Norman G. Ashton
Gregory H. R. Horsley

A rediscovered arkhisynagogos inscription from Thessaloniki, and an intriguing Iulia Prokla

Plates 1–6

Preliminary note

Written permission to publish the new inscription presented here was granted to us on 13 October 2015 by the Hellenic Ministry of Culture and Sports, per medium of Elena Kountouri, Director of the Directorate of Prehistoric and Classical Antiquities. Staff at the Archaeological Museum of Iraklion on Crete have been particularly helpful with much valued comment and advice — namely Stella Mandalaki, Director; Eirini Galli, Assistant Director; Charalambos Kritzas, Director Emeritus. Also, from the Archaeo­logical Museum of Thessaloniki, Styliana Galiniki, Evangelia Stefani, who provided photographs of four stelae (ΜΘ 1684 [sic; not 1689 as in Edson’s IG volume], ΜΘ 1694, ΜΘ 2186, Ρ 91) and, in response to our request dated 28 September 2015, per­mission to reproduce them here. We are grateful to the photographer of two of them (ΜΘ 1684 and Ρ 91), Orestis Kourakis.

On 3 February 2016 Pantelis Nigdelis (University of Thessaloniki) kindly provided us with a photo of IG X, 2.1.558, the sarcophagus still in situ elsewhere in the city (our pl. 4, fig. 7). After we sent him on 8 March 2016 a penultimate draft of our article and a photo of the stone, he clarified by email (received 29 March 2016) that the inventory number ΠΑ 5 on the top of the stone identifies it with a transcription of the inscription marked ‘inv. no. ΠΑ 5 Roman’ which he discovered in a catalogue of the Museum of Byzantine Culture, that is, ‘the former outdoor collection of Antiquities in the court of the church of Παναγία Αχειροποίητος’. He informed us, further, that following the liberation of Thessaloniki in 1912 ‘this church was scheduled to be used as the Byzan­tine Museum of Thessaloniki, a plan abandoned some years later’. With that message he also attached his publication of the text based on the transcription in the Museum catalogue, and a draft in Greek of his proposed updated publication in English of the text in the light of his own further research and drawing on the draft article and photo of the stone and of the squeeze which we had sent to him. These items are detailed in n. 12 below. [1]

In October 1993 Norman Ashton was contacted at the University of Western Australia by a local resident in Perth who had in his possession three fragmentary inscribed marble stelae which had been among the possessions of his long-deceased father, an erstwhile officer in the Australian military forces. On close examination of the stelae Ashton identified, on the reverse or top of each stone, what appeared to be painted Iraklion Archaeological Museum accession or inventory letters/numbers. After he informed the Greek Consul in Western Australia, the three stelae were duly returned by the Consulate to the Archaeological Museum in Iraklion on Crete.

Before sending them to Crete, the Consul informed the media, and in February 1994 articles with accompanying photographs appeared in Perth and Melbourne newspapers. One such photograph was of Norman Ashton with all three of the stones (pl. 1, fig. 1).

Of the three stelae one had not been published (the centre one in the photograph); over twenty years later this still remained the case until very recently. After the stones were sent back to Iraklion Museum, its erstwhile Director, appointed Director of the National Epigraphical Museum in Athens in January 1994, Charalambos Kritzas pub­lished a short article drawing attention to these returns, though he did not publish the previously unknown inscription, [2] whose text includes reference to an arkhi­synagogos and is the subject of this publication.

Obtaining precise information as to how the three stones came to be in the posses­sion of Major James Wilson [3], as stated in October 1993 by his now-deceased son, Derek Wilson, has proven fruitless. Newspaper reportage at the time relied heavily on verbal information provided by the son — as it turns out, not always accurate or reliable in his recollections of when and where his father had acquired the three stones. [4] All we can say is that at some point those stones were acquired by Mr Wilson (it is not known when, where, by what means, or as one group or separately), and remained in his possession until his death in 1963.

