Adverbial Clauses: Boyer

                                                       Grace Theological Journal 11.1 (1991) 71-96.

[Copyright © 1991 Grace Theological Seminary; cited with permission;

digitally prepared for use at Gordon College]




                        ADVERBIAL CLAUSES:

                        STATISTICAL STUDIES


                                                    JAMES L. BOYER


            This third article in a series of studies dealing with subordinate

clauses in the Greek NT will be concerned with the adverbial clauses.

The over-all classification is functional, based on the kinds of adverbial

modification made by the clauses. Only in the case of the conditional

clauses is it necessary to carry the classification further. Attention will

be given to the conjunctions or conjunctive relative phrases used to

introduce the clauses, to the moods used, and to the clause order. A

special feature of this series of studies is the attempt to give statistical

information at every level, so that the student may begin to appreciate

the relative magnitude of each structure.

                                                            *     *     *


JUST as adjectives modify nouns so adverbs modify verbs, limiting

and defining the circumstances under which the action of the verb

is to be understood. As adjectives answer the questions "who?" "what?"

"what kind?" so adverbs answer such questions as "when?" "where?"

"why?" "how?" "under what circumstances?" They may be single words

(as nu?n), or phrases (as dia< tou?to), or full clauses. The clauses are the

subject of our present study.

            They will be taken primarily in the order of frequency of occur-

rence in the NT, except that in a couple of instances similarity or

relationship between classes will bring two together out of the numeri-

cal order.


                                    CAUSAL CLAUSES


            As the name adequately indicates, causal clauses modify the main

verb of a sentence by stating the cause or reason for that main asser-

tion. Their meaning is reflected in the way they are translated into


72                    GRACE THEOLOGICAL JOURNAL



English. Using the NASB1 as point of comparison these clauses are

introduced by "for" (473 times), "because" (224), "since" (26) and a

variety of at least 16 other ways, each occurring less than six times.

            No attempt is made in this study to refine the classification further,

no sub-classification will be attempted.

            One problem of identification needs to be considered; the distinc-

tion between coordinate (main clause) and subordinate clauses. For

example, it is not always easy, or even possible, to decide whether ga<r

or even o!ti is introducing a subordinate or a main clause. Actually

GRAMCORD has listed 800 occurrences of ga<r as introducing main

clauses (CX)2 and only 241 with subordinate clauses (SC). The reverse

is the case with o!ti, 1291 are connected with subordinate clauses

(SN,SC,SR) and only 10 with main clauses (CG,CX).

            Ordinarily one would expect that a causal clause at the beginning

of a sentence would be either (1) subordinate to a main clause which

comes later, or (2) the explanation of something that is present in the

preceding context or to the mind. Unfortunately it cannot always be

known where a sentence begins. The lack of punctuation in the original

manuscripts and the tendency to hook long sentences together with

many subordinate clauses, complicates the problem, particularly in the

light of our precisely opposite modern preference.

            In a few instances in this study such ambiguous identifications are

called to attention, but usually a choice is made and that is followed.



Conjunctions Used

            These may best be shown in table form.

Causal Conjunctions,      NT       Mood                Before or After Main Verb

Conjunctive Phrases   Count      Used                > before            < after              ?

            o!ti                   439       Ind.                  16                     423

            dio<ti                21         Ind.                                            21

                 kaqo<ti        4          Ind.                                              4

            ga<r                 243       Ind.                  1                      241                   1

            e]pei<                  26         Ind.                  2                        24

                 e]peidh<          9        Ind.                  4                          5

            e]peidh<per          1        Ind.                  1


            1 Unless otherwise indicated all formal translations of the Bible text will be given

from the NASB version.

            2 These letters in parenthesis are coded tags used by GRAMCORD to identify the

various functional classifications of conjunctions. The first letter in the code indicates

whether the clause is coordinate (C) or subordinate (S). The second letter designates the

function: CG for interroGative, CX for eXplanatory, SC for Causal, SN for Nominal,

SR for Result. Others will be identified as they occur.


                        BOYER: ADVERBIAL CLAUSES              73


Causal Conjunctions,      NT       Mood                Before or After Main Verb

Conjunctive Phrases   Count      Used                > before            < after              ?

di ] h!n [ai]ti<an] 7         Ind.                                          7

a]nq ] w$n            5         Ind.                                          5

e]f ] &$               2         Ind.                                          2

ou$ ei!neken        1         Ind.                                          1

ou$ xa<rin          l          Ind.                                          1


!Oti and ga<r account for 93% of all the subordinate causal clauses.

e]pei< and its compounds are comparatively rare. I have already called

attention to the relative phrases which by antecedent or by context

become in effect causal conjunctions.3



In every instance the mood of the verbs within the causal clause is

indicative. This is to be expected, since causes and explanations are

characteristically simple statements.


Clause Order

The causal clause follows the main verb in 97% of the instances.

Even the e]pei< group, which show more tendency to precede the main

clause, are still 74% following. Again, it is more logical that causes and

explanations should follow that which is being explained.

Ga<r, here as elsewhere, is post-positive; it never stands as the first

word in its clause. Usually it is second or third in sequence, in three

instances4 it stands as the fourth word in its clause.


Other Causal Constructions


Beside these conjunctive and relative clauses there are other ways

of expressing what amounts to a causal clause in the Greek NT.


Adverbial Participles

The anarthrous participle very frequently functions as an adverb

in the sentence. While it may not technically be called a clause (there is

no finite verb in the construction) yet it clearly functions as one; in

most instances the best way to translate it is by an English clause. Of

these adverbial or circumstantial participles, 303 are causal in sense,

including 35 genitive absolutes.5


3 See my article, "Relative Clauses in the Greek New Testament: A Statistical

Study," GTJ 9 (1988) 233-56.

4 Luke 6:23, 26, 2 Cor 1:19.

5 See my article, "The Classification of Participles: A Statistical Study," GTJ 5

(1984) 163-79. At the time that article was prepared I did not have the computer facilities

now available for tabulating and collecting information, so the identification of the

adverbial functions expressed by the participles was not included. Later this inadequacy

74                    GRACE THEOLOGICAL JOURNAL

Articular Infinitives with the Preposition Dia<

The articular infinitive after prepositions, while no finite verb is

involved, is so completely clausal in character that it is impossible to

translate into English without converting it into a full clause. Those

which as causal in sense are dia< with the neuter accusative article and

an infinitive (32 examples); also, e!neken tou? and e]k tou? + an infinitive

(one each).6



An extensive and detailed consideration of the conditional sen-

tences has previously been published by this writer7 so this section will

be primarily a summary and collection of statistics. For a fuller dis-

cussion and support for some statements made here the reader is

referred to these article's.

The conditional sentences proper are composed of four classes.


