Current discussion on the nature and substance of covenant theology has centered around biblico-theological as well as systematico-theological questions. While these discussions will have some bearing upon the subject of this paper, I want to focus mainly on the philosophical and apologetical ramifications of the concept of covenant. More narrowly, I want to investigate into the relationship between the concept of the covenant and epistemology. I will argue that the theological concept of the covenant can provide some needed solutions to epistemological impasses in philosophy. If man is in everything everywhere and always “coram deo” and epistemology is at bottom covenantally construed, this will have consequences for our apologetics. The results of the best of Reformed covenant theology, both biblico-theologically and systematico-theologically, will then feed into a uniquely biblical epistemology which will, in turn, ground a powerful apologetic for our times and issue a covenantal call to repentance.

1. The Covenant as a Theological Concept

Broadly historically speaking, Reformed orthodoxy (and Reformed theology since then) has seen the history of redemption through the lenses of three covenantal arrangements: the covenant of redemption (pactum salutis) made in eternity past among the members of the trinity; the covenant of works made between God and Adam in the state of original innocence; and the covenant of grace made between God and the elect in Christ. All three lenses were understood to be guiding and magnifying our vision of the glorious gospel of Christ. And all three covenantal arrangements were seen to be working in concert, rather than in opposition to each other. Presently, in an attempt to reconstrue covenant theology, two of the three lenses have been challenged from within the Reformed camp. The covenant of redemption, on the basis that there is no biblical warrant for it, but that it is rather a philosophical and speculative concept;¹ and the covenant of works, on the same basis but also on the basis that it runs contrary to the gracious character of God’s dealings with man and brings an aspect into the history of redemption, namely the “works principle,” that is utterly foreign to it.² While full-fledged discussions of the covenant of redemption and the covenant of works merit separate papers, we need to establish the validity of these theological concepts before we can talk about the full extent and meaning of the covenant for theology and apologetics.

1. 1. The Covenant of Redemption (CoR)

In Reformed orthodox theology, the CoR (or pactum salutis) has been defined as “the pretemporal, intratrinitarian agreement of the Father and the Son concerning the covenant of grace and its ratification in and through the work of the Son incarnate. The Son covenants with the Father, in the unity of the Godhead, to be the temporal sponsor of the Father’s testamentum in and through the work of the Mediator.”³ What is the use of this doctrine? It is “to emphasize the eternal, inviolable, and trinitarian foundation of the temporal foedus gratiae.”⁴ Far from being a speculative concept, the pactum salutis grew out of reflections upon the origin of the covenant of grace and the relation of that temporal covenant arrangement to the eternal decree of God. In that sense, the CoR was and is understood to grant historical meaning to the covenant of works (CoW) and the covenant of grace (CoG). The concept is biblically grounded in passages such as Psa. 110, John 5:30; 6:38-40; Lk. 22:29-30; John 17 and Gal. 3:20. Kline explains that in general, “Messianic psalms reveal to us the eternal communion between the Father and Son, in which the Father covenants to the Son a kingship on Zion over the uttermost parts of the earth (Psa. 2:6-9) and grants him by oath an eternal royal priesthood (Psa. 110:4; cf. Heb. 5:6; 7:17,21).”⁵ Far from simple proof-texting, these exegetical considerations ground the CoR deeply within the redemptive-historical flow of covenant theology.

Some Reformed theologians (e.g. A.A. Hodge) have repudiated the eternal/temporal covenant construction and thereby done away with a pactum salutis altogether. Hodge argues that Our Standards say nothing of two covenants. They do not mention the covenant of redemption as distinct from the covenant of grace. But evidently the several passages which treat of this subject (Conf. Faith, ch. 7., s. 3; L. Cat., q. 31; S. Cat., q. 20) assume that there is but one covenant, contracted by Christ in behalf of the elect with God in eternity, and administered by him to the elect in the offers and ordinances of the gospel and in the gracious influences of his Spirit. The Larger Catechism in the place referred to teaches how the covenant of grace was contracted with Christ for his people. The Confession of Faith in these sections teaches how that same covenant is administered by Christ to his people.⁶

