Volume 46 - Issue 2
A Two-Dimensional Taxonomy of Forms for the NT Use of the OTBy Douglas S. Huffman
The Christian church has long embraced a commitment to the unity of the whole Bible, not least because the New Testament bears testimony that Jesus is the fulfiller of the Old Testament. There is a unity of Scripture, and somehow Jesus is that unity. And we must quickly point out that, even as the Jesus of the New Testament alone makes sense of the Old Testament, it is also the case that the Old Testament alone makes sense of the Jesus of the New Testament. This dual observation is evident from the regular references to the Old Testament made by the writers of the New Testament. Nevertheless, in studying the use of the Old in the New, many questions arise regarding how and why the NT authors utilize the OT Scriptures as they do.
Several factors are involved in assessing the use of the Old Testament in the New Testament. These include analysis of the form classification, analysis of any textual factors involved (i.e., LXX and/or MT dependence), analysis of the NT context where the referenced OT passage occurs, analysis of the OT context of the referenced passage, and analysis of other ancient uses of the OT passage. This article offers a new, two-dimensional taxonomy by which to analyze the forms used by NT authors to reference the Old Testament.1
1. Confusion of Terms
When we speak of the form of a passage, we are referring to the means by which a NT writer refers to an OT text. Scholars working in this field disagree about the kinds of forms observed and what to call them, and many readily admit that the terminology in use can be confusing because of ambiguous definitions.2 This is particularly the case when it comes to such form labels as citation, quotation, paraphrase, allusion, echo, and the like. Most might agree that the noun use is the broadest label and that it encompasses all of the forms descriptive of how the NT might utilize the OT.3 But other labels are also employed at this largest umbrella level including the terms citation and echo. There is even the proposal that the label allusion be the broadest umbrella term—equivalent to the term use—and that an introduced quotation be recognized as a specific kind of allusion.4 Thus, even though many scholars might prefer to employ the terms citation, quotation, paraphrase, allusion, and echo as specific labels for different, mutually exclusive categories underneath the broad umbrella term use, the English language allows for overlap in the semantic domains of these terms, even to the extent that each can be (and has at times been) employed at the highest level.
To address this kind of confusion, we suggest employing a refinement of the labels used for the different forms. Furthermore, we recommend thinking in terms of a taxonomy about the overlapping relationships of these different forms and suggest that this will be helpful to scholarly interaction. But before we discuss form taxonomies further—and recommend a particular two-dimensional taxonomy of forms—we first want to emphasize the importance of refining the definitions of the forms.
2. Disambiguation of Form Labels
The importance of distinguishing between forms is for gaining clarity on what one is attempting to study. If one wishes to study, for example, Luke’s use of the Old Testament (i.e., “use” in the broadest sense of the term), but then proceeds to examine only Luke’s explicit biblical citations, there is some self-deception afoot. Luke’s use of Scripture is much more involved than can be discovered by simply examining his limited explicit citations of it. While many scholars (of course, not all) dwell on citations of the OT in the NT, and while this may well prove valuable to the discussion, we must not assume that this gives us the whole picture concerning a NT author’s use of the Scriptures.5 To assume so is perhaps another version of the old assumption that NT authors always cite (i.e., in a quotation or a paraphrase) everything from the Scriptures that they are thinking about; nothing in the OT context is to be considered if the NT author does not explicitly cite it. But this proof-texting assumption is surely as wrongheaded as such proof-texting accusations. For example, the book of Revelation is clearly steeped in OT imagery and contains dozens of allusions and verbal parallels to the Old Testament even though it has not a single citation of OT Scripture.6 The NT authors utilize Scripture in a variety of forms to serve their purposes, so an assessment of their utilization of Scripture should be informed by that same variety of forms.
So, a certain amount of disambiguation of terms can be helpful to the general discussion of the New Testament use of the Old Testament. This is especially the case if one wishes to investigate a NT author’s particular kinds of uses of the Scriptures. Thus, a taxonomy of forms with clear definitions for the labels can only be of service to the discussion.
