This post was destroyed by the infection of my old blog. This is a recovery of what I think was an earlier version of it.
Over the last few days I have written over 15,000 words on the subject of limited atonement. Whilst limited atonement is an issue that I have discussed at considerable length on Internet fora on a number of occasions in the past, I have never properly expressed my opinion in a single blog post. I thought that it might be worthwhile to sum up my position on the issue, collecting together thoughts from my various comments and putting them into some sort of order. At the very least it would provide me with something that I can direct people to in the future and save me the trouble of outlining the whole of my position every time I enter into a new debate on the subject. I apologize for the rather rambling character of many of my points. As the following material is largely culled from comments it is unwieldy and hard to work into a clearly structured argument.
Personal Background
Before I start out it might be helpful for me to give a brief sketch of what has led me to my current position. A number of years ago I became convinced of the limited atonement position. I felt that it was the only way to hold that the cross actually achieved something, avoiding the ‘possibility-realization’ model that makes my personal response the deciding factor. I believed that the limited atonement model was demanded by the doctrine of penal substitution, which I was convinced was a biblical atonement model. I was also convinced that the scriptural doctrine of election demanded a limited atonement; universal atonement would be inconsistent with the fact that God has only chosen a particular number of individuals to be saved. Whilst I would happily grant infinite worth to the cross of Christ, I was not willing to say that it was God’s purpose to save everyone.
As time went on, I began to feel uneasy about the doctrine. The thing that made me feel most uneasy was the manner in which I had to construct elaborate explanations for so many biblical verses. The biblical authors seemed to feel no need to qualify statements that, at least on the surface, seemed to directly oppose limited atonement. I started to wonder whether anyone who held the doctrine of limited atonement could speak in such a manner, without feeling an irresistible urge to qualify. I also started to study particular biblical books in more depth, rather than just studying doctrines with their associated proof-texts and problem texts. As I did so I encountered proof-texts for limited atonement in their original context and they did not seem to carry the weight that I had put upon them.
I also found myself re-examining the issue of election, as I started to study the way that the NT spoke of the subject. As I arrived at new conclusions on this subject I found that new possibilities were opened up for my doctrine of the atonement, possibilities that would relieve much of the tension that I had felt between the doctrine of the limited atonement and the text. I also found that I could take on board most of the concerns on both sides of the familiar debates over limited versus unlimited atonement. In particular, I realized that it was not necessary to abandon a belief in penal substitution or an effectual atonement if I was to move away from the doctrine of limited atonement.
Although I have moved considerably in my understanding of the atonement and election over the last few years I feel a strong continuity in the whole movement. Throughout the whole movement I have been guided by particular biblical principles that I was not willing to sacrifice under any circumstances. I am convinced that a truly Christian theology is one in which these principles are given full expression.
In many respects this could be understood as an aesthetic sense of Christian truth. There have been many occasions when I have not been able to see any logical flaws in a particular position, but am convinced that such flaws are present because the position seems unbalanced or ill-proportioned in certain aspects. Some doctrines just are not as beautiful as they ought to be.
Limited atonement has always seemed to be one such an aspect within the ‘Calvinist’ system. There are facets of the Calvinist system that are truly beautiful. The focus on the priority of God’s grace and the centrality of God’s glory is truly beautiful. When you first see these things, their beauty almost takes your breath away. I might be wrong, but most people who become Calvinists don’t find themselves drawn to the position by the doctrine of limited atonement. They hold the doctrine of limited atonement because they believe that it is necessarily required by the other doctrines that are so dear to them.
What if it were possible to consistently stress such things as the priority of God’s grace, the centrality of God’s glory and an effectual atonement without actually holding a limited atonement position? I am suggesting that we can retain the beauties of the Christian faith that the Calvinist system highlights, whilst avoiding the sides of Calvinism that so many find problematic and ugly — things like double predestination, the idea that God purposed the Fall of mankind, the doctrine of a limited atonement and the charge of determinism that the Calvinist system finds so hard to shake.
