Bulletin for Biblical Research 13.1 (2003) 97-125   [© 2003 Institute for Biblical Research]

 

            Greek Grammar and the Personality

                          of the Holy Spirit

 

 

                                       DANIEL B. WALLACE

                                 DALLAS THEOLOGICAL SEMINARY

 

            The modern, broadly conservative articulation of the distinct personality

            and deity of the Holy Spirit has often included in its arsenal a point or two

            from the realm of philology. The Fourth Gospel has especially been mined

            for such grammatical nuggets, though Ephesians, 1 John, and sometimes

            even 2 Thessalonians have been claimed as yielding syntactical evidence in

            defense of the Spirit's personality. Two kinds of texts have been put forth

            in support of this supposition: passages involving grammatical gender

            and passages involving notions of agency. Those involving grammatical

            gender are used as an apologetic defense of a high pneumatology; those in-

            volving agency are simply assumed to prove the point. I believe that this

            grammatical defense for the Spirit's personality has a poor foundation. If

            it is indeed invalid, then to use it in defense of a high pneumatology not

            only damages Trinitarian apologetics but also may well mask an emerging

            pneumatology within the NT.

 

            Key Words: Holy Spirit, pneumatology, gender, personality, Greek

            grammar

 

 

         PASSAGES INVOLVING GRAMMATICAL GENDER

 

About half a dozen texts in the NT are used in support of the Spirit's

personality on the grounds of gender shift due to constructio ad sen-

sum ("construction according to sense" or, in this case, according to

natural as opposed to grammatical gender). That is to say, these pas-

sages seem to refer to the Spirit with the masculine gender in spite of

the fact that pneu=ma is neuter, and grammatical concord would nor-

mally require that any reference to the Spirit also be in the neuter

gender. Such gender shifts are attributed to the fact that the Spirit is

 

Author's note: An earlier version of this paper was read at the annual IBR meeting in

Denver, Colorado. Thanks are due to Dr. Buist M. Fanning, Prof. R. Elliott Greene, Dr.

Scott Hafemann, Dr. W Hall Harris, Prof. C. F. D. Moule, and Dr. David H. Wallace for

looking at a preliminary draft of the paper and offering their input.

 

98                     Bulletin for Biblical Research 13.1

 

a person, and hence the biblical authors naturally speak of him as

such, even though this manner of speaking is contrary to normal

grammatical convention.¹

            A word should be mentioned first about the use of natural gram-

mar in the NT. All exegetes recognize that natural gender is some-

times used in the place of grammatical gender in Greek. Robertson

notes that "substantives have two sorts of gender, natural and gram-

matical. The two do not always agree. The apparent violations of the

rules of gender can generally be explained by the conflict in these

two points of view."² For example, in Col 2:19 we see the construction

th_n kefalh/n . . . e)c ou{ ("the head . . . from whom"): the antecedent of

the masculine pronoun is a feminine noun. But in the context, kefalh/  

refers to Christ (see Col 1:18; 2:10). In Matt 28:19 the Lord instructs

the eleven to "make disciples of all the nations, baptizing them"

(maqhteu/sate pa/nta ta_ e@qnh, bapti/zontej au)tou/j): although "nations"

is neuter, the pronoun "them" is masculine because people are in

view. In Gal 4:19, Paul speaks of "my children, whom" (te/kna mou

ou#j), using the masculine relative pronoun to refer to the "children."³

In Acts 21:36 we read of "the multitude of the people crying out" (to_

plh/qoj tou= laou= kra/zontej): not only is there a gender shift but a

number shift too.⁴ There are even one or two indisputable texts that

refer to an evil spirit with the masculine gender. For example, in

Mark 9:26 the masculine participles kra/caj and spara/caj refer back

to the pneu=ma of v. 25.⁵ These examples could be multiplied⁶ and are

 

            1. By "normal grammatical convention" we do not mean prescriptive rules that

are imposed on the writers by modern researchers but merely the conventions of the

language—how it was used by real people. Such grammatical "rules" are thus descrip-

tive of what Koine speakers actually did rather than being prescriptive of what they

should have done. When a notable exception to such behavioral patterns is observed,

it can be called a violation of a grammatical rule.

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

Research (4th ed.; Nashville: Broadman, 1934), 410.

            3. See also, regarding te/knon, Phlm 10 (parakalw= se peri_ tou= emou= te/knou, o$n

e)ge/nnhsa); 2 John 1 (toi=j te/knoij au)th=j, ou#j [in which the antecedent of the masculine

ou#j is both a feminine singular and a neuter plural]). The word paidi/on is similar: cf. Mark 5:41

(krath/saj th=j xeiro_j tou= paidi/ou le/gei au)th|=), in which the feminine pronoun is bracketed

by paidi/on and to_ kora/sion; Mark 9:24-26 (paidi/ou . . . au)to/n . . . nekro/j).

            4. The neuter singular noun plh=qoj is followed by the masculine plural participle

kra/zontej. It will not do to say that the participle agrees with laou= since that is in the

genitive singular. This is constructio ad sensum, pure and simple.

            5. See also Luke 9:39-40 in P⁴⁵ (pneu=ma . . . au)to/n).

            6. Cf., e.g., Matt 25:32; Mark 3:8; 5:41; Luke 2:13; 10:13; 19:37; John 1:12; 6:9; 17:2,

24; Acts 5:16; 8:5, 10; 9:15; 13:48; 14:4; 15:17; 25:24; 26:17; Rom 2:14, 26; 4:9-11; 9:23-24;

Gal 1:22-23; 4:19; Eph 2:11; 4:17-18; Phil 3:7; Col 2:15; Phlm 10; 1 Pet 2:19; 2 Pet 2:17; 2 John

1; Jude 7, 12. For number shifts, cf. Matt 21:8; Mark 3:32; 5:24; Luke 5:29; 6:19; 8:40; John

6:2, 37; 7:49; 12:18; 17:2; Acts 6:7. Omitted from this list are the numerous examples

in Revelation because it is hardly representative of the literary level and style of Greek

found in the rest of the NT.

             WALLACE: Greek Grammar and the Holy Spirit                    99

 

common knowledge to anyone who works in the Greek NT.⁷ For our

purposes, the point to make is simply that, because such gender

shifts are unremarkable, if the NT authors indeed conceived of the

Holy Spirit as a person, we may well expect to see natural gender

taking precedent over grammatical gender in various passages that

speak of the Spirit.

            The passages adduced for this grammatical argument are John

14:26; 15:26; 16:7, 13-14; Eph 1:14; 2 Thess 2:6-7; and 1 John 5:7. These

fall into three different groups: the Upper Room Discourse texts all

involve a masculine demonstrative pronoun, Eph 1:14 employs a mas-

culine relative pronoun, 2 Thess 2:6-7 and 1 John 5:7 use a masculine

participle.

            As a preliminary consideration, one of these passages, 2 Thess

2:6-7, can be dismissed from consideration with minimal fuss. There

the exchange between to_ kate/xon and o( kate/xwn involves a long-

standing interpretive conundrum, in spite of the fact that within a

certain segment of Protestantism—namely, dispensationalism-

some interpreters have insisted that the Holy Spirit is in view and

that from this exegetical conclusion they can affirm the personality of

the Spirit on the grounds of Greek grammar.⁸ Even if the referent of

to_ kate/xon/o( kate/xwn is the Spirit, the fact that nowhere in the pas-

sage is pneu=ma a#gion even mentioned⁹ renders this passage worthless

for the purposes of explicitly grammaticizing the Spirit's personality.

On similar grounds, John 16:7 can be dismissed, since pneu=ma is not

 

            7. Turner calls the incongruence of gender or number that is due to constructio ad

sensum "good Greek" (Nigel Turner, Syntax, vol. 3 of J. H. Moulton et al., A Grammar

of New Testament Greek [Edinburgh: T. & T. Clark, 1963], 311). BDF, well known as a

grammar of exceptions, does not even list the use of masculine for neuter, presumably

because it is so common. They do list, however, feminine for neuter, masculine for

feminine, and neuter for persons, "if it is not the individuals but a general quality that

is to be emphasized" (pp. 76-77 [§138]). This lacuna has not been filled with BDR

(p. 115 [§138]). This instance of constructio ad sensum is also common enough in Clas-

sical Greek (cf. B. L. Gildersleeve, Syntax of Classical Greek from Homer to Demosthenes

[New York: American Book, 1900-1911], 2.204-7 [§§499-502] for numerous examples

of various kinds of pronominal incongruence).

            8. to_ kate/xon/o( kate/xwn has/have been variously identified as the Church, the

proclamation of the gospel/Paul, Elijah, an angel (especially Michael), the Roman Em-

pire/emperor, the Jewish state/James, God's will/God, the Holy Spirit, the mystery of

lawlessness/Satan, a false prophet, etc. Among the more recent treatments, see espe-

cially Paul S. Dixon, "The Evil Restraint in 2 Thess 2:6," JETS 33 (1990): 445-49 (es-

pousing the interpretation that the mystery of lawlessness/Satan is in view); Charles

E. Powell, "The Identity of the 'Restrainer' in 2 Thessalonians 2:6-7," BSac 154 (1997):

320-32 (arguing that the proclamation of the gospel/Holy Spirit are in view); Colin

Nicholl, "Michael, the Restrainer Removed (2 Thess. 2:6-7)," JTS n.s. 51 (2000): 27-53.

            9. Although two of the three instances of pneu=ma in 2 Thessalonians occur in this

"little apocalypse" (2:1-12), neither of them refers to the Holy Spirit (2:2 refers to a

prophetic utterance, while 2:8 refers to the breath of the vanquishing Messiah as that

which destroys the man of lawlessness).

100                    Bulletin for Biblical Research 13.1

 

explicitly mentioned in this chapter until v. 13.¹⁰ In any event, the

other texts in the Upper Room Discourse have been almost univer-

sally regarded as greater demonstrations of the Spirit's personality,¹¹

so no harm is done in removing John 16:7 from consideration. The

five remaining passages, however, deserve some attention.

            Many theologians treat these passages as a primary proof of the

Spirit's personality. Long ago, Charles Hodge gave a detailed expo-

sition of this viewpoint when he wrote:

 

            The first argument for the personality of the Holy Spirit is derived

            from the use of the personal pronouns in relation to Him.... Our

            Lord says (John xv. 26), "When the Comforter (o( para/klhtoj) is

            come whom I will send unto you from the Father, even the Spirit of

            truth (to_ pneu=ma th=j a)lhqei/aj) which (o#) proceedeth from the Father,

            He (e)kei=noj) shall testify of me." The use of the masculine pronoun

            He instead of it, shows that the Spirit is a person. It may indeed he

            said that as para/klhtoj is masculine, the pronoun referring to it

 

            10. In John 16:8, the only explicit antecedent to e)kei=noj is o( para/klhtojin v. 7. The

personal pronoun au)to/n in v. 7 also refers back to para/klhtoj. As Curt Steven Mayes

(Pronominal Referents and the Personality of the Holy Spirit [Th.M. thesis, Dallas Theo-

logical Seminary, 1980], 33) notes on this passage, "The fact that John often uses

e)kei=noj as the equivalent of a personal pronoun (= he or they) may be significant for

the Spirit's personality. But the question is, how is the masculine form in this passage

to be explained? Is it meant to teach theology or agree with para/klhtoj? Surely the lat-

ter is a grammatically sound conclusion." Mayes's observation leads to a further in-

teresting point: in 1 John, as R. Brown and others have repeatedly noted, the author

consistently uses the pronoun e)kei=noj, to refer to Jesus (as opposed to God the Father).