The two inscriptions, which had already been published with photographs, and were taken to Perth at an unknown date before finally being restored to Crete in 1994, are: A. Iraklion Archaeological Museum inv. no. E134 published 1935: M. Guarducci, I.Cret. I.xvi (Lato) no. 17 (II BC); and B. Iraklion Archaeological Museum inv. no. E156 published 1942: M. Guarducci, I.Cret. III.iv (Itanos) no. 5 (c. III BC). [5]

The then-unpublished, inscribed stone in Mr Wilson’s Perth garden was not pre­viously held in the Iraklion Museum, but has now been accessioned there. We felt that its provenance was not certain, but our initial hypothesis was a natural one: that it came from Crete in view of the firm provenance there of the other two stones which had also been in Mr Wilson’s possession in Perth. This hypothesis proved incorrect, for after Horsley gave a seminar on the inscription at his university in Australia on 26 February 2016, Paschalis Paschides made contact with us (email dated 28 March 2016) and advised that, despite differences in the transcription, this must be the same stone as Professor Nigdelis had published in 2015 on the basis of a defective transcription in a museum catalogue in Thessaloniki.

Iraklion Museum inv. no. E 433 (date of registration in the Museum: 13.10.1995), though once in the Museum of Byzantine Culture at Thessaloniki with the inv. no. ΠΑ5 (pl. 2, fig. 2 front and 3 top, pl. 3, fig. 4 squeeze).

Fragment of white marble, text complete left and right and below, but top portion broken off. No paint visible on the lettering. No ornamentation was carved on the pre­served fragment. On the broken top face has been painted ‘ΠΑ5’. Dimensions: 0.305(w.) × 0.345(h.) × 0.065(d.) m. Letter height: varies a little, but mostly 0.020 m. (some omikrons 0.015 m.). Interlinear space av. 0.008 m. Lettering very square.

‘[The Herakles(?)] association of the fourth (of the month), when Eula(n)dros son of Zosimos, citizen of Thessaloniki, was arkhisynagogos, for their fellow-member Tiberius Kathekon as a memorial; and Iulia Prokla for her relative, and Epigone for her husband.’

We have printed the text largely as re-edited by Professor Nigdelis in 2016 (see Preliminary note above), though have suggested a possible incipit for the text in the first two lines.

Margins are observed on both left and right sides, such that syllable division is not consistently observed between lines.

The lettering is very clear and the carving quite carefully executed. No lettering is lost in lines 2–10; but only the lower parts of several letters in line 1 are visible. The number of letters per line varies considerably due to ligatures, from 13–18; but 14–15 is the number of letters in the two full lines without ligatures (2, 7), and we have taken that simply as a guide for line 1, and for the line(s) which we infer must have preceded, though the original number of lines prior to line 1 cannot be determined with certainty. At a minimum, at least one line plus space for the top margin must have preceded, but possibly more.

Letter shapes: broken-barred alpha; eta andtheta with wavy horizontal bar; short oblique arms on kappa; epsilon (middle horizontal bar a little shorter than top and bottom bars), lunate sigma, mu and omega all carved very square, the latter two both with internal oblique arms half the height of outer perpendicular hastae.

Numerous ligatures: ΤΗϹ, ΤΕ (1; see below for further comment), Γ (3), Ϲ (4), ΗΚ (5), ΤϹ (the only instance of a ligature joining two words, as also in line 8 for the same letters), ΝΗ, ΜΝΗΜΗϹ (6), ΠΡ, ΤϹ, ΝΓ (8), ΤΗϹ, ΓΝΗ, Τ (9), ΗϹ (10).

Puncts at lines 2, 3, 5, 9; no leaf or other decorative feature to fill up the final line.

Prior to the first partially-surviving line, we suggest that the name of an association (implied by τῷ συνήθι in line 6) began the text; and in view of the unlikelihood that so little as the top half of one line (plus top margin) had broken off (but was not sawn off, however) so nearly horizontally, we infer a loss of at least a little more from the top of the stone. By analogy with IG X, 2.1.288 (Thessaloniki, AD 154; reprinted below, §III.A) and the more fragmentary no. 289 (pl. 1, fig. 5), both of which appear to relate to the same association, and a further Herakles association group published more recently, [6] we have proposed as a suggestion the wording given above, that this could be another association whose patron was Herakles.