1 All or Not All?
Limited atonement is a response to an ‘all or not all’ question about the atonement. It seems to me that this is a bogus question in many important respects. There are a lot of assumptions bound up in the question that need to be unpacked and examined. Having examined these assumptions we may well find that the question itself needs to be rejected.
2 Limited Atonement in Calvin and Dordt
The ‘all or not all’ question seems to have assumed a centrality in many modern Reformed understandings of the atonement that it did not have historically. Whether or not the question is a good one, the prominence that it has in modern Reformed thought is significant.
It has always interested me that, despite the amount that Calvin wrote on the subject of the atonement, no consensus can be reached on whether or not he held to the doctrine of limited atonement. Not only is Calvin’s theology not orientated to the answering of this question, it is not easy to ascertain how Calvin would have responded to the question had he been asked it.
I agree with Bavinck and Berkouwer in their claim that Dordt was not responding to an ‘all or not all’ question in its doctrine of the atonement. Dordt was opposing a ‘possibility-realization’ model of the atonement. It taught a doctrine of efficacious atonement. Its purpose was not to argue for limited atonement. Whilst it might affirm some moderate form of limited atonement, we should not see this as the main thrust of the document.
3 Forms of the Doctrine of Limited Atonement
It appears to me that there are a number of different forms that the doctrine of limited atonement can take and also some views that affirm the doctrine of election whilst denying the doctrine of limited atonement. There are some important, but fine distinctions that need to be drawn if we are to recognize the significant differences between these.
Just about everyone agrees that Christ’s death is of infinite value. When many people say that they believe that the atonement was sufficient to atone for the sins of the whole world they mean nothing more than this. It is like saying that the amount of ingredients in a particular type of baked cake is sufficient to make three other different types of cakes. Although the ‘ingredients’ of the atonement (i.e. the Incarnate Christ’s sinless life) are regarded as being sufficient to save the whole world, the actual atonement (the ‘cake’) precludes that even being a possibility. The ingredients have been committed to a particular form and the cake cannot be unbaked.
This position is held by those, for example, who argue from the idea of penal substitution for particular individuals, and rule out the possibility of double jeopardy. The notion is that on the cross the infinitely valuable life of Christ was sacrificed in the place of particular individuals, who cannot then be expected to pay the price themselves. Christ’s sacrifice is of potentially infinite value, but it is actually limited in its effect to the bearing of those sins that were imputed to Him, which were the sins of elect individuals alone. The sacrifice of Christ is of infinite value, but its provision is inherently limited to the elect.
For others the limitation of the atonement comes at the level of design. The provision of the sacrifice of Christ is not inherently limited, but is limited in regard to God’s design in the death of Christ. The sacrifice of Christ itself is like the ingredients for the cake in the previous illustration. There is nothing about the sacrifice of Christ considered in itself that rules out the possibility of God’s using it to save anyone or everyone. However, God has a set purpose for the sacrifice of Christ. He gave up His Son with the purpose of saving particular individuals rather than others. The provision of the atonement is not essentially limited, but is limited in the design for which it has been made.
A number who hold such a position will be quite comfortable speaking about Christ bearing the sins of the world, and using other such universal-sounding statements. There may also be a willingness to think in terms of multiple designs behind the atonement that greatly soften the particularism of an atonement governed by individual election. They will be able to acknowledge senses in which it is appropriate to speak of the cross truly being for all, whilst still maintaining that it is governed by the purpose of individual election.
There are others who, whilst believing in the doctrine of individual election, deny that the design of election should be regarded as the controlling design behind the atonement. The atonement is guided by a broader salvific purpose that is not limited (with respect to individuals). This need not mean that the design of the atonement is necessarily to save every individual — ‘unlimited’ atonement need not mean atonement with the purpose of saving each and every individual (although often it does). Such an atonement is ‘universal’, making provision that is sufficient for all and is not limited by any design to save one individual rather than another.
The fact that the atonement is limited with respect to its final result (i.e. in the individual elect) cannot be taken as proof that the atonement itself is limited. There is seen to be a need to hold universalism and particularism in tension in a way that limited atonement fails to. Such an approach, without denying individual election, seeks to deny it teleological priority. God has more important purposes than that of saving a particular set of individuals, even though that is one of His purposes.