First Class Conditions

Significance, Meaning

Its meaning is very simple: "If this. . . then that . . ." It indicates

nothing as to the actual situation, whether the condition is true or

false; in fact it is frequently used for both sides of a true / false

condition. Its use of the indicative does not in any way indicate that the

protasis is true, or even that it is "assumed for the sake of argument."

Sometimes it may be true that the English word since is a possible

translation, but it is never a "proper" translation. Since carries an

implication that the condition is true; the Greek first class condition

does not. If used to translate a statement which is actually true then the

translation would not be "wrong" or "untrue," but it would not be a

correct translation in that it would be saying something more than the

Greek says.


was met by my Supplementary Manual of Information: Participles. This is now available

by inter-library loan from the Morgan Library, Grace Theological Seminary, 200

Seminary Dr., Winona Lake, IN 46590.

6 See my article, "The Classification of Infinitives: A Statistical Study," GTJ 6 (1985)

29-48. Complete listings are available in the Supplemental Manual on Infinitives (see

previous footnote).

7 There are four articles in the series: James L. Boyer, "First Class Conditions: What

Do They Mean?" GTJ 2 (1981) 74-114; "Second Class Conditions in NT Greek," GTJ 3

(1982) 81-88; "Third (and Fourth) Class Conditions," GTJ 3 (1982) 163-75; and "Other

Conditional Elements in NT Greek," GTJ 4(1983) 173-88. No supplemental manuals are

available for these studies.


BOYER: ADVERBIAL CLAUSES                          75




First Class conditions use ei] with an indicative verb in the protasis;

the apodosis may be of any type.



*** Protasis                              Apodosis ***

Conjunctions Count Mood          Order               Sentence Type

ei]         302  ALL                                  S          --Statement                   138

M         --Command                     76

ei@ ge        5       I           Prot.                 RQ       --Rhetorical quest.           52

N         before               MR      --Request                        11

ei] mh<                D         Apod.               PR       --Promise                        11

I                                   AS       --In subordinate cl.            5

ei] mh<ti    1       C          >267                 O         --Oath                              5

A                                 P          --Potential                         5

ei@ pwj     1      T                                  TH       --Threat                            4

I           Prot.                 X          --Exclamation                    3

ei@per        6      V         after                 CH       --Challenge                       2

E          Apod.               HS       --Hortatory subj.                2

ei@te           1     S                                  Q         --Question                         2

< 43                  RC       --Rel. Clause equiv.           2

MN      --Emphatic negative           1

e]a<n          2                                          MP       --Prohibition                      1

( )         --(No apodosis)                 1



It will be noted that all except the last are introduced by the

conjunction ei] or a combination of ei] with another particle. Even e]a<n

is, of course, a combination of ei] + a@n, an indefinite particle.

The mood in every instance is indicative, even with e]a<n. The two

instances where e]a<n has the indicative, Rom 11:14 and Rev 11:5, seem

to be first class in sense, even though e]a<n normally is used in third class

conditions, sometimes there with the indicative.

The protasis precedes the apodosis in 267 out of 310 examples

(86%). There are 13 instances where the apodosis is missing.

A great variety of sentence types form the apodosis of first class



Second Class Conditions

Significance, Meaning

Probably the least controversial, its significance is clear: The

protasis sets forth a condition which is not true or is thought to be not


76                    GRACE THEOLOGICAL JOURNAL


true, and the apodosis states the potential consequence if it had been

true. "If this were the case (which is not), then that would be. . . . "



Second Class conditions use ei] with a secondary (past) tense of the

indicative verb in the protasis; the apodosis characteristically is some

potential construction such as a secondary tense indicative, usually but

not always with a@n.

            The conjunction used is always ei], sometimes with the negative mh<

added (10 times); once it also has de<, in ei] de> mh<.


Analysis of Verb Forms: Statistics8

Protasis:                                                            Apodosis:

Tense:                          Order                           Tense:

     Imperfect     21         Proto before                  Imperfect          21

     Aorist          14             > = 42                       Aorist               18

     Pluperfect      6        Proto after                    Pluperfect           3

      [. . . ]            7            < = 6                        [. . . ]                  5

     Total            48                                             ----                     1


            The mood of the protasis is always indicative. The apodosis is

always some potential construction, almost always a secondary tense

of the indicative, usually with a@n (31 times).


Third Class Conditions

Significance, Meaning

            This is properly labelled the Future Condition. It always deals

with a future potential, uncertain (subjunctive) because it hasn't hap-

pened yet.9 The subjunctive does not indicate the degree of uncertainty,

only the fact of uncertainty by reason of futurity.


            8 Some symbols appearing in this and following charts are codes I have used for

abbreviation and convenience:

            [. . . ] = Verb is not present; left to be supplied

            ----    = There is no apodosis present

            >       =  Protasis precedes the apodosis

            <       =  Protasis follows the apodosis

            9 This statement seems to ignore a large number of condition sentences which use

e]a<n with subjunctive, the so-called "present general conditions". I have already given

extensive treatment of these elsewhere in my article on Third Class Conditions, GTJ 3

(1982) 172-75. The "general" or "whenever" idea always introduces potentiality or fu-

turity, and to the Greek mind was expressed naturally by this construction.

                        BOYER: ADVERBIAL CLAUSES              77



            Third Class conditions use e]a<n with a subjunctive verb in the

protasis, the apodosis may be of any type, usually future in its time-



Conjunctions Used

            Almost always it is some form or combination of e]a<n; the simple

e]a<n (213 times), a@n (a contracted form, 3) ka@n (by crasis for kai>  e]a<n,

14), e]a<nper 3; it has the negative added (e]a>n mh<  52); total e]a<n, 286

times. Ei] is used 5 times; simple ei] once, ei] pwj; once, ei@te (twice, in

correlative clauses), and e]kto>j ei] mh< once.


Analysis of Verb Forms: Statistics

            Verb in the Protasis:

            Tense:                                                 Mood:

                 Present                  105                 Indicative                        2

                 Future                        2                  Subjunctive                284

                 Aorist                    177

                 Perfect                      2

                 [. . . ]                           4                  [. . .]                                 4

                 Present. (1?3)           1*                Indicative                        1*

                 Perfect (1?3)            1*                Indicative                        1*


Verb in the Apodosis:

Tense:                                                             Mood:

     Present                  129                             Indicative                    218

     Future                      97                              Subjunctive                  21

     Aorist                      42                              Optative                          1

      Perfect                     7                              Imperative                    33

                                                                        Infinitive                         2

      [. . .]                         12                              [. . .]                               12

        ----                          3                              ----                                  3

      Present (1?3)           1*                            Indicative                        1*

      Perfect (1?3)           1*                            Indicative                        1*


[Explanation: * = Double or doubtful entry; also counted elsewhere


Order of Clauses:

            > =  Protasis precedes Apodosis                 241

            < =  Protasis follows Apodosis                     48

            -=    No apodosis                                               3


78                    GRACE THEOLOGICAL JOURNAL


Sentence type of Apodosis:

            > AS                8          Within a subordinate clause

            > M                 42        Command

            > MN              11        Emphatic Negation

            > MP              2          Prohibition

            > MR              1          Request

            > P                  3          Potential

            > PR               22        Promise

            > RQ               32        Rhetorical question

            > S                  158     Statement

            > TH               11        Threat


Fourth Class Conditions

            Fourth Class conditions use ei] with an optative verb in the protasis

and supposedly (from the ancient pattern) a@n with an optative verb in

the apodosis. But there are no complete examples in the NT, only a few

(9) protases. It has the same significance as the Third Class, only stated

a bit less dramatically.