All three previously known inscriptions and the new one mention an arkhi­synagogos as the implied leader of the group. A further similarity between the two IG texts and the new inscription is the presence of the funerary cliché in mid-text rather than at the end of the wording (the text mentioned in n. 6 is too fragmentary to allow us to draw any conclusion in this regard). A more general factor is the importance of Herakles in Makedonia. An association of Herakles is not the sole context in which the noun synetheia occurs: IG X, 2.1.291 (Thessaloniki, c. AD II fin.) is an epitaph for a citizen of Thyateira provided by ἡ συνήθεια τῶν πορφυροβάφων ‘of eighteenth street.’[7] However, the wording and certain other features of the new inscription exhibit notable similarities with those two Herakles association texts from that city, as we shall see later.

1 The lower part of several letters towards the right side of this line is all that survives, apart from traces earlier on the line. From the photo and the squeeze we[8] read [ . . ] . . . . . . Ο̣(?)Γ̣(?)ΑΡ̣(?)Ϲ̣Ε̣. Working backwards from the right edge, the final two letters printed seem fairly certain as sigma followed by epsilon; before sigma occurs a gap too wide for iota to be proposed for the upright hasta. Ρ or Τ appear to be the only possibilities — (Π in ligature with sigma is excluded because we should expect psi to have been carved) — and we suggest that rho (or perhaps tau) before sigma is the best fit for forming a word. To the left of that letter the remains of the letter make alpha certain. A narrow, upright hasta precedes that alpha: so either iota or possibly gamma; and, to the left of it, the first visible part-letter on the line which, with its rounded base, allows only omikron or theta. These remains did not yield any obvious wording to us.

We considered whether the genitive form of a name, whose lettering concludes at the start of line 2 and whose nominative is inferred to be -τραϲ, should be proposed. Yet no names with this ending are attested in Dornseiff/Hansen or in Zgusta. A search on PHI yields two names with this ending: Εὐλείτρας (Mysia), and Σωπάτρας (Knidos).

If a word other than a personal name could be entertained, one possibility considered by us was τετρᾶς, ‘quadrant’ of a circle (LSJ, s.v.), the sole attestation in Kretschmer/Locker. Its applicability in the inscription seemed so doubtful that we excluded it. In our view, the best option was to consider the genitive of τετράς ‘fourth’ [day of such and such a month], the sole suggestion in Buck/Petersen. [9] This may imply that a date had been included in the text. Yet against this proposal was the presence of the fairly certain sigma as the second-last identifiable letter in our line 1.

However, Professor Nigdelis whose initial publication in 2015 also identified solely the same last two letters, was able to improve on this substantially and largely convincingly with the aid of the photos of the stone and of Ashton’s squeeze and our transcription in our draft article sent to him. In particular, he drew the inspired inference that the crossbar of a tau in ligature with the final epsilon must have been carved, but is no longer visible due to the loss of the upper half of the letters on that line. So Τ̣Ε̣|ΤΡΑΔΟϹ is confirmed, and our reading of the letters fits well with his conclusion (e.g. his three-letter ligature TΗϹ for our ΡϹ or ΤϹ as the second and third last letters in the line), which drew more sense out of the traces than we were able to.

4 At the end of the line lambda is squeezed in smaller, due to a flaw on the edge of the stone.

5 -ΕΟϹ lapis, for -ΕϹ

5 A superior bar occurs over TI, marking abbreviation: Τιβερίῳ or (less likely) Τίτῳ.[10]

6 -ΘΙ lapis, for -ΘΕΙ

7 Iota smaller between alpha and epsilon

8 -ΓΕΝΙ lapis, for -ΓΕΝΕΙ

9 -ΓΝΗ lapis, for -ΓΟΝΗ.

In view of a catalogue copy of the text found in a now-closed museum in Thessalo­niki by Professor Nigdelis (see our Preliminary note, above), that city or the region nearby must have been the original provenance. This is confirmed by the affinities of the wording and other carving features which it shares with another inscription from Thessaloniki (IG X, 2.1.288, reprinted below at §III.A); this had been suggested to Ashton some years ago by Noy. These other features in common include the notable number of ligatures, and the very square lettering which facilitates that; and Edson’s careful commentaries in the IG volume indicates that this particular text from that city is by no means the only one on which these characteristics in the carving are present.

Nevertheless, it is odd that a citizen of Thessaloniki would so designate himself if the text were being erected there, for that would seem otiose. However, as we shall see, this is not unparalleled: see below on IG X, 2.1.288 (§III.A).

