Where the world comes to study the Bible

The Witness of the Spirit in Romans 8:16: Interpretation and Implications

A comprehensive examination of the usage of summarturevw in Greek literature, as well as other exegetical evidence, leads to the conclusion that this verb means “bear witness to” in Rom 8:16. The passage thus is affirming that the Holy Spirit has an ongoing witness to our inner being. One implication of this is that we have assurance of salvation not just because of the word of God but also because of the testimony of the Spirit to our hearts. There are also implications for perseverance of the saints and for believers working at the highest levels of scholarship (e.g., it is dangerous to think that mere exegesis will give one certainty; ultimately, a vibrant relation to God through the Holy Spirit must be at the heart of loving God with one’s mind).

Introduction

By way of introduction to the discussion of Rom 8:16, I want to talk about a larger issue. Consider this introduction as the saddle burr.

The Church at the beginning of the 21st century is facing several crises, many of them of its own making. Among these are the crisis of the Spirit and the word. Non-charismatic evangelicals typically give their allegiance to the word; charismatics, to the Spirit. It’s like being back in Corinth: “I’m of Paul”; “I’m of Apollos”; “I’m of the Bible”; “Yes, but I’m of the Holy Spirit.” This is not just an issue of authority, but even of how one defines what it is to be a Christian, both individually and communally. The name says it all: “Bible Church”; “Holy Ghost Cathedral of Praise.” One church spends the majority of its time studying the Bible; another spends most of its time praying and singing. One church sees Christianity through primarily cognitive lenses; the other, through emotional lenses.

Perhaps I have exaggerated the portrait just a bit. Still, it is true that we have two very different brands of evangelical Christianity today. Ironically, both brands, I believe, mimic culture. The first is, to some degree, a product of the Enlightenment in which reason reigns supreme. This is true even in many so-called “Bible churches.” “Bible” becomes a synonym for a particular set of interpretations about the Bible. Sometimes in our zeal to be biblical we forget what it means to be spiritual. Truth is prized more than love; interpretation takes the place of application. (And too often there are particular pet views that are unshakeable, in spite of the evidence to the contrary.) One lady in my church touched the heart of the problem when she facetiously declared, “I believe in the Trinity—the Father, Son, and Holy Bible.”

The second brand of evangelical is more a product of postmodernism—even before postmodernism was in vogue: personal experience reigns supreme. Ironically, many who were leaders in the Bible church movement are now spearheading the signs and wonders movement.

Now that I have succeeded in stepping on everybody’s toes, let me proceed. The topic of this essay is “The Witness of the Spirit in Romans 8:16.” I will offer both an interpretation and several implications from this text. I may be wrong in both. I don’t live in a vacuum either. Culture has impacted me, too—although I’m sure that it has not impacted me as much as it has impacted you: I’m from southern California, where there is precious little culture!

I ask you to hear me with an open mind and an open heart. We are all creatures of our culture; we are driven by our own presuppositions. As Bultmann noted, half the battle in exegesis is being able to ask the right questions, being able to challenge our own presuppositions. (Whatever Bultmann’s faults were—and they were legion—he did know how to ask the right questions.) The other half, though, is bowing to the evidence, rather than manipulating the data. After all, virtually all heterodoxy is based on what is possible, not on what is probable.

The question that this passage raises is this: What is our authority today? What is it that guides us and to which we subject ourselves? What gives us comfort and assurance? On what do we base the assurance of our salvation? Is it the word? Or is it the Spirit? Or is there some symbiotic relationship between the two that deserves exploring?

Put another way, Is our faith totally objective? Is Christianity only a logical and empirical reality, or is there not a mystical element to it as well? The answer that I will propose is not new, but it is increasingly being abandoned, especially among cessationists today.

Interpretation

Romans 8:16 reads simplyaujtoV toV pneu'ma summarturei' tw'/ pneuvmati hJmw'n o{ti ejsmeVn tevkna qeou'. There are two possible translations at the crucial juncture. Either “The Spirit himself bears witness with our spirit that we are God’s children,” or “The Spirit himself bears witness to our spirit that we are God’s children.”

Grammatically, the issue is simply this: Is tw'/ pneuvmati (to„ pneumati) a dative of association (“with our spirit”) or a dative indirect object (“to our spirit”)? Exegetically and theologically, the issue may be far deeper: If a dative of association is in view, then our spirit joins God’s Spirit in bearing witness that we are God’s children, that we are saved. But to whom is this witness made? Many argue that such a witness is made to ourselves (thus, “the Spirit bears witness along with our spirit to us that we are God’s children”). On the other hand, some argue that such a witness is made to God. In this construct, there is no witness of God’s Spirit to us. Rather, both “spirits” testify Godward; both are advocates of our status before the great Judge.

If, on the other hand, a dative indirect object is in view, then God’s Spirit is testifying to our spirits, that is, to us. In this case, believers are the recipients of the testimony of the Spirit.

The first view (what we will call the associative view) may imply that the Spirit has nothing to do with the believer’s assurance of salvation.1 This is especially the case if the witness is Godward. The second view (what we will call the indirect view or indirect object view) certainly implies that the Spirit’s testimony to the believer is an important aspect of assurance. The first view allows one to claim assurance based directly on the objective data, the word. The second view opens the doors to a soft mysticism, suggesting that though the word is essential to assurance, God’s Spirit is also essential to offer such comfort.

Arguments for “with our spirit”

Scholars are divided on this issue, even evangelical scholars. However, it seems that the predominant view is the dative of association view (“with our spirit”). Most translations take this view. Thus, “with our spirit” is the reading of the av, asv, nasb, rsv, nrsv, esv, nkjv, hcsb, niv, tniv, jb, njb, and Moffatt. It is also adopted by many commentators today such as Stifler, Shedd, Hendricksen, Dunn, Fitzmyer, Schreiner, probably Moo, and possibly Stuhlmacher. Others, too, seem to be in favor of this view.2

The most cogent argument for this view is found in Gordon Fee’s massive work, God’s Empowering Presence: The Holy Spirit in the Letters of Paul. He argues as follows: To take the dative as an indirect object is an “unnecessary expedient that abandons Pauline usage for the sake of a prior theological concern that is not involved here.”3 He draws the negative conclusion that “This means that those who make much out of the concept of ‘the inner witness of the Spirit’ are probably also missing Paul’s point.”4

The main arguments for the associative view are as follows:

1. In keeping with Deut 19:15, two witnesses are needed to establish the truth of a matter. Thus, the Holy Spirit and our spirit must give a combined testimony to confirm our salvation.5

2. To argue that the Spirit bears witness to our spirits seems to presuppose that the moment of salvation is in view. But the present tense (summarturei', summarturei) argues against this; this is an activity that is ongoing in the life of the believer. Hence, “to our spirit” cannot be the meaning of the text here.

