The Analysis of Verbs
Home Up The Analysis of Nouns and Pronouns The Analysis of Verbs The Analysis of Adverbs The Analysis of Adjectives The Analysis of Determiners The Analysis of Prepositions The Analysis of Conjunctions The Analysis of Particles

 

The Analysis of Verbs

Verb tags usually consist of eight symbols. Due to the deletion of final hyphens, tags for regular infinitives have four symbols; those for articular infinitives, five.

Mood 5.1

Tense(-Aspect) 5.2

Voice: Deponency 5.3

Case, Gender, Person, and Number in Verbs 5.4

Transliterated Verbs 5.5

Periphrastic Constructions 5.6

Complex Verb Tags 5.7

5.1 Mood

The first division among verbs is that of mood (mode). Since the first-level analysis is according to form rather than function, the moods as well as all other verbal distinctions are determined by form apart from context. If a given form permits more than one analysis, then the proper analysis is determined from the context. An analysis will not be in contradiction to the context.

5.1.1 Subjunctives

Subjunctive verbs preceded by m often function as the aorist imperative of prohibition. They are tagged as in this example: … m (ab) fobjqte (vsao--2p^vmao--2p) mjdš (cc) taracqte (vsap--2p^vmap--2p) (1 Peter 3.14). As noted earlier, ambiguous cases that may be read as either "subjunctive" or "subjunctive used as an imperative" are given only the latter analysis. Indirect commands following na (or a conjunction acting similarly) are left as simple subjunctives. No indication of the imperatival force of indirect commands is given. Hortatory subjunctives are not differentiated from other first-person plural subjunctives.

5.1.2 Infinitives

Simple infinitives are analyzed as vn followed by tense and voice symbols; for example, poisai (vnaa). Articular infinitives have an additional symbol to show case, as does poisai in this phrase: ev (pa) t (dans) poisai (vnaaa) (Hebrews 13.21). It seemed less complicated to indicate the articular infinitive by giving the infinitive analysis a case symbol than to indicate the construction on the tag for the preceding article, already marked for case. This is advantageous because, when two or three infinitives follow a single article in this construction, every infinitive is marked. (Note that this convention is unlike that for the articular participle, in which the construction is noted on the tag for the article; see 8.3 below for reasons.)

Articular infinitives, appearing as they do in construction only with neuter singular articles, must themselves be neuter and singular. Because gender and number are predictable they are not included in the infinitive tag. All cases except vocative are included in this construction. In at least one instance (Luke 17.1) a genitive article determines the case of the following infinitive to be genitive even though the construction is used where a nominative case would be expected grammatically.

We chose to analyze each occurrence of the articular infinitive for two reasons. First, the construction is not always obvious because the article and infinitive are often separated by intervening material. Second, we wanted articular infinitives to be grouped separately in the concordance volumes.

Infinitives, whether articular or not, figure in grammatical constructions. The most frequent has the infinitive serving as the object (complement) of a finite verb or even of another infinitive. Clear examples of both occur in Luke 5.34: M (qt) dnasqe (vipn--2p) … poisai (vnaa) njstesai (vnaa). Infinitives also serve as subject complements of other verbs. The impersonal verbs de and žxestin usually have infinitive clauses as their subjects: "To do such and such is necessary," "To do this or that is lawful." (This is best translated into English as: "It is necessary to do such and such," "It is lawful to do this or that.")

In Greek de is sometimes tied to a preceding clause by way of a relative clause headed by . This relative pronoun is not nominative and the subject of de, but is the accusative subject (as in, e.g. Revelation 4.1) or object (as in Luke 12.12) of the accompanying infinitive. Then the whole infinitive clause is the subject of de. In Acts 3.21 the relative pronoun is unambiguously accusative and thus not to be mistaken as the subject of de. In cases where the infinitive is present in the semantic structure but lacking in the surface grammatical structure, we analyze the former subject or object of the infinitive as the subject of the impersonal verb. For example, p€nta (ap-nn-p) moi (npd-1s) žxestin (vipa--3s) (1 Corinthians 6.12). The semantic structure is "For me to do all things is lawful," with the infinitive subject complement of de intact. At the surface level, however, it is optionally missing. In its absence p€nta becomes the surface subject and is appropriately given the nominative case tag. One further example awaits discussion by way of its working analysis: (apran-p^apdan-p&apran-p) m (ab) de (vipa--3s) (Titus 1.11). Though the infinitive is missing, we have still analyzed the relative pronoun as an accusative object due to the presence of m. (See 7.6.2 below for details on the working analysis of ; the simplified relative tag is -apran-p.)

