ABSTRACT: Speaking in tongues potentially includes three subcategories: (1) known language; (2) unknown language; and (3) language-like utterance—an utterance consists of language-like sounds but does not belong to any actual human language. Category (3) occurs today in charismatic circles. Given that the church in Corinth was permissive, it can be inferred that category (3) may have occurred at Corinth. Moreover, each of the three categories can occur either in inspired, infallible form or noninspired, fallible form. Thus, it is possible to hold a cessationist view of inspiration (no more infallible utterances) and a continuationist view with respect to noninspired forms.
1. Preliminary Observations
The issue of tongues is sometimes a matter of controversy and heat.1 As a result, let me state my intent at the beginning. I want to put forward an argument for the scope of speaking in tongues in the first-century church. But I do so in a tentative way. I hope not to stir up heat.
1.1. Tongues in Acts
Let us start with Acts 2. There are several interpretive views.2 For simplicity, we follow the majority view. It says that Acts 2 involves distinct languages, mutually unintelligible, rather than merely distinct dialects. But even if they were just dialects, the main point is that the utterances in Acts 2 were in natural human languages. We know that because hearers competent in the various languages were able to identify them.3
1.2. Tongues at Corinth
Now we proceed to 1 Corinthians 12–14. For illustrative purposes, we may imagine ourselves sitting in the place of a member of the Corinthian church. What would we hear when other members spoke in tongues? Perhaps on occasion someone was present who recognized the utterance as belonging to a language that he already understood. Then he was able to interpret. That kind of case leads us back to the instances in Acts 2. The language in question was identifiable.
But the letter of 1 Corinthians seems to indicate that at Corinth such an identification of the language was the exception rather than the rule. Most interpretation of tongues seems to have taken place not because a listener confidently understood the language, but because of a special spiritual gift for interpreting tongues (12:10, 30; 14:13). The ordinary listener at Corinth heard utterances that sounded like a communication in language. But he did not know the meaning (14:2). Even the speaker did not know the meaning (14:13–14). For practical purposes, from the point of view of a naive listener, anything that sounded like speaking in tongues was speaking in tongues. “Speaking in tongues” is a loose category that easily covers every kind of language-like utterance in the church service that does not belong to any of the major languages spoken in the church.
It might seem natural to infer that every instance at Corinth belonged to some natural human language.4 But that inference does not reckon fully with the flexibility that belongs to ordinary human use of terms. What happens when people are forced to develop a kind of standard designation for comparatively new phenomena in their midst? Anything that sounds like language will for convenience be loosely designated as an utterance in “language.” The ordinary person does not get fussy with a technical analysis such as a trained linguist might propose. He needs a short, convenient term, and “speaking in tongues” will do.5
The upshot is that we do not know exactly what happened at Corinth. Maybe all the instances belonged to natural human languages. Maybe only some did. Maybe almost none did. What we do know is that they were all “language-like” in some fairly flexible sense. The naive listener thought to himself, “It sounds like a foreign language.”6
2. Types of Language-Like Communication
So now the situation has become complicated. We have three possibilities when a speaker makes utterances without having first learned the language. (1) A known language: the utterance is identified by a listener as belonging to a language that he knows. (2) An unknown language: the utterance is in some human language, but not identified. (3) Language-like utterance: the utterance is language-like, but not belonging to any extant human language. Case (1) corresponds to what happened in Acts 2. Case (2) is what many biblical interpreters have seen in the Corinthian church. What about case (3)? My previous argument about the flexibility of common use of terms supports the conclusion that we cannot a priori exclude case (3) from the instances at Corinth.
Within each of these three categories it is possible to subdivide, and distinguish between two subcategories: (a) utterances inspired by the Spirit and therefore infallible; and (b) utterances not inspired by the Spirit, and therefore fallible. Subcategory (b) might still include utterances influenced by the Spirit. What kind of influence? Consider a modern Christian preacher who wants to be faithful to the Lord. He hopes and prays that the Spirit would fill him and guide his utterances when he preaches. But he does not claim to be inspired and infallible. He hopes for the Spirit’s influence.
