DEPONENCY AND THE MIDDLE VOICE
For Greek, then, what needs to be laid aside is the notion of deponency.
— Bernard A. Taylor
A paradigm shift is taking place in our understanding of Greek voice, with particular reference to the concept of deponency. A small but powerful movement is calling for the displacement of deponency, and the arguments put forth require careful consideration. The purpose of this chapter is to provide a brief review of the discussion thus far, since an awareness of the development of ideas aids our comprehension and assessment of the current state of affairs. Most of the chapter surveys the contributions of key scholars, with space given toward the end for assessment of the central arguments, remaining challenges, and some suggested ways forward.1
To understand why the validity of deponency is being questioned — or outright rejected — it is helpful to see that this trend has grown slowly over the course of the twentieth century, beginning with some well-known grammarians who raised question marks about the category, and climaxing with the Greek Language and Linguistics Unit of the 2010 SBL conference in Atlanta, at which the presenters uniformly rejected the legitimacy of deponency.
In the 1908 third edition of his grammar, Moulton begins his discussion of the middle voice with verbs that are found with active only or middle only forms, both of which should be given the “unsatisfactory name ‘deponent,’ ”2 if that name is retained for either. In other words, why call middle-only verbs deponent and not also active-only verbs?
On the relation between middle and passive voices, Moulton believes that there is little that distinguishes them: “the two voices were not differentiated with anything like the same sharpness as is inevitable in analytic formation such as we use in English.”3 “The dividing line is a fine one at best.”4 This virtual indistinguishability between middle and passive voices will have implications for the following discussion.
In the 1934 fourth edition of Robertson’s grammar, his initial discussion of deponency is found in the section ominously entitled “The So-Called ‘Deponent’ Verbs.”5 He agrees with Moulton’s assertion that the term should be applied to all three voices if any, but goes a step further by saying that “the truth is that it should not be used at all.”6
Drawing on the parallel of Sanskrit, Robertson points out that in Greek “some verbs were used in both active and middle in all tenses . . . some verbs in some tenses in one and some in the other . . . some in one voice only.”7 Consequently, such verbs are better regarded as “defective” rather than “deponent.” Thus, Robertson concludes that “the name ‘deponent’ is very unsatisfactory. It is used to mean the laying aside of the active form in the case of verbs that have no active voice. But these verbs in most cases never had an active voice.”8
After Robertson, there were a few voices of discontent with the term “deponency,” such as K. L. McKay, who described the label as “not entirely necessary in terms of ancient Greek itself.”9 In an appendix on deponent verbs in Friberg’s lexicon,10 published in 2000, Neva Miller did not set out to challenge the notion of deponency in an explicit fashion, yet her analysis nevertheless drives in that direction. Miller’s main concern was to revitalize our understanding of the middle voice, which has been neglected because of the assumption that verbs without active voice forms are deponent rather than properly middle.
By exploring a representative selection of so-called deponent verbs, Miller demonstrated that these could be understood as middle in voice, which thus suggests “an alternate way of thinking about them,”11 in which “the self-involvement of the subject in the action or state expressed in the verb is highlighted.”12 While Miller did not attempt an exhaustive study, and she put her conclusions tentatively,13 nevertheless she suggests that “if the verbs in the above classes are understood as true middles — and if active forms could not have expressed such concepts — then it may be that categorizing such verbs as deponent is no longer relevant.”14
It was not until 2001 that Bernard Taylor issued a direct challenge to the concept of deponency. Taylor’s paper presented at the 2001 SBL conference challenged not only the term “deponency” but also the phenomenon it supposedly describes, and his arguments reached fuller expression in his contribution to the 2004 Danker Festschrift.15
Taylor points out that the label “deponent” is a Latin term, which arose from the fact that “by the Renaissance, Latin grammar and terminology had become the norm and were used to describe and delimit other languages.”16 Taylor contends that deponency was a Latin notion transferred to Greek and had not existed in that language before.17
Taylor also points out that with the importation of the Latin category, an assumption about the nature of the Greek verbal system is smuggled in, namely, that a verb is not “normal” unless its forms occur in the active voice.18 Thus he objects that “the subject of middle verbs is still the doer of the action, even if the subject bears a different relation to the result of the action than a verb in the active voice.”19 These middle verbs have not “laid aside” a form, since the middle is as old as the active. “Nor have they laid aside meaning, since the middle shares with the active having the subject perform the action.”20
In 2002, classical Greek scholar Carl Conrad argued that “the conception of deponency is fundamentally wrong-headed and detrimental to understanding the phenomenon of ‘voice’ in ancient Greek.”21 The key to Conrad’s discussion is that “the fundamental polarity in the Greek voice system is not active-passive but active-middle.”22 As such, the middle voice needs to be understood in its own right and ought to regain the significance that the artificial category of deponency took from it.