But does the teaching of the Westminster Confession really contradict a pactum salutis? We think not because in the place, which Hodge mentions (VII.3), the confession only talks about the CoG. And the Standards themselves teach that this CoG, according to the Larger Catechism (Q. 31) “was not always administered after the same manner.” While we do think that the CoR is a subset of the CoG, the two concepts must not be collapsed into each other. Kline says thatthough interlocking, these two redemptive covenants, the eternal and the temporal, are nevertheless to be clearly distinguished from each other for they differ in several most basic respects. In the eternal covenant, (1) the Son is assigned the role of covenant servant; (2) the second party is the Son in his status as second Adam and thus, included along with him, the elect whom he represents, and them exclusively; and (3) the operative principle is works. Contrariwise, in the series of historical administrations of the gospel, (1) the messianic Son is Lord and mediator of the covenant; (2) the second party is the church, the community of the confessors of the faith and their children, including others beside the elect; and (3) the operative principle is grace (Kline, Kingdom Prologue: Genesis Foundations for a Covenantal Worldview, 138).

Thus, the CoR and the CoG differ with respect to time (eternal vs. temporal), with respect to the parties involved (Christ as servant vs. Christ as Lord and Mediator), with respect to the beneficiaries (Christ and the elect vs. the church consisting of both elect and non-elect) and with respect to the principle at work (works vs. grace). Therefore, we agree with Kline in his summary: “Enough of the evidence has been cited to show that the biblical theologian will certainly want to identify these eternal commitments between the Father and Son as a covenant.”⁷ And he will also want to distinguish clearly between the CoR and the CoG, the former being the ground of the latter, the latter being the historical outworking of the former.

1. 2. The Covenant of Works

The importance of the doctrine of the CoW becomes clear when it is pondered that “Covenant theologians have generally taken the position that the covenant concept can accommodate the entire history of the kingdom of God.”⁸ If there was no CoW, obviously, the concept of covenant could not be used as a unifying concept for biblical theology and hermeneutical considerations.⁹ It is essential to our project of defining a covenantal epistemology that we establish the biblical validity of the doctrine of the CoW.

We must begin our inquiry by asking: What is a covenant? What are the minimal requirements, the non-negotiables that make something a covenant? How could there be a covenant in Gen. 1-3 when the word does not even appear? Obviously, there must be more to it than simply doing a concordant search for all the occurrences of the word tyîrV;b (berith) or diaqhkh (diatheke) respectively. As Walther Eichrodt so cogently observed, in discussions on covenant: “The crucial point is not - as an all too naive criticism seems to think - the occurrence or absence of the Hebrew word berit.” ¹⁰ For the Reformed orthodox, the doctrine of the CoW was not derived from simple proof-texts out of Genesis 1-3 at all. On the contrary, if anything the Reformed orthodox theologians got the concept of the CoW mainly from Romans 5, working their way backwards biblico-theologically into the prelapsarian situation.¹¹ Systematic theology from Calvin onward has drawn from the well of biblical theology in order to conceptualize and systematize the biblical teaching on the covenant(s).

We may take Francis Turretin as an example of this rule. In his Institutes of Elenctic Theology, he takes up a common, minimalist definition of covenant. Covenant is “a mutual agreement between two or more persons concerning the mutual bestowal of certain goods and offices for the sake of common utility.” ¹² This agreement involves three particular features: (1) equality of status; (2) ability to fulfill the things required; and (3) the fact that without the particular agreement there is no duty to perform those things. According to this definition, then, Turretin concludes that there is not and indeed cannot possibly be a covenant between God and man, strictly speaking.¹³ But here we are immediately introduced to Turretin’s subtle and nuanced writing. When he says there is no covenant between God and man “strictly speaking,” he does not mean there is none at all. ¹⁴In fact, he is adamant in pointing out that God indeed entered into covenant with man. However, this covenant is (1) not one of parity and equal status, (2) man brings no absolute natural ability with him, and (3) man is always in duty bound to obey God - with or without covenant. ¹⁵