Now there are significant shifts (albeit subtle ones) in the terminology between the

Gospel of John and 1 John, but I wonder if the common thread here is the concept of

the ascended Christ as Spirit. If this were the case (and I admit it's an if), the author

would tend toward the masculine, not because of a view of the Spirit's personality, but

because of a view that the Spirit was identified somehow with the ascended, exalted

Christ (who would naturally be thought of as masculine).

            11. G. B. Stevens (The Johannine Theology: A Study of the Doctrinal Contents of the

Gospel and Epistles of the Apostle John [New York: Scribner's, 1899]) provides a notable

exception to this. He argues that "as soon as pneu=ma ceases to be the immediate ante-

cedent of pronouns designating the Spirit, masculine forms are employed" (pp. 195-

96). After discussing John 14:26 and 15:26, he states: "It is obvious that, in John's usage,

as soon as the necessity of referring to the Spirit by neuter pronouns which arises from

the immediate antecedence of to_ pneu/ma, is removed, he instinctively adopts masculine

designations. Accordingly in all the passages where the neuter word pneu=ma is not

used, we find the masculine pronouns au)to/j and e)kei=noj employed (xvi. 7, 8, 13, 14). . . .

It thus appears that John, when not prevented from so doing by the grammatical gen-

der of pneu=ma, uniformly designates the Spirit by masculine pronouns implying per-

sonality" (p. 196). But Stevens's premise is wrong: John is not prevented from using the

masculine pronoun in close conjunction with pneu=ma, as we have shown already with

routine uses of natural gender replacing grammatical gender. If his premise is wrong,

then his conclusion is not valid. But in any event, John 16:13 is the stronger of the two

texts, even on Stevens's reasoning, and is therefore discussed below.

             WALLACE: Greek Grammar and the Holy Spirit                     101

 

            must of course be in the same gender. But as the explanatory words

            to_ pneu=ma intervene, to which the neuter o# refers, the following pro-

            nouns would naturally be in the neuter, if the subject spoken of, the

            pneu=ma, were not a person. In the following chapter (John xvi. 13, 14)

            there is no ground for this objection. It is there said, "When He

            (e)kei=noj), the Spirit of truth, is come, He will guide you into all

            truth: for He shall not speak of Himself; but whatsoever He shall

            hear, that shall He speak, and He will show you things to come. He

            shall glorify me (e)kei=noj e)me_ doca/sei): for He shall receive of mine,

            and shall show it unto you." Here there is no possibility of account-

            ing for the use of the personal pronoun He (e)kei=noj) on any other

            ground than the personality of the Spirit.¹²

 

Other theologians have followed in Hodge's train, making this a

primary argument in their defense of the Spirit's personality. For ex-

ample, Walvoord writes, "The only explanation for the masculine [in

John 15:26 and 16:13-14] is that the pronouns refer to a person. Rel-

ative pronouns are used in the same way in Ephesians 1:13-14. These

indirect evidences confirm that the Holy Spirit is commonly re-

garded as a person in the Scripture."¹³ Erickson states,

 

            The first evidence of the Spirit's personality is the use of the mascu-

            line pronoun in representing him. Since the word pneu=ma is neuter,

            and since pronouns are to agree with their antecedents in person,

            number, and gender, we would expect the neuter pronoun to be

            used to represent the Holy Spirit. Yet in John 16:13-14 we find an

            unusual phenomenon. As Jesus describes the Holy Spirit's ministry,

            he uses a masculine pronoun (e)kei=noj) where we would expect a

            neuter pronoun. The only possible antecedent in the immediate con-

            text is "Spirit of Truth" (v. 13). . . . [John] deliberately chose to use

            the masculine to convey to us the fact that Jesus is referring to a per-

            son, not a thing. A similar reference is Ephesians 1:14, where, in a

            relative clause modifying "Holy Spirit," the preferred textual read-

            ing is o#j.¹⁴

 

            12. Charles Hodge, Systematic Theology (New York: Scribner, 1871), 1.524. I am still

in the process of tracing the roots of this argument. Calvin does not use it, nor do the

ancient Fathers (so far as I have been able to tell). It is also largely an argument found

among English-speaking scholars. It did not originate with Charles Hodge, but he was

a prime mover in getting the philological argument onto center stage of conservative

thinking about pneumatology.

            13. John F. Walvoord, The Holy Spirit: A Comprehensive Study of the Person and Work

of the Holy Spirit (Wheaton, Ill.: Van Kampen, 1954), 7 (italics added). This is part of his

third argument for the personality of the Spirit called "Use of Personal Pronouns Affirms

Personality" (p. 6).

            14. Millard J. Erickson, Christian Theology (Grand Rapids: Baker, 1985), 3.859-60

(italics added).

 

102                  Bulletin for Biblical Research 13.1

 

Dabney, Smeaton, Kim, Conner, Berkhof, Chafer, Thiessen, Pache,

Pentecost, Ryrie, Green, Williams, Packer, Sproul, Grudem, Fergu-

son, Reymond, and Congar make similar claims.¹⁵ Thus, the argu-

ment from natural gender often plays a large role in theologians'

defense of the Spirit's personality. An examination of these texts is

therefore in order.

 

Masculine Demonstrative Pronoun

 

Three passages in the Upper Room Discourse seem to speak of the

Spirit in masculine terms. They are as follows (the key terms are in

italics):

            John 14:26       o( de_ para/klhtoj, to_ pneu=ma to_ a!gion, o$ pe/myei o( path_r

                                    e)n tw|= o)no/mati/ mou, e)kei=noj u(ma=j dida/cei pa/nta kai_ u(po-

                                    mnh/sei u(ma=j pa/nta a$ ei}pon u(mi=n  [e)gw/]

                                    The Holy Spirit . . . he . . .

John 15:26                   #Otan e!lqh| o( para/klhtoj o$n e)gw_ pe/myw u(mi=n para_ tou=

                                    patro/j, to_ pneu=ma th=j a)lhqei/aj o$ para_ tou= patro_j e_k-

                                    poreu/etai, e)kei=noj marturh/sei peri_ e)mou=.

                                    The Spirit . . . he . . .

 

            15. R. L. Dabney, Lectures in Systematic Theology (originally published in 1878; re-

print ed., Grand Rapids: Zondervan, 1976), 195; George Smeaton, The Doctrine of the

Holy Spirit (2d ed.; Edinburgh: T. & T. Clark, 1889), 107; Seung Lak Kim, Pneumatology,

or the Doctrine of the Holy Spirit (Th.D. disseration, Dallas Theological Seminary, 1931),

37; Walter Thomas Conner, The Work of the Holy Spirit: A Treatment of the Biblical Doc-

trine of the Divine Spirit (Nashville: Broadman, 1940), 177; L. Berkhof, Systematic The-

ology (4th rev. and enlarged ed.; Grand Rapids: Eerdmans, 1941), 96; L. S. Chafer, Major

Bible Themes: 52 Vital Doctrines of the Scripture Simplified and Explained (rev. J. F. Wal-

voord; Grand Rapids: Zondervan, 1974), 89; H. C. Thiessen, Introductory Lectures in

Systematic Theology (Grand Rapids: Eerdmans, 1949), 144; Rene Pache, The Person and

Work of the Holy Spirit (Chicago: Moody, 1954), 13; J. Dwight Pentecost, The Divine Com-

forter: The Person and Work of the Holy Spirit (Westwood, N.J.: Revell, 1963), 12-13;

Charles Caldwell Ryrie, The Holy Spirit (Chicago: Moody, 1965), 14-15; Michael Green,

I Believe in the Holy Spirit (Grand Rapids: Eerdmans, 1975), 43 ("John breaks all the

rules of Greek by referring to the Spirit [a neuter noun in Greek] by the masculine pro-

noun"); John Williams, The Holy Spirit, Lord and Life-Giver: A Biblical Introduction to the

Doctrine of the Holy Spirit (Neptune, N.J.: Loizeaux, 1980), 25; J. I. Packer, Keep in Step

with the Spirit (Old Tappan, N.J.: Revell, 1984), 61; R. C. Sproul, The Mystery of the Holy

Spirit (Wheaton, Ill.: Tyndale, 1990), 17-18; Wayne Grudem, Systematic Theology: An In-

troduction to Biblical Doctrine (Grand Rapids: Zondervan, 1994), 232; Sinclair B. Fergu-

son, The Holy Spirit (Downers Grove, Ill.: InterVarsity, 1996), 31; Robert L. Reymond,

A New Systematic Theology of the Christian Faith (Nashville: Thomas Nelson, 1998), 314;

Yves Congar, I Believe in the Holy Spirit, vol. 1: The Holy Spirit in the 'Economy': Revelation

and Experience of the Spirit (New York: Crossroad, 1999), 57.

 

          WALLACE: Greek Grammar and the Holy Spirit                  103

 

John 16:13-14 o#tan de_ e!lqh| e)kei=noj, to_ pneu=ma th=j a)lhqei/aj, o(dhgh/sei

                                    u(ma=j e)n th|= a)lhqei/a| pa/sh|: ou) ga)r lalh/sei a)f  ) e(autou=,

                                    a)ll  ) o#sa a)kou/sei lalh/sei kai_ ta_ e)rxo/mena a)naggelei=

                                    u(mi=n. e)kei=noj e)me_ doca/sei, o!ti e)k tou= e)mou= lh/myetai kai_

                                    a)naggelei= u9mi=n.

                                    Whenever he comes, the Spirit of truth . . . he

 

            Several NT scholars have endorsed the view that the personality

of the Spirit is grammaticized in these texts. We have already noted

certain theologians. Among commentators on John, Lange, Godet,

Mortimer, Westcott, Bernard, Lenski, Hendricksen, Barrett, Behler,

Sanders and Mastin, Brown, Morris, Lindars, Newman and Nida,

Carson, and Beasley-Murray all use the grammatical argument in

one or more of these passages as evidence of the Spirit's personality.¹⁶

Indeed, this line of reasoning seems to be found more frequently and

more recently in exegetical literature than in theological literature.¹⁷

 

            16. J. P. Lange, Commentary on the Holy Scriptures—Critical, Doctrinal and Homilet-

ical: John (Grand Rapids: Zondervan, [1950] [translation of 1871 German work]), 469;

L. Godet, Commentary on the Gospel of John (New York: Funk & Wagnalls, 1893), 2.287;

A. G. Mortimer, The Last Discourses of Our Lord (New York: Thomas Whittaker, 1905),

226; B. E Westcott, The Gospel according to St. John (London: John Murray, 1908), 2.183;

J. H. Bernard, A Critical and Exegetical Commentary on the Gospel according to St. John

(ICC; Edinburgh: T. & T. Clark, 1928), 2.500; R. C. H. Lenski, The Interpretation of St.

John's Gospel (Columbus, Ohio: Lutheran Book Concern, 1942), 1013-14, 1090; W. Hen-

dricksen, Exposition of the Gospel according to John (Grand Rapids: Baker, 1954), 2.328;

C. K. Barrett, The Gospel according to St John: An Introduction with Commentary and Notes

on the Greek Text (London; SPCK, 1955), 402; G.-M. Behler, The Last Discourse of Jesus

(Baltimore: Helicon, 1965), 118-19; J. N. Sanders and B. A. Mastin, A Commentary on the

Gospel according to St. John (New York: Harper & Row, 1968), 329, 345; R. E. Brown, The

Gospel according to John XIII—XXI (AB 29a; New York: Doubleday, 1970), 639, 650, 689;

L. Morris, The Gospel according to John (NICNT; Grand Rapids: Eerdmans, 1971), 656

n. 70, 683-84 n. 63, 699 n. 26; B. Lindars, The Gospel of John (Greenwood, S.C.: Attic,

1972), 504 ("The pronoun is masculine, agreeing with the implicit ho Parakletos,

whereas Spirit is neuter, placed in apposition to it. Thus the personal character of the

Spirit is maintained." Curiously, Lindars makes the correct grammatical observation

here, but draws precisely the wrong conclusion from it); B. M. Newman and E. A.