M and an infinitive can sometimes be taken as a prohibition, standing alone as a stylistic alternate to the morphological imperative. Neither this nor any infinitive following as the object complement to a verb of commanding, whether its function is simple or derived, is analyzed here as an imperative.

5.1.3 Participles

Participles receive a straightforward analysis. We have added a 1 or 2 to the otherwise irrelevant person place in participle tags to show first- or second-person linkage, respectively. Our clue for this semantic information is either the personal ending on a finite verb or the person of a pronoun. For example, meqa (viim--1p+) dedoulwm™noi (+vprpnm1p) (Galatians 4.3). The participle tag includes a 1 for first person on the basis of its (periphrastic) link to the first-person finite verb. Another example comes from Mark 13.36: m lqn xafnjv er m‚v (npa-2p) kaqedontav (vppaam2p). The participle tag contains a 2 for second person because of its semantic tie-in with m‚v. When a hyphen appears in the person position of participle tags, it indicates what might, except for visual crowding, have been indicated by 3.

Our analysis of participles includes all those that have not been frozen as nouns. Among those analyzed by BAGD and us as nouns are ˆrcwn and okoum™nj. But participles themselves, even without articles, do function as nouns. Since these represent such a continuum from those that clearly act in particular contexts as nouns to those that may also have some verbal interpretation attendant to the governing verb, we have left all such participles whatever their function, as simply participles. Peinntav and ploutontav in Luke 1.53 are examples of participles that function as nouns. Articular participles are discussed in 8.3 below.

A special class of participles has been designated by second-position r rather than p. These appear in conjunction with imperatives and themselves have an imperatival sense. Not every adjacent imperative activates this imperatival sense. Sometimes, as Matthew 6.17 shows, the relationship between the imperative and the adjacent participle is that of contingency: "When you fast, anoint your head…." On the other hand, the participle is sometimes imperatival in concert with a morphological imperative (which usually follows the participle). Matthew 10.14 illustrates this case. Anticipating some inhospitable receptions for his disciples, whom he is about to send, Jesus does not say, "When you leave a house or town that has rejected you, however long after the inhabitants have become hostile, shake the dust off your feet as a sign against them…." Instead he seems to say, "Leave that house or town and shake…." In view of this, we tag the participle xercmenoi as vrpnnm2p (e.g. Luke 9.5). An r participle should be read as containing a potential choice: some instances may be interpreted either imperativally or otherwise, and the reader may opt for the latter.

The imperatival participles bear certain relations to the main imperative verb, of which we shall list several. A very common interpretation of an imperatival participle is commanded means. In Acts 22.10 ‡nastv (vraanm2s) is the means to obey the finite command poreou (vmpn--2s). First one gets up off the ground and then he goes. Commanded attitudes are frequent, especially in the letters. Colossians 3.17 has ecqristountev (vrpanm2p) as the attitude that should accompany the implied doing of all things. The imperatival participles in Romans 12.9-13 are the commanded specifics of the lead command or statement that love must be sincere. And as the initial example from Matthew 10.14 shows, there may be only a coordinate command, for it is possible to shake dust and not leave. As expected, these imperatival participles are in the nominative case. In 2 Timothy 2.15, however, we see an instance of an oblique case having this imperatival sense. There rqotomonta (vrpaam2s) has taken on the case of the reflexive pronoun seautn.

Observe that the examples given are all second-person imperatives and thus take a 2 in the participle tag to show the second-person link between the two verbs. 1 Corinthians 16.2 illustrates a third-person imperative with the expected third-person (-) imperatival participle.

