the time came for Tamar to give birth, there were twin boys in her womb. As she was giving birth, one of them put out his hand; so the midwife took a scarlet thread and tied it on his wrist and said, "This one came out first." But when he drew back his hand, his brother came out, and she said, "So this is how you have broken out!" And he was named Perez. Then his brother, who had the scarlet thread on his wrist, came out and he was given the name Zerah (NIV: Genesis 38:27-30)
Now that simple reductionism has gone out of fashion in theological circles, there can be a healthy re-analyzing of key Christian doctrines. Among such doctrines stands the ever monolithic Problem of Evil. The wind often bloweth where it will, but the problem of evil always bloweth us over. Let us confess we have no illusion of “solving” the problem of evil in this essay, or of giving a “proto-solution”, or of even saying anything particularly encouraging to those who continually kick against that old monolith, hoping to at least break something off and peer inside. We simply desire to scratch upon it an arrow, to suggest a new place to kick, as it were. There is more than just a little intuitive force to our suspicion that evil actions have a cumulative and lasting impact on systems of human culture. Systemic structures of evil suddenly and consistently appear in forms which are excruciatingly un-analyzable in terms of the mere actions of unholiness on the part of individuals. In this paper we count this as evidence for evil being an emergent property. By calling evil a property, we are counting it as a real feature of the world; and, this is in distinction from a concept, which usually is not considered as a feature of the world, but as a way we subjectively encounter the word. Ultimately, we want to suggest that evil is a supervenient property of human actions. To this end, we shall first develop a thought experiment to inform our intuitions about how evil acts operate systemically. Second, regarding the thought experiment, we shall explore three interrelated versions of supervenience as regarding acts and moral value (specifically evil). Finally, we shall discuss the limitations of viewing evil as harm to others (and other definitive statements about just what evil is), and move on to explore some of the various facets of systemic evil and how positions on this can (or can not) be held consistently with traditional views that what God creates is good. We close with a few remarks on our endeavors.


In the Old Testament we find an account of Tamar giving birth to twins. Suppose we were to interpolate a bit of the details here. It turns out that, as identical twins, Zerah and Perez share not only the same DNA, but the same type of behavioral dispositions. Where Perez is placed in combat with a 25 lb. battle staff and with two Canaanite soldiers assailing him, he maneuvers effectively and with key techniques to neutralize his foes. So too would Zerah, when likewise equipped and likewise placed, use those precise and exact techniques. Again, were Perez to see a man of Adullam, the first question he asks is whether that man knows Hirah, a friend of his father. So also were Zerah to see such a man would he ask the very same question, at the very same instant from their meeting. From the perspective of their acts, just when they are in indiscernible situations, Perez and Zerah are behavioral duplicates. (Of course, it was never the case that they were in indiscernible situations, so this curious fact was never noted by the Biblical writers, but God knew what counterfactually would have happened). Let us stipulate that they both grow up and enter Hebrew society, taking on all the standard tasks and responsibilities that living in a community of their type demands. We shall call all the states and events in which they find themselves their life domain.

If, in any given situation, were we to substitute Perez for Zerah, and they were to act in indiscernible ways, would they be morally indiscernible? Are behavioral duplicates necessarily moral duplicates as well? If you think so, then you are inclined to accept a value-action supervenience. Here no two persons can differ in a moral respect unless they also differ in some respect to action; indiscernibility of acts entail indiscernibility with respect to mental moral properties. This is the intuitive starting point for supervenience, but there are several alternatives for relating acts to moral value which are consistent with what we have so far noted.


As we move to tease out the various alternative for relating acts to moral value, let us incorporate into the notion of act both the behavior and the intention of the actor. Perez and Zerah behave the same way because, among other shared causes, they have the same intentions.

Depending on the domain they are in, we can attribute two kinds of properties to Perez and Zerah. One, $, consisting of their moral properties (e.g. evil); and the other, @, consisting of their action properties (e.g. neutralizing Canaanites in self-defense). With this shorthand notation of properties, we can ask: what is it for $ to supervene on @ in their life domain?