In fact, in somewhat of a reversal of standard thinking, Conrad argues that it is a mistake to understand the passive voice as an independent category with its own morphological paradigms. Instead, the distinction between “middle” and “passive” was not “sufficiently important to require distinct morphoparadigms to indicate the distinct notions.”23 In light of this, so-called “middle/passive” forms should be viewed as one set of forms that are secondarily capable of conveying passive function.24
This inevitably raises the question of the aorist stem that has separate “middle” and “passive” forms. To this, Conrad argues that “the -θη- endings were never essentially passive, even if they were often used and understood as indicating a passive sense to the verb in question; rather the -θη- endings are forms developed in the course of the history of ancient Greek to function for the middle-passive in the aorist and future tenses.”25 In other words, the paradigms currently labeled “middle-passive” and those labeled “passive” should both bear the label “middle-passive,”26 and context will determine whether a middle or passive sense is meant in each instance.27
Also in 2002, Ancient Greek linguist Rutger Allan’s dissertation on the middle voice was completed.28 The dissertation treats the middle voice, not deponency per se, but his conclusions have major implications for the discussion about deponency. Moreover, Allan explicitly rejects the label of deponency:
Media tantum (or middle-only verbs) are middle verbs that do not have active counterparts. They are sometimes called deponentia, a term borrowed from Latin grammar. This term is less adequate since it suggests that these verbs have “laid off ” (i.e. lost) their active forms. There is no historical evidence that this is what actually happened.29
The burden of Allan’s dissertation is to uncover a semantic element common to all uses of the middle voice (including those middles normally designated “deponent”).30 His conclusion is that “the middle voice expresses the presence of the semantic property of subject-affectedness.”31 This means that all Greek middle verbs — regardless of whether they are “middle-only” verbs (media tantum) or also have an active form — indicate that the grammatical subject of the verb is involved in the action somehow. This does not imply that active voice verbs indicate an absence of subject-affectedness; instead, “the active voice must be considered as neutral to the element of subject-affectedness.”32 Thus, active verbs may indicate subject-affectedness, but middle verbs are marked for it — they will always indicate subject-affectedness.
In a way, Allan’s position with regard to deponency is similar to Neva Miller’s conclusions, in that the category becomes unnecessary if a comprehensive understanding of the middle voice can be achieved. Allan is more forthright than Miller in rejecting deponency, but the logic is similar in that a cohesive understanding of the middle voice will render deponency irrelevant.
Following Taylor and Miller, Jonathan Pennington, first publishing on this topic in 2003 then again in 2009,33 points to the importation of the Latin concept of deponency and a weak understanding of the middle voice as the factors leading to the current situation. Moving beyond their lead, however, Pennington develops a defense against two potential problems for the “anti-deponency” position: the existence of active verbs that take middle future forms (i.e., “partial deponents”) and so-called “passive deponents.”
On the issue of middle future forms, Pennington refers to the cross-linguistic phenomenon in which there is a close affinity between the middle voice and future tense.34 Since the future tense can only present an event as a “mental disposition or intention,” the middle voice serves well to communicate that sense.35 This does not mean that all futures will be middle, of course, but it explains why some are.