5.2 Tense(-aspect)

In the indicative mood six tenses occur: present, imperfect, future, aorist, perfect, and pluperfect. The time element pertains only to the indicative mood. In the other moods, p represents durative or continuous action, whereas a stands for punctiliar action. These represent aspect. Thus at 2 Thessalonians 3.8, for example, rgazmenoi (vppnnm1p), the p ("present") in the third slot represents continuous action in the past. Future perfects appear only in periphrastic constructions, as in Matthew 16.19: žstai (vifd--3s+) dedem™non (+vprpnn-s). We have analyzed tense on the basis of form, not meaning; thus oda is perfect rather than present.

The future, like the subjunctive, is frequently used as imperative. This is limited to second- and third-person forms of the future and thus corresponds with the imperative forms. While the subjunctive used as imperative shows a correspondence between tenses, the future indicative used as imperative does not. So for every future used imperativally, we had to determine the tense of the imperative function. We did this item by item, deciding in each case the aspectual sense (punctiliar action, durative action, etc.) of the command. For example, o (qn) moiceseiv (vifa--2s^vmaa--2s) (Matthew 5.27) has the aspect associated with aorist tense, while ‡gapseiv (vifa--2s^vmpa--2s) tn pljson (Matthew 5.43) has the aspect associated with present tense.

Several short comments remain. Tense for periphrastics is assigned separately to each half of the construction, leaving the reader to determine for himself the tense of the whole. Tense is the parameter most affected by changes in accent (as opposed to the form itself), which requires the use of the exclamation mark symbol; for example, Luke 19.22: krin (vifa--1s!vipa--1s). In the few cases where alternate tenses possess identical form and accentuation and where we have been unable to determine the correct tense from the context, we have used a slash (/) and left the choice to others. In eighteen instances of žfj, for example, we have tagged the word as viaa--3s/viia-3s. (In the other twenty-five cases we were able to determine a unique analysis—either imperfect or aorist—from discourse signals.) In John 8.44 and Revelation 12.4 the choice presented in our analysis is not merely between tenses, but between tenses of different verbs, stkw and stjmi. Our analysis agrees with BAGD in giving a choice between perfect and imperfect tenses in John 8.44 despite the textual variation in the breathing mark.

5.3 Voice: Deponency

The matter of voice has received substantial attention in our analysis largely due to the problem of deponency. The three-way voice distinction itself is no problem; where middle and passive voices coincide in form in some tenses, considerations of meaning are usually sufficient to permit a choice between middle and passive. Deponency itself is the challenge. It is easy enough to say that deponency occurs when a middle or passive form of a verb takes on an active meaning, whether in all tenses, several tenses, or just one tense. It is more difficult to decide if deponency arises to fill the place of a missing active form with active meaning, or if verbs can have deponent forms (whether middle or passive) alongside active forms. Using our symbols (a = active, m = middle, p = passive, d = middle-form deponent, o = passive-form deponent), we can state the issue with more precision. Which of the following situations may represent deponency for a given verb: (1) a, m, p; (2) a, d, p; (3) a, m, o; (4) -, d, p; (5) -, d, o? The first is clearly not deponent, being the ideal, full-blown transitive verb. Some would answer, only 4 and 5; others, 2-5 and perhaps other situations as well. Before giving our answer, we will first briefly discuss the passive voice.

5.3.1 Passives as Intransitivizers

Passive voice is a grammatical construction that enables the speaker or writer to focus or topicalize the object of a transitive construction. If developing a discourse about the Book of Acts, in which the book is the topic of discussion, we are more likely to say (1) Acts was written by Luke or (2) It was written by Luke. In a discourse about the author, we would probably say instead (3) Luke wrote Acts. This is true of both English and Greek. But language, tool for communication that it is, is not bound to grammatical purity. Languages in general change the function or meaning of grammatical constructions to suit communication goals. A language may add meanings to grammatical constructions to suit its needs. In particular, the passive-voice verb in Koine Greek has more than one meaning or function: it may serve, as in English, to topicalize an object for purposes of discourse, but it may also function to "intransitivize" a transitive verb. Said another way (which may not be exactly equivalent), it may focus on the effect or result of an action while its active counterpart focuses on the causing of that action.