3. The verb itself suggests an associative idea: it is marturevw (martureo„) prefixed by the preposition suvn (sun). Such verbs regularly take datives of association.6

Arguments for “to our spirit”

The indirect object view is not without its representatives. For the most part, English translations are against it,7 but other translations (such as the Vulgate8 and Luther’s Bible9) are often for it. As well, several notable scholars have adopted this view, such as Luther, Calvin, Leenhardt, Godet, Hodge, Strathman (in TDNT), Morris, Murray, and Cranfield. The view is well represented, though more so among older, Reformed works than recent writers.

I think that this is the correct view. I wish to first interact with the arguments mentioned above; then, offer some further evidence on behalf of the indirect object position. The arguments are as follows.

1. To see Deut 19:15 as part of the background of this verse is unnecessary. It seems to presuppose that the Spirit’s testimony is not good enough if offered by itself. This also presupposes that our testimony, in combination with the Spirit’s, is good enough! But elsewhere in the NT a single testimony is often acceptable, especially one offered by God.10 Indeed, the vast bulk of texts in which the Spirit bears witness rest the claim made on that witness alone; no other witness is necessary.11

Further, if a second witness is sought in this text, there is a better candidate than our own spirits. Paul uses the verb marturevw only twice in Romans, only once prior to chapter eight. In what is his central passage on justification, 3:21–26, he notes that “the righteousness of God has been witnessed by the Law and the Prophets.”12 This righteousness is applied to “all who believe” (3:22). There, the justified state of the believer is witnessed by the scriptures; here, it is witnessed by the Spirit. It is precisely this twofold testimony—that of the Spirit and the word—offered at two key junctures in Romans, that constitutes the basis of the believer’s assurance.

But I do not think that that is Paul’s point here. I do not think that he is looking back five chapters to find a second witness. Let me repeat: the necessity of having a second witness is based on two assumptions: (1) that summarturevw (summartureo„) means bear witness together with, and (2) that the Spirit’s testimony is inadequate to confirm the truth of our salvation. The first of these assumptions is probably wrong, and the second is not in the picture here. But even if the first assumption is correct, this does not mean that “our spirit” needs to be that second witness.

2. It is erecting a straw man to say that the indirect object view only applies to the moment of conversion. To be sure, it does apply to that moment. But it also applies later. We should give the present tense, summarturei', its full force. But as such, it is rather broad. The present tenses in this chapter that refer to the Spirit consistently are used of the entire time period from regeneration to glorification. Note for example:

  • “You, however, are not in the flesh but in the Spirit, if indeed the Spirit of God dwells [oijkei', oikei] in you. Now if anyone does not have [e[cei, echei] the Spirit of Christ does not belong to him” (v. 9)
  • “the Spirit… lives [oijkei'] in you” (v. 11)
  • “all who are led [a[gontai, agontai] by the Spirit are the sons of God” (v. 14)
  • “the Spirit helps [sunantilambavnetai, sunantilambanetai] us… the Spirit intercedes [uJperentugcavnei, hupentugchanei] for us… the Spirit intercedes [ejntugcavnei, entugchanei] on behalf of the saints” (vv. 26, 27)

It is self-evident from these verses that the Spirit’s inner witness is part of the process of sanctification and encompasses the time from the spiritual cradle to the physical grave. Is not his intercessory ministry true for our entire lives, from the time we were converted? Does he not dwell in us from day one?

3. The lexical argument is the most compelling—namely, that suvn- prefixed verbs take datives of association. This is, prima facie, what the text is speaking about. This is the view that even first-year Greek students learn. But we need to nuance our view of the syntax here. Specifically, there are five problems with this assumption.

First, even if a suvn- prefixed verb does take a dative of association, this does not mean that it cannot take an indirect object or some other dative use. In my Greek Grammar beyond the Basics: Exegetical Syntax of the New Testament, I note that “not every dative following a suvn- prefixed verb is a dative of association…”13 A dozen examples are listed to prove the point, but several more could be mentioned as well.14

Second, a number of suvn- prefixed verbs have lost their associative force in Koine Greek. Sometimes the compound verb is weakened, becoming synonymous with the simple verb. At other times, the prepositional prefix functions much like other prepositional prefixes, viz., to intensify or strengthen the force of the verb. Students of Greek are well aware of examples such as katesqivw (katesthio„) as an intensification of ejsqivw (esthio„, “to eat” becomes “to devour”). This same kind of transformation occurs with suvn- prefixed verbs on occasion. Thus, for example sumbaivnw (sumbaino„) means “happen to,” not “go with” as in “it has happened to them [sumbevbhken aujtoi'", sumbebe„ken autois] according to the true proverb, ‘a dog always returns to its vomit’” (2 Pet 2:22).15

Third, the standard lexicon of the Greek New Testament, Bauer-Danker-Arndt-Gingrich’s Greek-English Lexicon of the New Testament and other Early Christian Literature (known as BDAG), says that summarturevw is an intensified verb. BDAG note that as early as the sixth or seventh century BC “the prefix suvn- [on this verb] has in the highest degree the effect of strengthening” the force of the verb.16

Fourth, although this particular verb occurs only three times in the NT (and not at all in the LXX), all of its occurrences are in Romans. (Remarkably, it is more frequent than marturevw in this book!) In its two other occurrences, it most likely has the force of intensifying the force of the verb;17 in the least, the evidence offers no comfort to the associative view. For example, Rom 9:1 reads, “My conscience bearing me witness [that] I am not lying” (or, “my conscience bearing witness to me [that] I am not lying”).18

Fifth, in light of the paucity of usage within the NT coupled with BDAG’s insistence that the prepositional prefix of this verb strengthens the basic force of the verb, an examination of its usage elsewhere is called for. Remarkably, most of those who argue for the associative view simply assume a meaning for this verb without examining the evidence—or, it seems, without interacting with BDAG. Before we look at the data, the parameters of our investigation need to be delineated.