Kritzas made enquiries among museums in Greece in 1994 when the three stones were returned to Iraklion, but no museum claimed the unpublished item.[11] However, Nigdelis has now convincingly clarified the inventory number when, as mentioned above, he came across a transcription of the inscription identified as ‘inv. no. ΠΑ 5 Roman’ which he discovered in a catalogue of the Museum of Byzantine Culture, that is, ‘the former outdoor collection of Antiquities in the court of the church of Παναγία Αχειροποίητος.’ He informed us, further, that following the liberation of Thessaloniki in 1912 ‘this church was scheduled to be used as the Byzantine Museum of Thessalo­niki, a plan abandoned some years later.’ [12]

The word in this new inscription which first catches our eye is ἀρχισυναγώγου (ll. 2–3). The mention of an arkhisynagogos immediately suggests as an initial hypo­thesis that the text is Jewish. However, among the more than fifty epigraphical attes­tations of this title several for non-Jewish associations do occur, mainly from Greece, Makedonia and Thrace (three from Thessaloniki, one from each of Pydna, Beroia, Olynthos and Sophia, plus Chios if correctly restored), which range in date from AD I–III, as well as several less certain instances from these regions and further afield — and across a wider chronological range. [13] This general concentration of all eight of the definitely pagan instances in roughly the north of Greece is a consideration to be given some weight in our determination of the provenance of this new inscription, which we regard as not Jewish.

Our fragment constitutes most of an inscribed text whose beginning is lost, reflecting formal honouring of a deceased member of what appears to be a private association. In our view it is not actually an epitaph ( pace Nigdelis), though it contains some wording typical of that kind of text. The placement of the phrase μνήμης χάριν in mid-text suggests this to us, as do the stone’s quite small dimensions. Furthermore, it is highly unlikely — perhaps even inconceivable — that immediate family such as a spouse would be relegated to a secondary role were it a funerary inscription. Some posthumous honorific inscriptions are devised on a grand scale, e.g. SEG 28.953 (Kyzikos in Mysia, second quarter of AD I). [14] Our stele in complete form was considerably more modest than that Kyzikos inscription, as can be determined by the quite small dimensions of the stone; perhaps it was erected on a wall in the location where the members of the group met. A fragmentary inscription (AD III, first half) of a private society from the city published a decade ago by Nigdelis appears to imply something closely akin to this.[15] We infer that associations such as this made the mounting of such plaques at the place where meetings occurred a way to remember their fellow members, since con­tributions towards the burial of members was commonly one function of such groups. The epitaph proper will most likely have been erected by Kathekon’s wife Epigone and other family members in the regular civic graveyard. A number of posthumous honor­ific inscriptions of similarly modest length are included in Edson’s IG volume for Thessaloniki; there seems to have been a vogue for this form of memorial in the mid-third century. Some of these are endorsed by the city council, and can be distinguished from those which are straightforwardly honorifica for living persons sometimes by the inclusion of a funerary formula, and occasionally by the addition of a word like γλυκύτατος when a woman erects the bomos for her husband or child. Examples (all about mid-III AD unless a specific date is included here) include IG XII, 2.1.188, 195, 196, 199, 200 (AD 261/2; n.b. line11, τιμῆς καὶ μνήμης ἕνεκα), 205, 206, 207 (note especially lines 14–15 where the mother provides the honouring memorial εἰς παραμυ|θίαν ἑαυτῆς). That subsidiary information (e.g. a fine for tomb violation such as on IG X, 2.1.588, reprinted below, §III.C) is not infrequently added after the ‘in memory’ wording on gravestones and other tombs.

Some of the names occurring in this inscription are prevalent in many regions (e.g. Zosimos, Epigone), while some others are far less frequently known.

Euladros (line 3) is attested rarely, known only in Hellespontic Phrygia. Related spellings include: Eulades (LGPN 1: Hephaistia on Lemnos, IV BC), Eulandros (LGPN 4: 7 examples from Makedonia, including one from Thessaloniki, IG X, 2.1.31.17, 27 BC–AD 14). Our instance of the name is most likely to have involved a small oversight of one letter by the mason for Eulandros, a proposal made independently by Nigdelis.

