The Validity of Ingressive Imperfects in the Greek of the New Testament

N.B. This essay was originally a response to a question raised by the translation committee of a recent English Bible. One of the translators on the committee insisted that there was no such thing as an ingressive imperfect in Greek. The style of the translation was in line with the RSV—that is to say, more of a literal or formal translation. I was an outside consultant. Although I think that translation is a faithful version, I also believe it is not as nuanced as the Greek is. For this kind of translation, the NET Bible is the best.

I will call this translator “Dr. Z.” Although this essay is a response to his objections to the ingressive imperfect, the reader should be able to tell easily enough what I am responding to.

Dr. Z has argued that there is no such thing as an ingressive imperfect. That is, the use of the imperfect tense to focus on the beginning of an action. Exhibit A in his argument is the RSV and NRSV translation. The RSV did not typically translate the imperfects ingressively (but see John 4.52; Acts 27.18), largely because the RSV did not typically translate most imperfects any differently than aorists. The translation principles behind the RSV were to be minimalistic because once the tenses are translated fully, the work begins to look too paraphrastic. A simple, streamlined approach was desired. Thus, it was not due to grammatical reasoning, but economy of verbiage, that has made the RSV what it is. (One argument that Dr. Z uses is that the RSV may have had more classically trained scholars than is often found today. I agree, but think that such is irrelevant for the issue at hand. Moulton lauded the RV for making careful distinctions of the tense uses—including the ingressive imperfect, and those scholars were certainly more classically trained than the RSV folks! As well, as I note below, the majority of the better classical grammars—as well as NT grammarians who were classically trained—argue for an ingressive imperfect.)

Back to the translation issue per se: On this score, I suggested in Exegetical Syntax that the glosses I suggest are often more for exposition than for translation. Indeed, to translate ingressive imperfects as “began doing” is sometimes quite cumbersome. A via media approach is to translate them as “began to do” which is often suitable enough, though not always.

Let’s take a look at some standard classical Greek grammars on this subject.

• Goodwin and Gulick: although they do not explicitly speak of an ingressive imperfect, they do mention the conative imperfect (268, §1255). The ingressive imperfect, of course, is but a step removed from this and, in fact, a step closer to what some translators see as the normative imperfect for it picks up the internal aspect a bit more.
• Similarly, A. N. Jannaris (436, §1848).
• So also Humbert, Syntax Grecque (139, §237)
• Smyth, 426, §1900 speaks of the inchoative imperfect, offering an excellent example from Thucydides 7.51. The inchoative imperfect is another label for the ingressive, just as inceptive is.
• Gildersleeve, 1.88 (§205) gives, as a routine translation of the imperfect, an ingressive notion. This comes in his opening remarks, with his lone example from Lysius 1.11: τὸ παιδίον ἐβόα “The baby was squalling, began to squall, squalled.” He says that “The imperfect tense denotes continuance in the past. It is the tense of evolution, of vision.” The example from Lysius then follows. What is most significant is that in the detailed discussion of the imperfect, involving 11 pages and 21 categories of usage, not once does Gildersleeve offer a category such as “ingressive, inchoative, inceptive” imperfect. Yet he gives this as a routine translation in his general definition. I submit that the ingressive imperfect is so routine that many grammarians simply do not comment on it; that may well be why Goodwin and Gulick do not have this as a distinct category. (This is a point that Fanning also makes in his Verbal Aspect.)
• Kühner-Gerth: §§382, 383 discuss the present and imperfect respectively; in both they note that the imperfect can be used without any stress on the continuation of the action, with the focus more on the beginning of the action.
• Schwyzer-Debrunner, 277, deny its use (this is the only grammar, off-hand, that I know of which does this)

Among NT grammars, note the following:

• Burton (1896) discusses the conative imperfect, but not the ingressive per se (12, §23). Again, if there is a conative imperfect, it presupposes that there is an ingressive imperfect.
• Moulton, Prolegomena (1908), has a decent discussion that may help to clear the matter up. On pp. 128-29, he discusses what he calls the conative imperfect. He begins by noting this: “Especially important, because more liable to be missed, is the conative imperfect, for which we might give the graph (______             ).” He notes, in his definition, that “our linear graph may either be produced beyond our vision, or reach a definite terminus in view … , or stop abruptly in vacuo.” In other words, the graph (and hence, the conative imperfect) puts the emphasis on the beginning of the action without mentioning whether the action was actually performed or concluded. Thus, both the conative and ingressive imperfects (in our terminology) are included. Again, this graph clearly shows that the ingressive imperfect is, if anything, closer to the ‘unaffected’ meaning of the imperfect than the conative imperfect is. If the latter is a bona fide category, then so is the former.
• Robertson lumps the inceptive/ingressive imperfect in with the conative (885).
• BDF, §326: what they call conative imperfect includes, naturally enough, the ingressive imperfect. Their example from Acts 27.41 confirms this, for they translate ejluveto as “ ‘began to break loose’ or ‘broke up more and more’.” This is repeated in BDR (1976). What is significant here is that the ingressive notion in Acts 27.41 clearly is not merely conative in the sense of an initial act that was not accomplished. Indeed, several translations opt for the second rendering which comes close to an iterative notion. Both BDF and BDR list Heb 11.17 (“Abraham tried to offer/began to offer” Isaac). It should be noted that there is little difference here between Abraham beginning to offer his son and trying to offer his son, since in both renderings it is clear that he did not complete the action (as the Genesis account notes) but did place him on the altar. I submit, once again, the basic linguistic argument: The conative imperfect presupposes the legitimacy of the ingressive imperfect, since the ingressive imperfect is closer to the unaffected idea of the imperfect than the conative is. To attempt an action without actually getting to it is one step removed from beginning an action that is not completed which, in turn, is one step removed from doing an action without completing it.
• Hoffmann and von Siebenthal (1990) give much time to the inchoative imperfect, discussing whether it is better to treat it under the conative or the durative imperfect (320, §198).
• Moule, Idiom Book, 9, calls the inceptive imperfect “frequent in the N.T.” He distinguishes it from the conative imperfect, but only gives one example (presumably because it is so routine). (Incidentally, Moule was weaned on classical Greek; his expertise in such matters is unquestionable. This will become relevant in a few moments, as I respond to one of the complaints about modern-day translators.)
• Zerwick, Biblical Greek (1944, 1963), speaks of the conative imperfect, but uses Luke 8.42 as an illustration, showing that in his view the ingressive imperfect is a part of the conative (92, §273).
• Turner, Syntax, 65 speaks of “Incohative” (sic) as an alternative to the conative imperfect option in Acts 27.41, but nowhere else discusses the option. Nevertheless, the fact that he so casually mentions it suggests that he does not rule out the usage.
• Funk cites Gal 1.24 as an example of an “inceptive or ingressive imperfect” (Hellenistic Greek, 2.625), calling this one of the “major nuances of the imperfect” (ibid.). He repeats this point, citing Luke 5.30, on the next page.

In sum: Although Dr. Z said, “There is something unsatisfying about this lack of agreement,” referring to the discussions of the ingressive imperfect in both classical and NT grammars, his case is overstated. His conclusion, after looking at the examples in my grammar, is: “But there is no such firmly established usage, as the disagreements among the advanced grammars attest. Moreover, the examples all have alternative explanations. The idea of an ingressive imperfect is an artifact of the lack of attention to alternatives, especially alternatives having to do with discourse structure. ‘Ingressive imperfect’ as such does not exist.” This again is very much of an overstatement. I submit that the “disagreement” that he claims among the advanced grammars is either due to labeling (grouping ingressive and conative imperfects) or to lack of discussion of routine uses. But none of the grammars he cited says explicitly that such a category does not exist, even though all of them feel free to mention the non-existence of categories elsewhere. Further, what is lacking in this translator’s discussion is any grammatical principle as to why the conative imperfect is permitted and the ingressive is not. My argument—and, I believe, the implicit argument of almost all grammars of ancient Greek—is that if the conative imperfect is a bona fide category, then so is the ingressive. Let’s return to the grammars again for a moment. Three points are to be made here.

(1) Although Dr. Z claims that Burton, Turner, BDF, and Goodwin-Gulick do not speak of the inceptive imperfect, this argument has overlooked the clear statements in Turner and BDF that confirm this category. And, as for Burton and Goodwin-Gulick, I have argued above that several grammars lump the two together, though giving examples only of the conative because if that is a true usage, then the ingressive imperfect must surely also be. Second, several grammars were not mentioned by Dr. Z, many of which are clear standard-bearers of both classical and Koine usage. Among them are Gildersleeve, Kühner-Gerth, Schwyzer-Debrunner, Jannaris, Humbert, Moulton, Moule, Hoffman and von Siebenthal, BDR, Funk, and Zerwick. Of these, the overwhelming majority explicitly include either discussions of the ingressive imperfect or give glosses of such; several of them speak of the ingressive imperfect as common or routine. That Gildersleeve gave this as his gloss in the overall definition of the imperfect, even though he never again addressed the usage is quite telling: the usage is so commonplace that many grammars simply do not discuss it. Its omission in BDF (except by way of illustration) is due to the nature of that grammar: it is a grammar of exceptions, not of regular usage.