Nida, A Translator's Handbook on the Gospel of John (New York: United Bible Societies,

1980), 497; D. A. Carson, The Gospel according to John (PNTC; Grand Rapids: Eerdmans,

1991), 510; G. R. Beasley-Murray, John (2d ed.; WBC 36; Nashville: Nelson, 1999), 261.

            17. The reason for this may be that theologies are increasingly getting away from

detailed (or exegetical) interaction with scripture. Among the works that may be char-

acterized as more theological than exegetical, Buswell, Montague, Moltmann, Pannen-

berg, Rahner, Hanson, Oden, Garrett, and Bloesch make no mention of Greek grammar

in support of the Holy Spirit's personality, even though all these authors seem to em-

brace it. To be fair, it is possible that one or more of these writers disagree with the phil-

ological argument and do not use it for that reason.

 

104               Bulletin for Biblical Research 13.1

 

            As well, a few specialized studies make similar claims. George

Eldon Ladd's Theology of the New Testament is representative: "where

pronouns that have pneuma for their immediate antecedent are found

in the masculine, we can only conclude that the personality of the

Spirit is meant to be suggested."¹⁸ After affirming this grammatical

phenomenon in John 14:26 and 15:26, Ladd then says, "The language

is even more vivid in 16:13: 'When the Spirit of truth comes, he (ekei-

nos) will guide you into all truth.' Here the neuter pneuma stands in

direct connection with the pronoun, but the masculine form rather

than the 'normal' neuter is employed. From this evidence we must

conclude that the Spirit is viewed as a personality."¹⁹

            It is not only exegetes and theologians who view these texts in

this way; one or two grammarians also consider them a s evidence of

the Spirit's personality. For example, Robertson argued that in John

16:13 the evangelist "is insisting on the personality of the Holy Spirit,

when the grammatical gender so easily called for e)kei=no."²⁰ More re-

cently, Young has also affirmed the philological argument in these

texts.²¹

            There is thus a large company of scholars who view the Upper

Room Discourse as affording syntactical evidence for the Spirit's per-

sonality. This august body has collectively argued that the masculine

pronoun is unusual in these verses, and that it can only be explained

by natural gender. Thus, if a masculine noun can be found in these

texts that can reasonably be considered as the antecedent to the pro-

noun, then these verses ought to be excised from the standard Trin-

itarian arsenal.

            The first two passages, John 14:26 and 15:26, can be handled to-

gether. In both of them, pneu=ma is appositional to a masculine noun,

rather than the subject of the verb. The gender of e)kei=noj thus has

nothing to do with the natural gender of pneu=ma. The antecedent of

e)kei=noj, in each case, is para/klhtoj, not pneu=ma.

 

            18. G. E. Ladd, A Theology of the New Testament (Grand Rapids: Eerdmans, 1974),

295. Cf. also Stevens, Johannine Theology, 196; H. B. Swete, The Holy Spirit in the New Tes-

tament: A Study of Primitive Christian Teaching (London: Macmillan, 1909), 292; K. H.

Schelkle, Theology of the New Testament, vol. 2: Salvation History-Revelation (Collegeville,

Minn.: Liturgical, 1976), 235.

            19. Ibid. The edition revised by Donald Hagner 19 years later (G. E. Ladd, A The-

ology of the New Testament [rev. D. A. Hagner; Grand Rapids: Eerdmans, 19931, 331) has

altered nothing in this paragraph. Later references are to this edition of Ladd.

            20. Robertson, Grammar, 709.

            21. Richard A. Young, Intermediate New Testament Greek: A Linguistic and Exegetical

Approach (Nashville: Broadman & Holman, 1994), 78: "the masculine pronoun e)kei=noj

is used in John 14:26 and 16:13-14 to refer to the neuter noun pneu=ma to emphasize the

personality of the Holy Spirit." Cf. also Robert Hanna, A Grammatical Aid to the Greek

New Testament (Grand Rapids: Baker, 1983), 178.

 

                 WALLACE: Greek Grammar and the Holy Spirit             105

 

 

 

                        Fig. 1. Alternate diagrams of John 14:26.

 

            This can best be seen if the texts are diagrammed.²² John 14:26 can

be diagrammed in one of two ways, depending on whether one re-

gards para/klhtoj as a nominativus pendens (see fig. 1A) or e)kei=noj as a

pleonastic pronoun (see fig. 1B).²³ John 15:26 can also be diagrammed

 

            22. The method of diagramming I am using is that of J. D. Grassmick, Principles

and Practice of Greek Exegesis (Dallas: Dallas Seminary, 1974).

            23. Mayes, Pronominal Referents, 28, takes the first approach, while the second ap-

proach is mine. See D. B. Wallace, Greek Grammar beyond the Basics: An Exegetical Syntax

of the New Testament (Grand Rapids: Zondervan, 1996), 329-30 (discussion of pleonastic

pronouns), 51-53. There is much overlap between these two classifications; the basic dif-

ference I see is that the nominativus pendens is the logical but not grammatical subject of

the sentence, for it is picked up by a pronoun in an oblique case. Since e)kei=noj is also

nominative, I would regard the construction to fall under pleonasm. But there is no real

objection to seeing nominativus pendens followed by a pronoun or participle in the nom-

inative. Either way, the idiom is most likely semitic. Cf. B. K. Waltke and M. O'Connor,

An Introduction to Biblical Hebrew Syntax (Winona Lake, Ind.: Eisenbrauns, 1990), 76-77

 

106                   Bulletin for Biblical Research 13.1

 

 

 

                    Fig. 2. Alternate diagrams of John 15:26.

 

in two different ways (see fig. 2).²⁴ With either diagram for these two

verses, it should be evident that the masculine demonstrative pro-

noun, e)kei=noj, stands in relation to o( para/klhtoj, not to to_ pneu=ma. In

____________________________________________________________________________

(§4.7), 128-29 (§8.3a); and Gen 3:12, C(h-Nm yl-hntn )wh ydm( httn r#) h#)h. (The

Semitic nature of the construction in John 14:26 is disputed by E. C. Colwell, The Greek of

the Fourth Gospel: A Study of Its Aramaisms in the Light of Hellenistic Greek [Chicago: Uni-

versity Press, 1931], 37-40.) John uses e)kei=noj 75 times (more than any other NT book),

52 of which are in the nominative case; 48 of the nominative uses are masculine. Ex-

cluding John 14:26 and 15:26 from the discussion, of the 50 remaining verses, the pro-

noun is pleonastic 11 times (John 1:18, 33; 5:11, 37; 6:57; 9:37; 10:1; 12:48; 14:12, 21;

17:24)—or 22% of the time; in the remaining 39 instances, it is syntactically unneces-

sary in virtually every instance (with possible exceptions in 7:11; 9:12; 18:15; 21:7, 23).

Thus, a known technique of the evangelist's is to employ e)kei=noj in a resumptive or

redundant fashion.

            24. The first diagram of John 15:26 is that of Mayes, Pronominal Referents, 31; the

second is mine.

 

               WALLACE: Greek Grammar and the Holy Spirit                  107

 

14:26, the noun clause—"the Holy Spirit whom the Father will send in

my name"—is in apposition to  o( para/klhtoj. How do we know that to_

pneu=ma is the appositive rather than o( para/klhtoj? Because it follows

o( para/klhtoj.²⁵ Appositives function routinely in a clarifying capacity

and thus naturally follow the substantive they are clarifying. The ap-

positional clause here can therefore be regarded as parenthetical:

"The Counselor (the Holy Spirit whom [o#] the Father will send in my

name) will teach you all things. . . ." Furthermore, appositional clauses

can normally be removed from a sentence without destroying the

structure of the sentence. In this case, the verse makes good sense as

follows: "The Counselor will teach you all things and will remind you

all that I told you." The rules of concord actually expect e)kei=noj rather

than e)kei=no, since the true antecedent is para/klhtoj. Thus, this verse

should be omitted from the roster of philological proofs of the Spirit's

personality.²⁶

            In 15:26, the situation is similar: the appositional clause headed

by to_ pneu=ma is parenthetical: "Whenever the Counselor comes (the

Spirit of truth who is coming from the Father), he will testify con-

cerning me." This appositional clause could be removed without

affecting the structure of the sentence: "Whenever the Counselor

comes, he (e)kei=noj) will testify concerning me." Although Morris ar-

gues that pneu=ma is the antecedent of e)kei=noj, based on proximity,²⁷

this is hardly an adequate basis, both because o( para/klhtoj agrees in

gender with e)kei=noj and because pneu=ma is appositional rather than

being the subject of the sentence. As Mayes argues, "That a referent

which is not in concord, but a few words nearer in the text, should be

chosen over a noun which agrees strictly and gives just as good sense

is nearly indefensible. Pronominal referents by no means have to be

the nearest noun. . . . It is difficult to avoid the suspicion that the-

ology has unduly influenced (perhaps unconsciously) the grammat-

ical analysis of this verse (as well as the others involved)."²⁸ If we

 

            25. Robertson, Grammar, 399: "Sometimes the word in apposition precedes the

other, though not usually." E. A. Abbott (Johannine Grammar [London: Adam and

Charles Black, 1906]) defines apposition as "a method of expressing the phrase 'that is

to say' without writing it, by 'apposing' a second word with a case-ending to the first

word with the same case-ending . . ." (p. 36 [§1928]). What should be noticed in such

a standard definition is that the appositive is the second word.

            26. It is rightfully so omitted by Robertson, Grammar, 709: "In 14:26 . . . the rela-

tive o# follows the grammatical gender of pneu=ma.   )Ekei=noj, however, skips over pneu=ma

and reverts to the gender of para/klhtoj."

            27. Leon Morris, The Gospel according to John (rev. ed.; NICNT; Grand Rapids:

Eerdmans, 1995), 606 n. 64: "The masculine e)kei=noj is noteworthy, for to_ pneu=ma.

is nearer than is Para/klhtoj." (The wording here is slightly stronger than the first edi-

tion; all other references to Morris's commentary on John are to the first edition.)

            28. Mayes, Pronominal Referents, 27.

 

108                      Bulletin for Biblical Research 13.1

 

applied the proximity principle in John 6:71, the result would be that

Jesus, not Judas, was the Lord's betrayer (ou{toj ga_r e!mellen paradi-

do/nai au)to/n, ei{j e)k tw=n dw/deka ["for he was about to betray him, one of

the twelve"])!²⁹ Further, the reason for the masculine pleonastic pro-

noun is that it is resumptive, and as such it is intended to reach back

to the masculine noun, para/klhtoj. Indeed, one of the major uses of

icci.voc in John is to refer back past the immediately preceding word,

phrase, or clause to the true antecedent.³⁰

            These two verses are similar to Col 3:4: o#tan o( Xristo_j fanerwqh|=

h( zwh_ u(mw=n, to/te kai_ u(mei=j su_n au)tw|= fanerwqh/sesqe ("When Christ,

your life, is revealed, then you also will be revealed with him"). Al-

though zwh/ is closer to the masculine pronoun au)tw|=), zwh/ is in appo-

sition to o( Xristo/j; there is no need to appeal to constructio ad sensum

here, as the grammatical antecedent of au)tw= is obviously o( Xristo/j.³¹

If this text is unremarkable syntactically yet is parallel to John 14:26

and 15:26, can we legitimately get theological mileage out of the

grammar of the Upper Room Discourse?³²

            Thus, contrary to the supposition that the proximity of pneu=ma to

e)kei=noj in John 14:26 and 15:26 demonstrates the Spirit's personality,

because the pneu=ma is appositional, it becomes irrelevant to the gender

of the pronoun. Had the evangelist wanted to show the Spirit's per-

sonality, he would in fact have written something like   #Otan e!lqh| to_

pneu=ma th=j a)lhqei/aj, o( para/klhtoj, e)kei=noj marturh/sei peri_ e)mou=. The

fact that pneu=ma and not para/klhtoj is the appositive renders the

philological argument in these two texts void.³³

            John 16:13-14, on the surface, seems to make out a better case

than the other two passages of the Upper Room Discourse. Indeed, it

is the major NT prooftext for the grammaticization of the Spirit's

 

            29. Cf. also John 7:45.

            30. Cf. John 1:18, 33; 2:21; 3:30; 7:45; 10:1; 13:25, 30.

            31. Cf. also Ps 64:6-7 (LXX): e)pa/kouson h(mw=n o( qeo_j o( swth_r h(mw=n, h( e)lpi_j

pa/ntwn tw=n pera/twn th=j gh=j, . . . e(toima/zwn o!rh e)n th|= i)sxu/i au)tou= ("Heed our God, our

Savior, the hope of all the ends of the earth, who establishes the mountains by his strength").