Kathekon (line 5) is very rarely attested as a personal name: two instances from Lakonia are listed in LGPN 3A, dated c. AD 70–100 and c. AD 160, and one Latinised version of the feminine Kathekousa from Imperial-period S. Italy. [16] Noy has also drawn our attention to an instance from Kos: R. Herzog, Koische Forschungen und Funde, Leipzig 1899, no. 97, where the name of the deceased is carved as Κατήκων. In our new inscription the deceased was a Roman citizen.

Iulia Proc(u)la (lines 7–8) was born a Roman citizen since she possesses a Roman nomen. Prokla is well attested in inscriptions: e.g. LGPN vol. 1 lists the name as attested on Cyprus, in Cyrenaica and on Lesbos, and vol. 5B has 15 instances from several provinces in Asia Minor. From Thessaloniki, Edson’s indexes in his IG volume attest a little over half a dozen males and females with this name (nos. 386, 581, 737 [a brother and a sister], 743, 828, 851), all dated AD II or III.

The Epigone in our inscription was not a Roman citizen, but presumably a local citizen (since she was married). Solin, GPR 2.964–65 lists 37 individuals so-named at Rome (over half incerti, but nearly one-third servile or freed). [17]

At least three inscriptions from elsewhere contribute to the interpretation of the new text. Brief notes are included on each prior to proposing how they elucidate the new inscription. First, the wording of the new inscription shows marked affinities with a well-preserved posthumous honorific inscription from Thessaloniki, already alluded to above.

A. IG X, 2.1.288 (Thessaloniki, AD 154, perhaps January) [18]

[19]

Metronymics (lines 5 and 9) in Makedonia are not especially unusual.[20] The arkhisynagogos is here the president of the association, and other office-bearers are listed. This same man, Kotys, is also attested holding the same office in the much more fragmentary IG X, 2.1.289 (pl. 1, fig. 5), which Edson dated c. AD 155. Sufficient lettering survives on that stone to connect it with the same association of Herakles. In no. 288 ἐπιμελητής can refer to a variety of functions, as LSJ attests. Here, ‘financial officer/ manager’ seems the most appropriate to us, though other options could be con­sidered.[21]

Inscription no. 288 has a notable oddity about it. Although it appears to have been found at Thessaloniki, one of those named identifies himself in an entirely otiose manner as a citizen of that city. With two exceptions, this is the only instance in Edson’s volume where a stone emanating from there carries such an identification. The addition of Thessalonikeus three times on IG X, 2.1.38B (pl. 6, fig. 9) is entirely explicable as an exception since this is a list of victors in the local games honouring Pythian Apollo in AD 252/3, and each of the 25 winners has his ethnic (two individual winners possess double citizenship) attached to his name. The other exception is IG X, 2.1.68 (AD I fin.), a list of members (over three dozen of them) of a private association, only one of whom is identified (line 28) as Θεσσ(αλονικεύς). This same individual appears in no. 69.28 (AD I, second half), a list of mystai of Theos Hypsistos; but he is not stated there to be a citizen of the city.[22] So, while we do not question the accuracy of Edson’s reporting of the find-spot of no. 288 in the city, a suspicion is raised whether it had been brought to the city at some time from elsewhere in the region. (The same is perhaps true for no. 68.) This may provide a basis for proposing that the new inscription is from the same location (given the similarity of wording with the Herakles association text): it was not originally from the city of Thessaloniki itself, but was brought in to the city by someone else or on a separate occasion from no. 288, and hence was deposited with another collection of inscribed monuments, where it was given what we take to be an early number (ΠΑ5) in its catalogue. From this we might infer that it arrived in the city only a little before that museum was in the process of being established.

Although they are not identical in every respect, inscription no. 288 and the just-published one have several common features: the locating of the posthumous honorific wording in the centre of the text, the title of the current leader of the association, and the terms for ‘association’ and ‘association members’ which are employed. [23] In these two texts from Thessaloniki (i.e. no. 288, and accepting Edson’s restoration by analogy in no. 289) the sense of the plural must be something like ‘club members’, i.e. members of the Herakles association. In view of the similarities (not fully identical, however) in phraseology and club officials’ titles, with due caution we infer that the new inscription may also have had Herakles as the group’s patron and focus. That terminology is not confined to the Herakles association or to religious associations in the city, however, as is shown by IG X, 2.1.291 (dated AD II fin.) noted above, an epitaph provided by the guild of purple dyers for one of their members. The similarly very square lettering and use of many ligatures in both texts — and cf. the related but fragmentary no. 289 with our pl. 1, fig. 5 — is also worth mentioning. We have inferred that the new text may also be reflective of a Herakles society. Nigdelis is right to give weight to τ̣ε̣τράδος, the word he convincingly read in lines 1–2; so we may have a ‘kindred’ association, rather than the very same one.