On the issue of passive deponents, Pennington regards these as middle in voice but with passive forms.36 The reason these middle verbs have passive forms is that “the middle voice form was losing ground to the passive,” such that the passive forms are gradually replacing the older middle forms.37 Labeling this phenomenon as a “slipping of register,” Pennington suggests that in such cases, the aorist middle should have been employed, but because such forms are disappearing, “it is not uncommon for authors to accidentally utilize the more common aorist passive forms.”38
In his dissertation completed in 2010, Stratton Ladewig has spoken in defense of the term “deponency.”39 Since his is the only work that gives an up-to-date defense of deponency at a serious level, it is necessary to engage his key theses. While Taylor has pointed out that the term “deponency” was not applied to Greek until the Latinization of the Renaissance period, Ladewig argues that, while the term is not found in the earliest Greek grammars, the concept is.40 If this claim were to be substantiated, it would, of course, offer a powerful argument in defense of deponency.
In his reading of Dionysius Thrax and Apollonius Dyscolus, Ladewig detects their wrestling with the middle voice. Dionysius says: “There are three voices: active, passive, middle; active such as τύπτω, but passive such as τύπτομαι, but middle representing at one time active and at another time passive.”41 Ladewig interprets this statement to mean that Dionysius here acknowledges a discrepancy between form and function; in other words: a verb may be middle in form, but active or passive in meaning.42 This apparent mismatch between form and function is what the later term “deponency” addresses.
However, I would suggest that this is a misreading of Dionysius Thrax. It is more likely that he is acknowledging that the middle voice is very broad in its meaning, capable of active and passive functions and everything in between. The same can be said about Apollonius Dyscolus.43 Thus, Ladewig’s most powerful argument — that while the term did not exist, the concept of deponency was understood by the Greeks — is not, in my opinion, successful.
The most serious problem with Ladewig’s work, however, is that he assumes a narrow understanding of the middle voice that is largely limited to reflexive and reciprocal functions. If one assumes such a narrow set of functions for the middle voice, then it is natural that one would detect deponency ubiquitously, since middle forms are apparently active in many instances. If, however, one accepts the possibility that the middle voice is much broader in its meaning — as Allan asserts, and as I suggest Dionysius and Apollonius assert — then there is no need for the category (nor the term) of deponency, and in fact it proves to be a mistaken understanding of Greek voice.
In November of 2010, a session of SBL’s Biblical Greek Language and Linguistics Unit was devoted to the question of deponency. The four presenters (Stanley Porter, Bernard Taylor, Jonathan Pennington, and I) each argued that the category of deponency ought to be abandoned. Notably, Taylor provided a history of the term “deponency,” proving its genesis in the Renaissance period.44 And Pennington revised his views about passive deponents, falling in line with Conrad.
It is rare at SBL to find four presenters who completely agree on a controversial topic, and the session seemed to have historic importance. So uniform were the conclusions, in fact, that two years later an SBL session was dedicated to the middle voice, playfully entitled “We Killed Deponency. Now What?” While not all Greek professors are yet convinced that deponency should be abandoned, there is a consensus among most scholars working in the field that it should.
Since a survey of several different scholarly contributions can become confusing, especially when the developments between them do not proceed in a linear fashion, it is worth drawing out a few threads from the previous discussion.
4.3.1 Terminological Reservations
First, there are those who take issue with the term “deponency,” but do not necessarily offer a comprehensive alternative for the category that the term represents. Moulton, Robertson, and McKay fit here, and while the former two scholars press the category hard, they do not go so far as to abandon it.
4.3.2 Reconstituting the Middle Voice
Second, Miller comes close to displacing the category of deponency simply by allowing the possibility that verbs that are middle in form may be middle in meaning. By reading so-called deponent verbs as genuine middles, she witnesses the category of deponency virtually vanish before her eyes. She does not, however, formulate a reconstruction of the voice system in the wake of deponency’s disappearance but simply demonstrates that a reconstituted appreciation of the middle voice seems to make deponency redundant.
Third, there is the group of scholars who wholly reject the label and category of deponency. Taylor, Conrad, Allan, and Pennington wish to see the term disappear and require the reformulation of our understanding of the voice system to better reflect the history and usage of the language. This latter requirement is noteworthy, since displacing deponency is not quite as simple as removing the concept from our reconstructions of the voice system.