One of the fundamental problems with this verb is that in certain contexts the meaning of “bear witness with” someone can mean almost the same thing as “bear witness to” someone. This is one of the reasons why there is confusion in Rom 8:16. For example, even in the indirect object view, there are various permutations:

1. “bear witness to” a jury or a judge

2. “bear witness to” the truth, act, opinion, etc.

3. “bear witness to” the defendant, either for/on behalf of (dativus commodi) or against (dativus incommodi) him; this also shades off into “assure.”

The first of these would be a pure indirect object usage: the jury or judge is neutral and is hearing the case. The second kind of bearing witness is a confirmation of the truth, etc. This would certainly not involve an associative idea unless that which bears the truth-witness is also cut from the same cloth, or if truth is personified. The third permutation, that of bearing witness to, for, or against a defendant is the kind of indirect usage I see in Rom 8:16. It is thus also a dative of interest.19 But this is the closest of the three permutations to an associative idea. So, how can we distinguish the two in other texts?

There are some guidelines that can be used. First, if the meaning fits in with (1) or (2) above, then the verb obviously does not carry an associative notion. But in instances where a dative of interest could be detected, other tests need to be employed. We suggest two. First, we need to compare summarturevw with marturevw. If the latter could be substituted for summarturevw without an alteration in the meaning, then summarturevw will be regarded as having an intensifying force. If, however, the substitution would alter the sense, then summarturevw is considered to bear an associative idea. It should be noted at the outset that marturevw regularly occurs with dative indirect objects that sometimes shade off into dativus commodi or dativus incommodi, but not with datives of association.20 The mere presence of a dative substantive with summarturevw, therefore, is not a sufficient basis for taking the dative as associative since dative substantives frequently occur with marturevw.21

Second, when the verb occurs without a dative substantive in the context, this semantic situation will usually indicate that the verb is an intensifying form of marturevw. There will be some rare exceptions to this (such as cryptic expressions in poetry as in Sophocles, Electra 1224), but the vast bulk of suvn— prefixed verbs that bear an associative nuance are found with an explicit dative of association in the context.22 The parameters are thus clear.

Summarturevw is, by any standard of measurement, a rare word. It occurs in extant Greek literature, from Homer to AD 1453, certainly no more than 200 times. A search of TLG, the published volumes of the Oxyrhynchus Papyri, Tebtunis papyri, and the digitized collections of papyri from Duke University and the University of Michigan—a grand total of more than 60 million words of Greek literature23—revealed only 164 instances.

I have not examined every instance in detail, but can at least offer some representative texts.24 These will be listed diachronically within two groups: instances in which a dative substantive is found, instances in which no dative substantive is found with the verb.

The relevant data are included in an endnote.25

In sum, I have found summarturevw predominantly to take dative indirect objects rather than datives of association.26 BDAG’s statement that “the prefix suvn– [on this verb] has in the highest degree the effect of strengthening” the force of the verb is largely vindicated. At the same time, some of the texts were ambiguous, and most likely, some that were not examined could possibly display an associative force. Nevertheless, the prima facie lexical assumption—viz., that since the verb here is prefixed by suvn– it must take a dative of association—is clearly wrong. The lexical argument, then, though plausible on its face, seems to fall apart upon closer scrutiny. Along these lines, it may be noteworthy that Fee does not even raise the lexical issue, but prefers to argue against the indirect object view by insisting that it is only a theologically-motivated interpretation without substance. Just the opposite seems to be the case.

In addition to these three counter-arguments, a few others may be put forth on behalf of the indirect object view.

4. The associative view involves too many complications. If, for example, the Godward associative view is adopted (viz., that our combined testimony is to God), then we have a couple of problems. First, where else does the Bible promote such a radical inequality in the co-witnesses? (This would be like Walter Cronkite and some three-year-old concurring on the details of a newsworthy event.) In particular, the Holy Spirit’s witness does not require a second opinion. Our testimony is unnecessary. To suggest otherwise is like saying, “God says—and I can back him up on this…”! Second, as Cranfield notes, “What standing has our spirit in this matter? Of itself it surely has no right at all to testify to our being sons of God.”27 In the context our spirit especially seems to have little say. The self-doubts expressed in Rom 7—whether that chapter is autobiographical or of a more universal nature—stand in bold relief to Rom 8. Whatever else chapter 7 is saying, it is arguing that the unaided inner self is defeated by sin and makes no contribution to sanctification. But thanks be to God that the law of the Spirit of life has set us free!

5. So much for the Godward witness. But if the dual witness is manward, there is another problem. If “our spirit” refers to our “inner person,” as almost all commentators take it, then what is the difference between “our spirit” and “ourselves”? If there is no real difference, what does it mean that “the Spirit bears witness with our spirit to ourselves”? Does this mean that we witness to us? This sounds as if the responsibility to convince myself of my salvation is myself. This interpretation, of course, is refuted on its face.

6. Positively, we can argue from two vantage points: context and correlation. The context of Rom 8 involves especially two themes—assurance of salvation and the role of the Holy Spirit in the believer’s sanctification. These two are not unrelated. The assurance offered seems to come from two sources: inner testimony and external fruit. The one, in fact, seems to be the prerequisite for the other. Notice the following verses:

