by J. W. Scott

The translators of the New International Version moved away from the principle of literal translation (i.e. “formal equivalence”) more than most evangelicals had previously been willing to move. They adopted a more idiomatic, and sometimes even paraphrastic approach, seeking to convey the meaning of the original in contemporary English style (i.e. moderate “dynamic equivalence”). There is certainly much to be said for such an approach. However, a high level of exegetical and literary skill is required for the successful execution of it. If translators do not understand the original text particularly well, a free translation of it will ordinarily convey less of the original meaning, and introduce more spurious meaning, than will a literal one. Furthermore, if the original grammatical structure and vocabulary are simplified in order to clarify the general sense (without paraphrastic expansion), some loss of meaning is inevitable. In other words, the reality of dynamic equivalence may turn out to be more dynamics than equivalence. A more “readable” translation may sacrifice substance to style.1

These dangers are, one regrets to say, all too often realized in the NIV As a result of simplification and paraphrase, the fine points of Scripture are sometimes lost, and these details are sometimes theologically important. For example, a serious problem is posed by the NIV’s handling of texts in the book of Acts that pertain to the scope and mode of baptism. We would argue that the NIV has consistently, though no doubt unintentionally, removed from these texts details that extend the scope of baptism beyond believers (especially to their children) and show that a small amount of water was sufficient for a proper baptism. No one would develop his doctrine of baptism solely from the book of Acts, but the fact remains that in that book the NIV provides substantial support for the doctrine of believers’ baptism.

Paedobaptists point to the apostolic practice of household baptism (see Acts 16:14–15, 30–34; 18:8 (by implication); 1 Cor 1:16; cf. Acts 11:14, but note 10:24) as evidence supporting infant baptism. The advocates of believers’ baptism, however, argue that each person in these households became a believer before being baptized, which would imply that there were no infants in them.2 In some passages there is little or no indication of the extent of faith in the household baptized. But Luke does comment on the extent of faith preceding the baptism of the Philippian jailer’s household (Acts 16:30–34) and the baptism of Crispus’s household (18:8).

According to the NIV, everyone in these two households professed faith before being baptized. In its original edition (NT, 1973) the NIV at Acts 16:34 related that the Philippian jailer’s “whole family was filled with joy, because they had come to believe in God” before being baptized. This was subsequently reworded (in the 1983 edition) to “he was filled with joy because he had come to believe in God-he and his whole family.” In either version we are told that everyone in the household believed.3 The NIV at 18:8 similarly credits an entire household with faith (prior to baptism): “Crispus, the synagogue ruler, and his entire household believed in the Lord.”

But the original Greek (as reflected by the more literal translations) is somewhat different—and more complex—in each passage. Where the NIV says that both the head of the household and his household believed, the Greek tells us that the head of the household believed, and that his household was associated with him in his faith. Acts 16:34 actually says that the Philippian jailer believed “with his whole household” (panoikei), and 18:8 similarly says that Crispus believed with his whole household” (sun ēolō tō oikō autou). The Greek expressions translated “with his whole household” are different, and a third is used at 16:32, but they are virtually identical in meaning.

The difference between “X and his whole household believed” and “X believed with his whole household” may seem slight at first, but it is highly significant. The former statement declares that each member of the household was a believer, but the latter one does not. The word “with” indicates that each member of the household was associated with the head of the house as he professed his faith, so that the household could be received together into the community of faith by baptism, but it does not specify the manner in which each member was associated with him as he embraced the Christian faith. Adults and older children would no doubt have joined with him in making a profession of faith. Younger children and infants, if any, would have joined with their believing parents in a largely or even completely passive manner, just by being brought along with them as their children.4 An infant’s status as the child of a believer would have constituted sufficient association with the head of the house who believed (cf. 1 Cor 7:14).5 Since infants could have been among those who were associated with the Philippian jailer and Crispus as they believed, they could also have been baptized in household baptisms. Acts 16:34 and 18:8, then, imply that baptism was available to the infant children of believers, and not (as the NIV implies) that it was precluded by the requirement of individual faith. By simplifying and supposedly clarifying these passages, the NIV has inadvertently changed their teaching on the scope of baptism.