We begin by considering a case where Perez acts to murder Hirah out of envy. Alas–though Hirah was the friend of Perez’ father (Judah), and though Hirah had never done anything to provoke Perez, and was otherwise a righteous man to all who knew him–the close friendship of Hirah to Judah was more than Perez could take. Of Perez and somebody else, say Mr. X, we may suggest the following:

Weak supervenience: Necessarily (that is in every hypothetical world that God ponders), if Perez and Mr. X (in their life domain) are indiscernible in acting to murder Hirah, (‘@-indiscernible’ in shorthand), Perez and Mr. X are also indiscernible in being evil (‘$-indiscernible’).

On the weak value-action supervenience, no hypothetical state of affairs contains two individuals who are alike in their actions but not alike in their moral properties. What if Mr. X were Zerah? If Zerah, instead of Perez, were put in the same position, Zerah would act the same and likewise would be evil. Now if acting to murder Hirah out of envy were so comprehensive that it took the exact location, the exact weapon, the exact time from when either brother was born, and the exact spelling of the name of the murderer, etc., to qualify; then neither could have murdered Hirah in precisely the same way. But then weak supervenience would not come to much, for there would not be a way for it to properly obtain. The precise and exact conditions are so stipulated that the conditions for indiscernibility never come about in the right way! A useful version of weak supervenience has to be modified. (For example, the spelling of the name of the murderer would have to go; for, Perez spells his name one way and Zerah spells his name another, so Hirah’s murderers would not qualify as indiscernible in every hypothetical situation.)

Being careful not to violate The 2nd Commandment, let us pretend to step into God’s view of this matter for the moment. God is pondering (akin to Leibniz’ possible worlds) the various ways the world might have been. If God decides to stick to weak supervenience, God cannot place in the same world actor-twins like Perez and Zerah that are not also moral-twins–i.e. person of equal moral merit. But nothing about weak supervenience restricts God from creating two actor-twins that are not also moral twins so long as he would place them in separate worlds. Weak supervenience makes its demands on what can be within a single hypothetical situation, but it does not make demands on what can be the case across the whole set of hypothetical situations that God ponders. Just because God ponders that moral values of actions are distributed one way in one hypothetical world does not restrict God to saying that the moral-value of actions could not be another way in a different hypothetical world. God thinks how there are several options he has when sticking to weak supervenience: (1) God could ponder a hypothetical world exactly like the actual world where Perez acts the same way but where Perez is not evil. (2) God could ponder a hypothetical world exactly like the actual world where Perez acts the same way and in which Perez is evil in the say way. (Again, it would be a hypothetical world just like ours, but it lacks just one property: existence.) (3) Or God could ponder another hypothetical world, again just like where Perez acted, but where Zerah acted instead, but where Zerah was not being evil; (4) Or God could ponder another hypothetical world, again just like where Perez acted, but where animals and angels, though no people, have moral value; after all, one only supposes of God that God stick to the weak supervenience, and all four options noted are perfectly consistent with it.

Of course the last option, if not earlier ones, shows us that weak supervenience does not give us a strong enough connection between action and value to map our moral intuitions. We want facts about our actions to determine facts about moral judgements within our life domain. When Perez or anybody acts a certain way, we want a consistent way to judge the moral merit (or iniquity) of his actions. But (1)-(4) are all allowable ways on weak supervenience. Clearly then, we need God to pare down the criteria of moral judgement over the domain of Perez’ life. We want such actions in Perez’ domain of life, i.e., where he murders Hirah, to be evil in the sight of God in any hypothetical situation.

We must reconsider the case of Perez’ murdering Hirah out of envy. There are two other versions of supervenience we might use:

Global supervenience: Any two hypothetical worlds that God ponders that are indiscernible with respect to Perez murdering Hirah out of envy are indiscernible with respect to Perez being evil, which is to say Perez murdering Hirah out of envy in two hypothetical situations cannot differ in how moral merit gets attributed to Perez.

Strong Supervenience: For any two individuals, say Perez and Zerah, and any hypothetical worlds that God ponders, say W1 and W2, if Perez in W1 is @-indiscernible from Zerah in W2 (i.e., Perez and Zerah both murder Hirah in their hypothetical life domain), then Perez in W1 is $-indiscernible from Zerah in W2 (i.e., Perez is evil in W1 and likewise Zerah is evil in W2.)