(2) Dr Z suggests that the discourse features are never noticed in these grammars and that such argue against the ingressive imperfect. But he does not note my discussion of discourse features which is a part of my presentation (Exegetical Syntax, 544). Further, his argument is confusing on this score. On the one hand, the older, classically-trained scholars behind the RSV did not (presumably) embrace the ingressive imperfect. On the other hand, older, classically-trained grammarians don’t discuss discourse analysis and therefore do not understand that discourse analysis would remove this category from consideration. It seems to me that he wants to have his cake and eat it too.

(3) It may, therefore, be profitable to look at two of the most recent and respected NT grammarians on the verb: K. L. McKay and B. M. Fanning. These men do not agree entirely on the meaning of aspect, but they agree on the ingressive imperfect. And both are well acquainted with the current trends in discourse analysis. (It should be noted as well that McKay is a classical Greek scholar who has an interest in NT Greek; Fanning teaches a course on discourse analysis at Dallas Seminary/Summer Institute of Linguistics [formerly co-taught with Stephen Levinsohn, a leader in the field].)

McKay, New Syntax of the Verb (1994), has a half page discussion on “the conative and inceptive imperfect” (44, §4.3.2), He notes that such uses are “merely the result of the interaction of the aspect of the verb and the context…,” a point I made in my grammar as well.

Fanning, in his Oxford monograph, Verbal Aspect in New Testament Greek (1990), has a full page of discussion of this usage (252-53). Several of his examples are the same as mine. He also notes discourse features: “The context in these instances involves the close collocation of two verbs denoting sequenced situations such that the first indicates the beginning-point of the second…” (252). Fanning also notes that just because most grammars do not include a separate label for this usage, this fact does not mean that they deny its existence (ibid., n. 120). Fanning then discusses the inceptive force of the tenses on pp. 191-92, under the larger umbrella of “the effect of discourse features on aspectual function.” Fanning’s distinctive contribution to the field is to show that Aktionsart—specifically, lexical intrusion—has impacted the ‘ontological’ force of the tenses so that a new combination occurs in various discourses.

A final few comments are in order. First, Dr. Z’s treatment of my examples strikes me as strained. I stand by my exegesis of those texts in general, though certainly one or two may be debatable.¹ Many of them are found in other grammars as well.