Let us look more carefully at category (1) (known language). For someone to make utterances in a language that he has not learned is supernatural. Conceivably it might happen through a counterfeit miracle of demonic origin. But we are talking about instances where Christians are empowered by the Holy Spirit.7 It is easy to assume that such an utterance must be inspired and infallible. But this conclusion does not necessarily follow. The Holy Spirit might supernaturally empower an utterance without necessarily guaranteeing and authenticating every detail of its content. Suppose, for example, that a missionary wants to share the gospel with someone, and finds no common language. He prays for help. He suddenly bursts out in an utterance that carries the content (not infallible) of what he already wants to say, conveyed in a language unknown to him. For our purposes, it does not matter whether this kind of event has ever happened. What matters is the possibility of it happening. Supernaturalism is not always identical with inspiration.
By similar reasoning, we can see that categories (2) and (3) can each be subdivided into (a) and (b), infallible and fallible forms (see the table below).
Table 1: Categories of Tongues
|Categories||(a) Inspired, Infallible Tongues||(b) Noninspired, Fallible Tongues|
|1. Known language||1a. Known language, infallible message||1b. Known language, fallible message|
|2. Unknown language||2a. Unknown language, infallible message||2b. Unknown language, fallible message|
|3. Language-like utterance||3a. Language-like, infallible message||3b. Language-like, fallible message or no message|
When we take everything into account, there is quite a range of possibilities concerning the details of what might be taking place in the first century church under the broad label of “speaking in tongues.” But, so far, some of these are no more than possibilities. We know in the case of Acts 2 that we are dealing with either 1a (known language, infallible) or 1b (known language, fallible) or both, because the languages were recognized. In the case of the Corinthian church, we seem to have mainly some combination of 2a (unknown language, infallible), 2b (unknown language, fallible), 3a (language-like, infallible), and 3b (language-like, fallible). But without further information than what 1 Corinthians 12–14 supplies on the surface, we cannot easily discriminate between the four possibilities.
3. Modern Free Vocalization
Let us now attempt to advance our understanding by looking at modern instances of speaking in tongues. Our first challenge is to find an adequate label for the phenomena. To call the modern phenomena “speaking in tongues” could easily be seen as a question-begging move—does such a label already assume commonality between modern phenomena and the phenomena in the NT? Are the modern phenomena really the same, from a theological point of view or from a linguistic point of view? In order not to appear to prejudge the question, let us temporarily use the label “free vocalization.” As a rough definition, we could say that “free vocalization” designates the human act of producing a stream of vocal sounds, subject to two conditions: (1) to a naive listener the stream sounds something like a foreign language; and (2) the speaker himself cannot identify or understand words or larger linguistic units within the stream.8
Free vocalization is attested outside the bounds of the Christian faith,9 as well as within it. What we are focusing on are those instances where Christians produce free vocalization in the context of an intention to worship or to speak to God or for God. The label “free vocalization” is not meant to prejudge the spiritual meaning or value of the act. The label itself does not specify whether or in what way the Holy Spirit is involved. The label is compatible with instances in which people may be exercising a genuine gift of the Holy Spirit. It is also compatible with instances where people may be merely playing vocally in a certain psychological state, or where there may be a demonic source. (Again, we reject the demonic option when Christians are the participants.)
When we survey instances of modern Christian free vocalization, what do we find? In a few instances, people claim to have spoken in a language that they did not learn, but which was recognized by a listener. Such instances, if true, would fall under category 1b (known language, fallible message). For theological reasons, we exclude 1a (known language, infallible message). There are good arguments that the canon of Scripture is complete, and that there are no more instances of infallible verbal special revelation. Some prominent continuationists agree with this restriction.10 (For the same reason, in the modern context we exclude the other options involving infallibility, namely 2a [unknown language, infallible message] and 3a [language-like, infallible message].)
Surveys of modern free vocalization show that instances of a recognized foreign language, if they exist, are rare.11 Most instances are not readily identified. So that leaves us with possibilities 2b (unknown language, fallible) and 3b (language-like, fallible). Careful analysis by linguists has persuaded them that most instances do not have all the features belonging to natural human languages, so they fall into category 3b (language-like, fallible).12 But we should note that confident discrimination between unknown languages (2b) and language-like utterances (3b) can take place only on the basis of technical linguistic expertise. To naive listeners, instances in category 3b (language-like, fallible) still “sound like” a foreign language.