3. Taxonomies of Forms
Some scholars take up the discussion of the New Testament use of the Old with the assumption that everyone agrees with their definitions for terms like citations and allusions, without explicitly defining their use of form labels. Some suggest—or rather, appear to be working with—two basic forms: quotations and allusions.7 With more overt intention and by making more fine-tuned distinctions, some scholars distinguish three forms of reference (e.g., quotations, allusions, and echoes)8 or four forms of reference (e.g., quotations, allusions, recollections, and motifs).9 And of course, some have more complex taxonomies that discern various subcategories.10 Employing subcategories can be helpful for clarifying definitions. For example, a broad two-fold division can be spread into a more complex six-fold taxonomy with the category of “quotations” divided into three subcategories (formulaic citation, unintroduced citation, and paraphrase) and the category of “allusions and verbal parallels” divided into three categories (allusion, recollection, and thematic echo).11
This six-fold taxonomy can be placed on a linear continuum or cline with formulaic citation as the most explicit and thematic echoes as the least explicit.12 With such a continuum, the categories overlap or blur into one another, such that some scholars might label a particular usage a paraphrase while others label the same passage an allusion. These labeling differences are more tolerable when the linear continuum is recognized (see Figure A). A linear continuum acknowledges that some passages can be clearly labeled paraphrases and other passages can clearly be labeled allusions, while a third set of passages is best mapped to the border between those two labels, with scholars pushing them to one side or the other (and thus the diagram in Figure A represents the borders between categories with permeable dotted lines).
Nevertheless, while such a diagram is helpful in some respects, this linear system of classification fails to account for some passages that seem to fit two categories that are not next to one another on the linear continuum. For example, sometimes a paraphrase of an OT passage occurs after the NT author makes a formulaic introduction of the text (e.g., Amos 9:11–12 in Acts 15:15–18). Thus, such a citation can be labeled formulaic, but it is not technically a quotation; the passage falls into two quite separated sectors of the linear continuum. We would like to propose a two-dimensional taxonomy that can address this problem.
4. Recommended Labels for a Taxonomy of Forms
Before describing our recommended two-dimensional taxonomy, we need to define our form labels. In commenting on his taxonomy of terminology (which is much like the linear continuum just discussed), Stanley Porter remarks, “I have tried to define the categories used to describe the use of the OT in the NT. I realize that these categories themselves are problematic, yet I offer them as an incentive for further discussion of what continues to be a problematic area of NT studies, both methodologically and in terms of the actual results.”13 What we offer here is a taxonomy that is similar to Porter’s, but a bit more specific at points. Ours utilizes a few more categories (eight over against Porter’s five) and suggests some changes in label titles. Naturally, we want to utilize some of the same labels already utilized in the discussion; but in hopes of greater clarity at the lowest levels of taxonomy of labels, we have tried to avoid using single-word labels that might be mistaken for an unintended category. Furthermore, our recommendation is that the taxonomy be envisioned on a multi-level continuum so as to overcome some of the drawbacks of envisioning the taxonomy on a linear continuum. In doing all this, we mean to take up Porter’s invitation to offer further refinement of the language used in this field of study.
We distinguish eight basic form classifications for the New Testament’s use of the Old clustered under the two broad categories of citations and allusions and verbal parallels. We will provide here brief definitions for each of these broad and more specific categories.
4.1. The Broad Form Category of Citations
Citations are quotations and paraphrases of prior texts. We use the term citation to describe when a NT author, in order to move his argument or narrative forward, was specifically setting out to cite a particular passage of the Old Testament. Most citations in the New Testament are introduced in some fashion (e.g., “It is written …” or “David said about him …”), which confirms the author’s intention to cite a specific OT passage. But some citations are not formally introduced.
Furthermore, we label some citations quotations where the OT vocabulary and word order are largely preserved; and other citations are labeled paraphrases where the NT author appears to be rephrasing the OT passage using synonyms, different verb tenses, altered word order, etc.14 This has led us to four subcategories for the citations form classification section of our taxonomy with two kinds of formal citations (i.e., those with introductory formulae) and two kinds of informal citations (i.e., those without introductory formulae).15
- Formulaic Quotations largely retain the wording of the source text and are introduced by the NT author as citations of a prior text. For example, the citation of Isaiah 61:1–2 in Luke 4:17–19 matches the OT text precisely (LXX) and has a rather formal introductory statement (i.e., “the scroll of the prophet Isaiah …”).