Second, parallels between the synoptic gospels show that the ingressive imperfect is a routine use. For example, although Mark 2.23 and Matt 12.1 introduce the pericope of the grain-plucking on the Sabbath with ἤρξαντο τίλλειν/τίλλοντες, Luke 6.1 has the simple imperfect ἔτιλλον. In the pericope of the healing of the Gerasene demoniac, Mark 5.20 says that the healed man went away and “began to proclaim” (ἤρξατο κηρύσσειν) which Luke replaces with the present participle κηρύσσων. Cf. also Mark 14.19 with Matt 26.22. Mark 6.1 tells of Jesus’ return to Nazareth where he “began to teach” in the synagogue (ἤρξατο διδάσκειν), which Matthew 13.54 replaces with the simple imperfect ἐδίδασκεν. Since Matthew has this verb linked to the aorist participle (ἐλθών), it is difficult to see Dr. Z’s point that “the [imperfect] is much more likely to be used in a paragraph where, elsewhere in the paragraph, another action is introduced that overlapped with the first in time.” Or again, “My [Dr. Z’s] explanation is that the imperfect is often used in historical narrative to describe main events that overlap in time with other events described in the immediate context.” This is his basic discourse argument for the use of the imperfect, but it simply won’t do here. Jesus did not continually come to Nazareth, while teaching in the synagogue there! There is, in fact, no overlap between the participle and the verb. The idea is ‘after he came, he began to teach.’ The imperfect picks up where the aorist left off, and is thus naturally treated as an ingressive imperfect (as the parallel in Mark demonstrates). The reason for the imperfect instead of the aorist is that the imperfect puts an accent on the ongoing nature of the enterprise, while the aorist simply introduces a new activity.² (It should be noted, however, that an ingressive or constative aorist sometimes replaces ἄρχομαι + infinitive, for the emphasis of the former is not so much on the continuation of the activity but only its beginning or its fact. Cf. Matt 14. 35 with Mark 6.55; Mark 6.7 with Luke 9.2; Mark 10.28 with Matt 19.27 and Luke 18.28; Mark 10.32 with Matt 20.17 and Luke 18.31; Mark 10.41 with Matt 20.24; Mark 10.47 with Matt 20.30; Mark 11.15 with Matt 21.12; Mark 13.5 with Matt 24.4 and Luke 21.8; Mark 14.65 with Matt 26.67; Mark 14.69 parallels a historical present in Matt 26.71; Luke 11.29 with Mark 8.12 and Matt 12.39.) Similarly, Luke 9.10 has ἐλάλει after the aorist participle ἀποδεξάμενος, clearly giving the force “after he welcomed them, he began to speak of the kingdom of God.” The parallel in Mark 6.34 has ἤρξατο διδάσκειν. Mark 15.18 says that the soldiers “began to salute” Jesus in a mocking gesture (ἤρξαντο ἀσπάζεσθαι), while Luke replaces with the simple imperfect, ἐλέγεν. Mark 8.15 tells us that Jesus “warned his disciples, saying…” (διεστέλλετο...λέγων) which Luke 12.1 expands into Jesus “began to speak to his disciples” (ἠ῎ρξατο λέγειν). Mark 11.9 tells us that the followers of Jesus “cried out” (ἔκραζον), an imperfect which Matt 21.9 duplicates. But Luke 19.37 clarifies and expands the verbal idea, saying that the disciples “began to rejoice and praise God” (ἤρξαντο χαίροντες αἰνεῖν).³ (Cf. also Matt 4.17 with Mark 1.14.)

Third, as much as I respect Dr. Z’s scholarship, I must confess that his sense that the RSV can hardly be improved upon because (in his view) NT scholarship has not advanced much in the last 50 years is somewhat of a surprise. It is similar to the KJV-Only kind of argument in many of its features. I’m sure he doesn’t want to fall into that trap, for it is methodologically anti-Protestant and anti-Reformed.

Finally, I would reiterate: It may well be that the best course for you to take on your translation is to flatten out the verbs, reducing the aspectual force in most places. But the reason for this is surely not that the RSV translators were trying to give verbal aspect its due consideration. The RSV was a minimalistic translation when it came to syntax and lexeme; the mantra was “as literal as possible, as free as necessary.” It would be a serious error, I believe, to read into this that grammatical more than economic principle was driving the RSV. You may well wish to follow their suit, but do so for the same reasons as they actually followed.

The above response can be followed up with a listing of some of the ingressive imperfects that are seen by the NET Bible translators. There is a place for a variety of Bible translations for the Christian today. But one must not mistake a more reserved translation—such as the RSV, ESV, NASB, etc. as necessarily more accurate. Often avoidance of paraphrasis drives the translators’ decisions more than a desire to represent the Greek text fully. Consider the following passages, and compare the NET to other translations.

Matt 4.11              angels came and began ministering to his needs.

Matt 5.2              he began to teach them

Matt 8.15              she got up and began to serve them

Matt 13.54              he came to his hometown and began to teach the people

Matt 15.36              he broke them and began giving them to the disciples

Matt 16.7              they began to discuss this among themselves

Matt 20.11              they began to complain

Matt 26.16              Judas began looking for an opportunity to betray him

Mark 1.21              Jesus went into the synagogue and began to teach

Mark 3.6              the Pharisees went out immediately and began plotting with the Herodians

Mark 5.40              they began making fun of him.

Mark 5.42              The girl got up at once and began to walk around

Mark 11.17              Then he began to teach them

Luke 1.21              they began to wonder why he was delayed in the holy place.

Luke 2.38              At that moment, she came up to them and began to give thanks to God

Luke 2.44              Then they began to look for him among their relatives and acquaintances.

Luke 4.15              He began to teach in their synagogues

Luke 6.11              But they were filled with mindless rage and began debating with one another what they would do to Jesus.

Luke 7.16              Fear seized them all, and they began to glorify God

Luke 9.30              Then two men, Moses and Elijah, began talking with him.