Global supervenience applies indiscernibility considerations to hypothetical situations taken in themselves rather than to Perez or Zerah within those hypothetical situations; furthermore, Global supervenience requires that worlds that are indistinguishable in terms of actions do not differ in terms of moral value. Still, Global supervenience allows Perez to be evil and Zerah to be evil, yet for different reasons–as when Perez murders Hirah out of envy and Zerah murders Hirah for fun.

It turns out, however, that global supervenience is too weak a notion to capture the idea, that vices are dependent upon, or determined by, actions, since on the global supervenience position, there might be a hypothetical world which differs from the actual world in some most tiny way (e.g., Perez clears his throat while acting) but which is entirely devoid of moral value, or has a radically different, perhaps totally irregular, distribution of moral value over its actors (as when Perez and Hirah are and do exactly the same, but where they are accounted without moral value at all, must less as evil or innocent, while the surrounding rocks, sand, and bushes are all evil or innocent.). Global supervenience would allow the least differences in actions between any two worlds to support as arbitrary, and as large, a moral value difference between them as is logically possible. This would seem to threaten the traditional Christian position that God’s creation is somehow inherently good.

Luckily, strong supervenience implies both global and weak supervenience, so it allows us to maintain a consistent moral value system within any given world, and yet preserves the traditional Christian position that God’s creation is inherently good. The advantage of strong supervenience is that it maintains moral consistency across individuals in the same world, yet preserves the goodness of God’s creation in this way: whatever God happens to create, as inherently good, will support like evaluations for like acts across hypothetical *worlds, thus avoiding logical situations as noted above where Perez and Hirah do exactly the same thing, but where they are accounted without moral value while the surrounding, non-acting objects could yet be accounted with moral value.

Up until now, we have been varying the metaphysical systems within which moral actions may be (or might not be) exercised by agents, trying to map our moral and theological intuitions by means of adjusting our thought experiments with Perez, Zerah, and Hirah. It now remains to discuss how evil emerges and influences actors within a system of action and moral valuation. So at this point, and with three inter-related conceptualizations of supervenience, we are left wondering how the following might operate as a productive statement concerning evil:

(1) Evil is a supervenient property

This has a nice ring to it. But (1) is a dreadfully general statement. What do we mean by evil? What, precisely, does evil supervene upon?


The first question (i.e., “What do we mean by evil?”) is tough. Evil might be defined as harm to others, but there are some worries here. First just as there are victimless crimes, perhaps there are, oxymoronically, harmless evils, wherein a person is not directly harmed, but somehow misses out of what might have been the case. For example, if Mr. X knew there was pearl of great price in his field, he would not have sold it to Mr. Y, and thus unknowingly cheat himself out of practicing great charity to Mr. Z., a blind man. In this case, both Mr. X and Mr. Z have missed a potential for experiencing a great good. Second, although martyrdom entails a harm, it has often been accounted a great good, for perhaps what some intended for evil, God designed for good; thus, one need not necessarily count the sufferings of this age as comparable to the glory of things to come. Other definitions of evil can be advanced, but one of the traditional problems in analyzing evil is that tightening the definition to cover only and all evil acts and situations has proven elusive. So we must, to some degree, leave it as undefined and hope that we share enough intuitions on it to move our discussion of systemic evil forward. We shall, however, draw some clarifying distinctions among ‘sin’, ‘vice’, and ‘evil’. Vice is a kind within the category of evil. And sin is a particular instance of vice. Thus, for example, we might say that a particular event of stabbing someone to death is a sin, yet this sin is a particular event which is an instance of a kind, namely the vice of murder. There are other vices as well, but the full set of vices, their relations, and effects (if any) within the context of moral value we term evil.

The second question (i.e., “What, precisely, does evil supervene upon”) is easier to answer, though in the end harder to analyze. Evil supervenes on what people do, or to propositionalize it:

(2) Evil is a supervenient property on human actions.