Conjunctions Used

            The conjunction is always ei], once with the indefinite particle pw<j



Verb Forms in the Protasis

            The protasis in all 9 instances is regular, with an optative verb, six

are present tense, 3 are aorist.


Verb Forms in the Apodosis

            In every instance the apodosis is either incomplete, irregular, or

missing. In four the apodosis is an infinitive in the predicate of the

main clause (Acts 17:27, 20:16, 27:12, 39). In two the protasis is in

effect a single word, a parenthetic expression adverbially attached to

the verb of the main sentence (1 Cor 14:10, 15:37: ei] tu<xoi, "if it should

turn out so"; translated in NASB by "perhaps"). In only three in-

stances is there an actual apodosis present. Two of these leave the verb

unexpressed so it is not possible to tell mood and tense (1 Pet 3:14, 17).

Conceivably an optative (ei@hte v 14, ei@h v 17) might be supplied in

conformity with the normal fourth class pattern, but the sense is not

right for that. Probably it is better to supply the indicative, as this

makes good sense. The third does have a complete apodosis (Acts

24:19). The verb is imperfect indicative, not the optative expected in

                        BOYER: ADVERBIAL CLAUSES              79


fourth class conditions, but possibly it may be taken as a potential

indicative equivalent to an optative. The absence of a@n is not a problem.


Clause Order


            In the seven examples where an apodosis can be identified it

stands after the protasis twice, before it five times.



            Exceptive clauses are a form of conditional clause. They use the

conditional conjunctions, particularly the combination ei] mh<, and in-

volve a special kind of conditional situation. They are treated separately

here because they represent a sizable group in themselves and have

several distinctive features.

            The name reflects the fact that these clauses usually are translated

into English by the word "except." They point to a general situation

which is not true (the apodosis) except for (ei] mh< = "if not") some

specific case (the protasis). Usually the exception is a part of the

general, but the parallelism is not always precise.

            The conjunction used is ei] mh<, so in form they are first or second

class conditions. Ei] mh< is not always exceptive; in 10 instances it is

simply a negative second class condition. There are two of the first

class passages which are extremely elliptical and the construction is


            Is e]a>n mh< ever exceptive? There are 43 third class conditions which

use ei] mh<. Only one of these shows the structural pattern of exceptive

clauses.11 All the rest are simply negative third class conditions.


Structural Classification


            In Greek, these clauses may be grouped into four classes on the

basis of their structure.




            I have used this term to describe the first group because the

conditional phrase used becomes in effect an adverbial introduction to

the "apodosis" or main clause of the sentence. Example: Matt 6:1 ei] de>

mh< ge, misqo>n ou]k e@xete para> t&? patri> u[mw?n t&? e]n toi?j ou]ranoi?j

"otherwise you have no reward with your Father who is in heaven."

There is an extreme ellipsis involved. Starting with ei] mh< if not, the


            10 1 Cor 7:17, 2 Cor 13:5.

            11  Mark 4:22.

80                    GRACE THEOLOGICAL JOURNAL


sense expanded may be "if the situation is not the one stated in the

context" then this is the result," or simply, "otherwise." I have listed 14

examples in this class.12

            The conditional phrases involved in this construction are ei] de> mh<

ge, (8 times), ei] de> mh<  (3), and ei] mh<ti (1). There is never a verb in the

protasis; in fact, there is no protasis at all except this phrase. The rest

of the sentence is the apodosis.


"No One. . . Except"

            The second group, 31 examples, shows a regular pattern. The

conditional conjunction is ei] mh<. The apodosis always stands before

the protasis. It uses the word ou]dei<j or mhdei<j and makes a statement

about "no one" or "nothing." Then in the following protasis it states

the exception to that blanket statement. Example: Phil 4:15 ou]demi<a

moi e]kklhsi<a e]koinw<nhsin ei]j lo<gon do<sewj kai> lh<myewj ei] mh>

u[mei?j mo<noi. "no church shared with me in the matter of giving and

receiving but you alone." The protasis is ei] mh> u[mei?j mo<noi; the verb is

always omitted, left to be supplied mentally: "if you [did] not."


"Not. . . Except"

            The third group, 33 examples, shows almost the same pattern,

except that the negative in the apodosis is a simple negation of the verb

rather than a negative pronoun. The negative particle ou in one of its

forms is usually used, once it is ou]de<. Mh< is used 4 times, once in ou] mh<.

Again the protasis uses ei] mh<, it always follows the apodosis and there

is no verb stated in the protasis. Example: John 19:15 Ou]k e@xomen

basile<a ei] mh> Kaisara. "We have no king but Caesar," or more

precisely, "We do not have a king if [we do] not [have] Caesar."


"Who . . . Except?"

            The fourth group, 10 examples, follows the same pattern except

that the apodosis is stated as a rhetorical question: "Who fits this

situation except . . .?" Example: Luke 5:21 Ti<j du<natai a[marti<aj

a]fei?nai ei] mh> mo<noj o[ qeo<j; "Who can forgive sins, but God alone?"

Note again that ei] mh< is the conjunction, the protasis follows the

apodosis, the verb is omitted in the protasis, and the negative is implied

by the rhetorical question.


            12 This and all other listings referred to in this paper are available through a

Supplemental Manual of Information: Adverbial Clauses, by inter-library loan from

Morgan Library, Grace Theological Seminary, 200 Seminary Dr., Winona Lake, IN


                        BOYER: ADVERBIAL CLAUSES              81


                                    CONCESSIVE CLAUSES

A Class of Conditional Sentences

            There are two groups of concessive clauses. Some (3l) are a

special group of conditional clauses. They are to be distinguished from

other conditional sentences by their use of the conjunction kai< in

combination with the conditional conjunctions ei] and e]a<n, and by their

distinctive meaning.

            Usually translated "though," "although," "even if," these clauses

state a conclusion which is affirmed in spite of the condition stated:

"even if this is the case, the result follows". Sometimes the condition is

considered as an extreme, unlikely case; an objection in spite of which

the conclusion is affirmed. Sometimes the condition is treated as a

matter of little consequence (like our English "So what?”).13 These

clauses may be First Class (20 examples), Third Class (10), even one

Fourth Class.