Therefore, the qualified definition Turretin provides reads as follows: “Strictly and properly, covenant denotes the agreement of God with man by which God promises his goods (and especially eternal life to him), and by man in turn, duty and worship are engaged (certain external signs being employed for the sake of confirmation).”¹⁶ On this definition, the covenant under consideration is a bilateral, mutual one, but not because of parity of persons, but rather because of the mutual obligations of the contracting parties: for the first, God, to keep his promise, for the other, man, to be obedient.¹³ As Turretin has pointed out, there can never be a strictly mutual pact between God and man founded on a parity of persons. The Creator/creature distinction basic to Christian theology forbids this. Nevertheless, says Turretin, God did in fact condescend and enter into covenant with man.¹⁷

The line of proof Turretin offers for the existence of an antelapsarian covenant can be summarized as follows: (1) the “essential parties” are there: God and man. Turretin sees several things being given in that very relationship: God is the Creator of man and therefore must be and in the nature of the case is (a) his governor: the Sovereign administering the context of Adam’s life from the beginning; (b) his lawgiver: establishing the terms of life and of the mutual relationship between the two covenant partners; and (c) his rewarder: rewarding obedience to the preestablished terms (law) of the covenant.¹⁸ These are the natural aspects of the “metaphysical” situation from the beginning or, again, of the Creator/creature distinction;¹³ (2) the fact that the initial antelapsarian relationship between God and Adam involved the giving of a law, God imposing legal requirements upon Adam, “implies a federal agreement and contract.” Law, for Turretin, involves a covenantal arrangement, because it binds the one under the law to obey its rule and the one giving the law to rule according to the law given (versus arbitrarily); (3) Hos. 6:7: though exegetically he allows for the possibility (or even preferability) of construing this to mean generic “men” instead of Adam as a historical person, “nothing prevents [these words’] being referred also to Adam (that they may be said to have violated the covenant like Adam…)”¹⁹ In fact, if Adam was a historical person, which Turretin would certainly affirm, then he must be included in the group of people termed “man” in Hos. 6:7 and therefore he was in covenant with God also; (4) “Such a covenant was demanded not only by the goodness and philanthropy of God . . . , but also by the state of man and the desire of happiness impressed upon his heart by God.”²⁰

Muller states that there were three non-negotiables of the concept of covenant among the Reformed orthodox: “As instituted by God, covenant has three aspects - a promise, a condition, and a sanction.”²¹ The promise is eternal life, the condition obedience performed, the sanction death. All three are fully present in the address of God to Adam. Two contracting parties, though not equal in status, enter into a two-sided agreement in which both are obliged to fulfill their part, God placing his requirements upon Adam, threatening death upon breach of the covenant and, by implication, life upon obedience.²² If Muller is correct, these essential and non-negotiable elements of a covenant are according to Turretin most certainly present²³ in the antelapsarian arrangement between God and man. Thus, we are not only granted but required to speak of an antelapsarian CoW.

1. 3. Summary

Now that we have our basic triple-covenant structure (CoR, CoW, CoG) firmly in place, we can draw some theological conclusions. We can now say that the theological concept of covenant encompasses all of history, including pre-lapsarian and redemptive history. Moreover, it also provides a way to relate the pretemporal or eternal plan of redemption to the historical outworking of it under the umbrella of covenant. We can now agree with B. B. Warfield who remarked that “[t]he architectonic principle of the Westminster Confession is supplied by the schematization of the Federal theology, which had obtained by this time in Britain, as on the Continent, a dominant position as the most commodious mode of presenting the corpus of Reformed doctrine.”²⁴

Before we continue we must add an important caveat. We can do this by asking whether all we have said so far justifies us or even forces us to view the covenant as our Zentraldogma?²⁵ If Reformed theology is covenant theology must we not say that absolutely everything in our theology is covenantally conditioned or defined? Over against some theologians who have rejected the covenant as a means to describe the intratrinitarian relationships altogether, there have also been some who have taken it too far the other direction and turned it into a speculative concept.