Next, we focus on Iulia Prokla, one of the people involved in the memorialising of the deceased Tiberius Kathekon, himself a Roman citizen. Although the name Iulia is ubiquitous and Prokla is attested epigraphically in a number of provinces, as already mentioned, Prokla in combination with the gentilicium Iulia is exceedingly rare in Greek inscriptions, known by only two other instances, from Rome and from Thessa­loniki. Eleven further inscriptions, all Latin epitaphs from Rome, mention a number of women with this name. Whether there is any link to be discerned between any of those texts and the woman named in the new Greek text will be evaluated later in §V.

B. IGUR 2.2.732 (pl.); Rome, no date proposed by L. Moretti, but H. Solin, GPR 2.576 suggests AD II/III; (pl. 4, fig. 6) [24]

[25]

2 Laevia Ithake is regarded as an incerta by Solin, GPR ibid.; however, we propose below that at the time of her death she had been freed, and had then married and subsequently bore her second child, Iulia. The name Ithake is very rare at Rome (three instances listed in GPR).

6 mid-height punct either side of the name Nomeus, for which Solin, ibid. attests six instances: 1 freeborn, 2 liberti and three incerti (including this instance). However, we propose below that Nomeus was born to his mother while she was still a slave, and therefore was a libertus now at the time of this epitaph for her.

This monument is finely carved; it is not one commissioned by impoverished people. We infer that Iulia Prokla (and her brother?) had money. This is of some con­sequence for the argument which will be developed below.

C. IG X, 2.1.558 (Thessaloniki, c. AD mid-III [Edson]; pl. 4, fig. 7)[26]

(vacat)[27]

1 Ῥουτινιανῷ, lapis. For the more usual spelling of this man’s name see IG X, 2.1.204. Names in -ianus become common in the Roman East from AD II onwards, indicating variously adoption, or filiation, or neither but simply a pointer to Romanisation. [28]

2 Whatever σύγκληρος means, at least we can say that the two daughters of Philotera had sufficient stake in the burial of Rufinianus to be named in the inscription on his tomb. The adjective is not attested by LSJ with a substantival use, though the sense ‘relative’ (cf. LSJ, s.v., 1) would be perfectly appropriate here. That one of these women shares the same name as the deceased and his sister is suggestive of some familial connection with the deceased. [29]

3 The wording ἐκ τῶν ἐκείνου ἐκείνῳ is uncommon, and much rarer than the ubiquitous ἐκ τῶν ἰδίων, but clear in referring to the financial assets of the deceased which he authorised to be used for his proper burial. By the practice of setting out burial requirements as part of a will, the deceased circumscribes what is to be done, and how, and how much is to be spent on it, and can even set aside the money in the will to make it happen.