            Structurally the only signal that a clause is concessive is the use of

kai< in association with the ei] or e]a<n. But it is not a clear signal. Kai<

occurs frequently in conditional clauses when it is not concessive (104

out of 746, 14%). Here is a summary of my conclusions after tabulating

the information.

            (1) Ei] kai< is clearly concessive (18 times). There are a very few

exceptions (4),14 but they each involve another particle along with the

kai<  (ei] de< kai< 3 times, ei@ ge kai< once).

            (2) Kai> ei] is rarely concessive (3 times15 out of 22); it is usually

simply "and if."

            (3) Kai< with e]a<n, whether it stands before or after, does not signal

concession. Out of 45 examples only three16 are concessive. With ka@n

(= kai> e]a<n) 517 of the 14 occurrences are concessive.

            (4) Sometimes the sense is concessive when the form does not

signal it.18


            13 These two concepts are not mutually exclusive. It is the judgment of this writer

that A. T. Robertson's strong distinction between the two, particularly his association of

it with the position of kai< before and after the conjunction, breaks down when the actual

examples are studied. See his A Historical Grammar of New Testament Greek in the

Light of Historical Research (Nashville: Broadman, 1934) 1026.

            14 For full lists, see footnote 12 above. The exceptions are ei] de> kai< Luke 11:18,

1 Cor 4:7, Gal 3:4; ei@ ge kai< 2 Cor 5:3.

            15 1 Cor 8:5, Heb 11:5, 1 Pet 3:1.

            16 With e]a>n kai<: 1 Cor 7:11, 28; with kai> e]a<n: Gal 1:8.

            17 Matt 26:35, Mark 16:18, John 8:14, 10:38, 11:25.

            18 A very interesting situation occurs in the parallel accounts of Peter's dual remon-

strance to Jesus' announcement of his denial (Matt 26:33, 35 and Mark 14:29, 31. All

four statements are clearly concessive in sense. In both accounts the first statement uses

82                    GRACE THEOLOGICAL JOURNAL


Conjunctive Concessive Clauses

            The second group of concessive clauses19 are introduced by con-

cessive subordinate conjunctions, kai<per (5), kai<toi (2), and kaitoige

(1); translated in NASB by "although," "though," "and yet." They are

not conditional, and are included in this place because they are equiva-

lent in sense to those which use the conditional conjunctions. As a

possible link between the two, note that they both to some extent

involve the use of kai<; in these the kai< is compounded with other


            The verb in these clauses is indicative (3 times), a participle (4),

and once it is elliptical. Half of them are found in the book of Hebrews.


                                    COMPARATIVE CLAUSES


            Comparative clauses are quite common in the Greek NT, 331

examples. They augment the statement by comparing it to something

which presumably is understood. Often they come in pairs, as in

English ''as. . . so. . . . " the comparative clause is introduced by a

comparative conjunction, the one to which it is compared may open

with a correlative adverb.

            These clauses either describe or emphasize20 the thought expressed

in the principal clause. Following this lead I have attempted to assign

each comparative clause to one of these categories. The result was:

descriptive, 253 or 76%, emphatic, 78 or 24%. But there were many

where the choice was arbitrary.

            I have attempted another approach to classification which I believe

is more helpful in understanding the possible significances of these

clauses. It is based on an attempt to discern what was the point or

reason for the particular comparison chosen; the primary element of

that comparison which the writer wanted to call to attention. For

example, in Matt 6:2, "When therefore you give alms, do not sound a

trumpet before you, as the hypocrites do in the synagogue and in the

streets, that they may be honored by men," it does not appear that

Jesus was warning against the manner, or the place, in which they blew

their trumpets, but in the fact that they did so at all.


ei] and the second e]a<n. But in Matthew's account kai< is added to the first statement and

not to the second. In Mark's account the reverse is true, kai< is used with the second and

not with the first. In both instances the presence or absence of kai< makes no difference in

the concessive nature of the statement.

            19 There are eight; John 4:2, Acts 14:17, Phil 3:4, Heb 4:3, 5:8, 7:5, 12:7, 2 Pet 1:12.

            20 H. Dana and J. Mantey, A Manual Grammar of  the Greek New Testament (New

York: MacMillan, 1948) 275.

                        BOYER: ADVERBIAL CLAUSES                          83


            In studying the comparative clauses from this point of view, I have

chosen five categories into which these "main point concepts" seem to

fit. The first is the one just illustrated; the action or fact or situation

itself. The comparison points to a similar or parallel situation "in

accord with which" the other is to be seen; a frequent example is the

comparative clause which says, " . . . as the Scripture says. . ." This is

by far the type found most frequently, 199 out of 331, or 60%.

            The others are more specific in their thrust and more easily de-

scribed. The second is a comparison in the manner of doing something

(75 examples); the third, a comparison in quality or character (37); the

fourth, a comparison in degree (17); and fifth, a "parable-type" com-

parison, amounting in effect to a short parable (only 3 such).21



Conjunctions Used

            The most frequently occurring subordinating comparative con-

junction is kaqw<j (175 times, plus one compounded with the particle

per, kaqw<sper. It is translated by the NASB as "just as'' (91), ''as''

(64), "even as'' (14) and by a few other phrases.

            The word w[j appears in the NT much more frequently than

kaqw<j, but often in relations other than the one we are presently

considering. As a comparative conjunction it occurs 108 times, plus 29

times compounded with the particle per, w!sper. To translate it the

NASB uses ''as'' (93), "just as'' (16), "even as'' (6), "like" or "just like"

(10) and several other phrases.

            Kaqa< occurs in this construction 11 times, plus once as kaqa<per.

It is from kaq ] a!, the neuter plural accusative of the relative pronoun

o!j ("according to which things," or "after the pattern of these things").

            Kaqo< is found 4 times. It is from kaq ] o!, the neuter singular

accusative of the relative, with meaning similar to kaqa<. Once it is used

along with the indefinite particle e]a<n.

            Kaqo<ti (from kaq ]  o! ti, neuter of the indefinite relative) occurs

twice as a comparative conjunction, both times with the indefinite

particle a@n.


Correlative adverbs used

            In 66 instances (20%) the comparative clause is countered in the

main clause by the use of a correlative adverb (cf. English ''as. . .

so . . ."). They occur in many combinations and in either sequence. The

list of correlatives, with counts and clause order noted, is as follows:


            21 Matt 25:14, Mark 4:26, 13:34.

84                    GRACE THEOLOGICAL JOURNAL


                                                Before the Comparative       After

            ou!twj                                                16                     31

            kai> ou!twj                                             1                      1

            kai<                                                          1                    12

            ou!tw                                                      1                      1

            o[moi<wj                                                  1                      1


            It should be noted that kai< occurs many more times than appears

in this list, and is used not only with the correlative but also with the

comparative. However, it usually is simply the adverbial kai<. Only

when it appears to be a part of a pair has it been tabulated as



Mood of the Verb in Comparative Clauses

            The mood is almost always indicative. There are four exceptions.