- Introduced Paraphrase citations are introduced by the NT author as citations of a prior text, but reformulate the source text by means of synonym substitutions, altered word order, etc. For example, the citation of Amos 9:11–12 in Acts 15:15–18 clearly paraphrases the OT text (scholars debate whether the paraphrase is dependent upon the LXX and/or the MT) and yet the citation has a formal introductory statement.
- Unintroduced Quotations largely retain the wording of their source texts but lack any introductory formulae identifying the citations as coming from a prior author; the NT author simply quotes the prior text directly as part of his own text. For example, Jesus’s citation of Psalm 118:26 [LXX 117:26] in Luke 13:35 matches the OT text precisely (LXX) but lacks an introductory statement.
- Unintroduced Paraphrase citations are places where a source text is clearly utilized in a reformulated way by means of synonym substitutions, altered word order, etc., but lack any introductory formulae identifying the citations as coming from a prior author. For example, Jesus’s last words on the cross in Luke 23:46 are a recognizable paraphrased citation of Psalm 31:5 [LXX 30:6] even though they have no introduction.
4.2. The Broad Form Category of Allusions and Verbal Parallels
Allusions and verbal parallels are places in the New Testament that indirectly borrow upon the words and/or ideas of Old Testament passages.16 As a broad grouping, allusions and verbal parallels in our taxonomy is divided into two categories that are each divided into two subcategories with more precise definitions that are of greater importance. The first two subcategories are kinds of recollections; the last two subcategories are kinds of allusions.
What we call recollections are passages that do not have any blatantly shared language structures with the prior text, but where the NT author, nevertheless, makes his intentional reference to the prior text clear enough with something of an introduction. Thus, recollections are something akin to a kind of authorially intended cross-reference, and the category is further divided into the subcategories of Scripture summaries (focused on teaching content) and reminiscences (focused on people and/or events).
Unlike recollections, which are kinds of introductions without any blatantly shared language structures with the prior text, the last two subcategories belong to a converse grouping that we call allusions, which do have limited borrowed language from a prior text but without any introductory statements. Allusions are shorter and subtler references to prior texts than citations are. Some scholars suggest that, in order to qualify as an allusion, a text must contain at least three but no more than five words in a combination unique to the prior text.17 But instead of counting the number of words, we follow the more general suggestion of Beale: “The telltale key to discerning an allusion is that of recognizing an incomparable or unique parallel in wording, syntax, concept, or cluster of motifs in the same order or structure.”18 The two subcategories of allusions that we recommend have the labels specific allusions and thematic echoes. Our suggestion is that a specific allusion contains parallel wording to a specific prior text (similar to a citation but less extensive), whereas a thematic echo more subtly shares an idea or thematic parallel with a set of multiple prior texts.
Here then are the more precise definitions of the four form classifications under the broad category of allusions and verbal parallels (numbered here as 5–8 so as to distinguish them from the four form classifications of the broad citations category that are numbered 1–4 above).
- Scripture Summaries are recollections of the teachings of prior Scriptures with a very limited use of similar language and yet some kind of introductory statement. For example, Luke 24:25–27 records Jesus’s interaction with the disciples on the road to Emmaus: “‘O foolish ones, and slow of heart to believe all that the prophets have spoken! Was it not necessary that the Christ should suffer these things and enter into his glory?’ And beginning with Moses and all the Prophets, he interpreted to them in all the Scriptures the things concerning himself” (cf. 24:32, 44–47). Paul similarly tries to convince the Jews in Rome “about Jesus both from the Law of Moses and from the Prophets” (Acts 28:23).19
- Reminiscences are recollections of people and/or events recorded in prior texts with a very limited use of similar language and yet some kind of introductory statement. For example, in Luke 6:3–4 Jesus introduces a particular event (i.e., “Have you not read what David did …”) reminiscing about David eating the bread of the Presence (see 1 Sam 21:1–6), but does so without any quotation or even paraphrase of that prior Scripture.20
- Specific Allusions involve intentional references to specific OT passages by means of borrowed phrases or similar wording but without any introductions. For example, when Jesus references “the Son of Man” coming in “a cloud” with “power and great glory” in Luke 21:27, it is difficult not to think specifically of Daniel 7:13–14, even though Jesus is not quoting or even paraphrasing that passage. With this restricted use of the label specific allusion, if the collocation is not specific enough to identify just one OT passage, we give it the label thematic echo.
