"Aleipterion" [Greek] Foss, Clive Greek, Roman and Byzantine Studies; Summer 1975; 16, 2; ProQuest pg. 217
Clive Foss
AMONG THE DISCOVERIES of the recent excavations at Sardis are
two inscriptions from the Gymnasium which give consider
able prominence to the aleipterion. They adorned the so-called
Marble Court, the monumental entrance portal to the enclosed part
of the large Roman bathing establishment at the western end of the
city. The portal faced onto a rectangular colonnaded palaestra from
which the baths were approached. In their restoration of the Marble
Court, the excavators have replaced the inscriptions in locations
where no one entering the baths could have failed to notice them. The
first, a long dedication by the city to Caracalla and Geta, set up in
A.D. 211/212, runs along three sides of the court.l The following part
is relevant to the present discussion:
< ~ ~ ~ '\ \ -'_\ , \ \ ~ [\ D'
••• 7'} £.JapoLwwv '1TOI\LC ••• TO Wl.EL'1TT7'}PLOV CUV '1TWTt TC[J I\WLVC[J
'] I 'E 'D ~\ \" ., ~ KOCP.C[J KaTECKEuacEV... XPUCWU7'} oE TO EPYOV U'1TO TE T7'}C
1\ \ 'A 1 ~ Q' • ~ \ m\ .Q' IT \' 7TOI\€WC Kat VTwv£ac £.Jat'€£V7')c U7TaT£K7'}C Kat 'Yl\at'Lac .l101\£T'T7'}C . ~ U'1TaT£K7'}C.
A later inscription, the dedication of a late antique governor of the
IV-VI century, stands directly over the arched entrance to the baths:
, E'1T1. EEOU( Tj )p( ou) ELP.'1TALKtOU TOU Aap.'1T( POTCXTOU) K6p.( 7'}TOC)
'1TPWT( ou) f3aOJL( OU ) aL'7T( OVTOC) TIjv €'1Tapx( ov) igouctw Kat
TOUTO TO €PYOV TfjC &[AEL ]'1T'T7'}ptac W€V€W07'}. 2
1 For this inscription see BASOR 162 (1961) 42 and 177 (1965) 24f. It will be published by
Professor Louis Robert in a forthcoming volume of the Sardis Repcrrts, as will the other
Sardian inscription discussed here. In neither case does the present note constitute publica
tion of the inscriptions; I merely reproduce the text of one and part of the other as a
convenience for the reader. I am grateful to Professor G. M. A. Hanfmann, Director of the
Sardis Expedition, for permission to cite these texts, and to Professors Louis Robert and
Robert Renehan for their numerous helpful suggestions. Note that a third inscription was
found on the same site, carved around the three sides of the Marble Court (see below).
That was a late antique inscription in verse which mentions a gold-gleaming roof, marble
veneer and a mosaic floor. Since many fragments of it were lost, it is not possible to deter
mine whether it referred to the aleipterion, a suggestion which its location might encourage.
The inscription is reported in BASOR 162 (1961) 43.
2 For a transcription of this inscription see BASOR 187 (1967) 54; abbreviations in the
titulature of the dedicant have been expanded by Professor Robert on the basis of a close
217
218 AAEIlITHPION
Difficulty has arisen in identifying the aleipterion with some part of
the reasonably well-preserved Gymnasium. It was plainly not the
Marble Court itself, an open structure which in plan closely re
sembles the rooms used for the imperial cult in gymnasia at Ephesus
and elsewhere.3 At first it was suggested that the inscriptions re
ferred to the room behind the Marble Court, but further excavation
showed that that space was occupied by a swimming pool. The present
discussion has been undertaken in the hope of providing an answer
to this difficulty and explaining the significance of the inscriptions.