Three are subjunctives; in each the sense is indefinite, the indefinite

particle e]a<n is present in one. The other (Heb 7:9) is an infinitive, and

the whole expression is an old classical idiom, "w[j e@poj ei]pei?n, to put it

in a word or if one may say so, used to soften a statement.”22


Clause Order

            The comparative clause usually follows the main clause (241, or

73% of the time). It precedes the main clause (76, or 23%). The other 14

are instances where there is no main clause expressed.


Other Comparative Constructions

            Our present study is limited to clauses, so such structures as the

use of comparative particles with single words and phrases are ex-

cluded. But it is not always easy to decide whether a particular expres-

sion is a clause or not. If a verb is present, that decides it as a clause;

but if there is no verb it may be questionable. For example, Acts 8:32,

 [Wj pro<baton e]pi> sfagh>n h@xqh, kai> w[j a]mno>j e]nanti<on tou? kei<rantoj

au]to>n a@fwnoj, ou!twj ou]k a]noi<gei to> sto<ma au]tou?. "He was led as a

sheep to slaughter; and as a lamb before its shearer is silent, so He does

not open His mouth." The first w[j stands with the noun pro<baton,

there is no need to supply a verb, and the w[j is probably a comparative

particle; there is no comparative clause. But the second w[j while there

is no actual verb present, clearly introduces a statement and needs a

verb to be supplied; it is a comparative clause. Another example may


            22 W. W. Goodwin, Greek Grammar, rev. C. B. Gulick (Boston: Ginn, 1930) 323.

                        BOYER: ADVERBIAL CLAUSES                          85


not be so easy to decide. Matt 6:5 ou]k e@sesqe w[j oi[ u[pokritai<. "You

are not to be as the hypocrites;" Is w[j a particle? or is it a conjunction

introducing a clause, ''as the hypocrites (are)"? Perhaps the question is

as insignificant in Greek as it is in English.

A group of relative clauses functioning adverbially and expressing

degree or measure need also to be listed here, since they involve a

comparative sense. They are introduced by the relative o!soj (some-

times correlative with tosou?toj, they all use a verb in the indicative

mood. There are 10 examples.23





Final clauses, sometimes called telic (Gr. te<loj, end), point to the

"end" of the sentence action, the direction toward which the action is

directed. They involve the two closely related concepts of purpose and

result. Purpose is intended result; result is accomplished purpose.

Unfortunately for NT exegetes neither NT Greek nor modern English

is careful to distinguish between them. In the vast majority of cases

there is no clue in the Greek text to differentiate between them. Usually

the sense of the context will decide rather clearly; when that is incon-

clusive or controversial it will be so noted. There are some grammatical

structures which identify some result clauses, they will be indicated.

In this study 597 clauses have been identified as final; 521 are

classified as purpose clauses, 65 as result clauses, and 11 have been

listed as doubtful.


Conjunctions Used


 !Ina (398),  i!na mh< (81),  i!na mh<pote (1)


More than 8 out of 10 times (80.2%) the final conjunction is i!na or

i!na mh<. It is used in clauses classified as purpose (460), as result (10),

and as doubtful (9). Its normal construction uses the subjunctive mood

(464), but it occurs also with the indicative (11 times, 9 future and

2 present). In 4 cases the verb is omitted, hence the mood is not

discernible. It is translated as "that" (206), "in older that" (66), "so

that" (64), and by an infinitive (65). With the negative particle mh< or

mh<pote it is translated "in order that. . . not" (10), "that. . . not" (26),

"so that. . . not" (15), or by "lest" (18).


23 See my article on "Relative Clauses," 240.

86                    GRACE THEOLOGICAL JOURNAL


 !Opwj (38)

This conjunction is very similar to i!na in usage and meaning, but

much less common. It is used almost totally in clearly purpose clauses,

only twice in doubtful passages. It always has the subjunctive mood, ex-

cept once the verb is omitted. The same translations are used for o!pwj

as listed above for i!na, even approximately in the same proportion.


Mh< (8), mh<pote (15)

The use of the negative particle mh< as a conjunction equivalent to

i!na mh< in negative purpose clauses is rare (8 examples), only slightly

more frequent when strengthened to mh<pote (15 examples). They all

show the subjunctive mood, and the translations are simply the trans-

lations given for i!na mh< above. Neither of these occur in clauses that

have been classified as result clauses.


 !Wste (53)

This conjunction is the only one that is specifically associated with

result clauses. In classical Greek it was "used with the infinitive and

with the indicative to express result. With the infinitive (the negative

being mh<, the result is stated as one which the action of the verb

tends to produce; with the indicative (the negative is ou]), as one which

that action actually does produce.”24 Both constructions are found

in the NT, although the difference between them is no longer strictly


There are 53 clauses introduced by w!ste in the NT. All but two

have an infinitive verb. The two exceptions have an indicative verb.

They are most frequently translated by "so that" or "so as to" plus an



[Wj (4)

[Wj is a word with many and varied uses in the NT; its use in final

clauses is rare. Twice with an infinitive it is a purpose clause, twice with

the indicative it is a result clause.


24 Goodwin and Gulick, Grammar, 308.

25 The difference should not be construed as a complete change of the sense, but

rather as a weakening of the distinction between them. A result actually produced would

also be one tending to be produced, so w!ste with infinitive would be possible in every

case. In the NT the use of w!ste with the indicative has almost disappeared, and the sense

of actual results is taken over by the infinitive construction, which it could properly do.

But w!ste with infinitive still can express the result which tends to follow. It is not correct.

to insist that result clauses are always actually realized. This insight can help in the

understanding of such passages as Mark 4:37, Luke 5:7, Acts 19:10, 12, Rom 7:6,1 Cor




 !Oti (1)

!Oti introducing a result clause is surprising, and very rare; only

one is so identified in this study (John 7:35). Result fits the sense well,

and there seems to be evidence of other such uses.26


Moods Used

The mood in final clauses is almost always subjunctive, as is to be

expected in a clause which speaks of purpose, intent, of future out-

come. But there are some which do not use the subjunctive. The largest

group consists of 52 result clauses which show an infinitive as verb, a

normal construction from classical times and in no sense an exception.

The seeming exceptions are 18 examples which use an indicative

verb. Of these, 11 are future indicatives. Elsewhere27 I have shown that

the future indicative is very frequently in the NT a practical substitute

for the aorist subjunctive. It is identical in form in many instances, and

differs only in spelling in many others. Its sense is basically the same,

expressing the potentiality that always is associated with the future. In

the NT it is actually used in almost every syntactical construction that

ordinarily uses the subjunctive. In the light of these facts it should not

be considered strange or exceptional when a future indicative is used in

a final clause.

