AGAPH outside the New Testament

Steve Kearfott skearfott at juno.com
Fri Jun 7 14:10:18 EDT 2002


A little different take on your question, but still relevant I think:

Grimm (Greek-English Lexicon of the NT, corr. ed., 1889, p. 4) defined
AGAPH as a purely biblical and ecclesiastical word   However, since that
time, scholars alleged to have found occurrences of AGAPH in a number of
papyri which led to the conclusion that, on the contrary, AGAPH was indeed
a common Hellenistic Greek word.  But when one examines the most
frequently mentioned occurrences, one finds that the conclusions of these
scholars are not nearly so certain as they purport, and should be taken
with caution.  These four occurrences are:

1)  Paris Papyrus 49, dating between 164 and 158 B.C., in which Deissmann
originally attested to the usage of the accusative singular AGAPHN. 
However, based on Blass’s suggestion that the correct reading might
instead be TARACHN, Deissmann later retracted his reference to this
citation (Bible Studies, 1979 reprint, pp. 198-199).  While he persisted
in his support of the Hellenistic usage of AGAPH (Light from the Ancient
East, p. 75), he no longer used Paris Papyrus 49 as evidence.  The rather
wide difference between those two words suggests that there must be a high
degree of uncertainty in the original text.  Ramsay (ET 9, 1898: 568)
attributed Deissmann’s original error to his failure to observe that
“there is a gap before AGA as well as after it.  As soon as one observes
this double gap, one realizes that the restoration is far from being so
certain as Deissmann supposes.”

2)  PAS (Papers of the American School), ii, 57, discovered in Asia Minor
and published in 1888.  Lines 7-8 are rendered HN EPIBALLH PENYEI D’EIS
AGA[PH]N SE FILOMMEIDHS AFRODEITH.  The significance of this occurrence
was considered especially important by Hatch (JBL 27, 1908: 136) because
“it occurs in a notoriously heathen context.  It is clear therefore that
this word is not confined to biblical and ecclesiastical writers or used
only in a religious sense.  It was a general word for love in the
Hellenistic Period.”  However, as indicated in the above rendering by the
letters in square brackets, which indicate a lacuna, there is uncertainty
as to AGAPHN being the correct reading.  Furthermore, this gap occurs
across two lines of text and the right hand margin extends beyond the
characters in line 7 (which is where the opportunity for one to restore PH
is provided) to the extent that there may be as many as 4 missing letters.
 So one can see that while the restoration of AGAPHN is possible, to
render it “certain”, as Hatch did (135), is unsubstantiated.  As an
alternative, Deissmann (Light, p. 22, n. 3) considered it “more probable
that we ought to read AGA[QO]N, a reading that is equally legitimate both
orthographically and contextually.

3)  While ultimately not persuaded by the previous two examples, Deissmann
was persuaded by the usage of AGAPH in a text from Philo (Quod Deus ist
immutabilis §14) “and in such manner as to repel the supposition that
Philo adopted the word from the LXX.” (Bible Studies, 199).  Hatch also
attested to the usage here and reiterated Philo’s independence from the
LXX and thus “AGAPH must have been known to extra-biblical Hellenistic
Greek.” (135).  However, Ramsay (568) implicitly suggested that this line
of reasoning is unwarranted and explicitly declared that “this proof is
about as strong as the former discarded proof from the Paris papyrus.”

4)  The cornerstone of the argument is offered in Oxyrhynchus Papyri
#1380, otherwise known as the “Invocation of Isis,” dating from the early
2nd century.  The text consists of a litany of virtuous titles bestowed
upon Isis throughout the known world, two of which allegedly contain
AGAPHN.  The first of these occurs in line 28, which is rendered QWNI
AGAP[HN…..….]  (translated “at Thonis love…”).  However the word in
question has experienced considerable damage: the P is considered doubtful
(as indicated by a dot underneath); the HN has been restored (as indicated
by square brackets); and there may be as many as nine other letters
contained in the word (as indicated by the nine dots within the square
brackets).  The second of these occurs in line 109, which is rendered EN
ITALIA A[GA]PHN QEWN (translated “in Italy love of the gods”).  This word
likewise has experienced considerable damage: the initial A and the N are
considered doubtful; and the GA has been restored.  Grenfell and Hunt
conceded that the initial A might be a L, and that the word might be
L[U]PHN (sorrow).  The extent of damage in these two occurrences has led
scholars such as Stephanie West to conclude that AGAPH “has no legitimate
standing here, and should not be used as evidence for the pagan use of the
word,” and to prefer instead the reading of  AGAQH (Journ. of Theo. Stud.
18, 1967, p. 143).  However, this conclusion seems to be the minority
view.  In a direct rebuttal to West, R. E. Witt (JTS 19, 1968, p. 210)
charged that she “has not shown why the reading AGAPH here must be wrong,”
and concluded himself that “nothing prevents AGAPH QEWN from having been a
title of Isis.” (211).  Witt’s points are accurate but fall short.  The
point is not that AGAPH is impossible, but only that it is not certain.  A
more curious defense of AGAPH came from C. H. Roberts, who concluded with
specific regard to line 109, “here is an indubitable use of AGAPH in a
pagan text” (Journal of Egyptian Archaeology 39, 1953, p. 114.  Although
what Roberts sought was to defend the integrity of the editors’ original
reading, for him to describe this usage as indubitable is instead a direct
contradiction to the editors who themselves cast a certain degree of doubt
regarding four of the six letters of the word.  J. S. Banks also
considered this occurrence indubitable (ET 9, 1898, p. 501), but in his
assessment he conveniently neglected to mention its conjectural status as
well as Deissmann’s retraction, as Ramsay pointed out.  More broadly, with
general reference to their own publication, Grenfell and Hunt noted that
the condition of P.Oxy. 1380 made deciphering it difficult and that “the
scribe himself was not very accurate.”  All in all, this is hardly the
material that certain readings are based upon.

In the end, in each one of the four most prominent examples of the alleged
pagan usage of AGAPH there is a high level of uncertainty.  To repeat, it
is not being argued that AGAPH must not be present in these texts, only
that the evidence presented in these four texts is not strong enough to
support such a claim.

Steve Kearfott
Louisville, Kentucky



More information about the B-Greek mailing list