  • Verse 4—“so that the righteous requirement of the law may be fulfilled in us, who do not walk according to the flesh but according to the Spirit.”
  • Verse 9—“You, however, are not in the flesh but in the Spirit, if indeed the Spirit of God lives in you.”
  • Verse 14—“For all who are led by the Spirit of God are the sons of God.”
  • Verse 23—“we ourselves also, who have the firstfruits of the Spirit, groan inwardly as we eagerly await our adoption, the redemption of our bodies.”
  • Verses 25–27—“But if we hope for what we do not see, we eagerly wait for it with endurance. In the same way, the Spirit helps us in our weakness, for we do not know how we should pray, but the Spirit himself intercedes for us with inexpressible groanings. And he who searches our hearts knows the mind of the Spirit, because the Spirit intercedes on behalf of the saints according to God’s will.” Paul is stressing in these verses, it seems, that though we might waver, the Spirit does not. The Spirit helps us in our weakness and our doubts.
  • Paul concludes with vv. 31–39 in which he evidently is arguing against doubts that are caused by our own inner turmoil as well as by external forces. Verses 31–34 ask the repeated question, “If God is for us, who can be against us?” (v. 31). Verses 35–39 ask the question, “Who will separate us from the love of Christ?” The accusation of condemnation is something felt on the inside. The defense is also internal: because we are in Christ and because he is in us, we stand secure before God.
  • In this great chapter it seems that the Spirit produces the assurance we so desperately need. Because we have the Spirit, we have hope. Yet even in our hope, he helps our weakness and intercedes for us. Because we have the Spirit, we bear fruit for God. The order is significant: The Spirit dwells in us, thus enabling us to live for God. We can have assurance of our salvation the moment we are converted, and as much as our own hearts try to condemn us, the Spirit intercedes.

7. Finally is the argument from correlation. To see the Spirit of God working on our hearts, sustaining our belief both in God and in our relationship to him, is a theme found elsewhere in the NT. This theme suggests an inaugural fulfillment of Jer 31. As new covenant believers, the Spirit of God has come to reside in us. We each know God immediately. Thus, in Heb 10 the author picks up the prophecy of Jer 31 and essentially argues that God places knowledge of himself within us when he forgives our sin. The author begins the OT quotation in v. 15 with an introductory statement that looks like it was lifted right out of Rom 8:16: “And the Holy Spirit also witnesses to us” (marturei' deV hJmi'n kaiV toV pneu'ma toV a{gion.)Thus, in both passages we are told that the Spirit bears witness and that there is something placed within us by God in relation to securing our salvation. This is no accident; it is part of the NT authors’ understanding of the new covenant.

The connection might be even stronger. Paul begins the eighth chapter of Romans by ringing the chimes of freedom that the Spirit has wrought for us: “For the law of the life-giving Spirit in Christ Jesus has set you free from the law of sin and death.” But he does not develop this idea of the law of the Spirit further; rather, he replaces it with talk of the Spirit directly. Indeed, the Spirit gets top billing in this chapter, just like the law received it in the previous chapter. Why, then, does Paul begin with “the law of the Spirit”? In the least, he is certainly connecting his argument to chapter 7, perhaps as a kind of rhetorical bridge. But there may be more; it is distinctly possible that Paul has in mind Jer 31:33: “I will put my law within them and write it on their hearts and minds.” The internal witness of God in one’s heart is here proclaimed; the law written on the heart, the law of the new covenant, indeed seems to be the Spirit himself.28

As well, the whole of 1 John stresses the role of the Spirit in our assurance. First John 3:20 is right to the point—“if our conscience [Greek: heart (kardiva, kardia)] condemns us, that God is greater than our conscience [Greek: heart (kardiva)].” This is perfectly in keeping with what I believe Paul is saying in Rom 8:16. The associative view has an anthropological-hamartiological problem at this point: if our heart condemns us, in what sense could it be a witness to God on our behalf?29

In preparing for this essay, I came across a statement by Gregory of Nyssa which also implicitly links 1 John 3:20 with Rom 8:16. In one place he comments on Paul’s great pneumatological confession:

For it is necessary, according to the oracle of Paul, that the Spirit of God should testify to our spirit, but not that our spirits should be approved by our judgment. For he does not say, “The one who commends himself is approved, but the one whom the Lord commends is approved.”30

It is evident that Gregory’s understanding of summarturevw in Rom 8:16 is similar to ours—and that the link of Rom 8:16 with 1 John 3:20 was already done centuries ago.

To sum up: Rom 8:16 should be translated “The Spirit himself bears witness to our spirit that we are God’s children” because of sound lexical, contextual, and biblico-theological reasons. Further, to take it otherwise leaves too many loose ends and raises more questions than it answers. Although to be sure some of the arguments are on better footing than others, it seems to me that the most probable exegesis of this passage (not just what is possible) is that there is an inner witness of the Spirit which prompts in the believer that filial response of “Abba, Father!” leading him to the assurance of salvation.

Some Implications

1.Obviously, the main implication has to do with assurance of salvation: We know that we are saved because of the testimony of scripture and because of the inner witness of the Spirit. I know I am a child of God not just because the Bible tells me so, but because the Spirit convinces me so. The present tenses in relation to the Spirit in Rom 8 are used predominantly to suggest an ongoing state from regeneration to glorification (thus, customary presents). Because of the Spirit’s witness I have assurance at the front end and throughout my life. But the Spirit sustains in me not just belief, but fruit. Because he dwells in me, he can prompt me to good works. The cause-effect relationship here must be carefully noted: I am assured of my salvation, first, because the Spirit indwells me. Thus, I can have such assurance before I do any good works for God. But as I continue, because of the Spirit’s presence and power, I will persevere. My saved status thus receives confirmation by my works.

2.For perseverance of the saints: There are some today who argue for eternal security, but against the perseverance of the saints. This viewpoint, in its most rigorous form, argues that the perseverance doctrine makes assurance based on works and thus cannot offer such assurance at the point of conversion. This view also argues that even if a person believed only for a short time, and then stopped believing, he is still saved. In order to sustain this argument, one has to deny the inner witness of the Spirit. The only assurance is the objective word. This, to me, smacks of rationalism. It is a view that, ultimately, finds its roots in the Enlightenment, not in the revealed word.

The Spirit not only assures our hearts that we are saved; he also sustains that belief. True believers continue to believe because the Spirit energizes that faith. And he does more: he also energizes the fruit that results from that faith. Thus, the position that we can be eternally secure without persevering seems to embrace both a weak view of sin (in that we have the ability to sustain belief without the Spirit’s aid) and, consequently, a defective pneumatology.

3.How does the Spirit bear witness to our spirits? Certainly, he works on our hearts to convince us of the truth of scripture. But there is more. His inner witness is both immediate and intuitive. It involves a non-discursive presence that is recognized in the soul. This at least is the position of Calvin and the Reformers, and I can find nothing to contradict this either in Rom 8 or in my own experience. Indeed, except for periods of heightened rationalism in church history (such as we faced in the 20th century, with its still lingering effects), this seems to be the steady opinion of the majority of orthodox theologians.31 Thus, the inner witness of the Spirit is supra-logical, not sub-logical—like the peace from God that surpasses all understanding. There are elements of the Christian faith that are not verifiable on an empirical plane. This makes them no less true.