In evaluating these strands, all three have merit, but the final solution can only be found via the third. It is not enough simply to dislike the term “deponency,” nor is it sufficient simply to reconstitute the middle voice, though of course that will be part of the solution. If we are persuaded by these challenges to deponency, we must rethink our entire approach to voice.
In light of these critiques, we should seriously consider abandoning the category of deponency, notwithstanding Ladewig’s arguments in its defense. The questions that remain, however, will be how to assimilate the problems of so-called “mixed deponents” and “passive deponents,” and how to make responsible assertions about voice, given that the matter appears to be more complex than simply recognizing morphology.
It remains an open question as to why some active verbs have middle future forms. Pennington and Egbert Bakker have made similar attempts at a solution, linking the future tense with volitionality and intention, thus explaining why some verbs become middle in their future forms.45 While this seems possible, it is not overly compelling.
This kind of solution makes philological assumptions related to philosophical reflections about futurity. Just because — when we stop to think about it — the future is unknown and our speaking about it must, in the end, be speculative or volitional, this reality does not necessarily correspond with our use of language. The fact is that people often do speak of the future as certain, even if it is not. If I say that I will finish writing this chapter in a few minutes, my use of language does not reflect any uncertainty about this future reality, even though it is possible that an earthquake, a heart attack, or a raging bull might prevent it from coming to pass. Consequently, a solution that relies on the volitional nature of the future tense to explain why some verbs revert to the middle voice in the future is unlikely, since volition is not inherent to the future tense.
A more serious problem is the existence of so-called “passive deponent” verbs, and the solutions offered by Conrad and Pennington both have vulnerabilities. Pennington’s original solution is simplest,46 relabeling “passive deponent” forms as middles that employ passive endings. A problem here is that the suggestion is speculative and therefore may be a little too convenient. Having said as much, Moulton does come to Pennington’s aid when he acknowledges that “the -θην aorist was originally developed, according to Wackernagel’s practically certain conjecture, out of the old aorist middle.”47 However, Pennington’s diachronic solution runs in the opposite direction: while Moulton and Wackernagel have middle forms becoming passives, he sees passives (effectively) becoming middles again. Pennington’s solution may also create another type of deponency, in which the middle form has been laid aside and the passive form has taken its place, thus getting us back on to that merry-go-round.
Conrad argues that “passive” forms are really an alternate set of middle-passive forms, so that both sets of middle-passive forms can express either middle or passive meanings, depending on lexeme and context. His understanding of the Greek voice system is that there is a paradigm for the active voice (A) and two paradigms for the middle-passive voice (MP1 and MP2), as opposed to the standard configuration of paradigms as Active, Middle, Passive. MP1 and MP2 are therefore two paradigms for the same semantic meaning, analogous to first and second aorist forms. This finds some support in Moulton’s observation of the virtual indistinguishability between middle and passive voices, and Wackernagel’s conjecture that passive forms developed out of the middle.48
While this is a bold suggestion that holds some promise, two questions come to mind. First, what do we make of verbs that have middle and passive forms (traditionally understood)? Does not the existence of both forms for the same lexeme suggest a meaningful semantic difference between them?49 Second, is it true that some middle forms are actually passive in meaning? This may be the case, but if it cannot be established, then Conrad’s framework could be threatened.
It is possible that Pennington or Conrad is correct, or that some other approach will solve the problem, but that is unclear at this point. What remains clear is that the phenomenon of so-called “passive deponents” currently presents a challenge to the anti-deponency movement.
A more positive challenge remains in which the relationship between lexeme and voice requires further investigation. As Bakker and Conrad have acknowledged, there is a complex interweaving between lexeme and voice, perhaps parallel to that between lexeme and verbal aspect. If voice is not simply determined through morphological realization but is dependent upon lexeme and context, then there is much more work to be done here.
If deponency is to be removed from the way we understand, describe, and teach the Greek voice system, it will be necessary to begin thinking through the various ways forward. A few preliminary thoughts follow in anticipation of more robust discussions over future years.