- Thematic Echoes use themes or ideas or structures that occur in multiple prior texts. For example, the birth narrative of John the Baptist to Elizabeth and Zechariah in Luke 1 echoes the OT stories about divine intervention in the pregnancies of Sarah and Abraham in Genesis 18, of Rebekah and Isaac in Genesis 25, of Rachel and Jacob in Genesis 30, of Manoah and his wife in Judges 13, and of Elkanah and Hannah in 1 Samel 1–2.21 Thus, as the metaphor implies, thematic echoes in this narrow sense are ideas heard over and over again in prior texts.22
5. A Recommended Two-Dimensional Taxonomy of Forms
Given our eight specific classifications under two broad categories, for greater clarity we have employed a diagram to illustrate the relationships between the various form classifications in the taxonomy. The diagram attempts to represent the complexity of relationships between the forms on a multi-level continuum rather than on a line (see Figure B). That is, rather than merely moving left and right, classification also moves up and down. Like the simple linear diagram, this diagram has left-and-right movement according to the similarity of wording with the OT passage: quotations have much the same wording, paraphrases and allusions have similar wording, and recollections and thematic echoes have somewhat different wording. This diagram, however, adds another level of complexity based upon the presence or absence of an introductory formula, or (in recognition of a continuum) the degree to which the reference is introduced. The more extensive the introduction, the higher the citation would be plotted on the diagram. Finally, the chart’s dotted lines between classification regions acknowledge that scholars differ on such things as the number of same words required for using the label quotation instead of allusion and how formal an introduction must be in order to count as formulaic.
So, for example, because the citation of Isaiah 61:1–2 at Luke 4:17–19 matches the OT text precisely as we have it in the LXX along with a rather formal introductory statement, that citation would be firmly plotted in the “Formulaic Quotations” sector of the diagram.23 Similarly, the citation of Psalm 118:26 (LXX 117:26) in Luke 13:35b also matches the OT text as we have it in the LXX, but it lacks an introductory formula and would thus be firmly plotted in the unintroduced quotations sector. Nevertheless, both of these citations would be labeled quotations because they have the same wording as the LXX.
On the other hand, Jesus’s unintroduced citation of Psalm 31:5 (LXX 30:6) in Luke 23:46 might be labeled an unintroduced quotation by some, but because it has one tense form change when compared to the LXX text, it might be labeled an unintroduced paraphrase by others. Thus, it can be plotted near the border between the “Unintroduced Quotations” and “Unintroduced Paraphrases” sectors. Meanwhile, the more clearly paraphrased citation of Amos 9:11–12 in Acts 15:15–18 has an equally clear formal introductory statement, so it can be firmly plotted in the “Introduced Paraphrases” sector. Each of these passages—with introductions and without—fall into the same general area designating citations of the Old Testament (the gray zone of the diagram; see Figure C).
Outside of the citations area and within the allusions and verbal parallels area of the chart (the blue zone of the diagram; see Figure C), a text such as Luke 6:3–4 has an introduction but no citation (reminiscing about David in 1 Samuel 21:1–6), so it can be firmly plotted with “Recollections,” specifically in the “Reminiscences” sector. On the other hand, Luke 24:46–47 is less firmly plotted, for its clear the introductory formula has led some scholars to suggest it is paraphrase of Hosea 6:1–2 and others to suggest that it is a Scripture summary of the teaching of several OT texts (e.g., Isa 52:13–53:12 along with Hos 6:1–2). Thus, Luke 24:46–47 might best be plotted on the border between the gray zone and the blue zone, between the “Introduced Paraphrases” sector and the “Scripture Summaries” sector. Clearer is Luke 21:27, which has no introductory formula but enough shared vocabulary with Daniel 7:13 to recognize it as a specific allusion. And the birth narrative of John the Baptist in Luke 1 has a clear thematic echo of several OT birth narratives (e.g., those in Gen 18, 25, 30, Judg 13, and 1 Sam 1–2).