Such an approach tries to do justice to the fact that God desires all men to be saved and seeks to demonstrate that God’s grace is not as essentially particularistic as most Calvinists claim. Amyraldianism falls under this category.
4 Denial of Limited Atonement not Denial of Election
I don’t believe that any of the three positions listed above is necessarily inconsistent with itself. I think that the third position is particularly important. It demonstrates that limited atonement is not necessarily implied by a particularistic understanding of election. To affirm limited atonement is to affirm more than a belief in unconditional individual election demands of us. For example, I am quite happy to affirm God’s comprehensive sovereignty, but totally unwilling to affirm limited atonement. This is not a debate about God’s sovereignty, but about the teleological priority that we are going to give to God’s decision to save particular individuals.
The problem that I have with the doctrine of limited atonement is its anthropocentricity. By subordinating the cross to the design of individual election we make the cross and God’s work in history in general to focus far too much on me and my salvation. God’s expansive covenant and redemptive historical purposes are given the backseat, whilst individual salvation takes the front seat. This is one area where I believe that the spirit of the Reformed faith must oppose the letter of the Reformed faith — ‘L’ in particular. God has a great and glorious purpose in His world that is focused, not directly on me, but on His Son. As an individual member of the human race I have been caught up in the powerful flow of God’s purposes, and I am profoundly aware that I am not ultimately what it is all about.
My understanding of the atonement is similar to that of the third position outlined above. The one important difference is found in the fact that I deny Reformed doctrines of election. Whilst affirming that, in His sovereign grace, God saves certain individuals rather than others, I do not believe that this is what the Bible means when it talks about election. Further differences will become apparent as I proceed. I have written at greater length on the subject of election here, here, here and here (however, there are many aspects of my position that I left unexplored in those posts that will be explored in this one).
5 Denial of Limited Atonement not Denial of Effectual Atonement
A denial of limited atonement need not be a denial of effectual atonement. The atonement is not a mere hopeful shot in the dark on God’s part. It is undertaken for a purpose and it fulfils that purpose. That purpose may well be a broader purpose than that of saving a specific set of individuals.
I will be arguing that one of the major purposes for the atonement is that of bearing the old humanity to death so that a new humanity could be formed. This purpose is perfectly fulfilled by the cross. The new humanity that the cross aims to form need not be formed of any one particular set of individuals. Whilst it will certainly be formed of particular individuals it is not the purpose itself that it be formed of those individuals. Indeed, it could conceivably be formed of all individuals. It is amazing how easily this simple point is misunderstood.
Once one has recognized that an effectual atonement need not be a limited one we can appreciate that the choice that we have here is not as simply as a choice between Calvinist limited atonement and Arminian possibility-realization atonement, where everything comes down to man’s decision and the cross is merely God putting the ball into our court. There are further options open to us. There are plenty of theologians within other Church traditions who appreciate this. The dilemma that Reformed people all too often work in terms of is simply unnecessary.
6 The Object of the Atonement
The atonement does not have a particular set of individuals as its direct object. Rather, it has a personal solidarity and those who belong to it as its direct object. Christ dies for His body, the Church, His flock, etc. We only become the objects of Christ’s atonement when we are brought into this personal solidarity (I will later explore ways in which we must biblically say that Christ died for all humanity).
Some might argue for a ‘limited atonement’ with Israel or the Church being the direct object, and individuals the indirect objects as they come to participate in the body. I would not really disagree with such an understanding in principle. I would just point out that this is a relatively novel way of understanding what it means for the atonement to be ‘limited’ and is not really the same as the common Reformed position under that name (at least not in my reading).
There are others who seek to argue that both individuals considered apart from the personal solidarity and the personal solidarity that they belong to are direct objects of election and the atonement. This is a slightly nuanced form of limited atonement, but limited atonement nonetheless. Even whilst there might be an appreciation of the corporate character of the solidarities for which Christ dies, the solidarities are generally regarded in a manner that makes particular individuals essential to their constitution.