Five indicatives are used with conjunctions which properly use the

indicative.    !Wste occurs twice with the indicative, a normal usage from

classical times, and both expressing actual result.  !Wj occurs twice with

the indicative, again classical usage in result clauses.    !Wj universally

uses the indicative, and while its use for a final clause is very unusual,

its use of the indicative is not.

There remain two28 indicatives which are significantly unusual.

They are 1 Cor 4:6  i!na mh< . . . fusiou?sqe "in order that no one of you

might become arrogant in behalf of one against the other," and Gal

4:17  i!na au]tou>j zhlou?te" in order that you may seek them."

Grammarians suggest several explanations. A. T. Robertson29

discusses the orthography and the possibility that by NT times the


26 BAGD, 589; John 14:22 "ti< ge<gonen o!ti . . . ; what has happened, so that (= to

bring it about that) . . . ?" They suggest that this may possibly be the explanation also of

1 Tim 6:7 and Heb 2:6.

27 See my article, "The Classification of Subjunctives: A Statistical Study," GTJ 7

(1986) 16-19.

28 A third example is often listed, 1 John 5:20  i!na ginw<skomen. But there is textual

variance in this case, and the NA26 and UBS3 texts have decided for the subjunctive


29 A. T. Robertson, A Grammar of the Greek New Testament in the Light of

Historical Research (Nashville: Broadman, 1934) 203, 325.

88                    GRACE THEOLOGICAL JOURNAL


indicative and subjunctive of -ow verbs have become alike as they are

in -aw verbs. Blass-Debrunner suggest the same; "fusiou?sqe 1 Cor 4:6

and zhlou?te Gal 4:17 are subjunctives.”30 In another place, apparently

speaking more generally, they say "The present indicative after i!na is;

of course, only a corruption of the text.”31 No one suggests that there is

a difference in meaning intended by the use of the indicative after i!na.


Clause Order

Final clauses normally follow the main clause. This is true in all of

the result clauses, and in all but 15 of the purpose clauses (97%). In 4

instances there is no main clause.


Other Final Constructions

Relatively infrequent is the use of the adverbial or circumstantial

participle to express purpose. According to my count there are about

56 examples, five with the future participle which in classical was

generally used in this way, but 48 with the present participle. Only

three aorist participles are used thus in the NT.

Articular infinitives with certain prepositions are used to express

purpose; ei]j to< (73) and pro>j to< (11).32 At least one of these is

understood to be expressing result in the NASB (1 Thess 2:16 ei]j to>

a]naplhrw?sai au]tw?n ta>j a[marti<aj pa<ntote. "with the result that they

always fill up the measure of their sins."

Another structure indirectly expressing purpose is a substantive or

noun clause which structurally stands in apposition to a pronoun

whose antecedent refers to a purpose. Example: Eph 6:22 o{n e@pemya

pro>j u[ma?j ei]j au]to> tou?to i!na gnw?te ta> peri> h[ma?j "I have sent him to

you for this very purpose, so that you may know about us . . ." If the

words ei]j au]to> tou?to were not present, the i!na clause would be a

purpose clause; as it stands it is a noun clause in apposition to that

phrase, stating the content of that purpose.33





Temporal Clauses are those which modify the main clause by

relating it in some manner to the concept of time, answering such


30 F. Blass and A. Debrunner, A Greek Grammar of the New Testament and Other

Early Christian Literature, trans. and rev. by Robert Punk (Chicago: Univ. of Chicago,

1961) 188.

31 Ibid., 187.

32 See footnote 6.

33 See my article "Noun Clauses in the Greek NT: A Statistical Study," GTJ 10 (1990)


BOYER: ADVERBIAL CLAUSES                          89


questions as "when?," "how often?," "how long?," "before?," "after?,"

"until?," "while?," etc. A variety of conjunctions and conjunctive rela-

tive phrases occur, also more variety in the moods used.


Conjunctions Used

            The subordinate temporal clauses included in this study are of two

kinds. Most (360 or 86%) are introduced by proper conjunctions. The

rest are introduced, not by conjunctions per se, but by relative phrases

which function as temporal conjunctions, wither with or without the

antecedent (h[me<raj or xro<nou) of the relative being actually expressed.

If the antecedent is present the clause can of course be considered a

simple adjectival relative clause modifying a noun which is functioning

as a temporal adverb. But they seem by their number and frequency to

have become fixed expressions, virtual conjunctions.34 Sixty such

clauses are included in this list, and the relative phrase is listed as a



 !Otan (123),    !Ote (102)

             !Ote alone, and  o!tan, which is o!te + a@n, comprise 54% of all

temporal clauses. They are the less explicit in time relation, expressing

simple concurrence. Most often they are translated by the NASB as

"when" (197 times);  o!tan, reflecting the particle a@n, is also translated

"whenever" (10 times). Rarely they are translated more specifically as

"after" (o!te  5, o!tan 2), "while" (o!te 3, o!tan 2), ''as'' (o!te once), and ''as

soon as'' (once each), even "until" (o!tan 2). These more specific render-

ings apparently are derived from the context rather than from the

conjunction itself.

             !Ote normally is followed by the indicative mood, only twice does

the subjunctive appear; once the verb is left unexpressed.    !Otan, as is

expected with a@n, normally uses the subjunctive (5 with indicative).


 [Wj (69),  [Wj a@n (3)


             [Wj has many other uses, but as a temporal conjunction it occurs

72 times. These are translated with almost identical expressions as were

o!te and o!tan, most frequently "when" (w[j; 50, w[j a@n 1), also "while"

(10), "after" (2), "whenever" and ''as soon as'' (once each). Renderings

other than those used for o!te are ''as'' (5) and "since" (2).

            With  w[j the mood is indicative. When it is used with the indefinite

particle a@n the mood is subjunctive.


            34 As we have already seen, they are not limited to temporal clauses. Similar phrases

are found introducing causal clauses, and clauses expressing manner, degree or measure,

a total of 90 NT examples. See my article. "Relative Clauses," 238-40.

90                    GRACE THEOLOGICAL JOURNAL


 !Ewj (18),  !Ewj a@n (20),  !Ewj ou$ (17),  !Ewj o!tou (5)

            The conjunction e!wj occurs alone or with the indefinite particle a@n

38 times. As a relative phrase which functions as a temporal conjunc-

tion it is found 22 times. They are translated "until" (54), "while" (4),

and ''as long as'' (once). The basic meaning seems to refer to a period of

time up to a designated point.

            The mood is indicative 13 times, and subjunctive 45 times, once

the verb is unexpressed. This is one of a group of words with the

meaning "until" which involve some special rules regarding mood, and

will be discussed separately below, under "Moods Used."


 @Axri (5),  @Axri h$j (4),  @Axri ou$ (4),  @Axrij ou$ (4),  @Axrij  ou$  a@n (1)

            As a conjunction a@xri appears 5 times. More commonly (13) it

appears in a conjunctive relative phrase. It is translated "until" in every

instance except one, where ''as long as'' is used.