According to the generally accepted definition, an aAE£7TT~PLOV, a
name derived from the verb aAEtcPw ,xVTt T61t EYAa~&VTt rijc
7TaAatCTpac dpoc/>61ca[t] 7'[ov AOV7'p61v]a KanJ 7'0 7T[atJ!oaywy€LoV Kat 7'0
'\ I "'l: If;). " " 'R " al\€t7T7'7]ptOV Kat T7jV €r:,€Opav Ka£ T7jV CTOav T7jV Ka7'a,..oppov Kat 7'ac
" , f;). I 112 'A' ~, \R' f;). \ ~ , ~ , ac7'€YOVC CToac avo .•. ; V7'£KW£ 7'W£ £yl\a,..OVTt O£€I\ELV TOV TO£XOV 1'01-'
1 A _, \ I [ ]' A ~ I A' I I-'€COV TOV UI\€£7TT7jP£OV Ka £ TOV 7TatoaYWYEWV TOV apxatOV •••
5. MAMA VI noA (Laodicea ad Lycum, ea 200 B.C.): 'OAVILmxoc
TPWtAOV ! 7'0 aA€L7TrrJpwv Ka[L] I 1'01-' 7TpO alhov OtKoV ! 7'61£ 07]ILW&,
6. IG XII.3 suppl. 1314 (Thera, II cent. B.C.): IIpOKAdoac Eva[y6pa]
YVlLvaCLapX7]cac Kat. 'A [AK£I-'€OWV] IIpoKAEtoa wOYVlLvaCL[apx7]cac] 7'0
aAEL7TrrlPLOV fEpl-'a.£ Ka[l fHpaK]AE"i.
7. IG V.1 1390.106 ff (Andania, 92/91 B.C.): 'AAEll-'l-'aToc Kat AOVTPOV' 0
, , , 1\ " " • (] 1\ ! [R-. \ 1 J ' A. A ayopavop..oc E7TLI-'€I\€tav EXETW J 07TWC o£ EI\OVTEC ,..UI\avEVE LV EV TW£ £EPW£
, \ ~ 1 '\ 1 f;). I _. \ A " ~, ILT} 7TI\ELOV 7TpaCCWV7'£ TOVC I\OV0p..EVOVC OVW XUIIKWV Kat 7TapEXWV7'£ 7TVP Ka£
IL,xKpav EVKpaTOV KaL 7'0 Le Ka7'aKAv! [~op..€vOte v]OWp EVKpa7'OV. Kat 07TWC 0
, f;). l: I A l: 1\' '" -'-. \ I I i: 1\ l: \
EyOEr:, al-'Evoe 7'WV r:, VI\WV Tav 7Tapoxav €tC TO WlEt7T77JpLOV 7TapEXEt r:, vl\a r:, 7]pa
, 0 , - -'- \ I [.J.. I ] " I " 1 fI " 0Rf;). I Kat 'Kava TO te WlE' ~0I-'EVme KaT al-'Epav a7TO 7'ETap7ac wpae EWC E,..OOl-'ae.
~A\ f;).' (]' -'-.\ .J..I(] 0",' 0 "~~, , 'A oovl\oe OE 1-'7] EtC WlE£~EC W. Ot OE tEpOt EYOWOVTW 7'av 7Tapoxav TWV
gVA[ w]v I [7'61V €lc 7']0 aAEL7TrrJpLOV.
8. Strabo 3.3.6 (154): brove Esc. of the Lusitanians] ... AaKOVtK61C Ot,xYELV
.J.. I '\ I I f;).' , , '\ '(] f;). ,
~acLV. aI\EL7T7"1}pW£C XPWILEVOVC OLC Kat 7TvpLaLC EK I\t WV ota7Tvpwv . . .
9. Herodotus Medicus, ap. Oribasius, Coll.Med. 10.10.1 (I cent.): rT]v
8 References to all but nos. 10, 13 and 16 are to be found in Delorme 301f. As in the case
of the Sardis inscriptions, I have abridged freely. I have omitted two inscriptions,
MAMA VI no.56 (Tripolis on the Maeander) and no.355 (Acmonia), which are too frag
mentary to permit certain restoration.
220 AAElflTHPION
'" \ '" \ A '8 ' , __ \ , , A)_ \ ,.. A \
OE OUX 'TWV av paKWV TTVpLaV 7Tapat1.7]7T'TEOV €V 'TOtc at1.EL7T'T7}PLOLC, rJ rn 7TPOC
'TaSE [codd: T}'TOL 7Tp6C 'TU SLU] 'TOU f1a.\avElov E7Tt. 'TWV XpOVLOV aAYTJp.a
, ,
EXOV'TWV.