For example, gerw is an active, transitive verb. The aorist active is used of Jesus’ disciples rousing him from sleep (Matthew 8.25) and of Jesus lifting to his feet a boy whom he has just healed (Mark 9.27). All of these instances show the causing of an action. Let us now look at instances of gerw that are aorist passive. In Matthew 9.19 there is a construction that recurs elsewhere often: "Getting up or rising, Jesus followed Jairus." The emphasis is on the effect or result of an action; it is intransitive. How this passive meaning of the active may have developed can be shown by contriving the agent that raised Jesus: "Having been raised to his feet by the action of his leg muscles, Jesus followed…." The focus, however, is intransitive: "Jesus rose." (The passive of gerw can at least ambiguously mean "be raised by someone." John 2.22, for instance, can be understood as "when Jesus rose from the dead" if the focus is on the intransitive result, or as "when Jesus was raised from the dead" if the focus is on the transitive action of causing Jesus to transfer from being dead to being alive.)

This digression has shown that active meanings ("rise") of nonactive forms can coexist with active meanings ("raise") of active forms of the same verb. This lays the groundwork for our claim that such conditions do not constitute a middle or passive deponent of such verbs. Our analysis, then, excludes from the category of deponent verbs many forms frequently called deponent by others. But we believe that the definition of deponency that follows, results in a better and more consistent treatment of this controversial phenomenon: a verb (or tense of a verb) is deponent only if it lacks an active counterpart. Before elaborating our application of this definition, we will list and explain the voice symbols.

5.3.2 The Voice Symbols

The first four of the voice symbols are a for active, m for middle, p for passive, and e for either middle or passive. (See the chart following the introduction for mnemonic help.) A verb is marked a only if it is active in form. Several verbs that, semantically, are stative rather than active are thus marked active: for example, em and active forms of gnomai such as the perfect, g™gona. To be marked m a verb must have a corresponding active counterpart, be middle in form, and not be passive in meaning. Verbs marked p must have a corresponding active counterpart, be passive in form, and not be middle in meaning. Verbs tagged e are those whose form can be either middle or passive (in the present, imperfect, perfect, and pluperfect tenses only), which have an active counterpart, and whose meaning, in context, does not allow a clear-cut choice between the two.

The primary considerations for these symbols, then, are a verb’s form rather than its meanings and, for m, p, and e, the existence of an active counterpart. The requirements that a middle not be passive in meaning and that a passive not be middle in meaning, mean that for ambiguous forms (i.e. other than future and aorist tenses), lexical and contextual meanings have been consulted. One must remember that, for cases like gerw (see 5.3.1 above), not all passive forms carry strictly passive meanings. In the overwhelming majority of cases, forms that are ambiguously middle or passive are clearly one or the other in context. Only about thirty times did we have to use the symbol e.

The other three voice symbols are d for middle deponent, o for passive deponent, and n for either middle or passive deponent. A verb is marked d only if it has no active counterpart and is unambiguously middle in form (that is, in future or aorist tenses). To be marked o a verb must have no active counterpart and be unambiguously passive in form (that is, future or aorist). A verb is tagged n if it has no active counterpart and is ambiguously middle and passive in form (that is, present, imperfect, perfect, or pluperfect).

A verb as a whole is frequently designated in the literature a middle deponent verb if its aorist form is middle and a passive deponent if its aorist form is passive. Thus punq€nomai is called a middle deponent because its aorist is middle in form: puqmjn. And dnamai is called a passive deponent because its aorist is passive in form: dunqjn. Occasionally a verb is called a middle and passive deponent because in the aorist it has both middle and passive forms (and the aorist passive form is not a true passive). One example is gnomai, which has both an aorist middle (genmjn) and an aorist passive (genqjn). We have analyzed each individual verb according to its form. We have not followed the traditional practice of describing a verb as a whole as a middle deponent, passive deponent or middle and passive deponent, based on the form of the aorist or future.

Let us illustrate the difference between calling a verb as a whole a certain kind of deponent and calling a particular form of that verb a deponent. Dnamai, usually or traditionally called a passive deponent, has one form that is not passive in form but middle: dunsontai (future tense). Since it has no active counterpart, it is analyzed as d. Ercomai has been called a middle deponent. It has been so labeled not on the basis of an aorist middle form (for the aorist is active), but presumably on the basis of the future form, lesomai. In present and imperfect forms, however, we analyze this verb as n (middle or passive deponent).

