Asyndeton as unmarked connective

In his Discourse Grammar of the Greek New Testament, I’m thrilled that Steven Runge begins chapter 2 on “Connecting Propositions” with a discussion of asyndeton as the unmarked (default) connective. Asyndeton basically means that no connective is used to link consecutive propositions in a discourse. I believe that exegetes may often find huge benefits in recognizing asyndeton as the unmarked connective, especially if they ask why no other connective was used. This is a great example of the payload of the concept that Runge introduces in chapter one that “choice implies meaning” (pp. 5ff). Since the author had the choice to use various connectives, what is the specific meaning of this connective? And this is no less true with asyndeton when no connective appears.

However, I think Runge is still riding the fence a bit and perpetuating a misconception about asyndeton that contradicts the linguistic theory of markedness that he has already summarized. He’s right on when he says this:

The use of asyndeton indicates that the writer chose not to make a relation explicit. The relation must be gleaned from the context (p. 22).

And he’s also right to suggest that “asyndeton can be used at points of discontinuity” or “in contexts of close connection, such as moving from generic to specific” (pp. 22-23). But that is the extent of what Runge says about asyndeton used to express continuity. Nothing else. On the other hand, the weight of his discussion promotes the idea that asyndeton does NOT mean what other connectives mean:

Levinsohn summarizes the use of asyndeton in non-narrative by stating that since explicit connectives are used to indicate clause relationships such as strengthening, developmental, associative, or inferential, “the use of asyndeton tends to imply ‘not strengthening, not developmental, not associative, not inferential, etc.‘ ” (p. 23, citing Levinsohn, Discourse Features of New Testament Greek, 2nd ed., p. 119).

But that is only one side of a much more carefully nuanced description by Levinsohn:

Strictly speaking, the absence of any conjunction between sentences of a Greek text should imply only that the author offered no processing constraint on how the following material was to be related to its context… In practice, however, New Testament authors tend to use a conjunction whenever the relationship with the context concerned is strengthening (γὰρ), developmental (δὲ), associative or additive (καὶ), or inferential-cum-resumptive (ὅτι), etc. Consequently, asyndeton tends to imply “not strengthening, not developmental, not associative, not inferential, etc.” This is why asyndeton is often the norm when the relation of the following material to the context is not logical or chronological. [At this point, Levinsohn includes a footnote in which he suggests asyndeton as the norm for orientation, restatement, and associative (comment and parenthetical) relationships.]

Asyndeton is found in two very different contexts in non-narrative text:

  • when there is a close connection between the information concerned (i.e., the information belongs together in the same unit)
  • when there is no direct connection between the information concerned (i.e., the information belongs to different units).  (Levinsohn, p. 118)

At this point, Levinsohn includes a footnote in which he suggests that one may recognize when no direct connection between juxtaposed information is intended by the presence of vocatives and orienters (complement-taking predicates). I don’t agree with Levinsohn on this point. These devices may be included precisely to draw attention to the next development in the argument whether or not it relates to the previous passage. On this point, Iver Larsen argues that a vocative is

a rhetorical device, not a structural device, and it functions to establish a closer relationship with the hearers.” (“Boundary features in the Greek New Testament,” Notes on Translation, vol. 5, 1991:51)

The default assumption in any communication is that consecutive units do relate. I don’t believe that just because an author may typically use particular connectives to explicitly convey specific relationships that the use of asyndeton in other places means that those relationships are not implied. It may simply be the difference of whether those relationships are explicitly indicated with a conjunction or implicitly included with asyndeton. Sometimes making the relationship explicit says too much, or skews the argument, or betrays the persuasive power of the author too soon and ruins the chances for successful and convincing communication.

Thus, asyndeton is best understood as the unmarked form that may implicitly include a broad range of semantic relationships. But the way that Runge presents asyndeton, with the weight of its supposed significance falling on discontinuity, contradicts what he has already summarized about markedness:

The default option is considered ‘unmarked’ for the qualities found in the other members of the set. The quality may or may not be present.  The choice to use a marked form represents the choice to explicitly signal the presence of a quality that would only have been implicit if the default were used (pp. 11-12).