In this manner, our recommended charting gives space to plot clear examples of the New Testament’s use of the Old Testament as well as allows the illustrative plotting of debated passages. It is also worth noting that the two-dimensional diagram has intentionally made room for the specific allusion classification to touch directly the unintroduced quotation classification even as does the unintroduced paraphrase classification, as this area is a nexus of discussion for assigning particular passages.24 While this organizational charting of the forms for the New Testament use of the Old Testament may need further refinement, it attempts to go beyond overly simplistic explanations such as systems that suggest all quotations have introductions and those without introductions are to be classified as allusions and echoes.25
Generally speaking, the following broad guidelines provide helpful distinctions between the form classifications we have suggested for the New Testament’s use of the Old Testament. A citation (whether formally introduced or directly cited without introduction) is an intentional excerpt of a specific passage in a prior text—whether that is a word-for-word rendering (i.e., a formulaic quotation or an unintroduced quotation) or a reworded rendering (i.e., an introduced paraphrase or an unintroduced paraphrase) of that prior text.26 Over against the extended shared wording found in citations is the broad category of allusions and verbal parallels, which has two bifurcated subcategories. Recollections are references to prior texts—introductions, if you will—but without much borrowed language, and they come in two varieties: a Scripture summary is a reference to the teaching of a collection of OT texts, and a reminiscence is a reference to a person and/or event found in a prior text. Finally, the common and perhaps overly stretchy term allusions—variously vague references without introduction—has two varieties: a specific allusion involves enough minimal borrowed OT language pointing to a specific OT passage, and a thematic echo is less particular and carries forward ideas and themes found in multiple places in the Old Testament.
Not surprisingly, and as already indicated in the discussion above, some texts are difficult to classify because they have features that seem to qualify them for two different form classifications. Our layout of the forms on a multi-level continuum is an attempt to allow for some of this ambiguity. For example, how dissimilar must the wording of a citation be before it moves from being labeled a quotation to being labeled a paraphrase? How dissimilar must the wording of an unintroduced paraphrase be before it is classified as a specific allusion instead? How many borrowed words and phrases must a specific allusion have? Indeed, the nature of such questions about a particular text may push the charting of that text to the fuzzy boundary between two neighboring zones of the chart.
Indeed, the primary contribution of this article is the visual charting of form classifications of the NT authors’ use of the OT on a two-dimension continuum that clarifies the potential overlap between classifications. The two-dimension continuum charts the presence of introductory formula on one axis and the level of verbal similarities on the other axis. On the one hand, this layout of the forms on a multi-level continuum is an attempt to allow for some of the ambiguity that seems inherent in discussions of particular NT passages. On the other hand, this charting of the different form categories in their overlapping spatial interrelations may well help scholars see that their differences in classifying particular NT uses of the Old may not be as far apart as previously imagined. Thus, the recommended two-dimension continuum gives scholars the opportunity to map their disagreements in such a way as to discover more agreement. If nothing else, one benefit of the suggested multi-level continuum for plotting these forms is having a playing field for scholarly discussions regarding the proper application of their form classification labels.
 This article on forms stems from a paper presented online for the 2020 annual meeting of the Evangelical Theological Society (ETS). It follows on from a paper about functions, entitled “A New Taxonomy of Functions for the NT Use of the OT,” presented at the 2019 annual ETS meeting in San Diego.
 Regarding confusion of terms among scholars for various categories for the NT use of the OT, see ch. 1 of Stanley E. Porter (with Bryan R. Dyer), Sacred Tradition in the New Testament: Tracing Old Testament Themes in the Gospels and Epistles (Grand Rapids: Baker Academic, 2016), 3–25; see esp. pp. 16–22, where Porter cites some specific examples of debated and confused terminology and discusses whether to classify Philippians 1:19—which has five Greek words in the exact order as the LXX of Job 13:16—as a “citation” or merely an “allusion.”