Our interpretation of the expression “with his whole household” is confirmed by a similar expression in Acts 21:5, where it is said that as Paul and his party left Tyre, all the (male) disciples of the city “escorted” (propempontōn) them down to the harbor “with (sun) the women and children” (not “and their wives and children,” NIV). The whole Christian community brought Paul out of the city. The men apparently led the way, joined by the women and children. The women and older children were probably actively escorting Paul and his party, but the youngest children were doing so only in the sense that they were brought along by their parents.6 Just as heads of households can escort “with” infants who are only passive participants in the procession, so also can they become believers “with” infants who are only passive participants in household conversions.7

It is important to note that a “household” would be baptized as such, which implies that family membership (and not just individual faith) was an important factor. As we have seen, an entire household would be baptized if the head of the house believed and those under his authority joined with him in accordance with their status and ability. Thus, it is irrelevant whether the specific households mentioned in Acts and 1 Corinthians actually included any children (or others) incapable of making a profession of faith. Since households were baptized as such, anyone who could belong to a household could also be baptized. Since household baptism was evidently a fairly common apostolic practice, numerous infants must have been baptized.

Furthermore, the precise manner in which Luke connects the head of the house with his or her household in various situations is highly instructive. The head of the house “and” (kai) the household are, saved (11:14; 16:31) and are baptized (16:15, 33), while the head of the house fears God (10:2), has the gospel spoken to him (16:32),8 and believes (16:34; 18:8) “with” (sun or its equivalent) the household. The connective is “and” where everyone in a household can fully enter into the activity involved, but it is “with” where infants cannot do so. Infants can be saved and can be baptized, but they do not have the mental capacity to fear God, receive the gospel, or believe. Thus, Luke’s paradigm for household conversion leaves room, no doubt intentionally, for the inclusion of infants with their parents in salvation and baptism.

Now it would be too much to expect the translators of the NIV to have been aware of the theological significance of this subtle verbal variation in Acts. However, if they had been translating more literally, as in the KJV or the NASB, the variation between “and” and “with” would have been retained. They would have preserved more meaning than they were aware of. But the reader of the NIV has no opportunity to perceive any of this, for in each verse the head of the house is linked to his household with “and.” This illustrates the fact that the “looser” a translation is, the more likely it is to convey only what the translators realize is there. When translators have a limited understanding of a text, they will ordinarily convey more of the original meaning, however awkwardly, with a more literal translation. The epistles of Paul are not the only portion of Scripture that contains things that are difficult to understand, even for an apostle (2 Pet 3:16). Even for the native speakers of the biblical languages, the problem was the inherent profundity of God’s Word, so how much greater it must be for us today, even with all our vaunted tools of modern scholarship. Perhaps the translators of the NIV (not unlike many other modern translators of the Bible) were overly confident of their ability to grasp the full meaning of the Word of God.

Now some may think that the present writer has inferred too much from these verses. Perhaps Acts 16:34 and 18:8, even in the context of Luke’s wider paradigm of household conversion, are not really decisive on the question of infant baptism. But if these passages are not decisive in the original Greek, they should not be decisive in English translation, either. The trouble with the NIV at least is that it comes down, whether intentionally or not, strongly on the side of believers’ baptism. “With” may be ambiguous, but “and” is not. According to the NIV, baptism was administered to people who professed their faith, both heads of households “and” all those in them. Everywhere in Acts, then, the pattern in the NIV is that “many…who heard him believed and were baptized” (18:8b). Few baptists will be convinced of the validity of infant baptism on the basis of rather abstract principles of covenant theology—and many paedobaptists will begin to wonder—when they see in their own NIV Study Bible (assisted by the note at 16:32) that only believers were baptized in the apostolic church.