            The mood is indicative 7 times and subjunctive 11 times. For the

use of moods with words meaning "until" see below.


Pri<n (8), Pri>n  h@ (4), Pri>n h@ a@n (1)

            This is the only conjunction in the NT translated "before."

The usual construction with pri<n is with an infinitive following (11

times). Once it is followed by a subjunctive and once by an optative.


 ]Af ]  h$j (6),  ]Af ]  ou$ (3),   ]Af ] ou$ a@n (1)

            These are all relative phrases functioning as temporal conjunc-

tions. Literally translated, the meaning would be "from which day, or

time." The NASB translates them, 6 of the 10 times, by "since" (the

others are freer, paraphrastic renderings: "once," "ago," even "for" and

"that"). In every instance it is used in measuring time starting from a

specific point; when that starting point is in the past the mood is

indicative (9 times), when it is in the future the mood is subjunctive (1).


 ]Ef ]  o!son (5),   !Oson xro<non (1)

            Again this is a relative phrase functioning as a conjunction.  @Osoj

is a correlative expressing a quantitative concept "how much"; the

phrase thus means "for how much time." It is translated in every

instance by ''as long as'' (5) or "so long as'' (once), and it always uses

the indicative mood.


 ]En &$ (4)

            Another relative phrase in which the time word   xro<n& is to be

supplied, it carries the meaning "in (or during) which time." It is


                        BOYER: ADVERBIAL CLAUSES              91


translated "while" (3 times) and "until" (once). The mood is always



Me<xri (1), Me<xri ou$ (2)

            Once it is a conjunction, twice it is in a relative phrase. Like its

synonym a@xri it is translated "until." The mood is always subjunctive.


 [Osa<kij e]a<n  (3)

             [Osa<kij is made up of the correlative o!soj, "how much," and the

adverbial ending -kij, "times"; thus "how many times." It is translated

''as often as.'' It always has the indefinite particle e]a<n and is followed

by the subjunctive mood.


 ]Epa<n (3)

            Arising from e]pei< (or e]peidh<) plus the indefinite particle a@n, this

conjunction seems to be a full synonym of  o!tan (twice, in Luke 11:21,

22 and 11:34, they are used interchangeably in parallel sentences). Like

o!tan it is translated "when." It is indefinite and takes the subjunctive



 [Hni<ka a@n  (1),  [Hni<ka e]a<n (1)

            Both forms are identical, differing only in the spelling of the

indefinite particle a@n. Translated "whenever," it is indefinite and takes

the subjunctive mood.


 ]Epeidh< (1)

This conjunction is more frequently causal, but once it is temporal

(Luke 7:1), translated "when." The mood is indicative.


Kaqw<j (1)

            Kaqw<j is usually comparative or causal, but in one place it seems

to be temporal (Acts 7:17). The NASB translates it ''as.'' The mood is



Moods Used

Basic Principles

            In the review of conjunctions we have already given statistics of

the moods used with each. Unlike the causal and comparative clauses

which were predominantly indicative, and the final which were subjunc-

tive, these like the relative and conditional clauses freely use both

moods. Here we shall discuss the basic distinctions which govern the


92                    GRACE THEOLOGICAL JOURNAL


moods used, giving special attention to the seeming "exceptions" to

those principles.

            The choice between indicative and subjunctive in temporal clauses

is determined by the distinction between actual and potential, and

between specific and indefinite. If it speaks of a specific time past or

present the indicative is the natural choice. If the time is future and

thus unknown and doubtful, or if it is indefinite, the mood expected is

subjunctive. These same factors also control the conjunctions used, so

that there is a close correlation between the conjunction and the mood;

we expect, for example, that o!te will use an indicative verb and that

o!tan will use a subjunctive.

            Since these two conjunctions account for more than half of all the

temporal clauses in the NT, we will use them as examples to illustrate

the distinction between the moods.   !Ote is followed by an indicative

verb 102 times out of 104; once it introduces a clause which has no verb

so the mood is not indicated, once it occurs with a subjunctive verb.

Examining the indicatives, 92 times it is used of an actual historically

past event, in four instances it speaks of actual contemporary time.35

The other five are future in time (either a future indicative or a

futuristic present indicative), but in every case they are specific (several

have the form, "the hour is coming when. . . “).

            In just one passage the NA26 text shows a possible place where

o!te may be found with a subjunctive verb; in Luke 13:35. This will be

discussed later, in the section on "Until Clauses."

             !Otan, which is simply the same o!te with the indefinite particle a@n,

is almost as exclusively followed by the subjunctive. There are 5 in-

stances of o!tan with the indicative. Three occur with statements of a

general truth, "whenever," which usually (44 times) is expressed with

the subjunctive. One of these (Rev 4:9) has the future indicative, which

as we have seen is practically synonymous with the subjunctive. The

other two (Mark 3:11, 11: 19) refer specifically to that which had

happened repeatedly in past time, hence the indicative is appropriate.

            That leaves two instances of o!tan with the indicative which are

more difficult to account for. Mark 11:25 seems clearly to be general

and future, so that a subjunctive would be expected. There are textual

variants showing a subjunctive form. Rev 8:1, after six times stating

the same fact using o!te with the indicative, on the seventh of the series

changes to o!tan with the indicative. It is simple description of an actual

event, past from the vantage-point of the writer, and with the same


            35 In Mark 14:12 o!te to> pa<sxa e@quon "the day. . . when the Passover is [always]

sacrificed," (my rendering) seems to be the imperfect of customary action. The NASB

rendering, "When the Passover lamb was being sacrificed" is grammatically possible, but

it seems impossible that the actual sacrificing was going on at the time when the disciples

inquired about "preparing" for it.

                        BOYER: ADVERBIAL CLAUSES                          93


sense as the first six. It clearly calls for a verb in the indicative.

Apparently the problem is with the conjunction, and some texts read


            With several other of the conjunctions the addition of the in-

definite particle a@n or e]a<n means a change from indicative to subjunc-

tive, and for the same reasons. When the temporal clause is definite the

mood is indicative, when the a@n changes it to indefinite the mood

changes to subjunctive. This is true with w[j and w[j a@n, e]peidh< and

e]pa<n, a]f ]  ou$ and a]f ] ou$ a@n. Kaqw<j, e]f ] o@son, and e]n &# do not use a@n

and are always indicative; h[ni<ka and o[sa<kij always have a@n, and

always use subjunctive.


"Until" Clauses.

            From classical times some special consideration was given to

those temporal clauses which express the notion of "until." The con-

junctions involved were e@nj, e@ste (not found in NT), a@xri and me<xri.

When they meant "while, so long as'' they were regular in construction.