Even though e)lpi/j is closer to the masculine participle e(toima/zwn, it is in an apposi-

tional phrase; thus, constructio ad sensum does not need to be appealed to, since the

grammatical antecedent is obviously o( qeo/j. This is common enough: see Pss 17:3; 27:8;

Eph 5:23; Phil 2:15; 4:1.

            32. See also 1 Cor 4:17: although te/knon a)gaphto_n kai_ pisto/n is closer to o#j²,

Timo/qeon is the obvious antecedent of the relative pronoun.

            33. Mayes (Pronominal Referents, 31) comments: "The most obvious fact which pre-

sents itself through this diagram is that the chief assertion of the verse consists of two

clauses—one independent and one dependent—of which the grammatical subjects are

e)kei=noj and o( para/klhtoj. All the rest of the material simply describes or qualifies (o(

para/klhtoj, and could be omitted with no damage to the sense"; and (p. 32) "No con-

structio ad sensum exists in this verse. There are three pronouns (o#n, o#, e)kei=noj), all of

which agree with their referent—two with para/klhtoj and one with pneu=ma."

                    WALLACE: Greek Grammar and the Holy Spirit                109

 

personality.³⁴ Robertson, for example, argues that this passage is "a

more striking example" than John 14:26 because "one has to go back

six lines to e)kei=noj again and seven lines to para/klhtoj."³⁵

            In John 16:13-14 the immediate context is deceptive: o#tan de_ e!lqh|

e)kei=noj, to_ pneu=ma th=j a)lhqei/aj, o(dhgh/sei u(ma=j e)n th|= a)lhqei/a| pa/sh|:

e)kei=noj e)me_ doca/sei. ("whenever he comes—the Spirit of truth—he will

guide you in all truth . . . he will glorify me"). This text does not need

to be diagrammed, because it reveals essentially the same features as

the previous two passages. The difference here is that e)kei=noj in v. 13

is the subject, rather than para/klhtoj. (to_ pneu=ma is, once again, in

para/klhtoj is so far removed (explicitly mentioned in v. 7), the

apposition to the subject.³⁶) The philological argument is that, since

e)kei=noj is more naturally associated with the nearer noun, pneu=ma.

And since pneu=ma is neuter, this indicates that the evangelist thought

of the Spirit in personal terms. The not-so-subtle assumption is that

para/klhtoj is simply too far removed to have an impact on the gen-

der of the pronoun, and that therefore the only logical explanation

for the masculine gender is the natural gender of pneu=ma that follows.

            Is this really the best way to handle the gender of e)kei=noj? Two

lacunae in the discussion (for either view) are a tracing of the argu-

ment of vv. 7-13, and true grammatical parallels.³⁷ Both of these sup-

 

            34. Mayes notes (Pronominal Referents, 34): "These verses contain the primary evi-

dence for the grammatical argument for the personality of the Holy Spirit. Almost all

commentators and theologians who discuss the argument cite this passage. It is, in a

sense, the sine qua non of the argument."

            35. Robertson, Grammar, 709. For similar arguments, see Morris, John, 699 n. 26;

Lindars, John, 504; Ladd, Theology of the New Testament, 331; Erickson, Christian The-

ology, 3.859-60.

            36. Although translations of v. 13 such as that of the NRSV and REB may be mis-

leading as to what the subject of the sentence is ("When the Spirit of truth comes, he

will guide you . . ."), their objective is not to be a handbook for Greek students.

            37. An exception of this twofold lacuna is the work of Mayes, Pronominal Referents.

His treatment of parallels wil be discussed below; Mayes’s discussion of the flow of

argument is worth quoting at length here (p. 35):

 

            It is necessary to begin back in verse seven. There the Spirit is introduced as

            the para/klhtoj and becomes the subject of an extended discussion.  Au)to/n in

            verse seven refers back to para/klhtoj, as does e)kei=noj in verse eight. Then

            verses nine through eleven explain the work of the para/klhtoj (with respect

            to the world) which (work) was introduced in verse eight. Notice the depen-

            dency of verses nine through eleven on verse eight, as attested by the incom-

            plete sentences in the former. Verse twelve sets the stage for another

            statement about the work of the para/klhtoj—this time with respect to be-

            lievers.   )Ekei=noj is used in both verses thirteen and fourteen, probably with

            the same reference. On the basis of this sequence, then, it is this writer's con-

            tention that o( para/klhtoj is introduced in 16:7 as the subject of the passage

 

110                    Bulletin for Biblical Research 13.1

 

            First, regarding the flow of argument, it should be noted at the

outset that, although the para/klhtoj is introduced in v. 7 and is not

mentioned again by name, this Counselor never really disappears.

The intervening material (16:8-11) keeps the para/klhtoj ever before

the reader in a way that is impossible to miss, since vv. 8-11 consti-

tute one sentence in Greek, with e)kei=noj (v. 8) as the lone subject. The

ministry of the para/klhtoj is first described in terms of a threefold

peri/-phrase ("when he [e)kei/noj] comes, he will prove the world

wrong concerning [peri/] sin, righteousness, and judgment." This pro-

grammatic statement is followed by vv. 9-11, each of which is a prep-

ositional phrase linked together by the correlative conjunctions me/n

. . . de/ . . . de/. Yet, as soon as v. 12 disrupts the flow of thought ("I have

many more things to tell you, but you cannot bear them now"), the

Paraclete is immediately brought back into view by the resumptive

e)kei=noj, followed by his identification as to_ pneu=ma th=j a)lhqei/aj. Thus,

in spite of the distance between para/klhtoj in v. 7 and e)kei=noj in v. 13,

since the para/klhtoj never really fades from view throughout the

discourse, the masculine gender of e)kei=noj can easily be accounted

for on grounds other than the Spirit's personality.

            Second, are there other parallels to this text—passages in which

there is great distance between a pronoun and its antecedent? Mayes

notes one such parallel: "An example of a significant separation be-

tween pronoun and referent is found in Mark 14, where the pronoun

au#th is used in verse nine, and its referent is in verse three (gunh/)!

True, there are three intervening instances of the demonstrative. But

even so, the nearest is in verse six, approximately six lines above au!th

in verse nine."³⁸ In Mark 14, "just as the woman never leaves the

spotlight in that story, so the para/klhtoj never fades from view in

this discourse."³⁹ In terms of word-count, the distance between the

au#th of v. 9 and its nearest antecedent is 55 words; by comparison, the

distance between the e)kei=noj of John 16:13 and its nearest antecedent

is 54 words. They are thus comparable.

            One might think that such sustained absence of the substantival

referent could only occur in the better writers.⁴⁰ But Mark and John

hardly belong to the upper echelons of hellenistic literary art! Indeed,

_______________________________________________________________________

            and remains the subject through 16:15. e)kei=noj would then refer to para/klht-

            toj in each instance (vv. 8, 13, 14)—simple agreement, the general rule.

            38. Mayes, Pronominal Referents, 17-18.

            39. ibid., 36.

            40. The reason one might think this is that hellenistic Greek tended toward

greater explicitness and toward removal of subtleties, in comparison with the Attic

dialect. Cf. M. Zerwick, Biblical Greek Illustrated by Examples (Rome: Pontifical Biblical

Institute, 1963), 161-62 (§§480-84); Wallace, Exegetical Syntax, 19-20.

 

              WALLACE: Greek Grammar and the Holy Spirit              111

 

one of the most remarkable examples of an absentee referent is found

in Mark 6:31-8:26. In the span of ninety verses, "Jesus" is not men-

tioned once. Nor is any other identifying noun that refers to him. In-

stead, he is kept in view largely by pronouns. Yet, even here, the

pronouns are relatively sparse: they appear in only 29 of the 90

verses,⁴¹ or approximately once every 40 words.⁴² The point is that

referential distancing is not out of the ordinary for pronoun usage—

even in the less-refined writers.

            In sum, in John 16:13 the e)kei=noj is best explained as reaching

back to v. 7, where para/klhtoj is mentioned. Thus, since para/klhtoj

is masculine, so is the pronoun. Although one might argue that the

Spirit's personality is in view in the Upper Room Discourse, the view

must be based on the nature of a para/klhtoj and the things said

about the Counselor, not on any alleged grammatical subtleties. The

fact is that, in all of John's Gospel, the only time a masculine pronoun

is used concerning the pneu=ma is in relation to o( para/klhtoj. This sug-

gests that the philological argument in John 14-16 may be a case of

petitio principii.

            Before we look at the next pronominal proof text, a word should

be said about a more subtle argument for the Spirit's personality in

the Upper Room Discourse. Although he is somewhat persuaded by

the masculine pronouns, Swete also mentions the gender of para/klh-

toj: "Yet the choice of o( para/klhtoj, where to_ para/klhton (pneu=ma)

might have been written, is significant."⁴³ Cook expands on this line

of reasoning:

 

            When used of the Holy Spirit, para/klhtoj is used as a substantive

            rather than as an adjective. As an adjective it would have no intrin-

            sic gender. As a substantive, however, it could be expected to be in

            the neuter gender to extend the sense of to_ pneu=ma (the Spirit) were

            it indeed true that the Spirit is an impersonal force or influence.

            However, [Jesus] . . . did express the fact of personality through

            para/klhtoj by putting it in the masculine gender. Thus, when this

            title (para/klhtoj) of the Holy Spirit is the antecedent of to_ pneu=ma to_

 

            41. In addition, the third-person singular suffixes of finite verbs and the singular

participle are found in 46 verses, though in 15 of these verses there is also a pronoun.

For similar Gospel texts in which neither   )Ihsou=j nor ku/rioj is explicitly men-

tioned, see Mark 3:8-5:5; Luke 14:4-16:2; 20:45-22:32; John 15:21-16:18.

            42. There are approximately 42 pronominal referents (including the article func-

tioning as a pronoun) to Jesus in 29 verses in this section, out of a total of nearly 1500

words. Some sections are quite lengthy without any pronominal referents to Jesus: 148

words in 7:5b-17a (though with much discourse material); 115 words in 7:17b-25a; 114

words in 7:32b-8:2b; 109 words in 8:12b-19b; 87 words in Mark 8:30b-35a; 67 words in

7:28b-32a; 54 words in Mark 7:2b-5a (with parenthetical material in between, similar

to John 16:12).