4.For conflict in the academic realm: If the witness of the Spirit that I am a child of God is intuitive, then it is outside the realm of that which is objectively verifiable. This does not make it any less true. We are too much sons of the Enlightenment when we deny intuition and internal apprehensions any value. When you fell in love, what scientific means did you use to verify the state of your heart? None. As every mother tells her child, “You just know.” It’s an apt analogy because it is one of the last vestiges of the pre-Enlightenment era that we still affirm. No one challenges it because there are no scientific means for determining whether a person is in love. Yet, we send bright young students, armed with an M.Div. or Th.M. from an evangelical seminary, into battle at secular schools, telling them only, “Trust your exegesis.” Too many have become spiritual casualties because they suppressed the inner witness of the Spirit. Liberal theology can tamper with the meaning of the text and plant seeds of doubt about historical proofs. If I am trusting only in historical evidences, not realizing that there is still a step of faith involved, I have already lost the battle. Don’t misunderstand: we need to contend for the basic tenets of the faith with all the logical and empirical tools at our disposal. But by themselves, these evidences are not capable of proof. The noetic effects of sin are so powerful that without the Spirit’s aid, we, too, will begin to doubt. Doctoral students and academicians, more than anyone else, need to maintain a warm heart for God precisely because of the academic rigor of their chosen field. An academic life gives one no excuse for a lack of piety. Indeed, if it is to be for God’s glory, piety must be at the heart of that breathless pursuit of truth. We simply cannot risk giving God a partial offering: we cannot give him our minds while holding back our hearts.


1 . Not all take it this way, of course. Many Reformed scholars have assumed the associative view, but have read the text either as though following the indirect object view, or with the assumption that the Holy Spirit and our spirit combine to witness to us. In other words, many have translated the text “with our spirit,” but interpreted the text to mean “to our spirit.” To some degree, this is sloppy exegesis. But we can understand how they get this. If I say, “The archangel Michael fought with Satan,” the preposition certainly does not mean “alongside of Satan”! Yet, many interpreters assume that the same thing is going on in Rom 8:16. One simply has to read their comments to see that they embrace the inner witness view (“to our spirit”), but read it as though it were otherwise (“with our spirit”). The problem with this in English usage has to do with the governing verb. If it implies some tension or contention, then “with” really has the force of “against” (“Michael fought against Satan”). But if it implies collaboration, then “with” typically means “alongside of.” Even verbs such as “dialogue,” in which the tension is subdued, suggest a notion of less than full collaboration: “He dialogued with me on the issue,” may or may not involve hostility. It stops short of “alongside.” It may be because of such verbs used frequently with “with” that many scholars adopt the “with” translation in Rom 8:16 but the “to” exegesis. Hence, the data may well be skewed in favor of the “with” view (especially with translations where no explanation is found).

2 . Surprisingly, G. I. Williamson’s annotations in the Westminster Confession of Faith for Study Classes (Philadelphia: Presbyterian and Reformed, 1964) argue this point, but for a slightly different reason (133): “Some have maintained that the Holy Spirit communicates assurance to the soul of the believer immediately, that is, without the use of Scripture. And Romans 8:16 is claimed in support of this view…. It is true that the Spirit himself bears witness. But he bears witness with the spirit of man immediately, and not to it immediately. In other words God exerts an immediate influence upon the spirit of man, but not by speaking directly to man’s spirit apart from Scripture. Rather, the immediate influence is such that man and God speak together….”

3 . Gordon D. Fee, God’s Empowering Presence: The Holy Spirit in the Letters of Paul (Peabody, MA: Hendrickson, 1994) 568. The exact quotation is in the plural because Fee combines the argument of the dative indirect object with that of “a watered down sense” of the verb “(‘assures our spirits’)”: “But these are unnecessary expedients that abandon Pauline usage for the sake of a prior theological concern that is not involved here.” What Fee never elaborated on, as far as I could see, is what the Pauline usage is that is ignored. If he means lexical usage of this verb, he is clearly wrong (see later discussion). If he means theological usage of Paul’s overall meaning, that is something to be construed in the context under consideration. Yet, if so, then isn’t Fee guilty of a theological predisposition in this text himself?

4 . Ibid., 569.

5 . This is Fee’s argument (ibid., 567). It can, of course, be nuanced in more than one way. Some may argue that our spirit’s testimony before God is essential for our own salvation. Others might argue that our spirit combines testimony with the Holy Spirit before us.

6 . Fee offers other arguments that are not very compelling. For example, he argues by innuendo that the indirect object viewpoint is theologically motivated from the get-go and that therefore it should be rejected: “Disturbed by the supposed theological infelicity of Paul’s sentence… some… take the dative as an indirect object…” (568). He pushes hard on Cranfield especially, saying that he has “an extensive argument that manifestly begins with this theological agenda in hand” (ibid.). Yet Fee’s theological agenda is hardly less suppressed: “the Spirit has not come to ‘take over,’ as it were, so that our own human responsibility is diminished” (569). What matters is not whether a particular viewpoint involves presuppositions (for all do), but whether there is evidence in the text to support such a viewpoint.

7 . But see the NET Bible: “the Spirit bears witness to our spirit.”

8 . “Spiritus testimonium reddit spiritui nostro”: “The Spirit offers testimony to our spirit.”

9 . “Der Geist selbst gibt Zeugnis unserm Geist”: “The Spirit himself gives testimony to our spirit.”

10 . Cf. Acts 13:22; 15:8; Rom 10:2; Col 4:13; 1 Tim 6:13; Heb 10:15; 11:4; 1 John 5:9-10; Rev 1:2.

11 . 1 John 5:7 is an apparent exception to this, for the Spirit is among the three who bear witness. But the previous verse highlights just the Spirit’s testimony (“the Spirit is the one who testifies, because the Spirit is the truth”), suggesting that it is sufficient in itself.

12 . Translation is my own.

13 . D. B. Wallace, Greek Grammar beyond the Basics: Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996) 159.