But when they meant "until" they showed many peculiarities,

            The same situation is mostly true of their NT use, but without

some of the rather complicated "rules." The list of conjunctions include

some of the relative phrases we have described already, and the in-

definite particle a@n is not so strictly required, but the use is basically

unchanged. When these words mean "until" two constructions occur.

If the clause is referring to a definite past action the mood used is

indicative. If they refer to an indefinite future action the mood is

subjunctive. This statement of the case is probably over-simplified for

the classical, but it will serve quite well for the NT pattern.

            I have taken these conjunctions which are translated "until”36 and

in every instance evaluated the "point until which" intended, whether it

is definite and past to the outlook of the speaker or writer, or potential

(i.e., general or indefinite) future. The results were then compared with

the actual structure used. These observations are the result.


            (1) There are 76 passages in the list. Eighteen are indicative, and

58 subjunctive.

            (2) In 12 instances the "time until" has been judged to be actually

past. In every instance the mood used is indicative, and proper,

            (3) The indicative is used in one passage (Rev 17:17) where the

"time until" seems to be future from the standpoint of the persons

involved, which might suggest a subjunctive verb. Two explanations


            36 For the most part I have followed the NASB. In some instances they have used

"until" where in my judgment the sense should be "while." (For the lists in this section,

see footnote 12 above.)

94                    GRACE THEOLOGICAL JOURNAL


may be suggested. The time may be considered from the vantage of the

One whose purpose was being executed and that purpose was a past

reality. Or, it should be noted that the verb used is future indicative,

which is often in Greek a substitute for the subjunctive. My preference

is for this last explanation.

            (4) In 63 instances the temporal clause has been judged to be

future or indefinite, indicating the subjunctive should be expected. 58

of them are actually subjunctive, leaving 5 examples where the verb is

indicative when we might expect the subjunctive.


            Most difficult is Luke 13:35 ou] mh> i@dhte< me e!wj  [h!cei o@te] ei@phte

"you shall not see Me until the time comes when you say. . ." A textual

problem is apparent; does the verb ei@phte go with e!wj or o!te? If the

bracketed words are omitted it goes with e!wj and the subjunctive is

proper (the time is indefinite future). If the bracketed words are left in

ei@phte must be construed with o!te and an indicative is indicated. Also,

in that case e!wj governs h!cei and that verb should be subjunctive. The

NA26 edition puts the words in the text, but in brackets; the UBS3

edition gives the same text but gives the extra words a D (poor) rating.

Westcott and Hort rejected them. The principle that the more difficult

reading is more apt to be the correct one would argue for their

inclusion, but perhaps that principle is not 100% correct.

            The other four (Luke 19:13, John 21:22, 23, 1 Tim 4:13) are all of

a similar nature. All are examples of the futuristic use of the present

indicative of e@rxomai, "I am coming." In each case the sense is future

and the time is indefinite, indicating that a subjunctive verb should be

expected. The nature of this particular verb may help to explain the


            But there may be more than this involved. The first of these four

has the conjunctive phrase e]n &$, which nowhere else means "until."

Even the NASB has in the margin the rendering, "while I am coming."

If we follow this meaning, it suggests that these present indicatives may

be placing the emphasis on the meantime activity ("while") rather than

on the future point of termination ("until"). The last three examples use

a different conjunction, e!wj, but e!wj too frequently means "while.”37 If

that is the case here, it makes the special "until" rule inapplicable and

the present indicative are natural and proper.


Clause Order

            Temporal clauses precede the main clause 267 times, or 63%, they

follow the main clause 155 times.


            37 Robertson, Grammar, 975, says, "When the present ind appears with e!wj the

notion is while, not 'until.'" He lists another example, Mark 6:45.

                        BOYER: ADVERBIAL CLAUSES              95


Other Temporal Constructions

            By far the most frequently used temporal construction in the NT is

not the conjunctive clause, rather it is the adverbial or circumstantial

participle. At least 765 participles function as the equivalent of a

temporal clause. Of course these participles do not constitute "clauses"

in the technical sense, but they contain every element of a clause: --the

participle itself is a verb form, the subject of the clause is always the

word with which the participle is in agreement. In most cases it is best

to translate them by temporal clauses, in many cases it is impossible to

do it any other way.38 The precise temporal significance is not explicit

and must be gathered from the tense used and the context.

Another prominent substitute for a temporal clause in the NT is

the use of the articular infinitive with a proposition. Four prepositions

are used in the temporal sense in this construction.  ]En t&? + infinitive

is the most common (56 examples). It is always used of concurrent time

("in, or during, the time when," and is usually translated "while." Meta>

to< + infinitive occurs 15 times, translated "after." Pro> tou? + infinitive

is found 9 times, translated "before."   !Ewj tou? + infinitive, once only

in this construction, is translated "until." Again, while these may not fit

the definition of a clause, they function in every respect like a clause.

Indeed, it is impossible to translate them into understandable English

except by converting them into a clause.





Local clauses, or locational to use a term parallel with temporal,

are those which tell where the action of the main clause is located. It

answers the question "where?" They are the fewest in number of all the

kinds of clauses in the NT, but they still number 112.


Conjunctions Used


 !Opou (72),   !Opou a@n  / e]a<n (10)


The most frequent of the local conjunctions, it is translated

"where" (61) and with e]a<n "wherever" (11), plus a variety of renderings

once each, such as "from which," "in which, " "on which," "there," even

"since" and "whenever." The last two are unusual departures from the

normal sense, and to me seem unnecessary. In both instances (Mark

9:18, 1 Cor 3:3) a location in place is a natural and more vivid sense,

even in English.


38 Listings of these temporal participles are available from the library at Grace

Theological Seminary, in my Supplemental Manual, Participles.

96                    GRACE THEOLOGICAL JOURNAL


Ou$ (22), ou$ e]a<n (1)

Ou$, originally the genitive singular of the relative pronoun, came

to be used as a conjunction of place, "where," and with the indefinite

particle e]a<n, "wherever."


 !Oqen (7)

With the adverbial suffix -qen, place "whence" or "from which,"

o!qen is the equivalent of e]kei<qen o!pou "from the place where" or "from

which." It is translated "from which" (4), "from there" (1), and simply

"there" (2).


Moods Used

When these clauses are without the indefinite particle a@n or e]a<n

the mood is always indicative. When the indefinite particle is present

the mood is subjunctive with one exception. Mark 6:56 has o!pou a@n

with an imperfect indicative. The sense seems clearly to be indefinite or

general, the use of the imperfect with a@n for a potential or iterative

sense is classical.39


Clause Order

Local clauses follow the main clause 75% of the time. The count is

84 following and 28 preceding.


39 Goodwin and Gulick, Grammar, 275-76.





This material is cited with gracious permission from:

            Grace Theological Seminary

            200 Seminary Dr.

            Winona Lake,  IN   46590

Please report any errors to Ted Hildebrandt at: