Daniel P. Bailey DanPBailey at aol.com
Mon Feb 21 01:22:42 EST 2000

Dear ALL:

I've just joined this list. I notice that some of the earliest subject 
strings last year concerned "justification by faith" or the "dying for" 
formula and discussed Paul's understanding of the death of Jesus in terms 
of substitutionary atonement, propitiation, expiation, penal satisfaction, 
etc. More recently on Feb. 5, 2000, Steve Nelson contributed a note on 
"Redemption as Ransom" in Romans 3:24 (to which no one has yet responded). 
On March 31, 1999, I find a message from Bill Ross with my title 
"HILASTERION." Jim West was on the same subject thread. I've tried to link 
to these messages, so as not to break the subject thread.

This discussion has been rather wide-ranging. If anyone would like to focus 
it more narrowly, I've just written a Ph.D. dissertation (University of 
Cambridge, 1999) concentrating on the lexicography of the Greek term 
HILASTERION and the implications that the lexical data have for the 
interpretation of HILASTERION Romans 3:25 and 4 Maccabees 17:22 (codex S).

An abstract of the dissertation is appearing in Tyndale Bulletin 51.1 
(April 2000). If anyone would like a preview, it is included below, and is 
not too long (about 4 pages in journal form). Members may use this material 
in any way they like, but please cite the printed version for any formal 

My external thesis examiner, Prof. James Dunn (Durham), said he liked my 
lexicography but urged me to include more "theology." (There is indeed more 
"Semantics" than "Theology," to quote from my subtitle.) What do you 

Best wishes,

Dan Bailey

Daniel P. Bailey (Ph.D. Cantab.)
6977 N. Bethmaur Lane
Milwaukee, WI  53209

(to be printed in Tyndale Bulletin 51.1, April 2000)

Jesus as the Mercy Seat: The Semantics and Theology of Paul’s Use of 
Hilasterion in Romans 3:25

Daniel P. Bailey

Ph.D. Dissertation, University of Cambridge, 1999

Interpreters of Rom 3:25 and of 4 Macc 17:22 (codex S) commonly base their 
conclusions about hilasterion upon the immediate literary context coupled 
with vague notions of Jewish sacrifice and of the verbs hilaskesthai and 
exilaskesthai. Instead, scholars should consider first the more important 
linguistic evidence, namely, the concrete, non-metaphorical uses of the 
substantive hilasterion in other ancient sources. They should be wary of 
investing hilasterion with meanings that are otherwise unattested (even 
though they may make sense in Romans or 4 Maccabees) and of parallelling 
Romans and 4 Maccabees prematurely. Only concrete, inanimate referents of 
this term are actually found in the other ancient sources; a hilasterion is 
always a thing -- never an idea or an action or an animal. This suggests 
that the uses of hilasterion in Rom 3:25 and 4 Macc 17:22 are metaphorical, 
while further exegesis shows that the two metaphors must be distinct, 
reflecting two different concrete uses of the term.

Unfortunately, past studies of hilasterion have often allowed theological 
considerations to overshadow lexicography. Hence it was the doctrine of 
propitiation rather than the actual occurrences of the term hilasterion in 
ancient sources that dominated the English-language discussion of Rom 3:25 
in the twentieth century. C. H. Dodd reacted against this doctrine and 
argued that the root idea behind Paul’s use of hilasterion was one of 
expiation (of sin) rather than propitiation (of God). However, Dodd based 
his study not on hilasterion itself but on the use of the verb hilaskesthai 
and its cognates in the Septuagint. The result was an over-emphasis upon 
verb-based notions of a theological function, whether the propitiating of 
God or the expiating of sin, with too little attention to the concrete 
referents of the term hilasterion, such as the Old Testament mercy seat and 
Greek votive offerings. Neither Dodd nor most of his early opponents 
considered what hilasterion actually denoted in Paul’s day.

Admittedly, abstract notions of propitiation or expiation can be fitted 
into the context of Rom 3:21-26, causing centuries of debate. The problem 
from a lexicographical standpoint is that words ending in -terion seldom 
denote abstract verbal ideas, while hilasterion never does; the suffix 
-terion is very concrete.

Additional mistakes can be made by ignoring the available linguistic 
evidence. Since Paul elsewhere compares Jesus to an animal victim, as for 
example in Rom 8:3, where the phrase "peri hamartias" is standard 
Septuagintal language for the Levitical "sin offering," Heb. "hattat" (cf. 
NRSV mg.), many have mistakenly concluded that similar victim language must 
be present in Rom 3:25. Jesus is said to be a hilasterion; he is also said 
to have shed his blood. Therefore, it is commonly assumed that a 
hilasterion in the ancient world must have been something that could shed 
its blood, i.e. a sacrificial victim ("sacrifice of atonement," NIV; NRSV). 
This, too, fits the immediate context. But it is a false syllogism, since 
it assumes that the meaning of hilasterion can be determined by the meaning 
of "blood" (after all, blood is sprinkled on the hilasterion of the 
Pentateuch, i.e. the mercy seat, but this does not make the hilasterion 
into a "victim"). It is also unsupported by external evidence: hilasterion 
never denotes an animal victim in any known source.

In fact, there are only two main applications of the term hilasterion up 
through the middle of the second century AD. It can designate (1) the 
golden "mercy seat" or "kapporet" on top of Israel’s ark of the covenant 
(LXX Pentateuch; Heb 9:5; six times in Philo); or (2) durable votive 
offerings to the pagan deities, generally anathemata (so LSJ s.v. 
"hilasterios" II 2). There are also minor extensions of the Pentateuchal 
use in the prophets: the altar ledges in Ezek 43:14, 17, 20 and perhaps the 
altar or one of its parts in Amos 9:1. (The later uses of hilasterion by 
Symmachus to refer to Noah’s ark at Gen 6:15 MT [6:16 Sym.] and by 
Byzantine Christian writers to refer to churches, altars, monasteries, and 
saints’ tombs raise additional problems.)

