Vous êtes sur la page 1sur 14

60

JSNT 3 (1979) 60-72

90· Ibid« ρ·ΐ86· In fact, of course, the problem is of precisely the same proportions: a minor agreement on the 2-document· hypothesis becomes an instance where Mark has chosen to ignore

the common witness of both his sources on the Griesbach hypothesis;

a common omission by Matthew and Luke becomes an addition by Mark to them· Either way an explanation is required·

91. Ibid· ρ·201·

92· Usually for reasons similar to those given by Kummel, Introduction ρρ·5δ-6θ·

93· Wilke, op>cit. ρΛ^3 ·

TEXTUAL VARIANTS AND THEOLOGY: A STUDY OF THE GALATIANS TEXT OF PAPYRUS 46

Howard EslU^ugh

Hi Π crest United Presbyterian Church

15 Church Street Burgettstown, Pennsylvania 15021

Several recent studies have shown the theological signifi­ cance of textual variants. By a comparison of the Galatian text of P46 (the earliest extant witness of the Pauline corpus) with other witnesses, several variants (3:19, 17; 4:6,7; 1:6; 2:20) have been found to be theologically signi­ ficant.

Eshbaugh: Textual Variants and Theology

61

The primary task of textual criticism has been the quest of the original text. Hence variant readings have been explained as the result of scribal error, a slip of the pen, or harmonization with another text/1/. Consensus seems to be that textual criti- cism is atheological, a safe discipline that is purely scientific. My own introduction to the mass of variants in the Nestle text was cushioned by the word of assurance from the intructor that none of these variants made any theological difference. While that scholar today would deny the validity of this claim, there are many who still affirm it: "There is no essential historical or theological

point that is determined one way or another by textual variants/2/.'

Such thinking is challenged by

scribe as a theologian:

The earliest stage of transmission was marked by an attitude of freedom in theological interpretation. Dogmatic purposes were in view and constituted the basic attitude in the use of the gospel text/3/. Clark documents this problem of the relationship of textual criti- cism and exegesis by showing specific instances within the Pauline letters where variant readings make a theological difference:

Rom. 8:28; 1 Cor. 2:1; 6:20; 7:5; 10:19; 11:29; 13:3; 14:38; and

15:51/4/.

K.W. Clark, who pictures the early

A monongraph by Eldon J. Epp further explores the relation- ship of textual criticism and exegesis by examining a larger block

of scripture, the book of Acts, noting the differences between the

text has an

anti-Judaic tendency that

the Western text the Jews and their leaders are more hostile to Jesus, and they are assigned a greater responsibility for his death; (2) the Western text minimizes the response of the Jews and the importance of Judaism to the new faith; and (3) the Western text portrays the Jews, and especially their leaders, as more hos- tile towards the apostles and as persecuting them more vigorously

/5/.

Western and Neutral texts. Epp finds that the Western

is shown by a threefold thrust: (1) In

Another contribution to the study of the theological impor- tance of textual variants is a doctoral dissertation by M.R. Pelt, written under the supervision of K.W. Clark. In this work, 175 New Testament passages have been selected "in which an important difference of theological interpretation rests upon the choice between two or more variant readings/6/. The variants have been discussed under the headings: God, Jesus Christ, and the Life of the Christian Community.

62

JSNT 3 (1979)

A recent addition to the study of textual variants and theol­ ogy is a doctoral dissertation written by G.E. Rice under the supervision of E.J. Epp. Rice has examined the text of Luke in Codex Bezae, and identified as emphases of this manuscript: (l)The exaltation of Jesus, and (2) an anti-Judaic bias/7/.

My own work, also under the supervision of E.J. Epp, was an examination of the Western text (primarily Codex Vaticanus) and the Western text (primarily Codex Claramontanus) of the Pauline epistles/8/. The number of theologically significant variants was quite small (24, comprising only about 50 words). Several reasons may account for this relatively small number of theological var­ iants: (1) The genre of the epistles: narrative is easier to alter/9/; (2) the relatively pure state of the Pauline text/10/; and (3) a conservative methodology/11/.

Several conclusions were drawn from this study: (1) The Western theological variants are concerned with many of the signi­ ficant theological problems that confronted the early church; (2) the Western theological variants as a whole do not show any pattern supporting one particular type of theology; and (3) it is impossible to discern whether a Western reading is the source or the result of theological conflict.