            43. Swete, Holy Spirit in the New Testament, 292 n. 1.

112                 Bulletin for Biblical Research 13.1

 

            a#gion, Christ repeatedly used the masculine gender; and when this

            title is referred to pronominally, He used the masculine form of the

            demonstrative pronoun, e)kei=noj.⁴⁴

 

In other words, if para/klhtoj is considered to be an adjective rather

than a noun, its gender is not fixed, and the choice of the masculine

form demonstrates the personality of the Spirit.

            What are we to make of this argument? First of all, the very

subtlety of the argument may speak more for its ingenuity than its

veracity. That the vast bulk of scholars who embrace a philological

defense of the Spirit's personality do not mention it may be quite tell-

ing. Second, although para/klhtoj could be etymologically described as

an adjective, in actual usage the masculine noun form is virtually

alone. An examination of all Greek literature from the fourth century

BC to the second century AD⁴⁵ reveals 61 instances of the second de-

clension stem paraklhto-/paraklhtw-.⁴⁶ In all of them, as far as I could

tell from a cursory examination, the gender was masculine every

time. And if it occurs only as a masculine, one has to wonder whether

it was truly functioning as an adjective. If this etymological adjective

had thus become virtually fixed as a masculine noun for a few hun-

dred years prior to the writing of the Fourth Gospel, one has to won-

der whether the evangelist truly had any real gender option with this

term.⁴⁷ Not only this, but as an adjective the word took on a passive

nuance (e.g., "summoned"), while as a noun it was active.⁴⁸ Whether

it ever occurs as a neuter substantival adjective prior to the fourth

century AD is, in fact, doubtful.⁴⁹ The question, thus, is how a helle-

nistic Greek reader would understand to_ para/klhton in John 14-16.

Since the passive meaning of the adjective is inappropriate in this

context, and since the neuter substantival adjective is apparently un-

 

            44. W. Robert Cook, The Theology of John (Chicago: Moody, 1979), 62-63.

            45. As recorded in the database of the Thesaurus Linguae Graecae, E disk (the most

recent version, released in 2000).

            46. The search also uncovered two instances of paraklh/twr, but since this in-

volves a different lexeme these two examples were omitted from the count.

            47. Cook speaks of "the extensive (and, in the NT, exclusive) use of para/klhtoj as

a masculine substantive from the fourth century B.C. on" as a potential problem for his

view (Theology of John, 63 n. 43). LSJ give only two examples in which the word is used

as an adjective, once in Dio Cassius 46.20 (2d/3d century AD; as a masculine plural with

dou=loi), and once in BGU 601.12 (2d century AD papyrus; here it is also masculine).

            48. Behm, TDNT 5.800-801.

            49. Behm notes the adjectival sense of "comforting" with reference to the Spirit in

Hippolytus, Haer. 8.19.1 (to_ para/klhton pneu=ma) and Mak. Hom. 6.6 (to_ pneu=ma to_

para/klhton) as "obviously a development of ecclesiastical speech on the basis of Jn.

14:16f.; 15:26" (TDNT 5.805 n. 38). The first instance of a substantival neuter adjective

is apparently found in Eusebius, Eccl. theol. 3.5.11-12 (to_ pneu=ma th=j a)lhqei/aj to_ para/-

klhton). Indeed, when this word is used as a true adjective in relation to the Spirit, is

it ever masculine?

 

            WALLACE: Greek Grammar and the Holy Spirit              113

 

attested before Eusebius, how could the evangelist have chosen the

neuter here?⁵⁰ This etymological argument is thus muted by actual

usage of this word.

 

Masculine Relative Pronoun

 

In Eph 1:14 the masculine relative pronoun is used in reference to the Spirit.

Ephesians 1:13-14 reads as follows, with the relevant terms in italic type:

 

            )En w|{ kai_ u(mei=j a)kou/santej to_n lo/gon th=j a)lhqei/aj, to_ eu)agge/lion th=j

            swthri/aj u(mw=n, e)n w|{ kai_ pisteu/santej e)sfragi/sqhte tw|= pneu/mati th=j

            e)paggeli/aj tw|= a(gi/w|,  14. o#j e)stin a)rrabw_n th=j klhronomi/aj h(mw=n, ei)j

            a)polu/trwsin th=j peripoih/sewj, ei)j e!painon th=j do/chj au)tou=.

 

 This text does not receive as much ink as the Johannine passages

with regard to the Spirit's personality. Nevertheless, a few scholars

see the relative pronoun grammaticizing the Spirit's personality

here. Among theologians, Berkhof, Dabney, Ryrie, Walvoord, and

Erickson may be cited.⁵¹ Among exegetes, one of the most extensive

 

            50. Besides these three texts, one other passage from the Upper Room Discourse

could possibly be used to offer grammatical support for the personality of the Spirit.

John 14:17 reads to_ pneu=ma th=j a)lhqei/aj, o$ o( ko/smoj ou_ du/natai labei=n o#ti ou) qewrei=

au_to_ ou)de_ ginw/skei: u(mei=j ginw/skete au)to/, o#ti par  ) u(mi=n me/nei kai_ e)n u(mi=n e!stai in NA²⁷.

However, in some witnesses both instances of the personal pronoun are in the masculine in-

stead of the neuter (au)to/j instead of au)to/). Among these witnesses are P^66* D* L 579 ()²

W Y can be added to the list in that they have the masculine pronoun in the second in-

stance). As well, D L* add a third masculine pronoun after ginw/skei. None of these vari-

ants is likely to be original, for they are both lacking in external and internal support.

(In particular, although P⁶⁶ is early the scribe was often sloppy in his copying habits; cf.

E. C. Colwell, "Method in Evaluating Scribal Habits: A Study of  P⁴⁵, P⁶⁶, P⁷⁵,” Studies

in Methodology in Textual Criticism of the New Testament [NTTS 9; Leiden: Brill, 1969],

106-24.) Nevertheless, even if original, this text would generally approximate John

14:26 and 15:26 in its structure, since the antecedent of such pronouns could easily be

construed as the para/klhton mentioned in v. 16. The sentence structure here, however,

is a bit more complicated than in the other two passages (the first personal pronoun is

in a causal clause, while the second is in a new sentence), affording a bit more ambi-

guity in the pronouns' referent. But this most likely is what created confusion for the

scribes: those who wrote the masculine pronoun probably took the antecedent to be

para/klhton (and the relative clause to be explanatory of the appositional noun pneu=ma)

while those who wrote the neuter pronoun regarded pneu/ma to be the antecedent. Fur-

ther, evidence that these scribes were not thinking of the personality of the Spirit but

were simply following normal grammatical conventions can be seen in their transcrib-

ing of the relative pronoun that immediately follows pneu=ma: it is neuter (6).

            51. Berkhof, Systematic Theology, 96; Dabney, Systematic Theology, 125; Ryrie, Holy

Spirit, 14-15; Walvoord, Holy Spirit, 7; Erickson, Christian Theology, 3.860. Mayes (Pro-

nominal Referents, 38) suggests that it is only the theologians who employ Eph 1:14: "On

this the theologians appear to have no support from scholarly commentaries." But even

though their language is more guarded than some of these divines, Barth's and Sim-

pson's assessments show that Mayes has overstated the case (see below).

114               Bulletin for Biblical Research 13.1

 

defenses is found in Markus Barth's magnum opus. He opines: "Who

is meant by the pronoun 'He' (hos)—Jesus Christ or the Spirit? If

Ephesians were written according to the rules of classic[al] Greek, the

pronoun would refer to Jesus Christ rather than to the Holy Spirit. For

the noun 'spirit' (pneuma) is in Greek (just as in English) neuter."⁵² He

concludes that the pronoun refers to the Spirit, adding that "Eph 1:14

may be a verse that shows in exemplary fashion how the formation of

a special grammar for church use began. In church and theological

language the Holy Spirit is often and with good reason denoted as a

person. The Spirit is respected as 'he' rather than as an 'it.'"⁵³

            Among grammarians, Chamberlain argues that the masculine

pronoun here "probably indicates that Paul was thinking of the Holy

Spirit as a person."⁵⁴

            Most scholars who enlist this passage are more guarded in their

assessment. Simpson, for example, suggests that, although the o#j "has

been taken as proof of the personality of the Spirit" by some exegetes,

"it might be explained on grammatical grounds as the result of attrac-

tion to the gender of the masculine a)rrabw/n."⁵⁵

            Best is even more cautious, for he suggests that if the masculine

relative pronoun was original, it was "attracted into the masculine

through a)rrabw/n." On the other hand, "If the neuter was original

then the masculine is an idiomatic or stylistic improvement, or an at-

tempt to treat the Spirit as personal."⁵⁶ In other words, if o#j is orig-

inal, it does nothing to demonstrate the Spirit's personality; but if

it was a scribal corruption, it might have been added by orthodox

scribes because of their belief in the Spirit's personality.

            Wood puts a different twist on things when he says that, whether

the pronoun is o# or o#j, "The personality of the Holy Spirit is not jeop-

ardized by either usage."⁵⁷ That may be so, but the Spirit's person-

ality is also not supported by either usage.

 

            52. Markus Barth, Ephesians 1-3 (AB 34; Garden City, N.Y.: Doubleday, 1974), 95.

            53. Ibid., 96. Barth nevertheless notes that the reason for the masculine o#j is its at-

traction to the gender of the predicate nominative, a)rrabw/n. He thus seems to have two

bases for the masculine pronoun, though they are not complementary. But if the at-

traction to a)rrabw/n sufficiently explains the gender of the pronoun, how does this help

demonstrate the Spirit's personality?

            54. W. D. Chamberlain, An Exegetical Grammar of the Greek New Testament (New

York: Macmillan, 1941), 49.

            55. E. K. Simpson, "Ephesians," in E. K. Simpson and E E Bruce, Commentary on

the Epistles to the Ephesians and the Colossians (NICNT; Grand Rapids: Eerdmans, 1957),

35 n. 24 (italics added).

            56. Ernest Best, Ephesians (ICC; Edinburgh: T. & T. Clark, 1998), 151 n. 71.

            57. A. Skevington Wood, "Ephesians," Expositor's Bible Commentary (ed. F. E. Gae-

belein; Grand Rapids: Zondervan, 1978), 11.28. Gordon D. Fee (God's Empowering Spirit:

The Holy Spirit in the Letters of Paul [Peabody, Mass.: Hendrickson, 1994], 668-69 n. 36)

makes a similar point.

 

             WALLACE: Greek Grammar and the Holy Spirit                 115

 

            There are two fundamental problems with the use of this verse

for the Spirit's personality. First, there is the textual problem. Instead

of o#j (found in ) D 33 M) several witnesses have o# (P⁴⁶ A B F G L

P 81 1739 al).⁵⁸ Externally, the neuter pronoun is supported by the

greater weight of evidence. Internally, although Barth sees no good

reason why scribes would change the neuter to the masculine,⁵⁹ the

masculine could well have been motivated by attraction to the gen-

der of the a)rrbw/n following.⁶⁰ The editors of the UBS text have

vacillated here between the masculine and neuter, with the neuter

pronoun getting the nod since the third edition. Thus, because of the

textual uncertainty of the very word in question, any argument for

the Spirit's personality on the basis of the grammar of Eph 1:14 is sus-

pect even before the grammatical evidence is examined.