David VanDrunen, for example, has recently argued for the “centrality of covenant for Westminster systematics” in The Pattern of Sound Doctrine.²⁶ What he seems to mean by that is that the covenant is not only our principle in organizing the corpus of revealed truth, but actually the substance of all theology. But that leads us into dangerous waters because it clearly confuses aspects of the immanent trinitarian relations with ontological categories. We have said above that the pactum salutis is the eternal covenant of redemption between the members of the Trinity. So the covenant does penetrate even into eternity. In fact, we have said that the eternal covenant only gives meaning to the redemptive-historical covenantal arrangements. However, we must be very clear here: the pactum salutis is part of the opera dei ad extra, is the expression of his will with respect to the creation. Deus ad extra and deus pro nobis are important ideas but have distinctively soteric functions in view.²⁷

Once we define covenant as Zentraldogma, we want to see everything through this lens, and fail to make proper distinctions between God ad extra or pro nobis and God ad intra or in se, between the ontological Trinity and the economic Trinity.²⁸ We dare not construe the eternal generation of the Son, for example, along covenantal lines. We dare not construe the spiration, i.e. procession of the Spirit, as covenantal essentially. The concept of covenant is not exhaustive of theology proper because God is not covenantal essentially. Rather, as we shall see, the covenant is a function of the Creator/creature distinction and we read it back into the self-contained Trinity only on pain of disregarding this vital and basic distinction. Since covenant denotes the ontological and volitional freedom of God with respect to creation and redemption, it cannot apply to the ontological Trinity. God’s intratrinitarian relations (e.g. the eternal filiation) are not free but necessary.²⁹ In the final analysis, the self-contained ontologically trinitarian being of God is actually the necessary foundation for the covenant rather then being covenantal in itself. The archetypal being of God is the foundation to the ectypal being of man. The non-covenantal, self-contained God in se as Trinity is the most basic metaphysical “category,” and not the covenant or the trinity ad extra. Therefore, we must turn to metaphysical considerations under the next heading and show how the archetypal/ectypal and the Creator/creature distinctions ground the idea of covenant but go beyond it and effectively qualify it.