10. Herodotus Medicus, in RhM 58 (1903) 112:9 45v • 8ff XP7Jcop.E8a S€ E7T'
, \ ~ I 'f\'.,.. f' , _)~\ I 'TOV'TOLC Kat, opW7TaKL €V 'YJIUCfJ 7] V7TOKaVC'TOLC at1.EL7T'T7}PLOLC.
11. Inscr.Perg. 466 (Pergamum, II cent.): Sfjp.oc E'Tlp.7]C€
T,(f1'PLOV) IO.avSLOv OUE'T€pa 'T6V 7TPV'TavW Kat. K'Tlc'MJ [V] 'TOU MEL7T'MJplov
'TOU lev] 'T[WL] 'TWV vEwv yvp.vaclwL.
12. ]HS 11 (1890) 123 no.6 eCeramus, Trajan): ... rfj 7Ta'TptJ" av'TOC
A , \ pOSElrn
Kat. [-]; fr.2 AV'TOKpa'T0pL KalcapL B€ou TpaLavou ll[-]; fr.3 EvSap.ov
Kat. Z~VWVOC Kat. 1 EvSap.ov Kat. KaMlov 'TWV Kat AOV'MJPCLV ; fr.4 [Z~V ]wvoc
'TOU EvSap.ov Kat V7r€P 'A7Tc/>[l]W[Voc]; fr.5 [V7r€1cX€'TO 'T6 M€L7TnlPLOV ...
14. IGRR IV 1431 (Smyrna, Hadrian): 16-20 KAavSLOc 7TP&ravLC XpvccfJ 1-
CELV 'TOV C;poc/>ov 'TOU ML7TT7Jplov 1 Tfjc YEpovclac Kat ['T]o[v] €lc 'TOV XaIPL
CnlPLOV VEW[V] KElova cvv C7TE'L IPOKE'#ACfJ' •.• 33-36 Kat. clca E7TE''TVI X0P.€v
7Tap~ 'TOU KVp{OV Ka{capoc 1 04. My
thanks to Professor Renehan for reference to this passage.
10 O. Liermann, Analecta epigraphica et agonistica (Diss. Philo!. Hal. 10, Halle 1889) p.73
no.14 = CIG 2782, from the copy of Sherard. It is reproduced here with the emendations of
G. Derenne, "rYMNA£TlKA," AntCl2 (1933) 74, and J. and L. Robert. BullEpig 1973,475
ad fin.; cf RevPhil 55 (1929) 138 n.2.
CLIVE FOSS 221
baths was fundamentally altered.ll From these texts, an impression
of the aleipterion emerges.
The aleipterion was distinct from the bath (nos. 1, 2, 4 and 7), as
would be expected from the expression rJ),ELfLfLIX KIX~ AOVTp6v. It could
form part of a gymnasium (nos. 4 and 6),12 as a bath could also
(noA).I3 It was a closed space (no.I) which could be heated with dry
wood (no.7) probably in a hearth or a hypocaust, the latter a tech
nique perfected only under the Romans.14 The heat thus produced
could be sufficient to cause sweating (no.3). It was apparently a dry
heat rather than a steam bath, for the two are distinguished in no.S,
which refers to the practices of the barbarians ofLusitania who bathed
in the Spartan manner-that is, they took a sauna followed by a
plunge in cold water.I5
The evidence so far considered is not adequate to permit recon
struction of the size, shape or furnishings of an aleipterion. In several
cases (nos. 1-3,8-10), it is not possible to determine whether aleipteria
formed part of a public or private bath.I6 Archaeological investigation
has not provided any satisfactory supplement. The gymnasia of Thera
and Delos, where aleipteria are mentioned in the inscriptions, have
been investigated, but in both cases the identification of the room in
question is hypothetical. In Thera, a case has been made for calling
a circular chamber in the gymnasium an aleipterionP The evidence
11 Both Delorme and Ginouves end their discussions around the time of Augustus, though
they draw freely on later material. For the reasons, see the important discussion of Delorme
243-50.
11 Note that the terms 7TaAalcTpa and YV/LV&CLOV were used interchangeably: Delorme 253.
Hercules and Hermes (no.6) were the gods par excellence of the gymnasium: Delorme 340.
IS For baths in gymnasia see Ginouves 124-50.