Further analysis by linguists and psychologists has convinced many that free vocalization is fairly easy to produce. The capability is widespread in the human race. And in some cases it can serve as a kind of help.13
4. Expectations for the First Century Church
With these points in hand, we now can return to the situation of the first century church. What was happening there? The descriptions in Acts 2 and in 1 Corinthians 12–14, which are the instances with the fullest information, suggest that instances of speaking in tongues in the first century church consisted in acts of free vocalization. Speaking in tongues was more, of course, because it was an exercise of a gift of the Spirit. But not less. That does not by itself imply that speaking in tongues in the first century should be equated with free vocalization in the context of the modern church. We deliberately crafted the label “free vocalization” to be a broad category. It appears to be broad enough to cover both the early church and the modern church. But that is in principle compatible with two diverse conclusions: (1) the phenomena are the same; or (2) the phenomena are at a theological level completely different, because only in the first century was speaking in tongues a genuine gift of the Spirit.
But now our survey of the modern situation has some bearing. In particular, it is noteworthy that free vocalization is fairly easy. And it has some value at a psychological level. These features seem to be features that belong to human nature in general. It is not something peculiar about modern times that has made free vocalization what it is—though certainly a particular theological interpretation of tongues within the charismatic movement comes in and overlays free vocalization by giving to it a theological interpretation. It would seem plausible, therefore, that free vocalization was a possibility for human nature in the first century just as much as it is today. It would be possible for people to engage in free vocalization in the first century, with much the same contribution of psychological, neurological, and muscular factors that analysts observe today.
Now we can combine that possibility or capability with the situation in the Corinthian church. The Corinthian church was by no means an exemplary church. It was disorderly and unruly and immature in several respects. It was confused doctrinally. But it was even more confused in its practice. Given that situation, it seems likely that, if free vocalization of a modern kind occurred anywhere, it might have occurred at Corinth. Of course we cannot know that it occurred. But it might have occurred.
Since the Apostle Paul in 1 Corinthians was giving general principles for guiding the practice of corporate fellowship and worship at Corinth, we can infer that his principles were intended to cover not only what actually happened before he wrote, but in principle anything that might occur in church after he wrote. In other words, his principles covered instances of modern-type free vocalization in category 3b (language-like, fallible), because such instances were possible.
If instances in category 3b (language-like, fallible) cropped up at Corinth, the Corinthian Christians would immediately have classified them as instances of speaking in tongues. So Paul’s instructions about speaking in tongues cover these instances. Therefore, free vocalization in category 3b (language-like, fallible) is a form of speaking in tongues, in the way that the expression is used in 1 Corinthians. Therefore it is a gift of the Spirit. Therefore it is a gift of the Spirit today. Therefore the continuationists are right and the cessationists are wrong, with respect to speaking in tongues within category 3b (language-like, fallible).
We should be careful about this train of reasoning. It is not airtight. At several points, it might get derailed. Let us consider some objections.
5.1. Tongues Are No Longer a Spiritual Gift
First, we might wonder whether making free vocalization a gift of the Spirit in the first century automatically makes it a gift now. It continues now, but perhaps it belongs to a different category now. Perhaps now it merely offers a form of psychological release. The trouble with this argument is that it appears to make speaking in tongues an exceptional case, in comparison with other gifts of the Spirit mentioned in the relevant passages, namely in Romans 12, 1 Corinthians 12–14, and Ephesians 4:11–16. As examples, let us consider the gift of teaching, the gift of administration, and the gift of help. All three are mentioned in at least one of the passages. Surely these functions are still here today, and we still regard them as gifts of the Spirit.
Here we touch on the disputed question as to which gifts continue beyond the era of the apostles. What about the apostles themselves? The gift of apostles appears in two lists of gifts (1 Corinthians 12:28; Ephesians 4:11). Both times it occurs first, showing its prominence. Apostles are indeed an exception, as cessationists and some continuationists have argued.14 The original apostles still speak to us through their writings in the New Testament. But we have no new apostles with direct divine authority in their speech.