The application of hilasterion to Greek votive offerings was the normal or 
mainstream use in the first century AD. While generally pagan, it is also 
reflected in Jewish sources such as Josephus Ant. 16.182 and 4 Macc 17:22 
(see below). The hilasterion in Josephus is a marble monument. But the most 
famous hilasterion in the ancient world was the Trojan Horse. This was 
called a thelkterion or "charm" by Homer (Od. 8.509) but a hilasterion or 
"propitiatory gift" by Dio Chrysostom (Or. 11.121) and by two later 
commentators on Homer (anonymous scholia, ed. Dindorf [1855]; comm. by 
Eustathius of Thessalonica, ed. Stallbaum [1825]). The term hilasterion or 
its Rhodian variant hilaterion was customarily inscribed on other gifts 
dedicated to the gods. These include statues, monuments, stelae (Inscr. Cos 
81 and 347, ed. Paton and Hicks [1891]; Bullettino del Museo dell’Impero 
Romano 3 [1932], p. 14, no. 11, ed. Patriarca, printed as appendix to 
Bullettino della Commissione Archeologica Comunale di Roma 60 [1932]; 
variant hilaterion, Lindos II, no. 425, ed. Blinkenberg [1941]), drinking 
bowls (e.g. a "phiale" as a hilaterion, Die Lindische Tempelchronik, B49, 
ed. Blinkenberg [1915]), and tripods (e.g. a "tripous" as a hilasterion, 
scholion on Apollonius of Rhodes 4.1549, ed. Wendel [1935]).

Hilasterion (hilaterion) in all these extra-biblical occurrences can be 
glossed by "(sc. anathema) propitiatory gift or offering" (LSJ). Or, to 
adopt an ancient definition, pagan hilasteria are generally "ta 
ekmeilixasthai dunamena dora," "gifts capable of appeasing" (sc. the gods) 
(scholion on Apollonius).

Since this application to votive offerings was typical, it is a possible 
background to Rom 3:25. Yet no one has ever succeeded in showing how God is 
supposed to have presented humanity (or himself?) with a gift that people 
normally presented to the gods. Moreover, the mainstream use of hilasterion 
finds no parallel in "the law and the prophets" to which Paul appeals (Rom 
3:21). The general meaning "propitiatory gift" therefore fails to fit the 
context of Rom 3:25.

By contrast, a more specialized allusion to the biblical "mercy seat" 
(which is not a gift to the gods) does fit Paul’s context, with plenty of 
support from lexicography (cf. LXX Pentateuch). Paul focuses on "the law 
and the prophets" and more particularly on the Song of Moses in Exodus 15. 
The combination of God’s righteousness and redemption in Exod 15:13 
("hodegesas te dikaiosune sou ton laon sou touton, hon elutroso") closely 
parallels Rom 3:24 (dikaioo and apolutrosis). Furthermore, Exod 15:17 
promises that the exodus would lead to a new, ideal sanctuary established 
by God himself. God’s open setting out of Jesus as the new hilasterion -- 
the centre of the sanctuary and focus of both the revelation of God (Exod 
25:22; Lev 16:2; Num 7:89) and atonement for sin (Leviticus 16) -- fulfils 
this tradition.

Applying the biblical sense of hilasterion to Jesus in this theologically 
pregnant way would not have been be entirely unprecedented for Paul (contra 
D. Moo, Romans, NICNT [1996], 236 with n. 79), since Philo thought of the 
mercy seat as "symbolon tes hileo tou theou dunameos," "a symbol of the 
gracious power of God" (Mos. 2.96; cf. Fug. 100). Perhaps this shows that 
Philo traced the term hilasterion etymologically not to hilaskesthai (to 
propitiate or expiate) but to hileos, "gracious" or "merciful." This would 
then support the translation by "mercy seat," though the vaguer expression 
"place of atonement" is also in common use (NRSV mg. at Rom 3:25 and Heb 
9:5). The old objection that Paul cannot have alluded to "the" well-known 
hilasterion of the Pentateuch without using the Greek definite article is 
baseless, since Philo clearly uses anarthrous hilasterion to refer to the 
mercy seat (Mos. 2.95, 97; Fug. 100).

Finally in 4 Macc 17:22 the original text, preserved in codex S (codex A is 
secondary), contains the controversial expression "to hilasterion tou 
thanatou auton," "the hilasterion of their death" (referring to the 
martyrs). This can be interpreted by the same kind of appeal to established 
usage, only the results are different from those seen in Rom 3:25. It makes 
no sense to speak of "the mercy seat of their death" in 4 Maccabees; this 
imagery works, if it does, only in Romans. However, the mainstream Greek 
metaphor "the propitiatory offering of their death" or "their death as a 
propitiatory votive offering" is completely in keeping with the use of 
Greek heroic and athletic imagery elsewhere in 4 Macc 17:8-24. While Rom 
3:25 cannot be understood apart from a knowledge of the Septuagint, no such 
knowledge is necessary to understand 4 Macc 17:22. The language and imagery 
are essentially Greek, and the more Jewish or biblical-sounding translation 
"their death as an [act of] atoning sacrifice" (NRSV) is misleading, since 
hilasterion does not denote an act of sacrifice, nor are the martyrs 
compared with the victims of sacrifice (such as those on the Day of 

In sum, considerations of both lexicography and context combine to 
discourage the common practice of parallelling Rom 3:25 and 4 Macc 17:22. 
Different metaphors -- one biblical, the other mainstream Greek -- explain 
each passage.


More information about the Corpus-Paul mailing list