1& The suggestion of Delorme 303 that the aleipterion was heated by portable braziers of
charcoal receives no support from the sources; no.7, for example, specifies that dry wood
be used for heating. On that theory, see Ginouves 204 n.4 and, for the heating of baths,
135f, 204-09 with 206f for hypocausts and 210 n.4 for open fires.
15 Martial 6.42, Ginouves 200: the sauna could produce dry heat, as described by Martial,
or steam.
16 Compare e.g. Plin. Ep. 2.17.11, in a description of a bath in a private house: adiacet
unctorium, hypocauston, adiacet propnigeon balanei. The unctorium is probably an aleipterion;
comparison with the passage of Herodotus Medicus (no.10) suggests that the following
comma should be deleted and that a heated anointing room is in question. In the other
Latin passage, incidentally, which is sometimes cited among the testimonia for aleipterion,
Vitro De Arch. 5.11, the term employed, elaiothesium, seems rather to denote a room for the
storage of oil: Delorme 304.
17 Thera: Delorme 82-86, Ginouves 138-140; Delos: J. Delorme, Delos XXV: Les palestres
(Paris 1961) 137ff. The identification of the aleipterion of Thera is not certain because the
inscription which mentions it was found not in the room in question but nearby.
222 AAEIIITHPION
suffers from the additional deficiency that it consists of short texts
scattered over four centuries. This makes it impossible to follow the
evolution of the meaning of the word (if indeed it changed at all in
that time) or of its architectural realization. Such as there is, however,
is reasonably consistent and does nothing to upset or alter the ac
cepted definition of an aleipterion as a room for anointing the body
with oil.
With the establishment of the pax romana and the great current of
cultural influences between east and west, the Greek gymnasium and
baths underwent considerable change. The two gradually came to be
combined into one monumental complex, the thermae.18 This incor
porated an elaborate bathing establishment with the manifold aspects
of the gymnasium: exercise, education, worship and social functions. Ii
The terminology employed naturally changed with the architectural
development; under the Empire, yvp.vaaov and /3a.)..a.VE'ioV were used
interchangeably to refer to the new baths cum gymnasia, which were
sometimes also called (NpP.a.L. 20
Under such conditions, it would not be surprising to find some
change in the meaning of aleipterion; such, I believe, may be demon
strated from the later testimonia.
These suggest that the imperial aleipterion was a structure of some
importance. A citizen of Pergamum prided himself as the founder of
one in the gymnasium of the VlOL (no.ll).21 Similarly, citizens of
Ceramus and Aphrodisias built aleipteria for their cities (nos. 12 and
15 and apparently 13). A magistrate of Smyrna gilded the roof of the
aleipterion of the Gerousia, and the emperor Hadrian himself con
tributed numerous marble and porphyry columns to it or another
in the city (no. 14). The inscriptions of Sardis may be seen in this
context. According to the first, the aleipterion was built from its
foundations and dedicated to the reigning emperors; it was gilded
(a reference to its roof) by the city and two consular ladies. Some
centuries later, it was restored by a governor of Lydia.
The later references are less informative. The statement of Pollux,
18 See supra n.9, Ginouves 147-50 and the remarks of Ward-Perkins in A. Boethius and
J. B. Ward-Perkins, Etruscan and Roman Architecture (Baltimore 1970) 399.
11 On the various uses of the gymnasium, see Delorme 256, 272-361.
10 Ddorme 246, Ginouves 148,220 with n.5 on thermai.
11 The gymnasium of the wo, was the whole complex built in terraces in the Middle
City on the slopes of the Acropolis: Ddorme 182.
CLIVE FOSS 223
a writer of the late second century, that the aleiptenon formed part
of the bath, need not be taken into account here. He was commenting
on the passage of Alexis Comicus (no. 1 ), and in any case reproduced
Attic usage rather than the idiom of his own time. The papyrus frag
ment (no. 16) is part of a petition in a case of robbery with violence in
which the unfortunate victim was locked in an aleipterion. In spite of
the dramatic circumstances, the passage adds little to what is already
known.
The likelihood that aleipterion in the imperial period continued to
indicate merely one room in a larger building is diminished by the
pride with which the donors commemorate their benevolence, and
the eminence of some of them. Closer examination of the texts will
reveal further difficulties in accepting the normal definition.