 But we hasten to note that the noun use is employed not only in reference to the form of NT use of the OT, but it is also employed when scholars discuss the completely different taxonomical realm of function for the occurrence of an OT reference in a NT passage.
 David McAuley, Paul’s Covert Use of Scripture: Intertextuality and Rhetorical Situation in Philippians 2:10–16 (Eugene, OR: Pickwick, 2015), 73–74.
 So also Porter, Sacred Tradition in the New Testament, 17, 22–24.
 See Brian J. Tabb, All Things New: Revelation as Canonical Capstone, NSBT 48 (Downers Grove, IL: InterVarsity Press, 2019), 15–17; Jon Paulien, “Elusive Allusions: The Problematic Use of the Old Testament in Revelation,” BR 33 (1988): 37–53; Steve Moyise, The Old Testament in the Book of Revelation, JSNTSup 115 (Sheffield: Sheffield Academic Press, 1995), 11–20; and G. K. Beale, John’s Use of the Old Testament in Revelation, JSNTSup 166 (Sheffield: Sheffield Academic Press, 1998), 60–128.
 The UBS5 has one index of “quotations” of the OT (pp. 857–60 for OT order and continues on pp. 860–63 for NT order) and another index of “allusions and verbal parallels” (pp. 864–83). Interestingly, the “Introduction” to the UBS5 has a brief section describing the cross-references given at the bottom of each page of its text. Rather than two, three categories are numbered and described: (1) quotations, (2) definite allusions (“where it is assumed that the writer had in mind a specific passage of Scripture”), and (3) literary and other parallels (p. 56*).
 Perhaps most notable for distinguishing “echoes” from “allusions,” albeit with great flexibility, is Richard B. Hays, Echoes of Scripture in the Letters of Paul (New Haven, CT: Yale University Press, 1989), esp. 23 and 29; see also the three-part system in Michael B. Thompson, Clothed with Christ: The Example and Teaching of Jesus in Romans 12.1–15.13, reprint ed. (Eugene, OR: Wipf & Stock, 2011), 30.
 Without introduction or explication of the categories, Wilcox simply states, “Old Testament quotations and allusions, reminiscences and motifs appear in some profusion throughout much of the NT”; Max Wilcox, “On Investigating the Use of the Old Testament in the New Testament,” in Text and Interpretation: Studies in the New Testament Presented to Matthew Black, ed. Ernest Best and R. McL. Wilson (New York: Cambridge University Press, 1979), 231. See also the four numbered (but otherwise unnamed) “classes” utilized by Henry M. Shires, Finding the Old Testament in the New (Philadelphia: Westminster, 1974), 65–72.
 For example, Dennis MacDonald utilizes a taxonomy of seven form categories in “A Categorization of Antetextuality in the Gospels and Acts: A Case for Luke’s Imitation of Plato and Xenophon to Depict Paul as a Christian Socrates,” in Intertextuality of the Epistles: Explorations of Theory and Practice, ed. Thomas. L. Brodie, Dennis R. MacDonald, and Stanley E. Porter, NTM 16 (Sheffield: Sheffield Academic Press, 2006), 213–14; some of his seven categories (i.e., citation, paraphrase, reference, allusion, echo, redaction, and imitation) have subcategories (i.e., marked vs. unmarked citations, and conforming vs. transforming allusions, redactions, and imitations). Marshall offers a taxonomy of four large groupings (i.e., actual citations, allusions, echoes, and language) containing a total of nine form categories: (1) summary references, (2) citations with formulas, (3) citations without formulas, (4) paraphrases, (5) allusions, (6) echoes, (7) scriptural terminology, (8) language, and (9) motifs and structures; I. Howard Marshall, “Acts,” in Commentary on the New Testament Use of the Old Testament, ed. G. K. Beale and D. A. Carson (Grand Rapids: Baker Academic, 2007), 518–19; he admits that the boundaries between his echo and language groupings are rather fluid (p. 519).
 The category label “allusions and verbal parallels” is suggested by the UBS5 text; see note 7 above.