14 . Ibid. The following texts are noted in which the suvn- prefixed verb takes other than a dative of association: Luke 11:48; Acts 6:9; 8:1; 18:7; Rom 7:22; 8:26; 12:2; 1 Cor 4:4; Eph 5:11; Phil 1:27; 2 Tim 1:8; Rev 18:4.

15 . In six of its eight occurrences in the NT sumbaivnw takes a dative, none of which are datives of association (Mark 10:32; Acts 3:10; 20:19; 1 Cor 10:11; 1 Peter 4:12; 2 Peter 2:22).

16 . This overall assessment is not shared by Strathman in TDNT however. Nevertheless, he argues that the force is generally “agree with,” but still sees Rom 8:16 as “bear witness to.” The problem with this view is that (sum)marturevw is not oJmologevw. That is, although those texts where he sees “agree with” as the meaning of this verb have a certain plausibility in their contexts, “bear witness to” is equally valid. Further, the semantic domains of these two verbs seem to be quite different, even though in translation they may well look alike.

17 . So Moo, Romans 1-8, loc. cit., who nevertheless does not seem to adopt this view here.

18 . The translation of Rom 9:1 is my own. In Rom 2:15 no dative is found.

19 . Wallace, Exegetical Syntax, 143: “Dative of interest typically (but not always) belongs to the larger category of indirect object.”

20 . marturevw occurs 76 times in the NT. Eighteen (18) instances take a dative indirect object (John 5:33; 18:37; 1 John 1:2; 3 John 3, 6; Rev 22:18; incommodi in Matt 23:31; commodi in Luke 4:22; John 3:26, 28; Acts 13:22; 15:8; Acts 22:5; Rom 10:2; Gal 4:15; Col 4:13; Heb 10:15; Rev 22:16); twenty-five (25) instances take a prepositional phrase that emulates a dative of interest, usually dativus commodi (periv in John 1:7, 8, 15; 2:25; 5:31, 32 [bis], 36, 37, 39; 7:7; 8:13, 14, 18 [bis]; 10:25; 15:26; 18:23; 21:24; 1 John 5:9, 10; ejpiv in Acts 14:3; 1 Tim 6:13; Heb 11:4; katav in 1 Cor 15:15). This leaves thirty-three instances, thirteen of which are passive and thus not easily able to accommodate a dative indirect object (none of the passive forms of this verb in the NT take a dative or prepositional phrase that emulates it). Thus, of the 63 examples of marturevw in the NT in the active voice, two thirds of them take a dative direct object or the like. None occurs with a dative of association.

21 . Not only this, but marturevw never occurs with metav plus the genitive in the NT. However, suvn— verbs that take a dative of association often use this prepositional phrase for the same idea (cf., e.g., sullalevw in Matt 17:3 with Mark 9:4; sunesqivw in Gal 2:12).

22 . Cf., e.g., sunanavkeimai (Matt 19:10; Luke 14:10); sunanavkeimai (Mark 2:15); sullalevw (Mark 9:4; Luke 9:30; 22:4); sunapoqnh/vskw (Mark 14:31); sunanabaivnw (Mark 15:41; Acts 13:31); sugcaivrw (Luke 1:58; 15:6; Phil 2:17); sunantavw (Luke 9:37; 22:10; Acts 10:25; Heb 7:10); sunesqivw (Luke 15:2; Acts 10:41; 1 Cor 5:11); sunodeuvw (Acts 9:7); sunanapauvomai (Rom 15:32); sunanamivgnumi (1 Cor 5:9; 2 Thess 3:14); summerivzw (1 Cor 9:13); sunqavptw (Col 2:12); sunegeivrw (Col 3:1); sunapovllumi (Heb 11:31); sunoikevw (1 Pet 3:7).

23 . TLG, or Thesaurus Linguae Graecae CD ROM D, is a digitized database of Greek texts from Homer to AD 1453, currently comprising some 57 million Greek words (Los Altos, CA: Packard Humanities Institute, 1993). Produced under the auspices of the Packard Humanities Institute, it is now marketed by the University of California at Irvine. CD ROM E was released in February 2000; it contains 6,625 works from 1,823 authors, and a total count of 76 million words of text. The cover letter to TLG subscribers notes that “This is a significant expansion compared to the 57 million words (from 831 authors and 4,305 works) included in CD ROM D.” Nevertheless, the CD ROM D is surely a representative database for the usage of summarturevw. Besides TLG, I also examined the first 63 volumes of the Oxyrhynchus Papyri, the first two volumes of the Tebtunis papyri, and Packard Humanities Disk #7 containing the Duke University and University of Michigan papyri data. Only one instance was found in the sources outside of TLG. Moulton and Milligan mention an additional two instances (BGU 1.86.41 [AD 155] and PSI 6.696.5 [3rd cent. AD]), bringing the number of which I am aware to 166 instances.

The potential verb forms include summartur—, sunmartur—, sunemartur—, summemartur—, xummartur—, xunemartur—, and xummemartur—.

24 . It should be noted that a large number of texts are patristic quotations from Paul’s three usages of summarturevw in Romans, without further comment. These therefore do not contribute to the discussion.

25 . The evidence is as follows:

Sixth Century BC

Solon, Fragment 36:

    summarturoivh tau't j a]n ejn divkh/ Crovnou mhvthr megivsth daimovnwn jOlumpivwn

    “May the great mother of the deities of Olympus bear witness at the trial of Chronos.” There is no dative substantive here; Bauer rightly considered Solon to contain the earliest attestation of summarturevwas bearing an intensive force.

Fifth Century BC

Euripides, Helen of Troy 1080: Menelaus says to Helen,

    kaiV mhVn tavd j ajmfivblhstra swvmato" rJavkh

    xummarturhvsei nautikw'n ejreipivwn.

    “Yes, and these rags wrapped about my body

    Shall testify of the salvage from the wreck.”

    No dative is used; the verb has the intensive force of testifying.

Euripides, Hippolytus 286: Nurse to her queen, Phaedra:

    wJ" a]n parou'sa kaiV suv moi xummarturh'/"

    oi{a pevfuka dustucou'si despovtai".

    “So stand thou by and witness unto me

    How true am I to mine afflicted lords.” (The translation is that of A. S. Way in the Loeb Classical Library [LCL].) The dative “to me” obviously bears an indirect force.