Proposition (2) is the paramount issue. On any account, whether evil is the same as human actions, or distinct from them, we must be clear as to whether the issue applies to concrete events (e.g., individual behaviors by particular subjects at particular times) or to general kinds (classes of acts or behaviors) under which such concrete acts fall. To this end, a useful distinction is regularly drawn between tokens and types. Consider the following short list: blue, red, blue. Moreover, consider an inquiry: how many words are in that list? The question is ambiguous, as we could give reasons for there being either two words or three words therein. To speak precisely, there are three word tokens and two word types. This distinction allows us to now formulate some theories on the matter.

A token identity theory of systemic evil would hold that every concrete event (roughly, ‘behavior’) falling under a general kind can be identified with some action or other: instances of sin (herein murder), for example, are taken to be not only instances of a vice (e.g. murder) , but instances of some action as well (say, fatally stabbing one’s neighbor. Token Identity is weaker than Type Identity. A type identity theory of systemic evil would go so far as to claim that instances of moral value themselves just are actions of a particular kind. It has been noted that Token Identity is entailed by, but does not entail, Type Identity. Token identity is entailed by type identity because if types of moral value themselves are types of actions, then each individual instance of moral value will also be an individual instance of an action. Token identity does not entail type identity, however, because even if a concrete event falls under both moral value and an action, this contingent fact does not guarantee the identity of the kinds of moral values whose instantiations are constituted by the concrete events.

So an identity theory of systemic evil, taken as a theory of types rather than tokens, must make some claim to the effect that vices (and not just individual instances of sin, such as murder) are contingently identical with–and therefore theoretically reducible to–actions such as fatally stabbing someone. Depending on the desired strength and scope of moral value-cum-human action identity, however, there are various ways of understanding this claim.

It turns out a damaging objection can be made to a type identity theory of systematic evil which effectively retires such a theory from a privileged position in theological analysis concerning the relationship between evil and action. The argument can be outlined as follows: (1) according to a type identity theorist, for every vice there is a unique type of action such that a person can exhibit that vice if and only if one performs a given action. (2) It seems quite plausible to hold, as a hypothesis, that a person can exhibit the same vice without having murder, for example, be the same unique action (e.g. stabbing, clubbing, choking, etc.). (3) Therefore, it is highly questionable that the type identity theorist is correct.

In support of the second premise above, a "multiple realizability" hypothesis about moral acts, consider the following point: we now have at hand good reason to suppose that somewhere in the universe–perhaps in the actual world, perhaps in a counterfactual situation–there is a possible person capable of exhibiting vice X (e.g., capable of murder) without performing a given action Y (that is, without acting to murder by means of stabbing). A vice can supervene on a wide set of acts.

Interestingly, this multiple realizability of vice can be presented as a negative version of a Hauerwas-like perspective on communities of character. Vice similarities across kinds of communities may often reflect convergent cultural pressures rather than underlying dispositions for an individual to sin. On Hauerwas’ view, for example, the church’s inability to realize its vocation as a called community means that it fails to understand the ethical imperative in the Bible, and thus is not a true community. This flawed understanding leads to egoistic-individualism, self-indulgence, violence, and a host of other vices. These vices, in as much as they are maintained by a false community, allow a system of evil to emerge from humans acting upon flawed understandings of their relationships with one another. Communities, even false ones, can outlive their founding members, so the moral values which remain as emergent will have continuing effects on new (or newly joined) members of the community. If the systemic moral values are evil, then only a truly independent actor may break the cycle of sin and hence vice which allows the systemic evil to emerge and pervade. On Christianity’s traditional view, only the Christ could be such an independent actor.


In the end, we note that a token identity theory of systemic evil, realized under the form of strong supervenience, is fully consistent with the multiple realizability of vice and the subsequent emergence of systemic evil, which is hardly a controversial implication, as even Hosea notes, "Because Ephraim has made many altars for sin, they have become for him altars for sinning. I have written for him the great things of My law, but they were considered a strange thing" (8:11-12 NKJV)


[Image] "The Twins" by Jacob Wenzka, Strange Horizons Gallery (Accessed 11/08/2006)

[Authors] This paper was presented at the Wesleyan Theological Society in 2003 by myself and Robert J. Thompson



At 1:10 AM, Blogger Archikins said...

hey Brinticus - you should join the Philosophy Blog War!


Post a Comment

<< Home