5 The mention of denaria in this text raises the question whether the specified amount is an aid to dating it. [30] Research undertaken in quite recent years on the monetary value of fines on internally-dated funerary monuments in different provinces confirms the general impression of the fine amount increasing from circa the 270s as empire-wide inflation begins to be felt. Not all regions reflect this trend at the same pace, nor is there consistency in the increase of the fines to be given to the imperial treasury (as in the case here, in view of the phrase εἰς τὸν φίσκον προστείμου) or the local city. We may infer, then, that the determinations as to amount were made either locally on a city-by-city level by authorities there, or simply arbitrarily by those responsible for erecting the memorial. After all, the function of the fine sanction is that of a deterrent to ward off mistreatment of the dead and their burial places. In this respect, the sanction serves a similar role to curses on epitaphs, however differently articulated and motivated (curses implying a divine threat). The 2,500 denaria amount in this inscription is the sum most frequently occurring on internally-dated epitaphs from both Northern Greece and Asia Minor, particularly in the fifty-year period from AD 220, though by no means absent from both earlier and later years. That said, it remains true that there is much variation, smaller amounts in later AD III no less than larger amounts in AD I and II. Accordingly, since the motive for the fine clause is not primarily a money-making one by administrators (whether the fine is to be directed to the city or to the imperial treasury), but rather aims to ward off tomb violation, could it be that the amount of the fine is more a reflection of the importance which the family/tomb erectors place on them­selves and their own local status? This suggestion may merit consideration in particular because the amounts vary so much, and so are presumably not determined by civic or other authorities. Although it is beyond the scope of our discussion here, we raise the possibility that there may also sometimes be a correlation between the amount of the fine and the quality of the funerary monument. Were this so, it also could suggest something about the family’s/erector’s own per­ception of their social standing locally. In general, then, the size of the fine could be determined primarily not by date or region, but mostly by the status which the family perceives itself to merit in its local context. This is not to be considered always the sole determiner, of course, since there may be other factors which we cannot now retrieve. A smaller amount relative to a perceived norm need not indicate automatically that a family accepted a low standing for itself locally, for that would fly in the face of the deeply-embedded givens about rank and status in the Roman Empire. If this suggestion has some merit, it could follow that smaller amounts may also occur by the later third century when there is a trend towards an increase in the size of the fine sanction. In the case of this inscribed sarcophagus, the prevailing consensus is that the amount of money mentioned confirms a date of the first half of AD III (Stefanidou-Tiveriou) or mid-III (Edson, Nigdelis). We suggest that the initiators of the memorial (whether the still-living heirs or the deceased via his will) may have felt their profile as a locally elite family in Thessaloniki was sufficiently high not to need to assert their status by imposing a much larger financial sanction, and simply opted for the ‘default’ (i.e. most commonly occurring) sum.

While the original location of this locally-produced sarcophagus is unknown, in the 19^th century, at least, it was situated outside the city walls.[31]

The name Iulia Prokla is very uncommon in inscriptions using Greek. PHI identifies two Greek texts only, the ones we have included above in §III.B from Rome and §III.C from Thessaloniki. These two inscriptions have not previously been connected; and the adjacent dates of publication of Edson’s IG volume (1972) and Moretti’s IGUR volume (1973) understandably precluded either of them making a connection, even though both texts had long been known in other publications. Now with the new text (§II, above) we can identify a third stone of approximately similar date and inscribed in Greek on which this name occurs. We propose as a cautious hypothesis that we may have attes­tation of three stages in one woman’s life.

First, at Rome Iulia Prokla joins her brother in memorialising their mother (§III.B, above). Unlike them both, she is born a Roman citizen as she has no Greek name. How­ever, her brother and mother do, and they share the same gentilicium. It follows that Ithake (presumably a slave brought to Rome from that Greek island) fell pregnant to her owner Laevius while still in servitude; and the child born was Nomeus (‘herder’), suggestive of the servile contribution it was anticipated he would make on an estate for his owner. Both mother and son were later manumitted, and took their former owner’s name while retaining their Greek name in the regular manner. Now a liberta, Laevia Ithake married, and the issue from that marriage was Iulia Prokla. So her father, the husband of Ithake, was a member of the Julian gens. It follows that Iulia Prokla and Laevius Nomeus were step-siblings, and she is listed ahead of her half-brother on the epitaph even though he is older because she was free-born and he was not: in formal terms she has superior rank. These youngsters cannot have been very old at the time of their mother’s death aged 32. Yet they are not young children, since they have taken this initiative themselves. That Ithake’s husband Iulius is not mentioned may be due to his having predeceased his wife. The quality of the memorial for their mother suggests that at least one of the two children had inherited; this is more likely to be the freeborn teenage girl (in view of her father’s probable death prior to her mother’s) rather than her freed, formerly servile step-brother. At most she may have been 13–15, i.e. already an adult in Roman terms, and Nomeus slightly older. At the time of her mother’s death, Iulia Prokla was not yet married though of marriageable age (i.e. a parthenos). [32] That she (as we have inferred) expends money on a fine memorial for their mother (as pl. 4, fig. 6 shows) suggests that she was old enough to take such an initiative. No date is provided by Moretti for this inscription; but the lunate letters suggest at least second century (though probably later), and Solin has already proposed second or third century.