7 What is a ‘Personal Solidarity’?
What exactly is meant by a ‘personal solidarity’? A personal solidarity is a unity that is shared or participated in by many persons. The unity is not an impersonal unity, but a personal one. The unity of the personal solidarity is secured, not merely by some legal framework or personal attribute or property that its members have in common, but by the living person or persons from whom the shared life originates. The existence of the personal solidarity, however, necessitates the existence of more persons than those from whom the life originates.
A personal solidarity does not work in terms of a simplistic individual/corporate divide. The individual/corporate paradigm suggests a tension between the personal ‘individual’ and the impersonal ‘corporate’. The personal solidarity model works in terms of a one/many distinction. However, the ‘one’ is one of the ‘many’ and is a person, not an impersonal system or social construct.
Within the type of personal solidarity that I will be speaking of the one and the many are mutually constitutive. There cannot be one without the other. The following are some examples. A (nuclear) family is a form of personal solidarity, whose life originates from parents and is participated in by children. There can be no parents without children and no children without parents. Both parents and children are essential to the existence of a family and both are members of the family. Israel is another example of a personal solidarity. Israel’s common life flowed from the patriarchs and was participated in by their seed and by those who were brought in from outside.
The Church (understood as the totus Christus — Christ, Head and body) is a further example of a personal solidarity and is the particular case that I will focus on within this post. The life of the Church flows from Christ the Head to the members.
The important thing that needs to be recognized about such personal solidarities is that some members are integral to the identity and essential to the existence of the solidarity in a manner that others are not. Whilst a family doesn’t cease to be a family when one of its children is adopted into another family or when a child from another family is adopted into it, there would be no family apart from shared parents. Whilst parents by themselves are not sufficient to form a family, parents are essential to the existence of the family and integral to its identity in a way that a particular child within the family is not.
The same can be seen in the case of Israel. Israel’s identity is derived from the patriarchs. If it were not for the patriarchs Israel as such would not exist. However, individual Israelites could be cut off from the nation, but the nation would still persist.
This is crucially important in our understanding of the Church. Christ is integral to the identity and essential to the existence of the Church in a manner that I can never be. I could be cut off from the totus Christus and the totus Christus would not cease to be the totus Christus. If Christ were cut off from the totus Christus, the totus Christus would cease to exist.
8 Covenant-Election Distinction
It is through its failure to think in terms of a personal solidarity model that the Federal Vision’s covenant/election understanding collapses, in my opinion.
One of the problems with many Reformed understanding of the object of election is that the object of election is understood in terms of an individual/corporate polarity. The Church is seen to be much like a bag containing many individual marbles. The marble/bag dichotomy corresponds to the individual/corporate dichotomy. The marble is conceived of as an entity with independent existence from the bag. God’s choosing of the bag would not necessarily mean that He had chosen the marbles inside the bag.
It seems to me that the Federal Vision’s (hereafter FV) covenant/election distinction is merely a variation on this basic theme and fails to truly overcome the root problem, which is the polarity itself. The covenant more or less functions as the bag and election has reference to the marbles. God’s election is the choice of particular marbles within his chosen bag.
The covenantal way of speaking speaks in terms of that which is within and that which is without the bag, because ultimately it will be the contents of the bag that will be saved. Those within the bag are ‘covenantally’ elect, though they may not be elect on a ‘marble’ level. The ‘bag’ gives us a conditional promise of unconditional ‘marble’ election. Those marbles which faithfully remain in the bag can gain assurance that they are among the chosen marbles.
I don’t hold to the FV covenant-election distinction. I did for some time, but it ended up raising more questions than answers. I do not believe that there really is a election-covenant / marble-bag / individual-corporate polarity at all. I believe that this is something that we have imposed upon Scripture, rather than something that arises from Scripture itself.
The FV position is certainly moving in the right direction by paying more attention to the place of the Church and the covenant in our understanding of election. In many respects, what the FV has done is reverse the traditional election-covenant polarity and privilege the covenant pole over the election pole (understanding election through the lens of covenant, rather than vice versa). However, the decisive step still has to be taken: that of rejecting the polarity completely.