            Second, there is a grammatical problem with this argument: con-

structio ad sensum is not the only thing that could account for the

masculine o#j. It can also be accounted for on the basis of attraction to

the predicate nominative, "according to a usual idiom."⁶¹ The reason

an author sometimes shifts the gender of a relative pronoun forward

to the predicate nominative is probably to put greater focus on the

predicate noun.⁶²

            Typical examples that are cited for this phenomenon include Mark

15:16 (th=j au)lh=j, o# e)stin praitw/rion ["the palace, that is the Praeto-

rium"]; Gal 3:16 (tw|= spe/rmati/ sou, o#j e)stin Xristo/j ["your seed, that is

Christ"]); Eph 6:17 (th_n ma/xairan tou= pneu/matoj, o# e)stin r(h=ma qeou= ["the

sword of the Spirit, which is the word of God"]); and 1 Tim 3:16 (to_ th=j

eu)sebei/aj musth/rion: o#j . . .  ["the mystery of godliness, who . . .]).⁶³

 

            58. Itala MSS b d are also cited on behalf of the neuter reading in Nestle-Aland²⁷

but should probably be omitted from tabulation since the genders are exactly the op-

posite of the Greek for the two key terms (spiritus is masculine, while pignus is neuter).

Thus, the gender attraction would run in the opposite direction of the Greek. Because

of this, the neuter pronoun found in these MSS is just as likely to be a translation of the

masculine Greek pronoun—especially since rules of attraction in Latin are generally

the same as they are in Greek (cf. B. L. Gildersleeve and G. Lodge, Gildersleeve's Latin

Grammar [3(1 ed.; New York: Macmillan, 18951, 395 [§6141; 149-50 N211.51).

            59. Barth (Ephesians 1-3, 96) simply declares: "An original neuter would hardly

have been later displaced by the masculine relative pronoun," with no evidence to

back up this claim. Ironically, he argues against his own view by his claim that the

masculine would have affirmed the Spirit's personality: if so, would not that be moti-

vation enough for some scribes to change the neuter to a masculine pronoun?

            60. So Bruce M. Metzger, A Textual Commentary on the Greek New Testament (2d ed.;

Stuttgart: Deutsche Bibelgesellschaft, 1994), 533. Cf. also Fee, Empowering, 668 n. 36.

            61. T. K. Abbott, A Critical and Exegetical Commentary on the Epistles to the Ephesians

and to the Colossians (ICC; Edinburgh: T. & T. Clark, 1897), 23.

            62. So G. B. Winer, A Treatise on the Grammar of New Testament Greek (trans. and rev.

by W. F. Moulton; 3d ed., rev.; Edinburgh: T. & T. Clark, 1882), 207.

            63. So Abbott, Ephesians, 23; Robertson, Grammar, 712-13; et al. Although 1 Tim 3:16

does not involve a predicate nominative, it can nevertheless be lumped in with these

 

116                 Bulletin for Biblical Research 13.1

 

            However, these passages might not affirm the point being made.

First, relative clauses that have the neuter construction o# e)stin may be

due to a common idiom that is equivalent to the Latin id est (= i.e.).

Thus, for example, in Heb 7:2 Melchizedek is called "the king of Sa-

lem, that is, the king of peace" (basileu_j Salh/m, o# e)stin basileu_j ei)rh/-

nhj)—even though basileu/j is masculine. In Mark 3:17, James and John

are called "Boanerges, that is, sons of thunder" (boanhrge/j, o# e)stin ui(oi_

bronth=j)—even though the nouns in each clause are masculine plural.

Many such constructions with o# e)stin should probably be deleted

from consideration because they may be due to the id est idiom rather

than the attraction-to-predicate idiom.⁶⁴ Second, Gal 3:16 may involve

constructio ad sensum due to identification of the spe/rma as Christ; thus,

natural gender rather than attraction to predicate could explain the

o#j.⁶⁵ And third, 1 Tim 3:16 most likely has an entirely different reason

for the masculine relative pronoun—namely, because it is probably an

embedded hymn fragment, there is no real antecedent.⁶⁶

            If these examples are illegitimate, are there better ones that dem-

onstrate the point of attraction to the predicate's gender? There may

not be many in the NT, but they occur frequently enough in helle-

nistic literature as a whole.⁶⁷ But even within the NT note, for ex-

_______________________________________________________________________

other passages because of certain similarities. For reasons discussed below, it should

be excluded from the pool of examples, however.

            64. See also Mark 12:42 (lepta_ du/o, o# e)stin kodra/nthj); Col 1:27 (tou= musthri/ou

tou=tou e)n toi=j e!qnesin, o# e)stin Xristo/j); and Col 3:14 (th_n a)ga/phn, o# e)stin su/ndesmoj

th=j teleio/thtoj). Not all such constructions can be disposed of, however. Rather, only those

that use the relative clause to clarify the sense or the referent of the previous noun as

an appositive can be rejected. However, in 2 Thess 3:17 we read,   (O a)spasmo_j th|= e)mh|=

xeiri_ Pau/lou, o# e)stin shmei=on e)n pa/sh|= e)pistolh|= ("This greeting is in my own hand,

Paul's, which is a sign in every epistle"). The relative clause does not clarify the sense

of the greeting but, rather, explains its purpose. As such, it apparently does not follow

the id est idiom and thus may be included in the predicate attraction examples. See also

Matt 13:31-32.

            65. Admittedly, there is only a slight difference between these two here: it is pre-

cisely at this point in the narrative that the seed is identified as Christ. Thus, the focus

of the passage naturally gravitates toward Xristo/j. But, since Paul knew where he was

going with the argument, the o#j could be considered anticipatory because of the natu-

ral referent he has in mind. Either way, this particular text is of no help in the argu-

ment against the grammaticization of the Spirit's personality in Eph 1:14 because it is

impossible to tell which of the two reasons Paul had in mind in using o#j in Gal 3:16,

or even if he consciously distinguished between the two.

            66. For discussion, see Wallace, Exegetical Syntax, 341-42.

            67. In the LXX, note Prov 12:11 (fre/nw=n o#j e)stin h(du/j). In the papyri, note POxy

1485.4, which employs a frequent idiom (sh/meron h#tij e)sti_n q ["today, which is the

ninth"]) of using a temporal adverb (such as sh/meron or au!rion) substantivally (fre-

quently such adverbs are arthrous), followed by the feminine pronoun iinc, whose

gender is due to the implied h(me/ra in the relative clause. (In the NT, see Matt 27:62.

Even though e)pau/rion takes the feminine article here, it is due to the attraction to the

 

              WALLACE: Greek Grammar and the Holy Spirit                        117

 

ample, 1 Tim 3:15 (oi!kw| qeou= . . . h#tij e)sti_n e)kklhsi/a qeou= zw=ntoj ["the

house of God, which is the church of the living God"]) and Rev 4:5

(lampa/dej . . . a# ei)sin ta_ e(pta_ pneu/mata tou= qeou=⁶⁸ [lamps . . . which are

the seven spirits of God"]). These illustrations could be multiplied.⁶⁹

The attraction-to-predicate idiom is thus common enough that, even

if the verse were textually stable, Eph 1:14 should still be removed

from the prooftext bin for the Spirit's personality.

            To sum up all the passages that involve masculine pronominal ref-

erents, we can lay out the evidence in table form:⁷⁰

 

Passage                       Pronoun          Referent                     Construction

John 14:26                 e)kei=noj        para/klhtoj          simple agreement

John 15:26                 e)kei=noj        para/klhtoj             simple agreement

                                    o#n                para/klhtoj             simple agreement

John 16:7-8               e)kei=noj        para/klhtoj             simple agreement

                                    au)to/n             para/klhtoj             simple agreement

John 16:13-14           e)kei=noj (2)    para/klhtoj             simple agreement

Eph 1:13-14               o#j (?)              tw|= pneu/mati          assimilation to

                                                                                                predicate

 

So far, it is evident that no grammatical construction can be unam-

biguously marshaled in defense of the Spirit's personality. But there

is one passage remaining.

 

Masculine Participle

 

            The final passage that is used in a philological defense of the

Spirit's personality is 1 John 5:7-8. The text reads as follows:

________________________________________________________________________

implied h(me/ra in the relative clause, since adverbs obviously do not have a set gender.)

In Classical Greek a similar phenomenon occurs with the attraction of the gender of

the demonstrative to that of the predicate (see Gildersleeve, Classical Greek, 1:58

[§127]). It should be noted that most NT grammars are not very helpful on this issue

since they lump in the id est relative pronouns with predicate-attraction relative pro-

nouns. But see examples noted below.

            68. Some MSS (1006 1841 M^k) have the feminine plural ai# here.

            69. See also Acts 16:12; 1 Cor 3:17; Eph 3:13 (unless here we should read h@ ti/j in-

stead of h#tij); Eph 6:2; Phil 1:28; Col 3:5; Rev 5:8 (although ) 046 1006 1841 2050 2344

and a few other mss have a# here). In Rev 5:6 several witnesses (1854 2050 2329 2344

2351 M^k) have a# ei)sin instead of oi# ei)sin, in agreement with the predicate nominative

(ta_ pneu/mata) against the antecedent (o)fqalmou/j), but this reading is most likely not

authentic.

            70. The following table is taken from Mayes, Pronominal Referents, 40, with some

modifications.

 

118                  Bulletin for Biblical Research 13.1

 

            7. o#ti tre=j ei)sin oi( marturou=ntej,  8. to_ pneu/ma kai_ to_ u#dwr kai_ to_ ai{ma,

            kai_ oi( trei=j ei)j to_ e#n ei)sin

            There are three who testify—the Spirit, the water, and the blood; and

            these three are in agreement.

 

In this passage the masculine participle marturou=ntej is followed by

three appositives—to_ pneu/ma, to_ u#dwr, and to_ ai{ma—all of which are

neuter in gender. The question is thus naturally raised, What is to

account for the masculine participle?

            Various interpretations have been put forth for the gender shift

here. One of the most common, however, is that of constructio ad sen-

sum—that the Holy Spirit is obviously a person, and thus the mascu-

line participle is used. Among commentators who embrace this view

are Westcott, Plummer, Smith, Hiebert, Burdick, Marshall, and Smal-

ley. I. Howard Marshall, in his NIC commentary, is representative:

 

            It is striking that although Spirit, water, and blood are all neuter

            nouns in Greek, they are introduced by a clause expressed in the

            masculine plural: trei=j ei)sin oi( marturou=ntej. . . . Here in 1 John he

            clearly regards the Spirit as personal, and this leads to the person-

            ification of the water and the blood.⁷¹

 

All of these commentators say essentially the same thing: the Spirit

is truly regarded as personal and the water and blood are merely

personified. Raymond Brown criticizes this view as follows: "Plum-

mer . . . presses the gender too far when he states, 'The masculine

points to the personality of the Spirit,' unless one wishes to claim the

personality of the water and blood as well. . ."⁷²

            In reality, Brown's critique is probably overdone. Greek gender

usage is such that mixed groups—which may include men, women,

and children—would employ the masculine gender to address them;

hence, the routine use of a)delfoi/ to greet congregations in the NT let-

ters, when both "brothers" and "sisters" is meant.⁷³ If a group had

one man and several children, or one man and several women, pre-

sumably the masculine would still be employed.⁷⁴ In 1 John 5:7, then,

 

            71. I. Howard Marshall, The Epistles of John (NICNT; Grand Rapids: Eerdmans,

1978), 237 n. 20.

            72. Raymond E. Brown, The Epistles of John (AB 30; Garden City, N.Y.: Doubleday,

1983), 581. Brown also notes that A. Greiff ("Die drei Zeugen in 1 Joh 5,7f.," TQ 114 [1933]:

465-80, esp. 477-78) does in fact see all three witnesses as personal: "he sees the water as

the baptized Christian, and the blood as the martyr!" (Brown, Epistles of John, 581).

            73. BDAG give several indisputable examples in hellenistic Greek in which a)delfoi/  

meant "brother and sister" or "brothers and sisters" (s.v. a)delfo/j, definition 1, p. 18). See

also Rom 16:3 where Prisca and Aquila are called collectively sunergoi/.