Whereas there is a certain correspondence between m and d, p and o, and e and n, it is not complete. The differences between the first and second parts of the three pairs are greater than merely that the first is nondeponent and the second deponent. With the first set, m, p, and e, one does refer to meaning in deciding among ambiguous forms; with the second set, one does not. Though e occurs in the New Testament only a few times, n occurs more than 1,600 times. The former symbol (e) says, "We cannot be certain, even after consulting the context, whether to call this word m or p as to meaning;" the latter (n), "The words so marked are ambiguously middle or passive in form." Why refer to meaning in the first case and not in the second? There is usually a systematic difference between middle and passive forms when there is an active counterpart to consult. When with deponent forms there is no active counterpart, the deponent forms themselves often seem active in meaning. In the case of an ambiguous deponent form, one can do nothing but label it n. The verb dnamai illustrates the pitfalls of trying to decide what the "whole verb" might be.

5.3.3 The Rules for Judging Deponency

Certain rules for determining deponency have emerged in the course of this analysis. These ten rules, with commentary, follow:

Rule 1. If any active form of a verb is found in first-century Greek, or if it can be inferred for it (because it is found in both earlier Greek and later Koine), then any middle or passive present, imperfect, perfect, or pluperfect forms of that verb are middle or passive, not deponent.

By way of explanation for this rule, we must first explain why some of our rules are formulated in terms of "first-century Greek." Diagram 1 shows us the alternatives. Because it is well established that language changes, we should not allow classical usage, four hundred or more years removed from the New Testament, to determine whether a verb is deponent. It is possible that during the intervening years an active dropped out of use and thus established deponency for a given verb (or tense of that verb). Or a deponent verb may have developed active counterparts and ceased to be deponent. For the same reasons we should not rest our judgments concerning deponency on Christian-influenced Byzantine Greek. But neither should we say that a verb with no active counterpart in the Greek New Testament must be a deponent. The Greek of the New Testament was the Greek of the New Testament world. Just as the papyri have thrown new light on New Testament vocabulary, so can they aid greatly in the matter of determining deponency. Rhetorical choices laid aside, we have settled for the Greek contemporaneous with the New Testament, roughly that of the first century of the Christian era.

Diagram 1

Usage in the

classical era

Usage contemporaneous to the New

Testament (i.e. in about the first century)

Usage in the

New Testament alone

Lexicons cited earlier have proved invaluable in tracking down this contemporaneous usage. The lexicon of Liddell, Scott, and Jones, while supposedly giving lemmas on the basis of classical or even Homeric Greek alone, has been an excellent resource. BAGD, in our opinion the finest lexicon available for New Testament Greek, has one disturbing shortcoming: It does not explain in its introduction the criteria employed for selecting lemmas (i.e. the citation form of words). Do they date from the classical period, the Septuagint era, or that of the New Testament and early church? Using BAGD, we have examined every active lemma in the light of contemporaneous usage. We have similarly tested every nonactive BAGD lemma that contrasts with a corresponding active lemma in Liddell, Scott, and Jones. The results of those searches furnish the basis for our deponency judgments.

Rule 1 states that any tense of an active counterpart serves to establish the nondeponency of just those tenses in which middle and passive coincide with respect to form. An aorist active serves to establish the nondeponency of a middle or passive present, for example, but a present active does nothing to establish nondeponency for an aorist middle.

Rule 2. If an active form exists in either the future or the aorist tense, active forms are assumed to exist for all other tenses.

Deponency of one or more tenses, but not every tense, is semideponency or partial deponency. Deponency of the future and aorist tenses is, then, semideponency. (There are a few exceptions, usually involving a change of root; for example, žrcomai, lesomai, lqon.) Rule 2, therefore, states that an active form in either the future or aorist tense (the domain of semideponency), assures active forms in every tense and hence rules out any deponency, full or partial.

Rule 3. If any active future form of a verb is found in first-century Greek, or if it can be inferred for it, then any middle or passive future forms of that verb are middle or passive, not deponent.

Rule 4. If any active aorist form of a verb is found in first-century Greek, or if it can be inferred for it, then any middle or passive aorist forms of that verb are middle or passives, not deponent.

Rule 5. If the future passive of a verb is known to be either deponent or nondeponent, then the aorist passive of that verb is the same.

Rule 6. If the aorist passive of a verb is known to be either deponent or nondeponent, then the future passive of that verb is the same.