Another passage warranting infant baptism is Acts 2:39. In this verse Peter tells those who have responded to his preaching that the promise of the Spirit is “for you, and your children, and all (pasin) who are far off, whomever (hosous an) the Lord our God shall summon.” The promise of the indwelling Spirit is a special feature of the New Covenant (see 2 Cor 3:6; Heb 10:16). Since that promise belongs to the children of believers, they must belong to the New Covenant community.9 Entrance into it is marked by baptism (see Acts 2:38, 41), and therefore the children of believers, including infants, were and are to be baptized.

The advocates of believers’ baptism respond that the expression “your children” is qualified by the concluding clause of the verse, “whomever the Lord our God shall summon,” so that only responding, i.e. believing, children are in view.10 However, this qualifying clause modifies only the category immediately preceding it, “all who are far off”; the Lord would summon some, but not necessarily everyone, to come near for salvation. The word hosos is often used in this way to qualify pas (see Acts 3:22, 24; 5:36, 37). And if “your children” were also qualified, so would “you,” but “you” were already summoned (see v 37). In any case, the summoning in view is not exclusively effectual calling, but includes the summoning of converts’ children into the church. It is their parents’ responsibility to respond to the summons and bring them in (cf. Luke 18:15–16).

But once again the biblical support for infant baptism is obscured in the NIV, which translates Acts 2:39 in this fashion: “The promise is for you and your children and for all who are far off—for all whom the Lord our God will call.” By not repeating “for” before “your children,” the NIV does link children with their parents, but the dash before the qualifying clause and its introduction with “for” suggest that both “you and your children” and “all who are far off” are modified by it. Without a comma after “children,” it is unnatural to limit the qualification to “all who are far off.” And by translating “call” (with most English translations) instead of “summon,” the impression is given that the gospel call to those old enough to understand it is exclusively in view. Thus, the NIV follows the baptistic interpretation of this passage.

The quantity of water necessary for baptism can be reasonably inferred, though with some uncertainty, from the circumstances surrounding various baptisms described in the book of Acts.11 But in one passage, namely 10:47, the wording implies that a small amount of water was employed.12 Peter says to the Jewish believers who have come with him to Cornelius’s house, “Surely no one can withhold (kōlusai) the water, so that these people cannot be baptized.” This was said in anticipation of the baptism of Cornelius and the many other Gentile converts in the house (v 48). Peter did not want water to be brought in, only to have someone with Jewish scruples hold it back from being used on Gentiles. Now one can “withhold” only what is in one’s possession, and it is hard to see how any of the Jews visiting in Cornelius’s house could have been about to be in possession of enough water for immersions. A bowl of water could have been brought in to the room or courtyard, and then held back, but hardly a large, heavy tub of water. If immersion was necessary, the converts would have been taken to a pool of water. In that case Peter might have spoken of someone possibly trying to keep them away from the water, but not of trying to keep the water away from them.

But the translators of the NIV, apparently failing to perceive this, restructure the passage: “Can anyone keep these people from being baptized with water?” This rendering removes the implication in the original that baptism was not (or at least did not need to be) by immersion. And since the NIV refers to the restraining of people (as one would if baptism were by immersion), rather than the withholding of water, it is agreeable to (though does not require) immersion.

Thus we see that the verses in Acts which shed the most light on the subjects and mode of baptism are translated in the NIV in a manner favoring a baptistic understanding of baptism. Three passages that actually favor infant baptism and one that favors pouring (or sprinkling) are simplified in a manner that removes such implications.

As we have suggested, the erroneous doctrinal implications of the NIV’s translation of these passages are probably the result of the methodology employed by translators who were unaware of the full import of the original, rather than the result of deliberate theological manipulation. That this is so is indicated by the fact that the doctrine of Scripture also comes up short in the NIV’s rendering of Acts, despite the translators’ express commitment to the full inspiration and infallibility of Scripture. Passages touching on this doctrine are not consistently mistranslated, but faulty exegesis and careless rephrasing are often evident.