The injunction of Herodotus Medicus (no.9) shows that steam baths
could be taken in the aleipterion, which seems not to have been true
earlier.22 Here, however, mention is still made of a room in a bath
which could be heated by a hypocaust (no.IO).
Much more significant are the hints of descriptions of aleipteria in
Smyrna, Aphrodisias and Sardis (nos. 14 and 15 and the first Sardis
inscription). That of the Gerousia in Smyrna (and apparently that of
Sardis) had a gilded roof, which would hardly be appropriate to a room
which was heated and possibly full of vapor; neither dry nor damp
heat is especially beneficial to wood covered with gilt.23 Equally
curious is the number of columns presented to the aleipterion of
Smyrna by Hadrian: 72 of Synnada marble, 20 of Numidian and 6 of
porphyry.24 It is difficult to conceive of an anointing-room (or perhaps
any other) which would contain so many columns.25 In Aphrodisias,
the aleipterion of the Diogenian gymnasium (no.15) had within it an
II See supra p. 221. The text in Oribasius is evidently corrupt; I would assume that a
word has fallen out after Tjj 7TPOC Taa~, for here one might expect a reference to some other
part of the bath near the aleipteTion.
13 This problem was clearly felt by Delorme, who. though he believed that the text
showed the aleipterion to be a room (302). realised that the gilding would deteriorate under
such conditions and was led to propose his theory that the heat was provided by braziers
of smokeless charcoal. On that, see supra n.14.
U The number 72 depends on a restoration: the first edition, eIG 3148 (from the notes
of Selden and Chandler), printed an upSide-down omega, for which an omicron seems a
reasonable substitution. The exact number, however, is immaterial for the purposes of
this discussion.
85 An exception might be made for the room for the imperial cult, which contains vast
numbers of columns (though hardly 98' in restored drawings.
224 AAElnTHPION
olKof3aCLALK6v, apparently some kind of public building.26 These texts
suggest that a fundamental change had taken place in the meaning
of aleipterion.
The archaeological evidence connected with these various attesta
tions is also helpful. The inscriptions of Sardis have already been
considered; they were in a space which cannot be considered as an
anointing-room. The fragmentary architrave inscription from Aphro
disias (no.13) was excavated in a room on the south side of the Baths
of Hadrian. This was an open colonnaded space adjacent to a large
peristyle. It was not on the main axis of the baths, which were entered
on the east from the Portico of Tiberius, part of the agora of the city.27
Although the inscription, now lost, was fragmentary, it may reason
ably be considered to have had some connection with the room in
which it was set up. That space, open and separate from the rest of
the bath, could not be considered an aleipterion according to the
normal definition of the term.
Such evidence plainly demands a new definition. A clue is provided
by the entry in the Suda (no.I7). There, the aleipterion is equated with
gymnasia and baths.28
A definition of aleipterion in the imperial period as the whole bathing
establishment would suit all the evidence. In Pergamum, it formed
part of the gymnasium of the VlOL, a large complex which in imperial
times included a bath building. If the aleipterion were a whole building
rather than one room, the various dedications would take on more
significance. The gilded roofs of Smyrna and Sardis would cease to
present a problem, and the great number of columns which Hadrian
contributed in Smyrna would be seen as far more appropriate to an
26 For the basilica thermarum, see D. Krencker et aI., Die Trierer Kaiserthermen (Augsburg
1929) index s.v., and for the term OlKO{3Ctcl,Ju.KOV used here to describe it, see the works of
Louis Robert referred to supra n.10.
27 Description, publication and fragmentary plan in G. Mendel, "Fouilles executees a
Aphrodisias," CRAI 1906, 72; discussion in E. Will, "Les ruines d'Aphrodisias en Carie,"
RA 12 (1938) 231-33. A detailed plan of the whole was published by L. Crema in "I monu
menti architettonid afrodisiensi," MonAnt 38 (1939) 266. For the Portico of Tiberius, see
Kenan Erim, "Aphrodisias in Caria. Results of the 1970 Campaign," Turk Arkeoloji Dergisi
19 (1970) 58.
28 Actually, the 'clue' may be fortuitous. The definition of the Suda also appears, in the
same form, in an anonymous l:vvay~ M,£wv XfYTIclp.wv I.K &a.popwv co4>wv TE Kat p.r,r&pwv
1I'oMwv in L. Bachmann, Anecdota Graeca (Leipzig 1828) 187.11. Etym. Mag. 242.50fI presents
a similar definition. These suggest that the definition in the Suda is abridged from an earlier
common source, perhaps even Alexandrian, for the compilers of these lexica were con
cerned with pedantry, not contemporary usage.