Rule 7. If a simple verb is deponent or semideponent, then its compounds are also deponent or at least semideponent in the same tenses.

This last rule says, for instance, that since gnomai is deponent, paragnomai will also be deponent. The converse of this rule does not hold. That is, although paragnomai is deponent, it does not necessarily follow that gnomai is deponent, though in fact it is. HEpilamb€nomai is deponent; but lamb€nomai is middle or passive, depending on the context, for there is an active counterpart, lamb€nw.

Rule 8. If a compound verb is nondeponent in all or some tenses, then its simple equivalent is also nondeponent in at least the same tenses.

This rule states, for example, that since ‡nair™w is nondeponent in all its tenses, then so is ar™w. Again, the converse of this rule fails to hold. That is, though sp€w is nondeponent, it does not necessarily follow that perisp€w will be, though in fact it is. HEpisp€omai, on the other hand, is deponent. Rules 7 and 8 are compared in diagram 2. An arrow indicates an "implied" relationship in the direction it points. A slash through an arrow indicates a denial of the relationship. The diagram shows the four possible implications.

Diagram 2

 

simple verb

 

deponent

nondeponent

deponent

 

nondeponent

 

compound verb

 

Rule 9. If a verb is deponent or semideponent, and if there are unambiguously passive forms but no unambiguously middle, then all forms are passive deponent. If, however, at least one ambiguous form (middle or passive) or one middle form occurs with a direct object, and if all passive forms lack direct objects, then the ambiguous or middle form(s) with direct object(s) is middle deponent and the passive forms are passives of the middle deponent; any other ambiguous forms must be judged individually.

The first sentence of this rule is not one of our rules for determining deponency but is used by some to determine the label for a verb as a whole (middle deponent or passive deponent). We note it here because of the exception to it contained in the second sentence. We analyze individual verbal deponents by their form: d if unambiguously middle, o if unambiguously passive, n if ambiguous. In a number of cases a passive form of a deponent verb is a true passive. It is marked p (an example of the fourth of five instances listed in 5.3 above). By "true passive" we mean that construction in which the object of an active verb becomes the subject of the passive verb and the subject of the active verb, if retained at all, becomes the agent (expressed in a by phrase in English, usually a p phrase in Greek). "Acts was written by Luke," was our illustration in 5.3.1 above. It seems quite appropriate that if a middle or passive deponent has an active meaning, then that deponent, if transitive, can be passivized.

Rule 10. Except with a few individual verbs, a p agent phrase implies that a passive form is nondeponent. One exception is gnomai which, though deponent, can take a p agent phrase.

A rule that some scholars consider important in determining deponency is this: If a verb has both active forms and middle and/or passive forms, and if the semantic meaning of the former forms is radically different from that of at least some of the latter, then the latter are deponent. For us to accept this rule would mean that we would introduce a number of homonyms. Fanw would mean "shine" and fanomai "appear." We do not deny that homonymy is a common linguistic phenomenon, but we thought it better to allow the reader to determine when meanings are radically different. Two meanings that seem radically different to us may not have seemed so to a first century Greek-speaking person, who, after all, perceived the world quite differently. A Greek speaker may have agreed that fanomai meaning "appear" and fanomai meaning "be shined on" are homonyms, but he may instead have argued that the verb is unified, that something "appears" when it "is shined on" by something. When BAGD identifies a homonym by giving two or more separate entries (e.g. sneimi and sneimi), we accept that judgment. When it gives a single lemma and includes in the definition the different senses (e.g. fanw, ‡n€gw, krem€nnumi), we treat the meaning of passive forms as p and not a separate meaning o.

The application of these rules was rather straightforward. In a few cases there was too little evidence by which to decide. In those few, if the BAGD lemma was active, we called nonactive forms m, p, or e, as relevant; if the lemma was nonactive, then d, o, or n. In a few cases we concluded that some supposedly contemporary evidence was in fact Atticistic: these few we discounted in deciding deponency.

5.3.4 A Categorization of Verbs

List 1 at the end of this appendix contains five sections. The first consists of those verbs in the Greek New Testament only the future of which is (middle) deponent. In the case of a verb such as ‡kow or z€w whose future middle varies with a future active, the middle forms are analyzed as m.