This paper is a further effort to show the theological signi­ ficance of textual criticism by the study of P46. This papyrus is the earliest extant witness of Paul, dating about 200 AD/12/. This initial study will examine variant readings in Galatians.

Gal. 3:19

P46

d F G

τι ουν ο νομός των Πράξεων -1 "

^παραδόσεων D

"Τετεθη

d f g Ambrosi aster Irenaeus

UBS/13/

τι ουν ο νομός των παραβάσεων χάριν προσετέθη

This verse introduces the story of law (3:19-25). Woven into this story are the origin, function, and limitation of the law. The entire passage is regarded by G.S. Duncan as a "depreciatory account of the Law/14/. On the other hand, some, e.o. R.T. Stamm,

Eshbaugh: Textual Variants and Theology

63

find this text to be obscure, and that this obscurity has produced the textual changes/15/.

In the UBS reading of 3:19 the use of the verb form "it was

added" has been interpreted as showing that the law "is a mere ad­

dition

in P46 and the verb of the Western uncials, "was established," makes interpretation of the law as an insignificant, parenthetical

afterthought less likely, if not impossible. Deletion of the verb "was added" also makes this passage more in harmony with 3:15 that states "no one annuls even a man's will, or adds to it/17/.

to the main stream of God's

purpose/16/. The lack of a verb

The UBS phrase χάρι ν προσετέθ η has been interpreted in two ways: (1) The law was added in order that man might know

what sin is, to define sin by recognizing its sinfulness/18/. (2) The law was added in order that man might sin, to make him more sinful/19/. The reading of P46, "the law of deeds," also may be in­

terpreted in two ways:

(1) The law was established

as a result of

evil deeds in order to check and restrain them/20/. (2) The law was established in order that good deeds might be accomplished. The law in this latter interpretation does not have any negative func­ tion in relation to sin, but has the positive function of bringing about good. Ambrosiaster, whose text is the same as the Latin of Claromontanus (d), states that this "law of deeds" was established to instruct the people how to fear God. Some of the specific deeds mentioned by Ambrosiaster are sacrifice, primogeniture, and the tithe/21/. The items on this list of good deeds in Ambrosiaster are some of the traditions of Israel, and Ambrosiaster's commentary could also serve as an exposition of the Greek text of D.

While it is possible to interpret the UBS text as a disparage­ ment of law, it is impossible to do so with the text of P46 and its supporting manuscripts. The law in P46 is either to contain and check evil deeds or to produce good deeds. This latter interpre­ tation is supported by the Greek text of Claromontanus. This posi­ tive interpretation of the law is also found elsewhere in the Paul­ ine corpus: "The law is good" (Rom. 7:12, 16b); and the law is one of the privileged possessions given to Israel by God (Rom.9:5).

Gal. 3:17

P46

UBS

D d

G F

δε

διαθηκην

τούτ ο

λέγω

προχεχυρωμενην

δε

διαθηχην

τούτ ο

λέγω

προχεχυρωμενην

υπο

το υ

θεού

υπο

το υ

θεου^ει ς

χριστο ν

 

Th

d

f

g

64

JSNT 3 (1979)

In 3:15-18, Paul applies to the prior covenant in Abraham (3:15) a human illustration about the irrevocable nature of a man's will. Paul asserts that Christ is the seed of Abraham to whom the promises are addressed (3:16). The law is shown temporally to be after the promise to Abraham and, therefore, as a codicil, incap­ able of negating the prior promise (3:17). Paul concludes that the inheritance is by promise and not by law (3:18).

The basic intent of either of the readings of 3:17 is to show that the law was established after the covenant with Abraham. For Paul, in contrast to the rabbis/22/, the law has no pre-existence The text of P46 and the text of UBS presents the fact without any Christological reference, but the other text (a Western reading) relates Christ to the prior covenant with Abraham/23/. This text indicates the pre-existence of Christ and shows him as acting in history prior to the incarnation.

Gal. 4:6

 

P46

UBS

μεν

οτι* δε

εστε_υιοι

οτι

δε

εστε

υιοί

 

εξαπεστειλεν

ο

θζ

ε^απεστειλεν

ο

θεός

το

πνα

αυτού

το

πνεύμα

του

υιού

αυτού

εις

χαρδιαζ

ημών

 

εις

τας

χαρδιας

ημών

 

I n

Gal .