Happily, one of the key passages in Acts relating to the doctrine of inspiration, namely 1:16, is adequately translated in the NIV: “the Scripture had to be fulfilled which the Holy Spirit spoke long ago through the mouth of David.” This translation brings out that God is the primary author of Scripture: the Holy Spirit uttered the words of Scripture, figuratively speaking, through the mouth of the human writer. God did not just influence the thought of his prophets, or give them a basic message to develop in their own way, as one might read into “spoke through” without “the mouth of.” Rather, he went so far as to speak through their mouth; that is, he controlled their very words. Calvin appropriately comments on this passage that Luke’s manner of expression teaches “that David and all the prophets spoke solely under the direction of the Spirit, so that they themselves were not the source of the prophecies but rather the Spirit who used their tongues as an instrument.”13 The words “spoke through the mouth of,” then, make explicit what would be only implicit in “spoke through.” They underline the plenary verbal inspiration of Scripture.14

Luke keeps repeating this important point, but the NIV does not. It does preserve the reference to divine utterance conveyed through “the mouth of” a prophet at 4:25, but it drops these words, together with the implication of direct verbal inspiration, in 3:18 and 3:21, as it also does in Luke 1:70 (cf. 2 Chr 36:22). In 3:18, for example, what God “proclaimed beforehand through the mouth of all the/his prophets” is truncated to what he “foretold through all the prophets.” If the translators were uncomfortable with Luke’s manner of expression, they should have expressed the same idea differently. But it should not have been treated as though it were superfluous.

Elsewhere in Acts the divine authorship of Scripture is indicated in the original Greek, but disappears altogether in the NIV. At 2:16 the NIV changes “what was spoken [sc. by God] through (dia) the prophet Joel” to “what was spoken by the prophet Joel.” This is a simple mistake, for dia (in contrast to hupo: see Matt 1:22; 2:15; 22:31; Acts 13:45; 16:14; 17:19; 2 Pet 3:2; Jude 17) indicates that the prophet was the human instrumentality “through” whom God spoke (as in Acts 28:25, where the NIV gets it right). A similar mistake occurs at 13:40, where the NIV changes “what is spoken [sc. by God] in the Prophets” to “what the prophets have said.”

In Acts 13:34–35 God is the subject of the verb anestēsen, and thus of the following verbs introducing OT quotations, namely eirēken, “he has spoken,” and legei, “he says.” God is similarly the implied speaker of Scripture in v 40, and its stated speaker in v 47. On the other hand, some are of the opinion that the speaker in vv 34 and 35 may be “it,” i.e. Scripture, rather than God.15 In either case, however, God would be understood as speaking in the Scriptures.16 But the NIV removes the speaker by adopting more idiomatic passive constructions: “the fact…is stated” (v 34), and “it is stated” (v 35).17 Such impersonal expressions, as B. B. Warfield notes, are easily construed as introducing Scripture “as if the source of the quotation were unimportant and its authority insignificant.”18 After examining extrabiblical literature, furthermore, he concludes that there “would seem to be absolutely no warrant in Greek usage for taking λέγει… really indefinitely.”19

Another problem with the NIV’s handling of texts in Acts pertaining to the doctrine of Scripture arises at 7:38, where Stephen says that Moses received living logia on Mount Sinai to pass on to us. The Greek word logia refers to divine utterances delivered through a human intermediary, i.e. “oracles.”20 But the NIV translates logia by the insipid word “words,” which poorly conveys the fact that complete utterances were received by Moses, and which fails to convey the important point that the utterances in view were of divine origin. Unless one infers that “the angel” who spoke to Moses was divine (or at least a conveyor of divine utterances), the NIV has actually removed from this verse the implication that God was the author of the Mosaic law.

It is disturbing to see the term “oracles” removed everywhere else in the NIV NT, too. Logia conveys more than logoi (“words”), but at Heb 5:12 “the oracles of God” is replaced by “God’s word,” and at Rom 3:2 and 1 Pet 4:11 “the oracles of God” becomes (somewhat better) “the very words of God.” It is not altogether clear why the NT translators avoided the word “oracle,” for the OT translators used it frequently.