The second section of this list consists of verbs that, though they have active lemmas in BAGD, are, according to our analysis, truly deponent in first-century times. We give them here with nonactive lemmas. When the letter p follows a lemma on this list, it means that some forms of this verb occur as true passives. Where these would normally be marked n or o in our analysis, they have been marked p instead.

Section 3 consists of verbs that have active lemmas in BAGD but that are semideponents. These are all future/aorist semideponents and therefore have an active lemma. Again p means that a passive form may act as a true passive of the deponent.

The next section lists verbs for which BAGD gives nonactive lemmas but for which we find evidence of active forms contemporaneous with the New Testament. Thus we cite the verbs with active lemmas.

The more than two hundred remaining verbs cited in BAGD with nonactive lemmas we have accepted as deponents. Seventeen of these we have found to have some instances of true passives (p), and these comprise the final section of the list. All are middle deponents.

5.4 Case, Gender, Person, and Number in Verbs

Only participles and articular infinitives exhibit case. Both case and gender positions are empty (-) with finite verbs and nonarticular infinitives. With finite verbs person is indicated by 1, 2, and 3; with participles (the person of which is supplied from context) by 1, 2, and -. A vocative participle is redundantly marked 2.

5.5 Transliterated Verbs

Verbs that are transliterated have been analyzed on the basis of their translation equivalent. Effaqa is tagged vmap--2s, based on its translation dianocqjti (Mark 7.34). Qa (1 Corinthians 16.22) is analyzed as vmaa--2s.

5.6 Periphrastic Constructions

Periphrastic constructions (identified by a plus sign in the direction of the other member of the pair, v+ +v) have a base verb whose only purpose is to give grammatical information; it has no semantic content. In our analysis there are two kinds of periphrastic constructions. The first is an empty verb and a participle. The common empty verb is em, though in several instances in Luke’s writings p€rcw performs this function (Acts 8.16; 19.36); prop€rcw (Luke 23.12 and Acts 8.9) seems to bear the semantic component of "previously" and thus isn’t thought periphrastic. We examined possible instances of žrcomai and gnomai as the empty verb but found in each case that the potential base added some semantic content. The second kind of periphrastic construction is m™llw and an infinitive, although this construction indicates some sense of futurity (…was/is going to…). In both kinds of constructions the finite base may be either before or after the related participle or infinitive. Periphrastics range from moderately to highly certain. All constructions analyzed here as periphrastics may be read as having an implied choice. In one case, John 1.9, the choice is spelled out: the participle is either nonperiphrastic (in which case it is accusative and masculine) or periphrastic (nominative and neuter).

5.7 Complex Verb Tags

A few verbs require complex tags, some of which have already been noted. Having discussed voice, we may note that in cases of a future deponent used as an imperative, not only must the tense/aspect of the imperative be determined, but also the voice of the derived imperative. Estai is tagged vifd--3s. But when it is used as an imperative, the voice is active, for there is no deponency in present-tense em reflexes. Thus the tag reads vifd--3s^vmpa--3s (e.g. Matthew 20.26). Fobjwte (Luke 12.4) is analyzed as vsao--2p^vmao--2p, with deponency indicated in both tags because the verb is consistently deponent.

With a number of instances of carw (e.g. Acts 15.23) and one of žrrwsqe (Acts 15.29) we have added to the tags a functional ^qs on the grounds that the verb is used as a formula of greeting or of taking leave.

In 1 Corinthians 16.6 there is an instance of rare accusative absolute (tucn). It seems to function adverbially, but it is not given a functional analysis any more than is a reduced genitive absolute.

In a couple of places the imperatival force of ˆge seems diluted and so the verb is tagged vmpa--2s^qs (e.g. James 4.13). With de the analysis is either vmaa--2s (when the lemma is edon) or qs (when the lemma is de), whichever is appropriate. The difference between the analyses of ˆge and de lies in the former’s being exceptional and the latter’s being regular. Further, de pairs with do, which is entirely qs.

Finally, both dero and dete are tagged ab^vm in all but one instance (dero in Romans 1.13—ab). Had the verbal function been exceptionless, we would have tagged them all as simply verbs. Desiring to relate the lone nonverbal instance to the regular usage, we chose ab.