4:1- 6

Pau l

explain s

th e

benefit s

o f

sonship .

God has

bestowed

a twofold

gif t

upon

his

people:

the

sending

of

his

son,

and

th e

sending

o f

th e

Spirit .

The

consensus

text ,

by

use

o f

a

genitive

phrase,

clearly

indicates

that

this

"Spirit"

is

the

"Spirit "

of

his

Son."

This

passage

from the

 

UBS text

along

with

John

15:26

are

the

classical

proof

texts

for

the

doctrine

that

the

Spiri t proceeds fro m th e Fathe r and

th e

Son

(

filioque )/24/ .

Thi s

doctrin e has been a continuin g foca l poin t fo

r

theologica l contro ­

versy

throughout

the

history

of

the

church/25/.

 
 

I

n

contras t

th e

tex t

o f

P46

state s

merel y

tha t

God has

sen t

hi s

Spirit .

I f

P46

i s

a secondar y

text ,

it s

deletio n

may

impl y

tha t

thi s

was

done

t o

show

tha t

th e Spiri t

proceeds

onl y

fro m

th e

Father .

However ,

i f

i t

i s

original

,

i t

may

imply :

(1 )

The

Spiri t

does

not

come from

the

Son;

or

(2)

the

origi n

o f

th e

Spiri t

has

not

yet

been

considered

a theological

problem.

 

Eshbaugh: Textual Variants and Theology

65

Gal. 4:7

P46

UBS

ώστε ουχετι ει δούλος άλλα υιός ει δε υιός χαι χληρονομος δια θεού

.

.

.

1962 arm et h r o

ώστε

άλλα

χαι

θεού

ουχει

υιό ς

ει

ει

δούλος

υιός

δε

χληρονομος

630 syrpal

81

.

1739 Ι

.

.

ο

δι α

5 5

·

cop s a

Jerome

χριστού

δ ι α

ιησου

χριστού

X C C 2 gpt

D Κ Ρ 88 104 614* Byz Lect a l

.

.

θεού

δι α

χριστού

326

Theodoret

θεού

614 C 2127

.

2495 syrp,h

ινσου

eth p p

δι α

χριστού

.

.

cop „ b o

ms

δια θεού εν ιησου

.

χριστώ

1984 1985 Theodoret Theophylact • o · μεν θεού συγχληρονομο£ δε χριστού

Here as in 3:17 Ρ46 is in agreement with the UBS text. The other readings are not well attested and are generally considered to be secondary. However, theologically it must be noted that some of them indicate that Christ has some role in the bestowal of the inheritance. This is in contrast to the reading of P46 which states that it is through God alone that the inheritance is be­ stowed.

66

JSNT 3 (1979)

In Galatians Paul often shows Christ to be the agent of Christ's blessing: freedom in Christ (2:4), justified in Christ (2:17); Christ redeemed us (3:13), Abraham's blessing comes upon us in Christ, and in Christ all are sons of God (3:26). Hence, it seems strange that in 4:7 there is no reference to the agency of Christ and that God alone bestows the inheritance. P46 here is Christologically barren.

Gal. 1:6

P46

UBS

θαυμάζω οτι ούτω ταχέως μετατιθεσθε απο του χαλεσαντος ημάς ε]ν χαριτι

θαυμάζω οτι ούτως ταχέως μετατιθεσθε απο του χαλεσαντος υμάς εν χαριτιΓυριστού

ΓΓησου χριστο ύ

D d syr

^χρΑστού ιησου χ cop

*-

Jerome

Origen (lat)

•θεού

y

Paul begins the main portion of this letter by stating his concern that the Galatians have abandoned the gospel. Textual var­ iation occurs in the second phrase. The UBS text modifies grace by "of Christ." Similarly D 326 sy modify grace by "of Jesus Christ," and cop Jerome use of "of Christ Jesus." These readings have been interpreted in several ways: (1) Christ is the object of the preposition "from." The Galatians have turned from Christ, and it is Christ who has called them by grace ; (2) "In the grace of Christ" is instrumental; (3) The phrase is elliptical and stands for the one who called you to be in the grace of Christ; (4) Paul stands in Christ's grace when he (Paul) called the Galatians/26/.