The necessity for the OT to be fulfilled in the NT age is well brought out in the NIV at Acts 1:16, 21; 3:21; 17:2–3; 26:22. But it is lost at 2:24–25, where a period is placed at the end of v 24 and the conjunction “for” (gar) in v 25 is completely ignored. By severing the connection between v 24 and v 25, presumably in the interest of contemporary style, the doctrine of Scripture, not to mention the flow of the argument, suffers as a result. Lost is the important point that Jesus could not be held in the grave (v 24) because Scripture had foretold otherwise (vv 25–28). A reason is reduced to an unrelated fact. If a period must be placed at the end of v 24, and v 25 cannot begin with “For,” then it should begin with a paraphrastic expansion like “This was so, because.”

Thus we see that in no fewer than eight instances the NIV does not do justice to what Luke says or implies in Acts regarding the inspiration and infallibility of Scripture. Since the evangelical translators of the NIV can hardly be accused of theological bias against this doctrine, their failure so often to bring it out (or bring it out properly) in Acts must be attributed to other factors. We have identified those factors as a sometimes inadequate understanding of the Greek text and a tendency toward simplification.

In conclusion, we have observed certain theological problems with the NIV’s rendering of Acts (and other passages). As the result of seemingly minor verbal changes, its teaching on the subjects and mode of baptism is significantly altered, and its teaching on the inspiration of Scripture is often diminished or lost. This is apparently the consequence of trying to produce a dynamically equivalent translation without always grasping the full import of the text and without fully appreciating the problems created by simplifying its structure and vocabulary. To the extent that these factors were at work elsewhere in the NIV, one should expect to find similarly inaccurate translations, sometimes adversely affecting doctrine.

1 The problems resulting from a more radical understanding of dynamic equivalence are surveyed by D. A. Carson in “The Limits of Dynamic Equivalence in Bible Translation,” Evangelical Review of Theology 9 (1985) 200–213.

2 So, e.g., G. R. Beasley-Murray, Baptism in the New Testament (Grand Rapids: Eerdmans, 1973) 315, 319–20.

3 The NIV Study Bible (Grand Rapids: Zondervan, 1985) accordingly comments on Acts 16:32: “Paul and Silas explained the gospel more thoroughly to the jailer and to all the other members of his household, and they all believed in Christ and were saved (v. 34).”

4 Similarly, we should understand that when Paul and Silas spoke the gospel to the jailer “with all those who were in his household” (16:32), the whole group was addressed together, but each person in it was expected to understand only what he was capable of understanding. Infants, if there were any in the household, were spoken to only in the sense that they were present with their parents. But cf. n. 8, below.

5 We are not saying that the faith of the head of the house sufficed for the salvation of everyone in his house, as Beasley-Murray (in Baptism, 319-20) takes the paedobaptist interpretation to be. Clearly, the members of the jailer’s household were gathered together to hear the gospel (16:32) in order that as many as possible might find salvation through faith (cf. 11:14). To Beasley-Murray it is “nonsense” to think that “the infants were brought out of their beds at this early hour of the morning and listened to the instruction before being baptized” (p. 315). But he forgets that mothers are loath to leave their infants unattended for an extended period of time, and that Paul probably instructed the jailer to bring his entire household to the jail (anticipating a household baptism).

6 It should be noted that propempō, as used here (and in 20:38), means “escort,” and not merely (as the NIV has it) “accompany” (as do sunepomai in 20:4 and sunerchomai in 1:21 and elsewhere). They “sent” (pempō) Paul and his party “forward” (pro-), not by sending them (with assistance) on their way (as in 15:3 and elsewhere in the NT), but by providing a personal escort. Cf. LSJ, s.v. propempō, II: “conduct, escort, esp. a departing traveller.” Luke is indicating their respect and love for Paul, not merely listing those physically present as he departed. In this activity the babes in arms were wholly passive.

7 In Acts the same construction with sun occurs at 1:14;3:4; 4:27; 5:1; 10:2; 14:13; 15:22. In these passages sun introduces those who follow the lead of others and join with them in their activity, whether fully or partially, actively or passively. Only at 1:14, 5:1, and 15:22 does the NIV satisfactorily bring out the force of sun, although an attempt is made at 3:4 and 4:27.