CLIVE FOSS 225
entire large building than to a mere anointing-room. The /3actAtKoV
of Aphrodisias would make sense as part of a whole bathing complex,
though impossible to fit into anyone room. The dedication from the
Baths of Hadrian in the same city could then be taken to refer to the
construction of the entire building.
Although no text besides the Suda supports the definition proposed,
parallels are not lacking. BaAave'i:ov underwent a similar development
and came to mean both an individual room where the fire for a bath
was heated and the whole complex of rooms which made up a bath
building.29 In the Greek gymnasium, the 7TatDaYWYEtOV was a room
where pedagogue slaves awaited their charges; it eventually came to
be applied to the whole schoo1.30 nVptaT~ptOV, a steam bath, was
originally the name of one room, but Cassius Dio found it convenient
Greek for the Thermae of Agrippa in Rome.31 The fluctuation of
nomenclature is shown by the term fUp/Lat itself (which originally
meant 'hot springs'); it was used both as the eqUivalent of the Latin
thermae and to indicate that part of the complex which was heated.32
Such a variety of terminology need not occasion surprise in an age
which saw major cultural and architectural transformation of its
institutions for exercise and bathing.33
The aleipterion was the place where the action of the verb aAEtc/JEtI)
took place, and where the aAet7TT7)c performed his duties. Considera
tion of those words will support the definition proposed foraAEt7TT~PtOV
in the imperial period. Consultation of the lexica will show that
aAetc/JEtV came to mean more than merely 'to anoint'; it also meant
'to train, to encourage, to supply oil,' that is, to perform other func
tions appropriate to a gymnasium.34 Particularly significant is the
29 Harpocration, ed. W. Dindorf, I (Oxford 1853) p. viii S.v.
30 Delorme 331.
31 Casso Dio 53.27; Delorme 137, Ginouves 321f.
32 Ginouves 220 n.5.
33 The use of aA£m-n1ptov to mean a bathing establishment was, of course, not universal;
{JaACt.V€£OV, Io.OVTpOV and YV/Lvactov continued in current usage and were indeed more common.
Consideration of the testimonia will show that all the certain examples of the extended use
of cX/o.€L7T7'1}ptov come from a relatively restricted area of western Asia Minor. Although this
might suggest that the term was a local usage, it is more probable that the examples
reflect the state of the documentation; for the early imperial period much more evidence
of building activity is available for western Asia Minor than elsewhere in the Greek
speaking provinces.
34 The meaning became especially extended by Christian writers; see Lampe, Patristic
Greek Lexicon, s.v.
226 AAEIIITHPION
term ot a~Etcpop.EVO' which referred to all the users of the gymnasium
in general.35 Similarly, ana~El7T'T1]C (Lat. aliptes or alipta) was not just an
anointer or masseur but a trainer and a teacher.3s If these terms came
to have such broad denotations, it is hardly remarkable that their deriv
ative, a~Et7T'T7JPWV, should have taken on a more extended meaning.37
The inscriptions of Sardis, with which this discussion began, may
now be seen more clearly. The aleipterion (or aleipteria) to which they
refer is no longer a room to be sought somewhere in the recesses of
the building which they so ostentatiously adorn, but the whole bath
itself. The actions of an emperor and a late antique governor now
become more readily comprehensible in their expression: both did
work on an entire building and commemorated it in an appropriate
place, the entrance to the baths.
UNIVERSITY OF MAsSACHUSB'ITS, BOSTON
October, 1974
85 Delorme, index S.v.
86 In addition to the Greek and Latin lexica see Ginouves 145 n.5.
87 The proposed definition suits the texts presently known; the poSsibility need not be
excluded. however, that imperial texts may be discovered in which &.\E'7M'1fPWV means
merely 'anointing-room'. In the similar cases of 1TVpWT"fJPWV, 81pf«U and {3a),avE'ioV, the re
stricted and extended meanings were in use concurrently. Since such a development would
also be natural in the case of &.\E'7M'1fpwV, each new text should be judged according to its
particular context.