Thinking in terms of a personal solidarity gives us a new way of looking at this. Within the personal solidarity model the members of the solidarity do not exist independently of the solidarity. To be a member of a family is not to be an individual element within a set, but to be a person bound up in certain relationships. For me to be a member of my family, for example, is for me to be a brother and a son. Abstract me from my family and I cease to be a brother and a son. It would be meaningless to speak of me as a brother or a son in such abstraction from my family. God’s choosing of a family is a choosing of particular persons, but only as they are in relationship with other family members and, most importantly, the Head.
The FV covenant-election approach really follows on from the traditional Reformed understanding of election, which tends to think of the objects of election in terms of individual members of a set. Consequently, even though it is argued that they can’t be identified, it is considered meaningful to speak of elect people outside of the Church and non-elect people within the Church.
My claim is that the object of election is a personal solidarity — the totus Christus. If I am right then it is quite inappropriate to speak of elect people outside of the Church (except perhaps in some very clever and sophisticated sense).
Election is essentially the choosing of a family. It does involve the choice of particular persons, as every family is composed of particular persons. However, individuals are not the direct object of election. They are chosen as members of the family, not as individuals considered in abstraction from the family. We should also recognize that the family cannot be viewed in abstraction from its members either. The object of election is not an abstract ‘set’ or ‘bag’, but a concrete people.
Once one has appreciated the character of personal solidarities a few further things emerge. If the object of election is the personal solidarity of the Church, then the place that individuals have within it is quite different from that which they have within the marble-bag model. Within the personal solidarity model the individual member of the Church is neither integral to the Church’s identity nor essential to her existence. Individual persons can be removed from the Church or brought into it without the Church ceasing to exist. Only Christ is integral to the Church’s identity and essential to her existence. By seeing election as akin to the choice of a bag containing particular marbles, Reformed doctrines of election have made particular individuals intrinsic to the identity of the chosen people of God.
9 Election Mediated by Christ
What results is an election that is not entirely mediated by Christ. Election must be exhaustively mediated by Christ if we are to be consistent with the Scripture and have true assurance of our own status. We become elect when we are united to Christ in history. We would cease to be elect if we were cut off from Him. The Book of Life is Christ. Those in Him are elect; those outside of Him are not. We have no elect status apart from Him. By retaining a covenant-election distinction the FV acknowledges the existence of elect people (in some sense) outside of Christ. Such people’s election is not exhaustively mediated by Christ.
They are elected as individuals to be in Christ, rather than elect because they are in Christ. There is a very important distinction between these two positions. Some seek to get around this by arguing that we are elected in Christ as individuals before the foundation of the world. This seems to arise from a misreading of Ephesians 1:3-4. It also damages the doctrine of union with Christ by making it a decretal union at root. The elect have always been united to Christ in a decretal sense, even when they are alienated from Christ in every other sense.
The problem with this proposed solution is that Christ still does not exhaustively mediate election. What we have is a large static and fixed group of individuals in decretal union with Christ being the object of election. This seems to be the type of problems that you will be left with when you make the particular individuals that constitute the body of Christ essential to its identity. I am arguing that this is an unbiblical route to take. The Church derives its essential identity from its Head, not from its members. The object of election is the totus Christus, which is formed by personal union with Christ by the Spirit, rather than merely being a decretal union. If this is the case then we only become elect when we are brought into this personal union. At no point are we essential to the identity of this personal union.
What constitutes the decretal elect’s identity? To be the true group of the elect, according to Reformed theology, it would have to contain a fixed number of particular individuals in union with Christ. The presence of elect person Joe Bloggs is essential to the existence of the decretal solidarity. Remove Joe Bloggs and the decretal solidarity is no longer truly present.
Incidentally, understand election in the way that I am suggesting and you need no longer believe in double predestination; nor do you need to believe that the Fall is necessary for the fulfilment of the decree. God could have fulfilled the decree of election — understood as the election of a personal solidarity — without a single individual perishing. By giving such teleological priority to the choice of the individuals who will finally constitute the totus Christus Reformed theology has set the purpose of God on its head.