Is the gift of tongues also an exception? But with the apostles we are saying there are no new apostles. With tongues, we cannot say that. Free vocalization is still here today. We are considering whether it makes sense to say that free vocalization was a spiritual gift in the first century, but even though it is still around today, it is no longer a spiritual gift today, but something else—a mere psychological help, perhaps. So, by parallel reasoning, could we not conclude that teaching, administration, and helps are still around today, but that today they are mere psychological helps? No. It simply will not work. We may suspect that an argument that treats tongues differently is singling them out merely because some people have already decided that they do not want tongues to occupy an integral role in the church today.
5.2. The First Century Was Highly Supernatural, Different from Today
A second way of avoiding a close relation between the first century and now would be to heighten the emphasis on the supernaturalism in the first century church. This emphasis on the supernatural might also be combined with an emphasis on the fact that for most of the first century the church was in an “open-canon” situation. Canonical writings were still being produced. Infallible oral teaching was being given by the apostles.
How would this situation look with respect to speaking in tongues? We might picture the church in our minds as filled with spectacular miracles, miraculous healings, infallible prophecies, and beautiful, infallible messages in tongues. We idealize it. We erase from our minds the possibility of confronting anything so lowly and so uncomfortable to respectable people as unintelligible utterances from unsanctified people with mixed motivations. It takes a case like the Corinthian church to dispel the illusion by showing that the first century church was not always a model church. In my opinion, the strong presence of supernaturalism does not erase the possibility of more prosaic forms of free vocalization.
In other words, though we grant that there may have been infallible messages in tongues, within an open-canon situation, we may also allow that there may have been fallible messages. Moreover, the presence of spiritual gifts may be combined with instances of misuse. Tongues, like the gifts of administration or teaching, may be used in a fleshly way, or with mixed motives.
5.3. Paul Had a Special Conception of Tongues
A third possibility for separating out the first century tongues is to focus not on the Corinthian church but on Paul’s conceptions of spiritual gifts. We might observe that Paul conceives of tongues as functionally equivalent to prophecy (e.g., 1 Corinthians 14:5, 12–13). So we might argue that this equivalence implies that for Paul the category of “speaking in tongues” has built into it the feature of infallible divine authority.15
That is plausible. But there are difficulties.
First, in the context of 1 Corinthians 14, Paul also observes differences between tongues and prophecy. The tongue-speaker does not understand with the mind (verse 14). His speech does not edify others unless it is supplemented with interpretation (verses 4–5). It produces a different reaction from unbelievers (verses 22–24). The partial equivalence between prophecy and tongues is for the purpose of building up the body of Christ. An infallible message can build up the body of Christ, and so can a fallible message if it is in fact true to biblical doctrine. The point of the comparison between prophecy and tongues in 1 Corinthians 14 is not to develop a detailed theology of the nature of tongues, in terms of its intrinsic divine or human qualities, but to instruct the Corinthian church in a practical way concerning the unity of the body and the importance of serving one another by edifying communication. Tongues that are interpreted can serve; uninterpreted tongues cannot.
Second, even if we read a detailed theology into Paul’s conception of tongues, it does not help us on the level of what Paul actually communicates to the saints at Corinth. Paul’s communication is designed to help the saints. Let us hypothetically suppose that, according to Paul’s theology, true tongues as a gift of the Spirit are always infallible and inspired and are always instances of human languages. Anything else is not real. It is bogus. In this case, you would suppose that Paul would have to instruct the Corinthians on how to distinguish the true kind from the false kind.