Whether the relation is explicit or implicit is very different from saying that the unmarked use of asyndeton implies that those logical relations are not present at all. ‘Unmarked’ means that a feature is not explicitly included. It does not follow, however, that a feature is explicitly excluded.

In my recent thesis addressing discourse concerns in the Letter of James, I have discussed the extensive use of asyndeton in that letter. Scholarship in James has too often erred in assuming that asyndeton implies discontinuity. That idea follows the outdated approach to James made popular by Dibelius in the early 20th century that the letter is composed of a string of unrelated sayings and shorter discourses. Yet, it is very enlightening to interpret the possible functions of asyndeton in the same letter if we follow the default assumption of coherence (as humanity universally does with almost any communication).

Here is the description of asyndeton that I have proposed in discussing the Letter of James. It follows the theory of markedness more closely…

Although the interpretation of explicit conjunctions is often a matter of inferring implied semantic information, the prevalence of asyndeton (the lack of conjunctions), means that coherence and text organization must be determined even more frequently on the basis of implied logical relationships. With 80 instances of asyndeton after 1.1, other bases must be recognized for grouping many units together, including implied semantic relationships. This is also true for larger spans. Investigating the possible logical relationships where asyndeton shows up at higher discourse levels is often neglected. It is difficult since the possibilities for coherence are multiplied with larger spans of text. The major difficulty with analyzing asyndeton is that it may represent either of opposite ends on a scale of cohesion. The two units may be so closely related that no conjunction divides the thought. Or, the units may be so distinct that no conjunction is necessary. If asyndeton is considered to reflect continuity, the extent of each unit being related and the implied logical relation are also unspecified. These determinations must be made from other contextual clues. If anything, asyndeton may indicate the author’s desire to not emphasize any specific relationship. Asyndeton is often found between spans that have some kind of continuity and discontinuity, and for that reason, an explicit conjunction may have communicated too much. (Pehrson, “Mitigation and Intensification of Persuasive Discourse in a Koine Greek Letter: Coherent Macrostructure in the Letter of James,” pp. 58-59)

I see now that what I have described above is somewhat close to what Levinsohn describes for asyndeton. We both recognize the possibility of asyndeton being able to reflect either continuity or discontinuity. However, I am much more ready to find continuity (even if it is a more loose or broadly defined continuity) where Levinsohn may see a change in major or minor topics (cf. Levinsohn, p. 119).

Levinsohn also suggests that asyndeton and other connectives are used differently by different New Testament authors:

The ways in which καὶ and δὲ are used in John’s Gospel do not correspond exactly with how they are employed in the Synoptic Gospels and Acts. This is because two other forms of linkage are employed in John’s Gospel in contexts in which καὶ and δὲ would have occurred had the material been written in the style of the Synoptics. One of them is asyndeton (the absence of a conjunction), which is John’s default means of conjoining sentences (Poythress 1984:331), instead of καὶ. John’s other common marker of linkage is ὅτι; he uses it as a low-level development marker in certain contexts in which the Synoptics and Acts use δὲ. (Levinsohn, pp. 81-82)

Such variance of usage between different authors is a good argument that supports the idea that “choice implies meaning.” Yet with asyndeton, we must not take this too far and conclude that asyndeton cannot mean what other connectors mean. Rather than being a question of either-or, it is probably more of a scalar notion. Asyndeton may imply the same relationship as other connectives, but it does not express it explicitly. There may be a difference of degree for the particular relation, or asyndeton may be an intentional move in the rhetoric to even momentarily hide the relationship. Good argumentation is not always immediately clear. Asyndeton may allow the movement of the argument to be realized only after more of the story is heard, and then with the benefit of gaining a better hearing.

So, I disagree with Runge somewhat in the details. As he says in the preface,

The reader still bears the responsibility of synthesizing and interpreting the analysis and can choose to reject a claim just as one might with most any other scholarly resource. (p. xix)

But I’m still loving this book. Runge is a good conversation partner. And it’s good to go back to Levinsohn and other discourse studies while reading him.

Runge fills a gap

Of the eleven new books on my shelf, Steven Runge’s Discourse Grammar of the Greek New Testament is the one I’m most excited about. First of all, it continues the line of valuable Greek reference tools that have a green cover. So it’s obviously in good company with Bauer-Danker-Arndt-Gingrich, Daniel Wallace, Stephen Levinsohn, and the Loeb Classical Library.