These are but a few of the examples in the NET. In each instance, the imperfect is viewed as introducing a new action. Often it comes on the heels of an aorist verb, which sets the stage for the imperfect. Further, the imperfect verb does not indicate that the action began but stopped. Normally, the implication is that it began and continued. The reason for the aorist and imperfect together is that the aorist then sets up the imperfect action which begins and continues. For example, in Mark 1.21, “Jesus went into the synagogue and began to teach” involves an aorist participles (translated went and) and an imperfect indicative (began to teach). The aorist quickly moves the narrative to the realm of the imperfect, and here is where the action takes place. The imperfect governs (in sense, not grammatically) the narrative from v. 22 through v. 27.

In some of the examples above, there are parallels in other synoptic Gospels which use ἄρχομαι (“begin”) followed by an infinitive. This is clear evidence that the ingressive imperfect is a valid category. As well, on many occasions the same kind of expression (e.g., “began to teach”) vacillates within the same Gospel between an imperfect and ἄρχομαι plus the infinitive. This suggests that there may not be anything significantly different in either way of expressing this thought; stylistic variation may account for it.

---

¹ For example, in Matt 3.5 Dr. Z ignores the temporal indicator, τότε, a clear marker of a new activity. In Matt 5.2, he does not address the aorist participle ἄνοίξας which is a discourse feature indicating the new activity of Jesus teaching on this occasion—precisely as I suggested would fit the semantic situation of the ingressive imperfect. His point that an ingressive imperfect here would imply that “Jesus had not taught crowds before this point in his ministry” strikes me as a bit bizarre. If that were the case, then why does Mark 12.1 tell us that Jesus “began to speak to them in parables” (ἤρξατο ... λαλεῖν)—a text which even Dr. Z would have to admit focuses on the beginning of the activity. (Indeed, I am puzzled as to how Dr. Z would handle many such texts that employ ἄρχομαι + infinitive.) If we were to apply Dr. Z’s criterion for Matt 5.2, then we would have to say that Mark is in error in chapter 12 because he has clearly shown that Jesus taught in parables eight chapters earlier! Dr. Z’s overall treatment of the imperfect here seems to be acontextual, almost as though Matthew were attempting to make universal statements that are not historically conditioned. Does he really mean that once Jesus started teaching, the evangelist could never, for any given occasion, speak of “Jesus beginning to teach” in this context? Dr. Z’s discussion of Mark 9.20 ignores the syntax of the clause with the clear temporal marker of the aorist participle followed by the imperfect. The boy had to fall down before he could roll around. In each of these texts, Dr. Z mistakenly believes that I think the imperfect is used solely to accent the beginning of the activity. He then shows (accurately, I believe) that the activity continued on, was iterative, etc. But I do not deny that; rather, my understanding of the ingressive imperfect is that it focuses on the beginning of an activity (as opposed to other imperfects) that did indeed continue (as opposed to the ingressive aorist).

Overall, I would have to say that Dr. Z seems to ignore the clear discourse features that indicate a topic shift—the major clue that an ingressive imperfect is in view.

² Dr. Z argued, at one point, as follows: “But Wallace says that the ingressive imperfect ‘stresses beginning.’ Does it? Or does it rather indicate the continuation? We are, of course, dealing with the indicative mood. The imperfect and aorist are the main tenses used for the main verbs in historical narrative. In historical narrative, the imperfect derives its distinctiveness from its contrast with aorist and, rarely, perfect.  So does the contrast indicate that it ‘stresses beginning’? It cannot, if the same stress sometimes belongs to the aorist.”

But this is to misunderstand the context of my treatment of the ingressive imperfect. Indeed, Dr. Z has taken the simple definition without looking at the very next paragraph. In the simple definition, I was contrasting the ingressive imperfect with other imperfects, not with the aorist. When compared with the ingressive aorist, the imperfect adds the element of continuation that is lacking in the aorist, as I argued clearly in the next paragraph on p. 544, under “Clarification and Amplification”: “The difference between the ingressive imperfect and the ingressive aorist is that the imperfect stresses beginning, but implies that the action continues, while the aorist stresses beginning, but does not imply that the action continues.”

³ It should be noted that I have translated the imperfects above as simple aorists so as not to prejudice the reader about its force. This should not be taken to mean that I think the imperfect is aspectually flat or is not ingressive.