Each of these interpretations has a specific Christological reference. In contrast to the UBS text and its cognates, some minuscule manuscripts modify grace by "of God." This reading specifically precludes any Christological interpretation. God

Eshbaugh: Textual Variants and Theology

67

alone here appears to be the actor and agent.

Since the text of P46 has no genitive modifier, it is pos­

sible on the basis

interpret "grace" as "the grace of Christ"/27/. However, the si­ lence of P46 makes it more probably that the text of P46 has no Christological reference. Thus 1:6 as well as 3:17; 4:6; and 4:7 are not Christologically oriented.

of a pre-understanding of what grace is, to

Gal. 2:20

P46 Β D d G g

UBS

πιστει ζω τη του θεού και χριστού

εν πιστει ζω τη του

υιού του θεού

/28/

In 2:15-20 Paul presents the heart of his argument to the Galatians. He does not argue abstractly but relates his own per­ sonal experience. V.20 summarizes this experience: "I have to be crucified with Christ; it is no longer I who live, but Christ who lives in me; and the life I now live in the flesh I live by faith in the Son of God (in God and Christ), who loved me and gave him­ self for me."

Both readings are unique. While Paul many times speaks about "faith in Christ" and its cognates, he never elsewhere uses the phrase "faith in the Son of God." The P46 reading "faith in God and Christ" is the sole occurrence of a double object of faith.

While Metzger's statement that "Paul nowhere else expressly speaks of God as the object of a Christian's faith"/29/ is true, there is the possibility that the reading of P46 is a subjective genitive and should be translated "I live by the faith (fulness) of God and Christ/30/.

These readings,like those considered above, show a Christo­ logical difference. The UBS text by its use of "Son of God" shows a higher, more formalized Christology than the P46 text which makes God as well as Christ the object (subject) of faith.

THE SIGNIFICANCE OF CHRISTOLOGICAL VARIANTS IN P46

This paper has attempted to show the theological significance of readings in P46 without noting whether they or the texts with

68

JSNT 3 (1979)

which they are compared are the original readings/31/. While this quest is beyond the scope of this investigation, some perspective on the theological significance may be gained by viewing the Christ-

ological variants in P46, as a whole, first as secondary as original readings.

and then

If P46 is a secondary text, these variants suggest that P46 has a subordinationist Christology. 4:6 speaks of the sending of

the Spirit. In P46 this Spirit is God's Spirit not the Spirit of the Son/32/. In 4:7, P46 makes no declaration of the agency of Christ. The inheritance has come through the agency of God alone. Similarly in 1:6, P46 does not speak of "the grace of Christ." In

2:20, P46 has changed the genitive phrase "faith

to read "faith in God and Christ." The addition of the phrase "in God" suggests that the scribe-theologian/33/ saw a need for a role of God in this passage to support the work of Christ. 3:17 makes no mention of the agency of Christ in establishment of the cove- nant of Abraham. Together these readings suggest subordinationist Christology.

in the Son of God"

On the other hand if P46 is the original text, the other text has embellished these readings and established a broader, higher Christology. In 4:6 the scribe-theologian has added the phrase "of his Son." 4:7 shows the agency of Christ. In 1:6 it is the grace "of Christ." 2:20 has the Son alone as the object of faith that enables Paul to live. In 3:17 Christ is not only pre-exis- tent but works in history before the incarnation. Thus, if P46 is original, this later text has added material in order to present a higher Christology/34/.

Where does P46" fit into church history? Some scholars have noted a relationship with Marcion. The Nestle text (25th ed.) con- jectures that Marcion utilized the reading of P46 that omits the phrase "of his Son" in 4:6. Harnack, as cited by Blackman,"prob- ably" lists 1:6 as found in P46 as a Marcionite tendentious emen- dation/35/. However, no explanation is given for this omission. On the other hand, Schlier and others have noted that 3:19 of P46 is probably a reaction to the theology of Marcion/36/.

CONCLUSION This paper is the beginning of a complete study of the signi-

ficance of textual variants in P46. Galatians was chosen at random

as a starting point, and some theologically

have been found. Whether other letters will be as fruitful remains to be discovered.

significant readings

Eshbaugh: Textual Variants and Theology

69

In such a study the problem of a precise methodology must be resolved. What should be the basis of comparison? What is the relationship of P46 and text-types? How are text types defined/37/?