8 The Byzantine text reads “and” instead of “with” in Acts 16:32. If this reading is correct, the speaking in view would have to be understood simply as the addressing of the group, irrespective of the extent to which the sounds uttered were meaningfully processed in the brains of the recipients. (Cf. Luke 9:11. Matt 14:21 indicates that children, no doubt including infants, were in the huge crowd to which Jesus spoke.) But cf. n. 4, above.

9 The children in view are immediate family members. Future generations or descendants in general would very likely have been referred to as sperma, not tekna, as we find the distinction made in Acts 7:5. People can be referred to as the tekna of a distant ancestor (as in 13:33), but prospective offspring are ordinarily called one’s sperma. The point in 2:39 is that children are brought into a covenantal relationship along with their believing parents, not that those in future generations will also be summoned into one.

10 So Beasley-Murray, Baptism, 342-43.

11 See Joseph Addison Alexander, The Acts of the Apostles (2 vols.; Philadelphia: Presbyterian Board of Publication, 1857; reprinted, Edinburgh: Banner of Truth, 1963), on 1:5; 2:41; 8:36, 38, 39; 16:31–33.

12 So Henry Alford, The Greek New Testament (4 vols.; 7th ed. of vol. 2; Cambridge: Rivingtons, 1877; reprinted, Grand Rapids: Guardian, 1976), on 10:47; less strongly, Alexander, Acts, on 10:47.

13 The Acts of the Apostles, 1-13, in Calvin’s Commentaries (ed. D. W. Torrance and T. F. Torrance; Edinburgh: Oliver and Boyd, 1965) 41. E. Haenchen comments in The Acts of the Apostles., A Commentary (Philadelphia: Westminster, 1971) 159 n. 9: “It was really the Spirit that spoke in the psalm, David being merely his human mouthpiece.”

14 Although the word “mouth” (stoma) is sometimes used figuratively to denote speech, here it is used literally to denote the organ of speech, since the reference to speaking is supplied by the verb. One normally speaks through one’s own (literal) mouth, but God spoke through the (literal) mouth of his prophets (cf. Exod 4:10–16). The total expression, “to speak through the mouth of,” unlike the word “mouth” within the expression, is of course figurative, for the mechanics of speech are used to describe the process of inscripturation.

15 Cf. BAGD, s.v. legō, I.7.

16 Cf. B. B. Warfield, “‘It Says:’ ‘Scripture Says:’ ‘God Says,’“ reprinted in The Inspiration and Authority of the Bible (Philadelphia: Presbyterian and Reformed, 1948) 299–348, esp. pp. 299-300, 348.

17 The NIV also introduces quotations with “it is said” in 1 Cor 6:16 (pēēsin) and in Eph 5:14 (legei), though not 4:8. In Heb 7:17 the NIV’s “it is declared” follows the passive martureitai of the Alexandrian text, but the Byzantine text has the active marturei

18 Warfield, Inspiration, 307-8.

19 Ibid., 346. BAGD, s.v. legō, I.7, cites Epictetus 1.24.12 for an impersonal legei, but W. A. Oldfather disagrees in the LCL edition of Epictetus.

20 So BAGD, s.v. logion. The term indicates, as Warfield comments in “‘The Oracles of God,’“ reprinted in Inspiration, 351-407, on p. 406, that Scripture “not merely contains, but is the ‘utterance,’ the very Word of God,” being “nothing other than the crystallized speech of God.” L. Gaussen, in Theopneustia: The Plenary Inspiration of the Holy Scriptures (London: Samuel Bagster and Sons, 1841; reprinted as The Divine Inspiration of the Bible, Grand Rapids: Kregel, 1971) 68, asks rhetorically: “Was there a word that could more absolutely express a verbal and complete inspiration?”

Westminster Theological Journal Volume 48. 1986 (vnp.48.2.350-48.2.361). Philadelphia: Westminster Theological Seminary.

© 2010, Matt. All rights reserved.