If election is the choice of fixed number of particular individuals then it is essentially hidden, because its object is unknown. We do not know the particular individuals that God has decreed to save. Christ enacts a decree that remains hidden behind Him. Although we know that all of the members of the decretal union will one day be brought into full personal union with Christ we cannot truly see God’s purpose of election revealed in the personal union itself because the purpose is not the creation of a personal union per se, but the creation of a personal union with individuals B, F, Q and Y rather than others.
I am arguing that it is the creation of a personal solidarity that is God’s electing purpose. Seen this way the object of election is the totus Christus and its members. This decree of election is completely revealed in Christ. We know the object of God’s election. We can know that we are elect. We can also know that others are elect (Philippians 4:3). There is no election that lies behind the personal covenantal union that we enjoy with Christ. In retaining the FV covenant-election distinction we are holding to a hidden decree that is not fully mediated by Christ. It seems to me that it is as simple as that.
Incidentally, it should be noticed that I am not denying the equal ultimacy of the one and the many in the totus Christus. The One — Christ — is always constituted by the many, just as He Himself constitutes them. What is being denied is that the particular individuals that are within the one-many solidarity of the body of Christ have an equal ultimacy.
10 The Salvation of Particular Individuals
It might be questioned what difference all of the above makes for one who holds that God orders ever single event in history. Surely we will end up with the same thing: God in His sovereignty saves some and not others and we have no way of knowing which He will save and which He will not and, more importantly, which of the two categories we fall into. I believe that it still makes an important difference.
First, the great purpose of election that God is seeking to work out is fully revealed. Whilst we may not know the exact way that God is going to work particular events to that great purpose, we do know the final goal. The exact set of individuals who will be saved simply does not have the teleological centrality that it does in most Reformed systems.
Second, we do not have a fixed position relative to God’s purpose. We can either positively participate in its outworking in believing allegiance to Christ or we can resist it. Such a purpose calls for a response of faith on our part.
Thirdly, we can know where we stand relative to God’s election. We know this by virtue of our relation to Christ. There is no more fundamental decretal union hiding behind covenantal union with Christ. The Church is the elect; those outside the Church are reprobate. There are no elect outside of the Church and no non-elect within the Church. My status as elect or reprobate and my knowledge of that status are both completely mediated by my relationship with Christ. The hidden manner in which God’s providence guides my life is no threat at all, because I know that God’s providence is ultimately focused on the outworking of His good purpose in Christ Jesus, which is no secret. I do not need to question whether God’s purpose is my salvation or damnation.
Fourthly, I do not believe that the decree of God (understood as His determination of the outcome of all events) takes place in eternity past. It is not a closed and completed event and does not take place in our absence. Consequently, God’s providential ordering of my life is very much a relational reality and is not deterministic or fatalistic in the slightest. I have argued that the idea that God has determined the outcome of all historical events in eternity past tends to collapse into a form of deism (towards the end of the sixth comment here).
Understood this way, the God who secretly orders the events of my life is — if I am a Christian — a God who is known to me in Christ. I know where I stand with this God and I can be certain of His love for me, because He has expressed His love for me in Christ. I can be sure that His intentions for me are good. The problem with the covenant-election distinction is that I can never be exactly sure where I stand with the electing God. His love expressed to me in covenant seems genuine enough, but I can never be certain that I have not been chosen for damnation and His purposes for me are not good. The face of the electing God is always hidden from me.
The FV covenant-election distinction tells us not to speculate about the hidden things, but to concentrate on what God has revealed and recognize the manner in which we can come to a knowledge of our election by means of what God makes known in the covenant. This is an improvement on most popular Reformed positions, but such a position still has the face of the electing God obscured. No matter how comforting God’s words to me in covenant may appear, I know that the final word is the word of election and I don’t know what that word says. Behind all of the covenantal love of Christ we have the Deus Absconditus of election and this is terrifying. Both Calvin and the FV rightly insist that we must look to Christ, but until they show that the electing decree is open and revealed in Christ they are failing to solve the root problem.