1. It is interesting to note that Robert Letham’s highly acclaimed book The Holy Trinity makes mention of the covenant of redemption only once and in that instance only as a working hypothesis of Warfield which, supposedly, was never substantiated (Robert Letham, The Holy Trinity: In Scripture, History, Theology And Worship [Phillipsburg: Presbyterian & Reformed, 2005], 401).↩
2. Representative of the latter opinion are, among others, the following: John Murray, Lectures in Systematic Theology, Collected Writings of John Murray, , Vol. II (Edinburgh: Banner of Truth, 1978), 47-59; Herman Hoeksema, Reformed Dogmatics (Grand Rapids: Reformed Free Publishing Association, 1966), 214-226; and Daniel Fuller, The Unity of the Bible: Unfolding God’s Plan for Humanity (Grand Rapids: Zondervan, 1992), 181-82. See also the “Federal Vision” representatives, such as Rich Lusk and James Jordan; cf. James Jordan, “Merit vs. Maturity: What Did Jesus Do for Us?,” in The Federal Vision, eds. Steve Wilkins and Duane Garner (Monroe: Athanasius Press, 2004), 192; Rich Lusk, “A Response to ‘The Biblical Plan of Salvation,’” in The Auburn Avenue Theology: Pros and Cons. Debating the Federal Vision. The Knox Theological Seminary Colloquium on the Federal Vision, ed. E. Calvin Beisner (Ft. Lauderdale: Knox Theological Seminary, 2003), 119.↩
3. Richard A. Muller, A Dictionary of Latin & Greek Theological Terms: Drawn Principally from Protestant Scholastic Theology (Grand Rapids: Baker, 1996), 217.↩
4. Ibid., 217.↩
5. Meredith G. Kline, Kingdom Prologue: Genesis Foundations for a Covenantal Worldview (Overland Park: Two Age Press, 2000), 139.↩
6. Archibald A. Hodge, The Confession of Faith: A Handbook of Christian Doctrine Expounding the Westminster Confession (Edinburgh: Banner of Truth, 1964), 126-127.↩
7. Ibid., 140.↩
8. Ibid., 14.↩
9. While not building his case for the CoW upon it, Kline says “it is possible that the Bible itself, in later references back to Genesis 1-3, applies the term berith to the situation there.” (Ibid., 14)↩
10. Quoted in Thomas Edward McComiskey, The Covenants of Promise: A Theology of the Old Testament Covenants (Grand Rapids: Baker, 1985), 216.↩
11. Cf. Richard A. Muller, “The Covenant of Works and Stability of Divine Law in Seventeenth Century Reformed Orthodoxy: A Study In the Theology of Herman Witsius and Wilhelmus À Brakel,” in After Calvin: Studies in the Development of a Theological Tradition, (Oxford: Oxford University Press, 2003), 183.↩
12. Francis Turrettin, Institutes of Elenctic Theology, Vol. 1 (Phillipsburg: Presbyterian & Reformed, 1992), 574.↩
13. Cf. ibid.↩
14. Again and again, in the discussion on covenant and merit, we encounter Turretin’s strict definition of the word “strict,” and we do well to heed this little word.↩
15. We shall see later that this “in duty bound” is in fact the essence of the covenant from man’s perspective. So the thought of man relating to God (and vice versa) apart from covenant is hypothetical.↩
16. Ibid., 574.↩
17. Cf. ibid., 574. The Westminster Confession also speaks, in language reminiscent of Turretin’s, of “some voluntary condescension on God’s part, which He has been pleased to express by way of covenant.” (WCF, VII.1)↩
18. Concerning the reward promised to Adam, Turretin says that God would “hold forth a reward to him for keeping [the law] (although bound by no right to that).” (Ibid., 575) What he means by “bound by no right” is that God is not absolutely bound to promise and give a reward to man, for the reasons given above (that God is, by way of fiat creation already governor and lawgiver and man has no claim on him). But Turretin does not mean to deny a consequent necessity (rooted in God’s covenantal faithfulness) to bestow the reward once it has been promised.↩
19. Ibid., 576.↩
20. Ibid. What Turretin means by the CoW being “demanded” is, again, consequent necessity, not absolute necessity per se.↩
21. Richard A. Muller, “The Covenant of Works and Stability of Divine Law in Seventeenth Century Reformed Orthodoxy: A Study In the Theology of Herman Witsius and Wilhelmus À Brakel,” in After Calvin: Studies in the Development of a Theological Tradition, (Oxford: Oxford University Press, 2003), 180.↩
22. Turretin concludes by good and necessary consequence that the CoW involved both the threat of punishment (death), which is explicit in the text, as well as the promise of eternal life, which is implicit in the sacramental sign of the tree of life and further expounded as a principle in passages such as Lev. 18:5 (Cf. Turrettin, Institutes of Elenctic Theology, 576).↩
23. Even hyperfulfilled, because Turretin sees four requirements for a covenant being given in Gen. 1-3.↩
24. Benjamin Warfield, The Westminster Assembly and Its Work (New York: Oxford University Press, 1931), 56.↩
25. Otto Weber has argued for predestination as the Zentraldogma in Calvin’s theology (Otto Weber, Grundlagen der Dogmatik, Vol. 2 (Neukirchen: Neukirchener Verlag des Erziehungsvereins, 1962), 473). David VanDrunen thinks for Reformed theology the Zentraldogma should be the covenant (David Vandrunen, ed., The Pattern Of Sound Doctrine: Systematic Theology At The Westminster Seminaries: Essays in Honor of Robert B. Strimple [Phillipsburg: Presbyterian & Reformed, 2004], 195-220).↩
26. Ibid.↩
27. Thanks to Dr. [Lane] Tipton who pointed this out in class.↩
28. This same criticism applies to Karl Barth for whom revelation in Christ (das Christusereignis) is a kind of Zentraldogma, for all intents and purposes, which leads to an identification of the immanent and the ontological trinity (cf. Rahner’s rule).↩
29. Contra Barth again. Barth tries to get around the implications of his own voluntaristic trinitarianism, but this writer believes he does so unsuccesfully (cf. Karl Barth, Die Kirchliche Dogmatik, Vol. I:1, Die Lehre vom Wort Gottes [Zürich: Evangelischer Verlag Zollikon, 1955], 456).↩