 This is very similar to the suggestion of Porter’s continuum of five categories moving from explicit to non-explicit: i.e., formulaic quotation to direct quotation to paraphrase to allusion to echo; Porter, Sacred Tradition in the New Testament, 33–47. See also Hays, Echoes of Scripture in the Letters of Paul, 23; Hays has a three-realm continuum moving from more explicit to more subliminal: quotation to allusion to echo.
 Porter, Sacred Tradition in the New Testament, 46.
 We recognize that some differentiate between citations and quotations as parallel categories, with a citation having an introduction (e.g., “it is written”) and a quotation lacking such an introduction; e.g., Kenneth Duncan Litwak, “The Use of the Old Testament in Luke–Acts: Luke’s Scriptural Story of the ‘Things Accomplished among Us,’” in Issues in Luke–Acts: Selected Essays, ed. Sean S. Adams and Michael Pahl, Gorgias Handbooks 26 (Piscataway, NJ: Gorgias, 2012), 148. Nevertheless, we prefer to use the labels quotations and paraphrases as parallel subcategories of citations and to recognize the presence of an introduction as a separate factor.
 See the discussion of the labels formal and informal in Christopher A. Beetham, Echoes of Scripture in the Letter of Paul to the Colossians, reprint ed. (Atlanta: Society of Biblical Literature, 2008), 16–17.
 As already noted, the title of this broad category is drawn from the otherwise undefined index heading in the UBS5 (over against its separate “Index of Quotations”). See the definition of “allusion” in William Harmon and Clarence H. Holman, A Handbook to Literature, 10th ed. (Upper Saddle River, NJ: Pearson/Prentice Hall, 2006), 15. Porter identifies five significant elements in this definition: the reference (1) may be to historical or literary entities, (2) is indirect, (3) is intentional, (4) might go unrecognized by a reader, (5) is most effective when the author and reader share the body of knowledge involved with the allusion; Porter, Sacred Tradition in the New Testament, 37.
 See Beetham, Echoes of Scripture in the Letter of Paul to the Colossians, 17–20; cf. Porter, Sacred Tradition in the New Testament, 35–36.
 G. K. Beale, Handbook on the New Testament Use of the Old Testament (Grand Rapids: Baker Academic, 2012), 31 (emphasis original).
 Cf. Steve Moyise, Evoking Scripture: Seeing the Old Testament in the New (London: T&T Clark, 2008), 130. What we call recollections, Dennis MacDonald calls the “reference” category; MacDonald, “A Categorization of Antetextuality in the Gospels and Acts,” 212. Hays and Green have something similar to our recollections category that they title “summaries of OT history and teaching” (a title that apparently allows for two subcategories akin to what we call reminiscences for historical events and Scripture summaries for teaching); Richard B. Hays and Joel B. Green, “The Use of the Old Testament by New Testament Writers,” in Hearing the New Testament: Strategies for Interpretation, ed. Joel B. Green, 2nd ed. (Grand Rapids: Eerdmans, 2010), 127.
 Albeit in a broader sense, Bovon also uses the term “reminiscence” in distinction from “quotations”; François Bovon, Luke the Theologian: Fifty-five Years of Research (1950–2005), rev. ed. (Waco: Baylor University Press, 2006), 118. Likewise, Berding uses the term “reminiscence” more broadly, covering this sixth category together with our eighth category of “thematic echoes”; Kenneth Berding, Polycarp and Paul: An Analysis of Their Literary & Theological Relationship in Light of Polycarp’s Use of Biblical & Extra-Biblical Literature, VCSup 62 (Leiden: Brill, 2002), 32. Thompson (Clothed with Christ, 30) also lumps together “echo” with “reminiscence.” Our classification of recollections (with both “reminiscences” and “Scripture summaries”), as a subcategory of allusions is most similar to what Porter labels in a narrow sense “allusion”; Porter, Sacred Tradition in the New Testament, 36–37, “the indirect invoking of a person, place, literary work, or the like, designed to bring the external person, place, literary work, or similar entity into the contemporary material” (p. 39).