Euripides, Iphigeneia at Aulis 1158:

    summarturhvsei" wJ" a[mempto" h gunhv

    “You testify how blameless a wife she was”

    There is no dative; the testimony confirms what she believes, but is not in association with her. (Cf. also in Euripides: with dative substantive which is other than an associative dative: Fragment 319; Comparatio Thesei et Romulii 6.5; without dative substantive: Scholia in Hipp 577.)

Sophocles, Electra 1224:

    fivltaton, summarturw'

    “I agree [with you]: [it is] a happy [day].”

    The context here is minimal, and is poetic (and thus cryptic). Nevertheless, here is a probable instance of an associative summarturevwwithout a dative substantive.

Sophocles, Philoctetes 438:

    xummarturw' soi.

    Philoctetes responds to Neoptolemus:

    “In that I’ll bear thee out.” (F. Storr’s translation [LCL].)

    The force is obviously agreement with someone else. Here the associative and commodi uses of the verb shade into one another: “I testify on your behalf” is little different from “I agree with you.”

Aelius Aristides, ProV" Plavtwna periV rJhtorikh'" 72.11:

    Plavtwna aujtoVn summarturh'sai tw'/ lovgw/

    “Plato himself bears witness to the word”

    Clearly, this is the indirect usage. (For other fifth century BC texts which confirm the intensifying force of summarturevw, cf. Pindar, Scholia et glossae in Olympia et Pythia P. 1.87-93; idem, Scholia P. 1.87-90; Aelius Aristides, PeriV tou' parafqevgmato"397.21; idem, Scholium 72.7. The first three lack a dative substantive, while the third involves a dative indirect object. All clearly involve an intensifying force for the verb.)

Fifth-Fourth Century BC

Xenophon, Hellenica 3.3.2:

    sunemartuvrhse deV tau't j aujtw'/ kaiV oJ ajlhqevstato" legovmeno" crovno" ei ai.

    “And time also, which is said to be the truest witness, gave testimony that the god was right…” (C. L. Brownson’s translation [LCL].)

    This text is difficult to assess. Like other texts, the associative dative and dative of advantage are close to one another. The aujtw'/ seems to suggest “bore witness with him that he was right.” However, it could mean “bore witness for him.”

Xenophon, Hellenica 7.1.35:

    sunemartuvrei d j aujtw'/ tau'ta pavnta wJ" ajlhqh' levgoi oJ jAqhnai'o" Timagovra"...

    “And the Athenian, Timagoras, bore witness in his behalf that all these things which he said were true.”

    The idea here seems to be closer to indirect object—i.e., he bore witness for him. It crosses over into dativus commodi, which in contexts such as this, looks like association, but it still is indirect object. The idea, then, would be “he testified to him.”

Isocrates, Trapeziticus 42:

    AujtoVn toivnun Pasivwn e[rgw/ parevxomai touvtoi" summarturou'nta.

    “Pasion himself, moreover—in effect, at least—I will present as corroborating these statements.” (LCL translation.) That is, “I will show that Pasion himself has borne witness to [the truth of] these statements.” That this is indirect object is evident by the fact that marturou'ntacould easily be substituted here.

Isocrates, Panegyricus 31:

    tav te pavlai rJhqevnta toi'" parou'sin e[rgoi" summarturei'

    “the words spoken long ago confirm the practice of today”

    (George Norlin’s translation [LCL].)

    That is, “the words spoken long ago bear witness to the present deeds.” Again, the indirect discourse force is seen in that marturei' could have been substituted.

First-Second Century AD

Josephus, Antiquities 19.154:

    Klhvmh"...meqivhsin pollw'n met j a[llwn sugklhtikw'n dikaiosuvnhn th'/ pravxei summarturw'n kaiV ajrethVn toi'" ejntequmhmevnoi" kaiV pravssein mhV ajpodedeiliakovsi.

    “Clemens, together with many others of senatorial rank, bore witness to the justice of the deed and to the valour of those who had made the plans and shown no weakness in the execution of them.” (L. H. Feldman’s translation [LCL].)

Plutarch, Quomod adulator ab amico interno 64.C:

    ...gaVr dei' tw'/ fivlw/... summarturei'n

    “For it is necessary to support one’s friend”

    The idea here seems to be more “agree with” than “bear witness to,” though it could possibly have the force of “bear witness on behalf of.”

Plutarch, Quaestiones convivales 724.C.:

    ouj summarturou'sin o{ti th'" eij" toV nika'n kaiV kratei'n dunavmew" tw'/ qew'/ touvtw/ plei'ston mevtestin…;

    “Do not the dedications of arms and the finest of the battle-spoil and trophies at Pytho attest that this god has much influence in the realm of victory and the winning of powers?” (LCL translation.

    The meaning is clearly that the display of evidence strongly attests to the god’s abilities and record.

Plutarch, Moralia 786.F:

    ouj mhVn ajllaV kaiV eujmenhV" summarturou'sa toi'" e[rgoi"...

    “Yes, and moreover kindly gratitude, bearing witness to the acts…”

    (I. N. Fowler’s translation [LCL].)

    Again, a clear instance of indirect object.

Vettius Valens, Anthologiarum 9.120:

    ejaVn dev pw" kaiV oJ tou' DioV" summarturhvsh/

    “If perhaps the son of Zeus should testify”

    No dative substantive here; an intensifying use of the verb.

    (Other examples from the first and second centuries AD include: Vettius Valens, Anthologiarum 9.61; idem, Anthologiarum 9.116. In each of these passages, the datives are other than associative.)

Third-Fourth Century AD

Hippolytus, Commentarium in Danielem 1.19.8:

    summarturw'n th'/ ajlhqeiva/

    “bearing witness to the truth”

    A clear instance of indirect discourse.

Eusebius, Ecclesiastical History 1.11.3:

    oJ d j aujtoV" jIwvshpo" ejn toi'" mavlista dikaiovtaton...toVn jIwavnnhn, toi'" periV aujtou' kataV thVn tw'n eujaggelivwn grafhVn ajnagegrammevnoi" summarturei'…

    “The same Josephus admits that John was peculiarly righteous, and a baptist, confirming the testimony recorded in the text of the Gospels concerning him.”