4.5.1 Understanding the Middle Voice
Neva Miller has provided a good starting point for improving our capacity with the middle voice. Her work ought to be studied carefully and perhaps even reproduced in a more accessible form, independent of Friberg’s lexicon. Future grammars would do well to incorporate her analysis of the various uses of the middle voice.
For deeper reflection and research, Rutger Allan’s dissertation provides substantial grounds for understanding the middle voice in the absence of deponency. Future dissertations on the topic will necessarily engage his work as the most important treatment of the Greek middle voice we have seen for some time.
4.5.2 Developing Voice – Lexeme Sophistication
The relation between voice and lexeme is not as easily addressed, and I fear this will take a great deal of time and effort. In the meantime, a useful suggestion has been posed by my former student, Ben Hudson. He views voice as a pragmatic outworking of the combination of morphology, lexeme, and context. In this way, voice may be seen in parallel to Aktionsart, which is a pragmatic outworking of the combination of aspect, lexeme, and context.50
Perhaps most significant is the need to develop ways to teach Greek without the category of deponency. Incorporating Allan’s and Miller’s work on the middle voice will be useful here. It will also be necessary to acknowledge that some “passives” will actually be middle in meaning, even if it is not yet possible to explain with certainty why this is so.
It is welcome news that there are now at least two Greek grammars that do not use deponency in their teaching of voice. Porter, Reed, and O’Donnell’s Fundamentals of New Testament Greek teaches the middle voice with only the following brief note on deponency:
Some verbs do not have all three voice forms in every tense, and some older grammars label such verbs deponents. Deponency is understood to mean that the middle voice form (or sometimes the passive voice form) performs the function of the active voice. Some grammarians, however, have questioned the category of deponency, as do we. In our view, every verb expresses the meaning of its voice form, even when other forms — such as the active voice — may not exist. That is, in interpreting the meaning of a verb form, we should try to understand its voice.51
Rodney Decker’s Reading Koine Greek follows suit.52 Both grammars are notably sophisticated in their treatment of advances in Greek linguistics, including voice and the rejection of deponency. They have set a new standard for modern grammars of New Testament Greek.
Allan, Rutger J. “The Middle Voice in Ancient Greek: A Study in Polysemy,” Dissertation presented to the University of Amsterdam, 2002. http://dare.uva.nl/record/108528.
Conrad, Carl. “New Observations on Voice in the Ancient Greek Verb.” Unpublished paper, 2002. See www.artsci.wustl.edu/~cwconrad/docs/NewObsAncGrkVc.pdf.
Miller, Neva F. “A Theory of Deponent Verbs.” Pages 423 – 30, Appendix 2, in Analytical Lexicon of the Greek New Testament. Edited by Barbara Friberg, Timothy Friberg, and Neva F. Miller. Grand Rapids: Baker, 2000.
Pennington, Jonathan T. “Deponency in Koine Greek: The Grammatical Question and the Lexicographical Dilemma.” TJ 24 (2003): 55 – 76.
———. “Setting Aside ‘Deponency’: Rediscovering the Greek Middle Voice in New Testament Studies.” Pages 181 – 203 in The Linguist as Pedagogue: Trends in the Teaching and Linguistic Analysis of the Greek New Testament. Edited by Stanley E. Porter and Matthew Brook O’Donnell. New Testament Monographs 11. Sheffield: Sheffield Phoenix, 2009.
Taylor, Bernard A. “Deponency and Greek Lexicography.” Pages 167 – 76 in Biblical Greek Language and Lexicography: Essays in Honor of Frederick W. Danker. Edited by Bernard A. Taylor et al. Grand Rapids: Eerdmans, 2004.
1. Much of this chapter originates from my conference paper, “Defective Deponency? A Review of the Discussion So Far,” presented at the 2010 SBL conference in Atlanta.
2. Moulton, Grammar, 153.
3. Ibid., 162.
4. Ibid.
5. Robertson, Grammar, 332.
6. Ibid.
7. Ibid., 332 – 33.
8. Ibid., 811 – 12.
9. McKay, A New Syntax, 25.
10. Neva Miller, “A Theory of Deponent Verbs,” Appendix 2, in Analytical Lexicon of the Greek New Testament (ed. Barbara Friberg, Timothy Friberg, and Neva F. Miller; Grand Rapids: Baker, 2000), 423 – 30.