I believe that study of variant readings in P46 and other manu­ scripts is useful to both the exegete and the church historian. The exegete will gain some insight on the text as the variant shows how some scribe-theologian wrote how he thought the text was to be un­ derstood/38/. These variants whould also help the church historian as they show areas of theological controversy.

NOTES

/!/ Note the frequency of "accidentally" in B.M. Metzger, A Tex­ tual Commentary on the Greek New Testament (New York, 1971),passim.

/2/ H.C. Kee, F.W. Young, and K. Froelich, Understanding the New Testament (Englewood Cliffs, 1963), "Introduction (n.p.H Others similarly state: "There is not one (variant) affecting the sub­ stance of Christian dogma." L. Vaganay, An Introduction to the Textual Criticism of the New Testament, cited by K.W. Clark, "Theological Relevance of Textual Variation in Current Criticism of the Greek New Testament, "Journal of Biblical Literature, 85 (1966), 3: and "There is no essential historical or theological point that is determined one way or another by textual variants.", J.H. Greenlee, Introduction to Textual Criticism (Grand Rapids, 1964), p. 68.

/3/ "Theological Relevance,"7.

/4/ "Textual Criticism and Doctrine," Studia Paulina (Haarlem, 1953), pp. 69-80.

/5/ The Theological Tendency of Codex Bezae Cantabrigiensis in Acts (Cambridge, 1966), p. 165.

/6/ "Textual Variation in Relation to Theological Interpretation in the New Testament" (Ph.D. dissertation, Duke University, 1966)

ΠI "The Alteration of Luke's Tradition by the Textual Variants in Codex Bezae" (Ph.D. dissertation, Case Western Reserve Univer­ sity, 1974), p. 262.

/8/ "Theological Variants in the Western Text of the Pauline Corpus" (Ph.D. dissertation, Case Western Reserve University, 1975).

191 W. Sanday and A.C. Headlam, Romans (6th ed.; Edinburgh, 1964),

70

/IO/ G.

JSNT 3 (1979)

Zuntz, The Text of the Epistles (London, 1953), p. 263.

/Il/ In order to qualify as a Western reading the variant had to appear in one of the bilingual codices (Dd, Ff, Gg) and have attestation by a Western Father. This excluded about 75 readings.

/12/ B.M. Metzger, The Text of the New Testament (New York, 1964),

p. 252.

/13/ The Greek New Testament of the United Bible Societies, ed. Κ. Aland, M. Black, B.M. Metzger, and A. Wikgren (Stuttgart, 1969) will be used for comparison with P46. When P46 agrees with the consensus text of the UBS, other manuscripts will be used for com­ parison. Such a methodology will be employed in this preliminary paper. If a complete study of the theological textual variants is carried out, a more reasoned methodology, probably utilizing text types, will be used.

/14/ The Epistle of Paul to the Galatians (London, 1934), p. 115.

/15/ "Exegesis of Galatians," The Interpreter's Bible, ed. G.A.

Buttrick, et al

/16/ G.S. Duncan, The Epistle of St. Paul to the Galatians, p.115.

/17/ E.D. Burton, The Epistle to the Galatians (Edinburgh, 1921),

p. 168.

/18/ J.B. Lightfoot, The Epistle of St. Paul to the Galatians (London, 1896), p. 144.

/19/ R. Bultmann, Theology of the Mew Testament, trans. K. Grobel

(New York, 1953), Vol. X, p. 513.

(New

York, 1951), Vol. I, p. 265.

/20/ This is similar to Lightfoot's first interpretation "to check transgressions" (p. 144). Lightfoot gives no Pauline support but lists Clem. Horn. XI. 16.

/21/ Ambrosiastri qui Dici tur Commentarius in

ed. H.J. Vogels (Vindobonae, 1966), Vol. I, p. 38.

/22/ Sifre Deut. #37:76a; Gen. R. 8,2; ABN 31.

/23/ There are two grammatical possibilities for the phrase

Epistulas Paulinas,

εις χρ ι στο ν

: (1) The phrase could be taken temporally,

that is, the law is only to exist until the coming of Christ.