If you know anything about what my latest research has been in applying discourse linguistics to the interpretation of the New Testament, you’ll know why I’m so excited. Too often, exegesis is limited by a narrow view of the meaning of words and sentences without considering how those words are used in the wider contexts of whole discourses with patterns of use within a language community (and even patterns shared cross-linguistically). Many have touted the benefits of discourse linguistics for exegesis, but it looks like this might finally be the work to bridge the much-needed gap in introducing the theory to a wider audience in the academic world of New Testament studies.

After reading the foreward by Daniel Wallace, I immediately thought back to Wallace’s own introduction to his Greek Grammar Beyond the Basics (1996) and why he excluded discourse analysis from his treatment of Greek grammar. In Runge’s preface, he himself cites Wallace’s reasons for leaving discourse considerations out. But I think Runge got Wallace’s sentiments a bit wrong! In only one sense can Wallace be said to be in the camp with those who “believe that linguistics and discourse studies have overpromised and underdelivered.” Wallace was still waiting for the promise to be fulfilled, still waiting for the delivery, but not because he doubted the value of discourse linguistics. In fact, Runge only cites the first three of Wallace’s four reasons for excluding discourse. The fourth one was the most promising, and the one that kept me waiting for a book like Runge’s: “(4) Finally, DA is too significant a topic to receive merely a token treatment, appended as it were to the end of a book on grammar. It deserves its own full-blown discussion.”

I’ve only read through the first chapter of Runge so far, but already, I’m very pleased with the kinds of things that he is challenging NT scholarship with…

Choice implies meaning

If a writer chose to use a participle to describe an action, he has at the same time chosen not to use an indicative or other finite verb form. This implies that there is some meaning associated with this decision. Representing the action using a participle communicates something that using a different mood would not have communicated. Defining the meaning associated with the choice is different from assigning a syntactic force or from determining an appropriate translation. It requires understanding what discourse task is performed by the participle that would not have been accomplished by another verb form. (p. 6)

Semantic or inherent meaning should be differentiated from pragmatic effect

Most languages do not have specialized devices that are singularly devoted to prominence marking. It is far more common to find a nonstandard usage achieving specific pragmatic effects. Greek is no exception. The use of the historical present for forward-pointing highlighting exemplifies this. Using a grammatical construction in an ostensibly wrong or unexpected way has the effect of making something stand out. The pragmatic effect achieved is dependent upon the discourse context in which it occurs. The devices described in the chapters that follow exploit some departure from an expected norm to achieve a specific pragmatic effect. Distinguishing semantic meaning from pragmatic effect is critical to providing a coherent and accurate description of the device and its function within the discourse. Neglecting this distinction leaves you with “messy discourse”! (p. 9)

Default patterns of usage should be distinguished from marked ones

To summarize, markedness theory presupposes that one member of a set is the most basic or simple member, called the “default” member. All of the other members signal or “mark” the presence of some unique quality, one that would not have been marked if the default option were used. The marked options are described based on how they uniquely differ both from the default and from one another…. Some English conjunctions distinguish semantic continuity versus semantic discontinuity (and versus but). The conjunctions καί and δέ do not encode this semantic constraint, leading them to be listed under both connective and contrastive relations [in Wallace]. The messiness of this overlap is caused by the mismatch of the feature to the framework used, not by the overlapping features that are marked.  καί and δέ are unmarked for the feature of semantic continuity or discontinuity. (p. 11, 13)


The primary objective of using the various discourse devices is to attract extra attention to certain parts or aspects of the discourse–that is, to mark them as prominent…. Regardless of whether we are looking at a scenic view, a piece of visual art, or even listening to music, we are constantly making judgments about what is “normal” and what is “prominent” based on the devices used to signal prominence. (p. 13-14)


Since prominence is fundamentally about making something stand out in its context, marking prominence typically involves creating contrast with other things in the context. Contrast, in turn, presupposes that a person recognizes the underlying pattern. Even if we cannot verbalize the pattern, we can still perceive contrast. (p.16)

Thesis on coherence in James finished!

I took a break from posting here to finish my thesis before we head back to our translation work overseas.

I successfully defended the thesis on Tuesday at the Graduate Institute of Applied Linguistics, and now we’ve got less than a week before we start heading back to continue our work in language development in Papua New Guinea.

Anyway, I’ve added the following thesis title to the Recent James Scholarship page and to the James Bibliography page…

“Mitigation and Intensification of Persuasive Discourse in a Koine Greek Letter: Coherent Macrostructure in the Letter of James.”

Here’s the abstract…

Supervising Professor: Shin Ja J. Hwang

A longstanding debate continues regarding coherent structure in the Koine Greek New Testament Letter of James. I argue that multiple linguistic perspectives confirm the central theme of trust in divine grace and mercy as foundational to Christian behavior. Applying Lakoff and Johnson’s cognitive semantics theory to James, a faith-journey conceptual metaphor structures the life of faith according to the source-path-goal image schema with a born-of-grace conceptual metaphor reflecting the source. Using Longacre and Hwang’s discourse theory, I describe James in terms of discourse type, notional schemata, macrosegmentation, skewing, paragraph relations, verb/clause salience, and embedding. A prototype approach reveals James as a persuasive text with embedded hortatory and expository units. Movements of mitigation and intensification most clearly reveal the coherent structure within the text’s profile and peaks. The controlling theme is ultimately derived from the above investigations in terms of van Dijk’s theory of macrostructure.

This isn’t part of the abstract, but the macrostructure (author’s global semantic plan) that I suggest for James is this:

Trust God, for he desires both to give grace to live righteously and to judge sin with mercy for all who have true faith, which is shown in humble acts of love for God and others.

I also argue for the following discourse peaks in James…

4:1-6 – First expository peak

4:7-10 – First persuasive peak

5:1-6 – Second expository peak (persuasive through skewing)

5:7-12 – Second persuasive peak

Kamell on Grace and Imitatio Dei in James

I have added the following article to the James Bibliography and Recent James Scholarship pages.

Kamell, Mariam J. 2009. “The Nature of Eternal Security in James: Divine Grace Pairs with the Imitatio Dei.” Paper in the current open online volume 2 of Testamentum Imperium, 28 pages.

The pdf article is available online here.

Mariam and I are very much on the same page regarding the overall message in James. While it is so easy to focus on the commands in the letter and the believer’s responsibility to have faith with works, Mariam recognizes the crucial message in James that such a living faith ultimately comes from the grace of God through his word. It also includes the extravagant mercy of God that triumphs over judgment. Thus, James conveys a theology that spans election through eternity, and this perspective is foundational to understanding the imperatives in the letter.

It’s good stuff to think about at Easter. While the moral obligation in James is hopeless for humanity, everything is possible with God. It reminds me of a quote I saw entering the library the other day: “Do not abandon yourselves to despair. We are the Easter people and hallelujah is our song.” –Pope John Paul II

Thank you Mariam, and congratulations on nearing the completion of your program at St Andrews! We look forward to much more good stuff from you.

The Greek ‘the’ and the English ‘the’

Wayne Leman over at the Better Bibles Blog has posted on the translation of Greek articles here, focusing specifically on occurrences of “the house” in Matthew where a house has not previously been introduced in the text. One discourse pattern we often find is that definite nouns are used only after that noun has been introduced in the text in an indefinite manner. However, definiteness is not always dependent on the article in Greek. I have posted my response to Wayne in the comments of his post. I’ll only repeat part of that here…

When I read “the house” in the gospels about a house that I have not been introduced to yet, this communicates to me that there was a definite house that Jesus was going to. If the translation were to say “a house,” that would sound to me like Jesus was aimlessly meandering and randomly came across any house when he felt like it was time to stop. So in some of these cases, the ‘the’ doesn’t have to have the same discourse function that we often think of when it is used to refer back to a previously introduced noun. Rather, the article conceptualizes the noun in a certain way (perhaps even making it definite, although it is true that definiteness is not ultimately dependent upon the Greek article) for other reasons besides its previous occurence in the text.In Mt. 9:28, I like what the NLT has done here with “the house where he was staying.” That has the effect of communicating a certain definiteness, and it seems to be a very likely referent that is not too overly specific without other clues. Often times “going into the house” in Greek is the equivalent of our English “going home.” On the other hand, isn’t it possible that “the house” refers to Matthew’s house, the last house we hear of before Jesus was summoned to go to the synagogue leader’s house? Maybe not, since that interpretation would assume that Jesus stayed there for more than just dinner and was there for several days during which the disciples of John the Baptist came to him before the synagogue ruler came to him. Probably quite unlikely. Therefore, it seems that the best option is that Jesus is still in his own town (cf. 9:1), so “the house” is probably whatever house he’s staying in, perhaps even a family house, or ‘home’ as “the house” often means in Greek.

As for 13:36, it’s very possible that Jesus is back in his home town again, since 12:15 says that he left the area he had gone to after he left his home town. Also, his mother and brothers are back in the picture in 12:46. Mt. 13:1 refers to Jesus leaving “the house” and so 13:36 refers to him going into “the house.” It’s the same one he left, very definite even if we don’t want to go so far as to say it was his family home.

As for 17:25, this is Peter’s home town, (cf. Mt. 4:18) and we know that Peter’s mother-in-law had a house there (Mt. 8:14), so this is probably one of those places where “the house” is best understood as the definite idea of ‘home’.

Mt. 24:43 has “the house of him” because it has already referred to the ‘homeowner’.

Daniel Wallace discusses the uses and non-uses of the Greek article in Greek Grammar Beyond the Basics, pp. 206-290. I have summarized his discussion in 4 pages if anyone is interested.

As for the differences between the Greek and English uses of the article, it is best to try to identify why a Greek noun has the article in particular instances before deciding if the same meaning is communicated in English with or without the definite article.

Non-imperatives in Romans 12:9-13.

Why can’t we translate the non-imperative clauses in Romans 12:9-13 as something other than imperatives? I don’t see any reason why not. In fact, I believe that translating these non-imperative forms as commands puts too much emphasis on our human effort that just isn’t in the text at this point. The introduction of 14 commands in this passage also has the effect of hiding the more relevant theme that must be continuing in this passage of the Spirit’s control of our minds (cf. Rom. 8:4ff; 12:2,6). The tension within the ethics of the New Testament is that we are frequently commanded to do what we are only able to do through the gift of the Holy Spirit in our lives. The pages of the New Testament do also give plain descriptions of what Spirit-regenerated life looks like, and these should not all be reduced to the rhetoric of direct instructions.

Understood descriptively, vv. 9-13 offer a detailed picture of what love looks like after the preceding discussion of gifts (analagous to 1 Cor. 12-13). It is not a series of orders, but it would have been heard as an attractive description of what love does. Certainly this should motivate us to those actions, but the real power behind any of these loving behaviors is the control of the Holy Spirit in our hearts. These aren’t just mitigated commands. The expository nature of the text leaves more room for figuring in the role of the Spirit.

The real difficulty is how to translate this passage into English when there are some surprising subject-verb agreement features of the Greek text that can’t be automatically carried over into English. The first phrase says “Love (singular) [is] unhypocritical (singular),” but the following supporting phrases are all plural verbal participles (e.g. “abhorring (plural) evil, clinging (plural) to the good…”).

I think the main reason all our English translations go the way of expressing all 14 of these clauses as commands where there isn’t a single imperative verb is this: in trying to do justice to the apparent discrepancy of the lack of number agreement between the first singular clause and the following plural participles, making them all commands apparently solves this problem in English.  By supplying an implied command (“let be”) for the first clause (“let love be unhypocritical”), there is no longer any lack of agreement between the subject of the first clause and the assumed plural “you” subject of all the following participles. The subject now is always ‘you’.

The big problem, however, is that the grammatical subject of this whole paragraph is ‘love’. The traditional English solution loses that and shifts the entire focus to ‘you’. A second person plural subject is not expressed in any form in any of these 14 clauses in vv. 9-13 (nor in the previous 5 verses).