            74. 2 John 1 even goes beyond this: e)klekth|= kuri/a| kai_ toi=j te/knoij au)th=j,

ou#j ("to the elect lady and her children, whom"), for the feminine singular noun and neuter

plural are together picked up by the masculine pronoun! But if "lady" is a metaphor

for the church, the reason for the masculine pronoun is due to constructio ad sensum.

            WALLACE: Greek Grammar and the Holy Spirit                 119

 

if the Spirit and only the Spirit is viewed as a person, it would be

wholly appropriate to use the masculine trei=j ei)sin oi( marturou/ntej to

describe the witness of the Spirit, water, and blood.

            Nevertheless, is that the real reason for the masculine participle?

The fact that the previous verse speaks of the Spirit as a witness using

a neuter participle (to_ pneu=ma/ e)stin to_ marturou=n) suggests that it is

not the Spirit's personality that is driving the gender shift in v. 7. Fur-

ther, those who see the Spirit's personality in oi( marturou=ntej often do

so because they already saw such in the pronouns in John 14-16. (And

many have, naturally, written an earlier commentary on the Gospel

of John.) But if, as we have argued, the Gospel of John provides no

grammatical precedent for this interpretation, it is doubtful that per-

sonality is the reason for the gender shift here.

            What then is the catalyst for the change? Several suggestions have

been made, one of which will be mentioned here. It is possible that "It

is the personal character of the witnessing that is underlined by the

masculine numeral, as well as by the use of the pres. ptcp. ('those who

bear witness' rather than 'witnesses'): the three go on witnessing."⁷⁵

Therefore, to clarify that the witnesses were personal and thus valid,

the masculine participle was used. Taking this a step farther, it is pos-

sible that the masculine was used, almost subconsciously, because the

only legitimate witnesses in Jewish courts would be male.⁷⁶ Why,

then, was the masculine gender used only with reference to the three

witnesses and not to the Spirit in v. 6? Perhaps because as soon as the

number of witnesses shifted from singular to plural, the nature of

the witness shifted from impersonal (which was valid in a limited

sense)⁷⁷ to personal, and the Deuteronomic law of establishing the

truth of a testimony by two or three witnesses⁷⁸ thus came to the

 

            75. Brown, Epistles of John, 581.

            76. Josephus (Ant. 4.8.15 §219) says that women were disqualified because of their

inherent "vanity and rashness." See also m. Ketub. 1:6-9; Sipre Deut. §190; y. Sotiah. 6.4,

21a (where the testimony of a hundred women was worth no more than the testimony

of one man). Tal Ilan (Jewish Women in Greco-Roman Palestine: An Inquiry into Image and

Status [TSAJ 44; Tubingen: Mohr, 1995], 163-66) summarizes her research on the mat-

ter as follows: "We may conclude that the specific law disqualifying women as wit-

nesses was formulated as a general halakhic principle, just as in other matters such as

punishments, but that many exceptions arose from actual custom and practice. During

a normal trial in court, women's testimony was not sought out and was in fact avoided

whenever possible because 'no man wants his wife to degrade herself in court' (b. Ketub.

74b), but testimony which could not otherwise be obtained was by all means accepted"

(p. 165).

            77. Brown, Epistles of John, 581: "In Jewish tradition personal testimony can be

given impersonal witnesses, e.g., by a heap of stones (Gen 31:45-48), by heaven and

earth (Deut 31:28), by clouds and rain (Enoch 100:11)."

            78. Cf. Deut 17:6; 19:15. This multiple testimony to the truth is repeated in the NT

(cf. especially John 8:17; note also Matt 18:16; 2 Cor 13:1; 1 Tim 5:19; Heb 10:28).

 

120               Bulletin for Biblical Research 13.1

 

foreground.⁷⁹ The metaphor, rather than the Spirit's personality, is

thus driving the gender shift.

            Whatever the reason for the masculine participle in v. 7, it is ev-

ident that the grammaticization of the Spirit's personality is not the

only, nor even the most plausible, explanation. Since this text also

involves serious exegetical problems (i.e., a variety of reasons as to

why the masculine participle is used), it cannot be marshaled as un-

ambiguous syntactical proof of the Spirit's personality.

            In sum, none of the gender shift passages clearly helps establish

the personality of the Holy Spirit. In light of this, I would recommend

that an argument that appears to be a modern invention⁸⁰ be excised

from our theological textbooks.

 

                        PASSAGES INVOLVING AGENCY

 

As we said at the beginning of this paper, passages involving agency

are assumed to show the personality of the Spirit, but they are not

proof texts at all. These passages are used only because the person-

ality of the Spirit is assumed to be demonstrated on other grounds or

in other passages. All of these texts are of one sort: they involve the

expression (e)n) pneu/mati. Frequently in both theological and exeget-

ical literature, this expression is assumed to mean personal agency.

Two passages especially are of interest here, Gal 5:16 and 1 Cor

12:13. In Gal 5:16 Paul says, "walk by the Spirit (pneu/mati) and you

will not fulfill the lust of the flesh." In 1 Cor 12:13 he says, "by one

Spirit (e)n e(ni_ pneu/mati) we were all baptized into one body."

            Because of the already undue length of this paper and the major

focus on the masculine gender passages, we will not dwell here too

long. We begin with a survey of the grammar. Standard Greek gram-

mars note that the simple dative case is used for personal agency only

on rare occasions—and when it is so used, it is found with a perfect

passive verb.⁸¹ BDR, for example, cite only one possible instance

 

            79. This would not necessarily mean that the Spirit, by himself, was viewed as

impersonal, though this is possible. But it is just as likely that the personal and mas-

culine nature of the testimony was triggered in the author's mind by the plurality of

the witnesses that he introduced in v. 8.

            80. As far as I can tell, Patristic writers never use this argument either. Of course,

the Greeks do not usually comment on grammatical features of their own language,

and the Latin Fathers do not comment on the Greek! For what it is worth, Calvin does

not use this argument, nor does Gill. I have not yet traced the roots of the grammatical

argument, but my suspicion is that it began in the eighteenth or nineteenth century.

            81. BDR 154 (§191). Cf. also H. W. Smyth, Greek Grammar (rev. G. M. Messing; Cam-

bridge: Harvard University Press, 1956), 343-44 (§1488-94); Wallace, Exegetical Syntax,

163-66. Smyth notes that "the usual restriction of the dative to tenses of completed

 

               WALLACE: Greek Grammar and the Holy Spirit                 121

 

(Luke 24:15). But Jas 3:7 is also a likely candidate.⁸² The word e)n with

the dative for personal agency is just as rare, if not more so.⁸³ Suffice

it to say that every clear instance of this usage in the NT involves the

simple dative substantive with a perfect passive verb.⁸⁴ Thus, if one

wishes to argue that pneu/mati is used this way, clear examples with

other personal nouns must be brought forth to establish the usage. To

argue that pneu/mati\ is so used, even lacking association with a perfect

passive verb, is simply to beg the question. Applying this to 1 Cor

12:13, if we were to take the Spirit as the agent of the baptism into the

body of Christ this would inadvertently mask the fulfillment motif of

the apostle here. Mark 1:8 records John as saying, "I baptized you

with water, but he will baptize you with the Holy Spirit (bapti/sei u(ma=j

e)n pneu/mati a(gi/w|)."⁸⁵ The text of 1 Cor 12:13 repeats this prophecy, ex-

cept that it uses an aorist (e)bapti/sqhmen) instead of the future tense.

It is evidently meant to indicate the fulfillment of this prophecy in the

life of the church. If so, then Spirit baptism speaks of the divine ini-

tiative in salvation, rather than a second blessing later on, since "we

all were baptized . . . we all were made to drink of one Spirit." Ironi-

cally, those who see the Spirit as the agent of baptism in 1 Cor 12:13

inadvertently open the door to two Spirit baptisms—the initial one in

which the Spirit is the agent and a later one in which Christ baptizes

by means of the Spirit. But not only is it nearly impossible to read e)n

pneu/mati as indicating personal agency, but there is no linguistic dif-

ference between the prophecies about the Lord baptizing with the

Spirit and the statement in 1 Cor 12:13. All of the evidence points to

Paul consciously linking the Johannine prediction of Spirit baptism

with the ecclesiastical reality.

____________________________________________________________________

action seems to be due to the fact that the agent is represented as placed in the posi-

tion of viewing an already completed action in the light of its relation to himself"

(ibid., 343-44 [§1489]).

            82. This is disputed by BDR 154 (§191.3). See also John 18:15; Rom 14:18; 2 Pet 2:19;

and Jude 1 for other possible examples, all of which employ a perfect passive verb. In

the LXX, however, I have noticed Neh 13:26 (a)gapw/menoj tw|= qew|= h}n), where an imper-

fect periphrastic construction is used.

            83. Cf. Wallace, Exegetical Syntax, 373-74.

            84. The best candidate with the preposition is found in 1 Cor 6:2: e)n u(mi=n kri/netai

o( ko/smoj ("the world is to be judged by you"). But this is by no means certain. A. T.

Robertson and A. Plummer (The First Epistle of St Paul to the Corinthians [ICC; 2d ed.;

Edinburgh: T. & T. Clark, 1914], 112) suggest that it speaks of sphere/locality: "in your

court," "in your jurisdiction." So also BDR 178 (§219.1), noting parallels in profane and

Patristic literature.

            85. Cf. also Matt 3:11; Luke 3:16; John 1:33; Acts 1:5; 11:16, all repeating the same

Johannine saying (with bapti/zw e)n pneu/mati each time). In each of these texts, the for-

mula involves a futuristic Spirit baptism.

 

122                        Bulletin for Biblical Research 13.1

 

                                            CONCLUSION

 

Penultimate Conclusions

 

There is no text in the NT that clearly or even probably affirms the

personality of the Holy Spirit through the route of Greek grammar.

The basis for this doctrine must be on other grounds. This does not

mean that in the NT the Spirit is a thing, any more than in the OT

the Spirit (xwr—a feminine noun) is a female! Grammatical gender is

just that: grammatical. The conventions of language do not necessar-

ily correspond to reality.⁸⁶

 

Where Do We Go from Here?

 

One implication of these considerations is this: There is often a tacit

assumption by scholars that the Spirit's distinct personality was fully

recognized in the early apostolic period. Too often, such a viewpoint

is subconsciously filtered through Chalcedonian lenses.

            This certainly raises some questions that can be addressed here

only in part: We are not arguing that the distinct personality and

deity of the Spirit are foreign to the NT, but rather that there is

progressive revelation within the NT, just as there is between the

Testaments.

            Evangelical defenses of various doctrines occasionally are poorly

founded. We sometimes claim things to be true because we want

them to be true, without doing the exhaustive spadework needed to

support our conclusions. Regarding the personality of the Holy Spirit,

the quick leap to exploit Greek grammar in defense may actually

work against a carefully nuanced pneumatology. Taking our cue from

christology, we note that several biblical scholars working in that

field would argue for progressive development of the understanding

of the person and work of Christ. Not all would affirm that the apos-

tolic band embraced the deity of Christ shortly after the resurrection.

Some would argue that this understanding took years to develop.