While this interpretation is a grammatical possibility, its weak­ ness is that it places the covenant on the same temporal level as the law. This interpretation diminishes the thrust of Paul's

argument. (2) eir χριστο ν

could mean that Christ is the agent

of the covenant. This is based upon the fact that in Koine Greek

Èshbaugh: Textual Variants and Theology

II

the preposition ε κ can have the same meaning as the preposi­ tion ε ν . Hence, it is possible that the phrase εις χριστον shows Christ as a personal agent. This interpretation is shared by the Latin witnesses of the western text, for the Latin prepo­ sition ijn also* can show agency. This alternative is taken as more probable because of the united testimony of the Greek and Latin witnesses.

/24/ "Confession of Faith," The Constitution of the Presbyterian Church in the United States of America, 1919 (Philadelphia), p.20, n.q.

/25/ W. Walker, A History of the Christian Church(New York, 1950):

The Third Council of Toledo added the phrase to the Ni cene Creed in 589 (p.180).

/26/ Bligh, Galatians in Greek (Detroit, 1966), p. 80

/27/

p. 37, n.2.

/28/ The RSV as well as many commentators make this phrase an ob- jective genitive. Some as H. Ljungman, Pistis: A Study of Its Presuppositions and Its Meaning in Pauline" Use (Lund, 1964), p. 38; and M. Barth, "Galatians" (unpublished, Pittsburgh), p. 421 take this to be a subjective genitive and translate the phrase as "by the faithfulness of the Son of God." The possibility that the text of P46 is a subjective genitive will be discussed below.

H. Schlier, Der Brief an die Galater, 12ed. (Göttingen, 1962,

/29/ A Textual Commentary on the Greek New Testament (New York, 1971), p. 593

/30/

/31/ In 3:17 the UBS text is the same as P46 and other mss. are used for comparison.

/32/ P46 may have done this to keep from confusing the persons of the Trinity. E.g., The Shepherd of Hermas, Sim. 5,5,2 and 9,9,1 equates the Spirit and the Son.

/33/ The dual function was suggested by K.W. Clark.

/34/ Further research must be carried out in order to see if this pattern exists elsewhere in the Pauline corpus. If it does not exist in the corpus as a whole, it may be indicative that each letter has had its own textual history.

See Rom. 3:3 for this usage (the faithfulness of God).

72

JSNT 3 (1979)

/35/ E. C. Blackman, Marcion and His Influence (London, 1948), pp. 60 , 108.

/36/ H. Schlier, Der Brief an die Galater (12th ed.; Göttingen, 1962), p. 151.

/37/ Β is usually taken as the best representative of the Neutral (Alexandrian) text-type, but Sanday and Headlam, The Epistle to the Romans, p. lxvii, state: "and Β also (as we shall see) in the Pauline Epistles has a clear infusion of Western readings." Metzger, The Text of the New Testament (New York, 1964), p. 252, states that P46 is Alexandrian; but E.J. Epp, "The Twentieth Century Interlude in New Testament Textual Criticism," Journal of Biblical Literature, 93 (1974), 397, describes P46 as "midway between" the Neutral and Western.

/38/ Markus Barth describes variants as a "first commentary" (forthcoming in Colossians: The Anchor Bible).

^

s

Copyright and Use:

As an ATLAS user, you may print, download, or send articles for individual use according to fair use as defined by U.S. and international copyright law and as otherwise authorized under your respective ATLAS subscriber agreement.

No content may be copied or emailed to multiple sites or publicly posted without the copyright holder(s)' express written permission. Any use, decompiling, reproduction, or distribution of this journal in excess of fair use provisions may be a violation of copyright law.

This journal is made available to you through the ATLAS collection with permission from the copyright holder(s). The copyright holder for an entire issue of a journal typically is the journal owner, who also may own the copyright in each article. However, for certain articles, the author of the article may maintain the copyright in the article. Please contact the copyright holder(s) to request permission to use an article or specific work for any use not covered by the fair use provisions of the copyright laws or covered by your respective ATLAS subscriber agreement. For information regarding the copyright holder(s), please refer to the copyright information in the journal, if available, or contact ATLA to request contact information for the copyright holder(s).

About ATLAS:

The ATLA Serials (ATLAS®) collection contains electronic versions of previously published religion and theology journals reproduced with permission. The ATLAS collection is owned and managed by the American Theological Library Association (ATLA) and received initial funding from Lilly Endowment Inc.

The design and final form of this electronic document is the property of the American Theological Library Association.