But Paul does not do that. By not doing that, his words confirm the Corinthians in the naive assumption that anything that sounds like tongues is tongues. So Paul is instructing the Corinthians on a practical level that actually has no contact with his alleged neatly crafted theological conception. The theological conception does not actually get expressed in a practical way in 1 Corinthians 12–14. If we are Christians, it is the canonical document, the expression in 1 Corinthians 12–14, that governs us, not a hypothetical conception in Paul’s mind that does not get expressed.16
5.4: The Gift of Discerning Spirits Saves the Corinthian Church
A fourth route points to the gift of “the ability to distinguish spirits” in 1 Corinthians 12:10. Commentators discuss the meaning of this gift. There is some uncertainty. But it appears to be a gift for discerning between good and bad spiritual sources—the Holy Spirit and angels on the one side, and demons and the human spirits of false teachers on the other. We might think of the case in Acts 16:18 where Paul recognizes the spirit of divination in the slave girl. Conceivably the functioning of this one key gift could enable the Corinthians quickly to sort out and suppress anything that was not a “true” gift. And in some people’s minds that might include anything in the category 3b (language-like, fallible).
In my opinion, among the four objections this one is the most appealing. But there are some difficulties.
First, the presence of this gift does not seem to have enabled the Corinthians to have sorted out very serious problems that are addressed earlier in the letter and in 1 Corinthians 15. So it seems overoptimistic to assume that it would be the answer to the difficulties with discerning kinds of tongues.
Second, categories 1b and 2b seem undeniably to represent cases of positive activity of the Holy Spirit, because speech in another human language needs the power of the Holy Spirit. It is unclear why everythingin category 3b would be automatically excluded as unspiritual, since, in content, it might be just as “spiritual” as the content in the form 2b. The exclusion of everything in category 3b seems arbitrary, except as a move deliberately designed to cut the connection with modern tongues.
Now, if some instances in category 3b are fleshly or unsanctified, it does not follow that they all are. And that is exactly the approach taken by some modern continuationists, who recognize that some instances of modern free vocalization are fleshly in motive. There remain some instances in category 3b that may not be fleshly. And then we must recognize the presence of a spiritual gift in the modern situation.
The arguments in this article are tentative. But, granted this tentative status, we can still explore possible implications. If, in the end, we decide that free vocalization in the 3b category is still a gift of the Spirit—or that some instances are, while some are fleshly—what do we do?
It would seem that the further instructions about tongues within 1 Corinthians 12–14 are still relevant. First, 1 Corinthians 14:39b may be relevant to our own time: “do not forbid speaking in tongues.” But of course the same guidelines would hold today as we find in 1 Corinthians 14. Tongues-speaking in public should be accompanied by interpretation (v. 27). If we maintain that the gift of interpretation has ceased, then only private tongues-speaking should take place. Whether the gift of interpretation continues today is a topic that needs its own discussion. But some of the discussion of tongues might be suggestive by analogy.
Second, we may consider whether the debate about the nature of prophecy runs in some ways parallel to the debate about the nature of tongues. If tongues potentially come in an infallible form (type [a]) and a fallible form (type [b]), perhaps the same is true for prophecy. Does the naive listener at Corinth consider “prophecy” to be anything that sounds like prophecy?
That discussion is for another day.
 An earlier version of this article was delivered at the Annual Meeting of the Evangelical Theological Society (Denver, CO, 13 November 2018).
 Some commentators have proposed that what we have here is a miracle of hearing. The audience heard in their own languages, but the speakers were speaking in their own native language—Greek or Aramaic. But this proposal seems implausible, because Acts 2:4 indicates that the Spirit empowered the speakers, not the listeners. The same verse indicates that the speakers spoke in other languages, not that the hearers heard in other languages.
A second proposal says that we have merely different dialects of Greek, belonging to different regions of the Roman Empire. This interpretation is possible, since the key word διάλεκτος can designate either a dialect or a language (BDAG). But this proposal weakens the theological significance of the event. The day of Pentecost represents a reversal of Babel. The separation of people by languages is being overcome by the unity of renewed humanity in Christ, through the Spirit. See Craig S. Keener, Acts: An Exegetical Commentary (Grand Rapids: Baker, 2012–2015), 1:821–23.
 Tongues are also mentioned in Acts 10:46 and 19:6. There is no detailed information in either case about the linguistic nature of the phenomena. But one can see how both verses fit into larger purposes in the book of Acts. Acts has as a major theme the spread of the gospel to broader geographical areas and diverse ethnic groups (Acts 1:8). Acts 10:46 serves to confirm that the Gentile God-fearers are included when they believe. Acts 19:6 deals with disciples of John the Baptist. Both passages have links backward to the tongues on the day of Pentecost. “Extolling God” in 10:46 has a tie with 2:11, while “prophesying” in 19:6 has a tie with prophecy in 2:17–18.
 Some interpreters have claimed to find a clue to the nature of speaking in tongues in 1 Corinthians 13:1, which mentions “tongues … of angels.” But we can only speculate about what angelic languages might be. See Vern S. Poythress, “Linguistic and Sociological Analyses of Modern Tongues-Speaking: Their Contributions and Limitations,” WTJ 42 (1980): 367–88 [374–75], reprinted in Speaking in Tongues: A Guide to Research on Glossolalia, 469–89, ed. Watson E. Mills (Grand Rapids: Eerdmans, 1986). In this article we therefore confine ourselves to the question of human languages.
 See D. A. Carson, Showing the Spirit: A Theological Exposition of 1 Corinthians 12–14 (Grand Rapids: Baker Academic, 1987), 80, citing Cyril G. Williams, Tongues of the Spirit: A Study of Pentecostal Glossolalia and Related Phenomena (Cardiff: University of Wales, 1981), 26.
 See Vern S. Poythress, “The Nature of Corinthian Glossolalia: Possible Options,” WTJ 40 (1977): 130–35. Close to this is Max Turner, The Holy Spirit and Spiritual Gifts: Then and Now (Carlisle, UK: Paternoster, 1996), 314. In 1 Corinthians 14:21–22, Paul cites Isaiah 28:11, which in context prophesies conquest by a people of “foreign tongue,” an actual human language.
 There may be a third category, where a Christian produces utterances from fleshly motives. See below.
 For additional features, see Poythress, “Linguistic and Sociological Analyses,” 369–70; D. A. Carson, Showing the Spirit, 77–88; Turner, The Holy Spirit, 303–14.
 L. C. May, “A Survey of Glossolalia and Related Phenomena in Non-Christian Religions,” American Anthropologist58 (1956) 75–96; Keener, Acts, 1:817.
 E.g., Wayne A. Grudem, The Gift of Prophecy in the New Testament and Today (Westchester, IL: Crossway, 1988), especially 64–65.
 Poythress, “Linguistic and Sociological Analyses,” 374.
 David Hilborn, “Glossolalia as Communication: A Linguistic-Pragmatic Perspective,” in Speaking in Tongues: Multi-Disciplinary Perspectives, ed. Mark J. Cartledge (Milton Keynes, UK: Paternoster, 2006), 112–17; William J. Samarin, Tongues of Men and Angels (New York: Macmillan, 1972), 118–28.
 Poythress, “Linguistic and Sociological Analyses,” 370; Keener, Acts, 1:818.
 Grudem, Gift of Prophecy, 25–65.
 We cannot within the scope of this article take up the debated question of whether all “prophecies” in the first century church were either infallible or utterances of false prophets.
 We may profit much by considering the idea of speaking in tongues and 1 Corinthians 12–14 in the larger contexts: Paul’s theology of spiritual gifts, of the church, of the kingdom of God, and of salvation. When we do so, it is natural to consider Paul’s thinking from our own angle. It becomes easy to color Paul according to our unconscious preferences. If we care for sound doctrine and prefer precise terms (e.g., terms for “inspiration,” or for “speaking in tongues”), we imagine such precision in him. If we care for the distinction between canon and human fallibility, we imagine Paul as naturally paying close attention to it in his instruction to the churches. If, on the other hand, we incline to liberal or neoorthodox theological ways, we may feel suspicion toward “propositional” precision in past orthodoxy. We imagine Paul caring about spiritual vitality but not propositional doctrine.
I find myself between these two approaches because I care deeply for sound doctrine, but see it as built on a complex web of teaching in the whole Bible, rather than on a match between modern technically precise terms and the more ordinary and flexible modes of communication in Scripture. Scriptural communication is based ultimately on the mystery of trinitarian communication (Vern S. Poythress, In the Beginning Was the Word: Language—A God-Centered Approach [Wheaton, IL: Crossway, 2009]).