            At the same time, there is evidence that christology developed

more quickly than pneumatology. Take, for example, the epistolary

salutations: virtually all of the corpus Paulinum offers grace and peace

"from God the Father and the Lord Jesus Christ" (only Colossians and

1 Thessalonians are excepted). This expression almost implicitly puts

Christ on the same level with God, giving evidence that the apostles

were going through a binitarian transformation of monotheism.⁸⁷ But

 

            86. Cf. Wallace, Exegetical Syntax, 10-11.

            87. Occasionally, the Spirit shows up in benedictions along with the Father and

Son (e.g., 2 Cor 13:13; 2 Thess 2:13), but the syntactical structure of such benedictions

 

          WALLACE: Greek Grammar and the Holy Spirit                 123

 

where is the Spirit? It is only in the Apocalypse that the salutation is

from Father, Son, and Spirit (if indeed this is what "seven spirits"

means in Rev 1:4⁸⁸).

            Further, when we look at Acts we notice that water baptism is ap-

parently never done in the "name of the Father, Son, and Holy Spirit";

it is done in Jesus' name alone (cf. Acts 2:38; 8:16; 10:48; 19:5; 22:16).

What is to account for this? Either the Trinitarian formula in the Great

Commission (Matt 28:19) was a later accretion added either by the

evangelist or, possibly, by some ancient scribe;⁸⁹ or, more likely, there

was a lack of understanding on the part of the apostles when Jesus

gave the commission.⁹⁰ That baptism was apparently not done in the

Father's name either suggests that the apostolic band was wholly con-

sumed with Christ or that the Trinitarian formula made little sense to

them. In other words, their initial understanding of the relation of

Father, Son, and Holy Spirit may well have been rather fuzzy.

            If this is how it was with Christ, whom the disciples had seen,

how much more would it be this way with the Holy Spirit, whom they

had not?

            To extend this analogy, the work in Jewish sources in relation to

christology shows that a second prong in a high pneumatology is per-

haps overstated as well. Specifically, many of the NT passages that are

adduced to show the Spirit's personality or deity find parallels in Philo,

intertestamental literature, the Dead Sea Scrolls, or even the OT.⁹¹ Un-

less we can distinguish the NT from these other ancient sources more

______________________________________________________________________

does not as strongly put the Spirit on the same level with the Father as the salutations

do the Son.

            88. For a good discussion of the problem of the seven spirits, see Swete, Holy Spirit

in the New Testament, 272-74. Swete adopts the view that the Holy Spirit is in view.

            89. That the original form of Matt 28:19 did not have the Trinitarian baptismal

formula was the conclusion of F. C. Conybeare, "The Eusebian Form of the Text of Mt.

28:19," ZNW 2 (1901): 275-88, based on a faulty reading of Eusebius's quotations of this

text. The shorter reading has also been accepted, on other grounds, by a few other

scholars. For discussion (and refutation), see B. J. Hubbard, The Matthean Redaction of

a Primitive Apostolic Commissioning: An Exegesis of Matthew 28:16-20 (SBLDS 19; Mis-

soula, Mont.: Scholars Press, 1974), 163-64, 167-75; and Jane Schaberg, The Father, the

Son and the Holy Spirit: The Triadic Phrase in Matthew 28:19b (SBLDS 61; Chico, Calif.:

Scholars Press, 1982), 27-29.

            90. I am not arguing for ipsissima verba here but simply that the theology of the

dominical material often displays a greater sophistication than can be found in the

evangelists' narratives. Indeed, when it comes to the personality and deity of the Holy

Spirit, far and away the most impressive texts are dominical sayings. This may suggest

that the evangelists treated the traditional sayings of Jesus relatively conservatively.

Thus, ironically, the most advanced theology of the NT might often be found at the

most primitive layer—preserved as it was by the evangelists, whether they clearly

grasped its meaning or not.

            91. See, for example, Judg 16:19-20 (here, Samson's strength "left him" in one

verse; then he understands this to mean that the Lord "left him" in the next [rws used

 

124                   Bulletin for Biblical Research 13.1

 

clearly, we would either have to say that Philo was a Trinitarian (!) or

that the NT authors were not.⁹² It is not enough to say either that the

Spirit is presented as personal⁹³ or that he is sometimes not distin-

guished from God (as in Acts 5:3-4). What also must be done is (1) a

clear demonstration that language about the Spirit's personality can-

not be due to figurative rhetoric or circumlocution of the divine

name,⁹⁴ and (2) that where he is viewed as personal he is also viewed

as deity, yet, (3) in those same texts, is seen as distinct from both Father

and Son. That such passages are few and far between may indicate

something of an emerging pneumatological understanding within the

_____________________________________________________________________

each time]; in the least, this kind of text should give us pause about using Acts 5:3-4

for equating the Holy Spirit with God without sufficient nuancing); 2 Sam 23:2; 1 Kgs

22:21-22 (here an evil spirit is referred to with the masculine pronoun au)to/n in the

LXX, the very kind of evidence that is lacking in the NT for the Holy Spirit [but cf.

Mark 9:20, 26 for the masculine participle used with an evil spirit; Luke 9:39-40 in P⁴⁵

(pneu=ma. . . . au)to/n)]; in Patristic literature, note 2 Chan. 20:4 [pneu=ma mh_ o@n di/kaion]); Job

33:14 (this text seems to use synonymous parallelism between "spirit" of YHWH and

"breath of God"—certainly offering no comfort for those who see the Trinity in the

OT!); Ps 139:7 (equates God's "spirit" with his "presence"); Ps 143:10 (cf. Acts 5:3-4!);

Isa 30:1; 40:13; 48:16; 63:10, 14; Ezek 2:2; 3:24; 11:5; Wis 1:5, 7-9; 7:12, 22, 23, 24, 27; 8:8;

9:11; 10:1; Odes Sol. 3:10; 36:1; T. Jud. 20:5; 1QS 3:13-4:26; Philo, Mos. 2:265.

            92. After tracing the development of the sense of spirit in the intertestamental lit-

erature (including the DSS), John Breck (Spirit of Truth: The Holy Spirit in Johannine Tra-

dition, vol. 1: The Origins of Johannine Pneumatology [Crestwood, N.Y.: St. Vladimir's

Seminary Press, 1991], 160) summarizes: "Ruach . . . gradually developed from a ca-

pricious inspirational dynamis or charismatic power in primitive Hebrew thought into

the indwelling bearer of the divine Word. Thus ruach became a virtual synonym for

Yahweh in His act of self-disclosure." With reference to Philo, he notes: "In the writ-

ings of Philo we find an impressive attempt to draw together the Greek pneuma and the

Hebrew ruach, but the synthesis remains incomplete because the two spirit concepts

are basically incompatible" (ibid.).

            93. Apart from the grammatical argument that has been addressed in this paper,

the NT speaks of the Holy Spirit in personal terms, especially as the subject or object

of personal verbs (e.g., teaching, grieving, blaspheming, etc.). Many theologians and

exegetes appeal to such texts as though they demonstrated the personality of the Spirit

without showing how similar phenomena in Jewish literature do not demonstrate this. For ex-

ample, in Sir 39:28, pneu/mata (which, in this context, means "winds") is personified,

with the masculine pronoun au)tou/j, following.

            94. On several occasions in Jewish literature, "spirit of" is really a circumlocution

for the simple noun in the genitive. Thus, xwr / pneu=ma takes on the gender of the geni-

tive noun. "The spirit of God/YHWH" in the OT is frequent enough, occasionally even

being used as a circumlocution for God himself (cf., e.g., Pss 139:7; 143:10). In Job 33:14

the spirit of YHWH is the same as the breath of God. In Wis 7:7 we read of pneu/ma sofi/aj  

("the spirit of wisdom"), but this is immediately picked up by the feminine pronoun in

the next verse: "I esteemed her more than sceptres and thrones; compared with her, I

held riches as nothing" (proe/krina au)th_n skh/ptrwn kai_ qro/nwn kai_ plou=ton ou)de_n h(gh-

sau/mhn e0n sugkri/sei au)th=j). Verses 8-11 have pneu=ma (v. 7) as the grammatical antecedent,

yet the feminine pronoun is used ten times! Does this make the pneu=ma feminine here?

 

             WALLACE: Greek Grammar and the Holy Spirit                125

 

NT itself.⁹⁵ If we rush to a Chalcedonian view of the NT, simply be-

cause we know that it's right, perhaps we will overlook some of the

theological development and therefore rich tapestry of NT thought.

            In sum, I have sought to demonstrate in this paper that the gram-

matical basis for the Holy Spirit's personality is lacking in the NT, yet

this is frequently, if not usually, the first line of defense of that doc-

trine by many evangelical writers. But if grammar cannot legitimately

be used to support the Spirit's personality, then perhaps we need to

reexamine the rest of our basis for this theological commitment. I am

not denying the doctrine of the Trinity, of course, but I am arguing

that we need to ground our beliefs on a more solid foundation.

______________________________________________________________________

Or is it rather that pneu=ma functions almost with a genitive of apposition, and the gen-

der of the genitive is picked up in the following discussion?

            95. Although it is beyond the scope of this paper, it seems to me that a desideratum

of evangelical scholarship should be not only to compare NT pneumatology with the

OT and Jewish materials, but also to contrast the two bodies of literature. Several works

have looked at the Jewish writings as a source for the NT teaching on the Holy Spirit,

especially focusing in relation to para/klhtoj (cf., e.g., E. F. Scott, The Spirit in the New

Testament [London: Hodder and Stoughton, 1923]; S. Mowinckel, "Die Vorstellung des

Spätjudentums vom heiligen Geist als Fursprecher und der johanneische Paraklet,"

ZNW 52 [1933]: 97-130; idem, "The 'Spirit' and the 'Word' in the Pre-Exilic Reforming

Prophets," JBL 53 [1934]: 199-227; A. J. MacDonald, The Interpreter Spirit and Human Life:

A Study of the Holy Spirit in the Old Testament, the Wisdom Books and the New Testament

[London: SPCK, 1944]; 0. Betz, Der Paraklet: Fürsprecher int häretischen Spätjudentunt, im

Johannes-Evangelium und in neu gefundenen gnostischen Schriften [AGSU 2; Leiden: Brill,

1963]; G. Bornkamm, "Der Paraklet im Johannesevangelium," in Geschichte und Glaube:

Gesammelte Aufsätze, Part 1/vol. 3 [Munich: Chr. Kaiser, 1968], 68-89; G. Johnston, The

Spirit-Paraclete in the Gospel of John [Cambridge: Cambridge University Press, 1970];

A. R. C. Leaney, "The Johannine Paraclete and the Qumran Scrolls," in John and Qumran

[ed. J. H. Charlsworth; London: Chapman, 1972], 38-61; U. B. Müller, "Die Parakleten-

vorstellung im Johannesevangelium," ZTK 71 [1974]: 31-78; G. M. Burge, The Anointed

Community: The Holy Spirit in the Johannine Tradition [Grand Rapids: Eerdmans, 1987];

Breck, Spirit of Truth; S. Notley, The Concept of the Holy Spirit in Jewish Literature of the Sec-

ond Temple Period and "Pre-Pauline" Christianity [Ph.D. diss., Hebrew University, 1991];

C. S. Keener, The Spirit in the Gospels and Acts [Peabody, Mass.: Hendrickson, 1997]), but

few seem to attempt to analyze orthodox arguments for a high pneumatology in light

of such materials along the lines that Hurtado or Bauckham have done with christology.

            In general, I would agree with Alister McGrath (Christian Theology: An Introduc-

tion [2d ed.; Oxford: Blackwell, 1997], 294) on how to construct the Trinitarian doctrine:

"The doctrine of the Trinity can be regarded as the outcome of a process of sustained

and critical reflection on the pattern of divine activity revealed in Scripture, and con-

tinued in Christian experience. This is not to say that Scripture contains a doctrine of

the Trinity; rather, Scripture bears witness to a God who demands to be understood in

a Trinitarian manner." If this is so, then we must engage in careful thinking about what

the apostles consciously embraced about God, as well as what they were groping to un-

derstand and express