A better understanding of the apparent mismatch in the Greek subject-verb agreement of vs. 9 is that Greek normally allows the semantics of the situation to dictate the forms of the subject and verb. This is regularly seen in various disagreements for person and number when there are compound subjects, and for a variety of semantic reasons (see my summary of the issues here). In Rom. 12:9 the disagreement comes about because the true initial subject of the paragraph is the singular notion of love, but Paul is talking about love that is expressed by the multiple members of the body of Christ. The mismatch in number agreement happens when the singular abstract concept of ‘love’ is introduced and then described by participles that are plural due to the multiple agents in view.

Here’s my first attempt at a translation. Notice that the first command does not occur until vs. 14…

9 Love is unhypocritical: it is people abhorring evil, clinging to the good, 10 affectionate to one another with brotherly love, leading the way in showing honor to one another, 11 not shrunk back in eagerness, boiling over in the Spirit, serving the Lord, 12 rejoicing in hope, enduring suffering, persevering in prayer, 13 sharing their possessions for the needs of the saints, pursuing love between strangers. 14 Bless the ones pursuing you; bless and do not curse.

After translating this passage to more carefully reflect what is happening in the original text, a few things stand out that are not so apparent when all the dependent clauses are translated as separate commands…

The single sentence that includes vv. 9-13 starts off with the broad thematic content of the paragraph, namely, that love is unhypocritical. This theme is illustrated by a broad movement in the following participles from showing love to the brothers in the community who are called saints to a love that endures suffering and is sought after even between strangers.

The transition from the string of participial and adjective phrases in vv. 9-13 to the commands in vs. 14 is marked by the double use of DIWKW ‘pursue’ in vs. 13 in the sense of ‘hospitality’ (or more literally “pursuing love between strangers”) and in vs. 14 in the sense of “bless the ones pursuing (i.e. persecuting) you.”

Understood as a description rather than a series of commands, it also becomes more reasonable to understand the TW PNEUMATI as “the Spirit” rather than as the human spirit that one can manipulate. And that, I believe, is the whole point of Paul using participles in this paragraph rather than imperative verbs. It’s the Spirit’s work, first of all, before it is our own.

“Memorial offering” of Cornelius – Acts 10:4

What does it mean in Acts 10:4 when the angel of the Lord tells Cornelius that his prayers and charity to the poor went up as a “memorial offering” before God? Clearly, this is the language of acceptable sacrifice. But what is the significance of this particular kind of offering, a “memorial offering”?

This question came up because a first draft of a translation I am checking in the Arop-Sissano language in Papua New Guinea has it something like this…

“You often pray, and all the things you give to people with nothing, God has seen this and he thinks of you.”

Is that the intended significance of “memorial offering”—that God thinks of (or remembers) the person who has given the offering? Sort of.

The Greek word here is μνημόσυνον mnēmosunon ‘memorial’, something that enables someone to remember. So if the memorial goes up before God, then it makes possible sense that it functions as a memorial for God to remember something about the one who gives the memorial.

But this word was used in the Septuagint to translate the Hebrew אזכּרה ’azkārâ, the ‘memorial portion’ of the grain offering in Leviticus 2:2, 9, 16; 5:12; 6:15; Numbers 5:26. Driver (Journal of Semitic Studies, 1 [1956], 100) described it this way: “It is the sign whereby the worshipper is reminded or taught that the whole offering is in fact owed to God but that He is pleased to accept only a part of it as a ‘token’ while remitting the burning of the rest of it on the altar so that it may be otherwise consumed.” Thus, Driver puts the focus of the remembering on the worshipper, not on God.

Regardless of whether we think the memorial is more for prompting the worshipper or God to remember something, the particular thing that Driver identifies as the thing to be remembered may be key for understanding the significance of Acts 10:4. The “memorial offering” was only a portion of the grain offering. God was pleased to accept this small portion and allow the rest of the grain offering to be left for the priests to eat even though the whole offering was due him. In Acts 10 it soon becomes apparent that Cornelius and his household function in the story as a representative portion of the Gentiles. Just as the prayers of Cornelius and his charity to the poor arose as the “memorial portion” of an offering before God, Peter recognizes through the grace given to the one man Cornelius that God “accepts men from every nation who fear him and do what is right.”

Thus, in drawing attention to the piety of Cornelius as the ‘memorial portion’ of a worship offering to God, the angel of God anticipates how Cornelius will function later in the episode as a representative of men from all nations who receive grace and peace from God through Jesus Christ.