11. Ibid., 426.
12. Ibid.
13. Ibid., 430; “I do not claim that all the dust surrounding the issue of deponency is settled. . . . Much study remains to be done on the matter of deponency.”
14. Ibid., 429.
15. Bernard A. Taylor, “Deponency and Greek Lexicography,” in Biblical Greek Language and Lexicography: Essays in Honor of Frederick W. Danker (ed. Bernard A. Taylor et al.; Grand Rapids: Eerdmans, 2004), 167 – 76.
16. Ibid., 170 – 71.
17. Ibid., 171.
18. Ibid., 173.
19. Ibid.
20. Ibid.
21. Carl Conrad, “New Observations on Voice in the Ancient Greek Verb.” Unpublished paper, 2002. www.artsci.wustl.edu/~cwconrad/docs/NewObsAncGrkVc.pdf, 1 – 2.
22. Ibid., 3.
23. Ibid., 7 – 8.
24. Ibid., 8.
25. Ibid., 3.
26. Ibid., 11.
27. Ibid., 13.
28. Rutger J. Allan, “The Middle Voice in Ancient Greek: A Study in Polysemy,” Dissertation presented to the University of Amsterdam, 2002. Allan’s thesis is available for download at http://dare.uva.nl/record/108528.
29. Ibid., 2, n. 4.
30. Ibid., 3.
31. Ibid., 185 [italics are original].
32. Ibid., 19 [italics are original].
33. Jonathan T. Pennington, “Deponency in Koine Greek: The Grammatical Question and the Lexicographical Dilemma,” TJ 24 (2003): 55 – 76; Jonathan T. Pennington, “Setting Aside ‘Deponency’: Rediscovering the Greek Middle Voice in New Testament Studies,” in The Linguist as Pedagogue: Trends in the Teaching and Linguistic Analysis of the Greek New Testament (ed. Stanley E. Porter and Matthew Brook O’Donnell; New Testament Monographs 11; Sheffield: Sheffield Phoenix, 2009), 181 – 203.
34. Pennington, “Setting Aside ‘Deponency,’ ” 194 – 95.
35. Ibid., 194.
36. Ibid., 195.
37. Ibid.
38. Ibid., 195 – 96.
39. Stratton, L. Ladewig, “Defining Deponency: An Investigation into Greek Deponency of the Middle and Passive Voices in the Koine Period.” Unpublished doctoral dissertation, Dallas Theological Seminary, 2010.
40. Ladewig, “Defining Deponency,” 36, 41.
41. Thracis, Ars Grammatica, 46 – 53. The English translation is produced by Ladewig, 27.
42. Ladewig, “Defining Deponency,” 28 – 30.
43. Ibid., 32 – 33.
44. Taylor’s paper is being published as “Greek Deponency: The Historical Perspective,” in Biblical Greek in Context (ed. T. V. Evans and J. K. Aitken; Leuven: Peeters, 2015).
45. Egbert J. Bakker, “Voice, Aspect and Aktionsart Middle and Passive in Ancient Greek,” in Voice: Form and Function (ed. Barbara Fox and Paul J. Hopper; Amsterdam: John Benjamins, 1994), 29.
46. He later changed his approach to this question; see §4.2.9, above.
47. Moulton, Grammar, 161.
48. See §4.2.1, above, and ibid., 162.
49. This question was posed by my former student Ben Hudson, in his unpublished essay, “A ‘Paradigm’ Shift: Getting Pragmatic about Greek Voice,” Moore Theological College, Sydney, 2010.
50. Ibid.
51. Stanley E. Porter, Jeffrey T. Reed, and Matthew Brook O’Donnell, Fundamentals of New Testament Greek (Grand Rapids: Eerdmans, 2010), 125 [emphases are original].
52. Rodney J. Decker, Reading Koine Greek: An Introduction and Integrated Workbook (Grand Rapids: Baker, 2014), xxvi, 252 – 53.