 See Kenneth Duncan Litwak, Echoes of Scripture in Luke–Acts: Telling the History of God’s People Intertextually, JSNTSup 282 (London: T&T Clark, 2005), 55. Hays and Green use the label type-scenes for this specific kind of thematic echo; Hays and Green, “The Use of the Old Testament by New Testament Writers,” 127–28.
 While used in earlier works, the metaphorical term echoes for references to the OT found in the NT has come to the fore with the work of Hays, Echoes of Scripture in the Letters of Paul. The impact of Hays’s book is evidenced by the fact that it precipitated the publication of an interactive set of essays from the Scripture in Early Judaism and Christianity section of the Society of Biblical Literature: Craig A. Evans and James A. Sanders, eds., Paul and the Scriptures of Israel, JSNTSup 83 (Sheffield: Sheffield Academic Press, 1993). Even though Hays leans upon John Hollander’s The Figure of Echo: A Mode of Allusion in Milton and After (Berkeley: University of California Press, 1981), where echo is a distinct “mode of allusion” (see pp. ix, 75, 103, 125, passim; esp. 72: “Echo, allusion, and quotation, then, are forms of citation that are clearly related and clearly distinct”), Hays tends to use the label echo in a broader way as if it were synonymous to the label allusion; so also Litwak, Echoes of Scripture in Luke–Acts. By utilizing the more detailed labels specific allusion and thematic echo, we intend to maintain a distinction between these form classifications. This follows somewhat Beetham, who more clearly distinguishes between allusions and echoes; Beetham, Echoes of Scripture in the Letter of Paul to the Colossians, 20–24; cf. 34. See also Porter’s plea to distinguish between these labels; Porter, Sacred Tradition in the New Testament, 46–47.
 Luke 4:18–19 also contains a specific allusion to Isa 58:6; on this, see note 25 below.
 Of course, we readily admit that this charting has difficulty capturing all the complexities in the discussion of form classifications for the NT use of the OT. For example, the classifications of specific allusion and unintroduced paraphrase share space in the category chart with both describing unintroduced references of similar wording; the difference between these two classifications seems related to intention as much as to form.
 E.g., Peter R. Rodgers, Exploring the Old Testament in the New (Eugene, OR: Resource, 2012), 11; cf. Thompson, Clothed with Christ, 30; Travis B. Williams, “Intertextuality and Methodological Bias: Prolegomena to the Evaluation of Source Materials in 1 Peter,” JSNT 39 (2016): 175–76.
Another issue to be faced in charting a taxonomy of forms is the question of where to slot compressed citations, i.e., when a NT author quotes several key phrases from a particular OT source text while eliminating certain parts of the quotation. This notion of “compressed citation” comes from Berding, Polycarp and Paul, 31. Berding distinguishes between a “true citation” (what we call here a quotation), a “loose citation” (what we call here a paraphrase), and a “compressed citation” (which for convenience some consider as a kind of “paraphrase,” although perhaps some of these such passages would best be placed directly on the dotted line between Quotations and Paraphrases in Figures B and C).
Likewise, problematic is the question of composite citations or conflated passages, i.e., when a NT author cites two (or more) different passages (even from different OT books) as if they are one passage. Linking texts on similar subjects is not merely an ancient practice (commonly identified with rabbinic exegetical methods); modern interpreters regularly engage in this practice as part of “doing theology.” Pragmatically we treat such conflated passages here in separate fashion, e.g., the formulaic quotation of Isa 61:1–2 in Luke 4:18–19 has a specific allusion to Isa 58:6, so we acknowledge them each in their own categories.
 One rule of thumb might be to note that a citation is intentional enough that it would be placed within quotation marks (“”) in modern English translations. But, of course, there are differences between modern English translations on what OT citations are to be recognized in the NT with quotation marks.
Douglas S. Huffman
Doug Huffman serves as professor of New Testament and associate dean of biblical and theological studies for Talbot School of Theology at Biola University in La Mirada, California.
Other Articles in this Issue
This article contrasts two books on missiology: Amos Yong’s Mission after Pentecost and T’ien Ju-K’ang’s Peaks of Faith...