    This example uses a dative and seems to have a commodi force. One must keep in mind that there are sometimes subtle shades of difference between “agree with,” “bear witness with,” “bear witness to,” “bear witness for,” and “confirm.” However, is it really possible to substitute murturevwfor summarturevwhere? That seems a stretch, so this instance probably should belong with the associative group.

Eusebius, De laudibus Constantini 16.11:

    summarturei'n th'/ ajlhqeiva/

    “bearing witness to the truth”

Eusebius, Supplementa ad quaestiones ad Marinum 22.989:

    summarturei' deV aujtw'/ kaiV oJ iJeroV" ajpovstolo" w|de Korinqivoi" gravfwn, o{ti w[fqh Kefa'/, ei a toi'" e[ndeka.

    “And the holy apostle also bears witness to him here, in writing to the Corinthians: ‘He appeared to Cephas, then to the eleven.’”

Eusebius, Demonstratio evangelica 8.2.123:

    aujtaV dhV tau'ta kaiV oJ Fivlwn summarturei'

    “Philo also testifies of these same things”

    (For other fourth-century examples, cf. Eusebius, De ecclesiastica theologia 2.2.1; idem, Praep Evang 6.8.24; Chrysostom, In epistulam ad Romanos 60.428 [these three examples involve datives that are other than associative]. In the following three texts, there is no dative substantive: Basil, De baptismo libri duo 31.1561; idem, Orationes 31.1684; Chrysostom, Fragmenta in Jeremiam 64.905.)

Fifth Century AD

Theodoret, Eranistes 87:

    oJ mevga" Pevtro" th'/ ajlhqeiva/ summarturw'n

    “the great Peter bearing witness to the truth”

Theodoret, Interpretatio in Jeremiam 81.604:

    pravgmatav sou kathgorei', kaiV toV suneidov" sou summarturei' soi

    “your deeds condemn you, and your conscience bears witness against you”

Theodoret, De providentia orationes decem 83.569:

    e{tera, toi'" protevroi" summarturou'nta

    other things which bear witness to former things

Theodoret, De incarnatione domini 75.1428:

    jIwavnnh" summarturei' levgwn: jIde oJ ajmnoV" oJ ai[rwn tou' kovsmou thVn aJmartivan

    “John testified saying, “Behold the lamb who takes away the sin of the world”

    Theodoret uses summarturevwfor John the Baptist’s testimony about Jesus in John 1:29. Five times in John chapter 1 marturevwis used of John’s testimony, yet Theodoret here introduces his testimony with a Pauline word. It is evident that it bore the same essential meaning as marturevwand was selected, in all probability, because of its intensifying force. This instance is significant, too, because there is no dative substantive, and nothing in either Theodoret’s context or that of John 1 would suggest an associative notion here.

Sixth Century AD

Simplicius, In Cat 8.262:

    th'/ toiauvth/ crhvsei summarturei'

    he “bears witness to such great power”

    (Cf. also Simplicius, in Ph 9.18 [using a dative indirect object].)

Eighth and Ninth Centuries AD

John Damascene, De azymis 95.392:

    jIwavnnh" summarturei' tw'/ lovgw/

    “John bears witness to the word”

    This is followed by a quotation of 1 John 5:7. What is signficant is that the author of the Fourth Gospel and the letters of John uses marturevwforty-three times—more than 50% of all instances in the NT. The concept of John bearing witness to the truth or the word is, therefore, ready at hand. But John Damascene chooses instead to use the verb summarturevweven though this verb never appears in John. It is evident that it bore the same essential meaning as marturevwand was selected, in all probability, because of its intensifying force. That this is a very late text, however, softens the value of this. Nevertheless, we already saw Theodoret’s similar usage three centuries earlier.

    (Cf. also John Damascene, Passio sancti Artemii 96.1268 for a similar example. In addition, Nicephorus Gregoras, Historia Romana 3.361 illustrates the usage of summarturevwwithout a dative substantive.)

Nicephorus Gregoras, Historia Romana 3.430:

    summarturei' d j hJmi'n kaiV oJ mevga" ajpovstolo" Pau'lo", levgwn: nu'n meVn di j ejsovptrou ejn aijnivgmati blevpomen, tovte deV provswpon proV" provswpon.

    “The great apostle Paul also testifies to us, saying, ‘Now we see in a mirror dimly, but then face to face.”

26 . Forty-five of the 166 instances in extant Greek literature were examined. As much as one third of the remaining examples are quotations from Romans without further ado. By my count, at least thirty-eight of the forty-five examples should be labeled “intensive,” while two should be labeled “associative,” and three are somewhere in between. As well, Moulton-Milligan add two more instances to the associative side of things, bringing that to four. Certainly there are more instances of the associative category (as well as the intensive category), but the pattern is nevertheless clear with the representative texts we have examined.

27 . Cranfield, Romans (ICC) 1.403.

28 . This notion is picked up elsewhere in Paul’s writings: cf., e.g., Gal 5:14-25.

29 . Other pieces of evidence could be used to show that certainty of our convictions and knowledge of God and his will are directly linked to the presence of the Spirit in our lives. In particular, note the radical difference between the apostles’ convictions about the reality of Jesus’ resurrection before Pentecost (cf. Matt 28:17 [“they worshiped him, but some doubted”]) and after—even though Jesus went to great lengths to prove to them that he was truly risen from the dead (Acts 1:3).

30 . Gregory of Nyssa, De instituto Christiano 8,1.73 (dei' gaVr kataV toV lovgion Pauvlou toV pneu'ma tou' qeou' summarturei'n tw'/ pneuvmati hJmw'n, ajllaV mhV th'/ hJmetevra/ krivsei taV hJmevtera dokimavzesqai: ouj gaVr fhsivn, oJ eJautoVn sunistw'n ejsti dovkimo" ajll j o}n oJ kuvrio" sunivsthsin).

31 . Credit is due to M. James Sawyer for this historical summary. The Puritans did not hold to the immediate inner witness, but felt that through the long process of sanctification one came to this. Thus, they both reversed the dwelling and the doing roles of the Spirit and virtually held to a second blessing view of assurance! Warfield and many modern Reformed scholars also hold to a different view, shaped, it seems, by Scottish common sense and the Enlightenment.

Related Topics: Pneumatology (The Holy Spirit)