Second, IG X, 2.1.558 (§III.C above) gives us a further glimpse of (as we propose) the same Iulia Prokla in Thessaloniki at a slightly later stage of her life, perhaps only a few years after her mother’s death. She had moved from Rome to northern Greece, and had married, though apparently had no living children at the time of her husband’s death. Whether her father, whom we have inferred to have been a Iulius in view of her nomen, had familial connections in Thessaloniki, we cannot say. Yet it is noteworthy that she marries into a family where the Julian name is prominent. Her husband — we can clarify his full name from IG X, 2.1.204 as C. Iulius Rufinianus Artemidoros also called Melanthios — from the Julian gens and possessing more than the straightforward tria nomina, has a Greek name suggestive of an origin in the east of the Empire rather than in Italy. His Julian lineage via his mother appears to stem at least from his maternal grandfather who is named in IG X, 2.1.204. Since he has a Julian gentlicium, whereas his sister Aelia Naevia lacks it, we should allow the possibility that in the previous generation their mother Iulia Artemidora first married a Iulius with whom she had a son, the bouleutes who later married Iulia Prokla (IG X, 2.1.558). That first husband died, and subsequently (presumably within a couple of years, at most) [33] Iulia Artemi­dora remarried, this time to P. Aelius Rufinianus; and their child was Aelia Naevia Artemidora, to whom were given names derived from both her parents. The different gentilicia of the children are suggestive of their being step-siblings, just as we have inferred for Iulia Prokla and her step-brother Laevius Nomeus. The fact that P. Aelius Rufinianus is named in the inscription honouring C. Iulius Rufinianus when he becomes a bouleutes (IG X, 2.1.204; inv. no. Ρ 91; pl. 5, fig. 8) does not pose a problem for our surmise. We have suggested that his actual father had died, so the step-father is included in the congratulatory inscription — though it is noteworthy that the maternal grand­parents are inferred to be more prominent, as the ones who have taken that initiative, listing them­selves at the end of this rectangular bomos of white marble — and that grandfather is a Iulius.

As already mentioned, names in -ianus (such as Rufinianus here) enjoy a vogue from the second century; and by the third century Roman nomenclature in the Greek East especially becomes more varied and less systematised vis-à-vis the tria nomina. The fact that Iulius Rufinianus’s sister Naevia Artemidora (whose full name can be reconstructed in conjunction with IG X, 2.1.205 as Publia Aelia Naevia Artemidora) is also included in the sarcophagus text permits the inference that Thessaloniki was in fact the home city for that family. This is confirmed as well by IG X, 2.1.204 (mentioned just above), on which Rufinianus is honoured by his maternal grandparents for being appointed a member of the city council. Edson’s stemma for the family, included in his commentary on that inscription at p. 86, draws upon several inscriptions from the city (nos. 38B, 204, 205, 558) to identify four generations of the family. On the basis that no. 38B (pl. 6, fig. 9) is dated internally to 252/3, he dated to the mid-third century the other three inscriptions relating to the family. It is unclear whether Aelia Naevia Artemidora was already married at the time of her brother’s death. Unlike her sister-in-law, Iulia Prokla had no children when her own husband died. The fact that Iulia Prokla’s name is absent from IG X, 2.1.205 (ΜΘ 1694; white marble bomos, dated by Edson to AD mid-III; pl. 5, fig. 10), the tombstone which Aelia Naevia and her son provide for her husband, suggests that she was not in the city when her brother-in-law died.[34]

Others are also specified on no. 558, the funerary inscription for Rufinianus, though as mentioned above it is not certain what σύγκληρος means in this context. In any case, as apparently happened with her mother’s burial at Rome, Iulia Prokla is explicitly the one who makes the decisions about her husband’s funeral and burial.

Third, consideration may now be given to the new text (§II above). The name Iulia Prokla is once more present. This woman is involved with memorialising, together with the widowed Epigone, a ‘relative’ — hardly an otherwise unattested brother, as this would surely have been specified just as Naevia Artemidora is on the epitaph for Iulia’s husband at Thessaloniki (IG X, 2.1.558 = §III.C, above). We infer, then, that Tiberius Kathekon was a relative by marriage, i.e. a relative on her husband’s side. Here the two women are not the ‘main players’ in this memorialising of the dead man, close as their relationship to him undoubtedly must have been. For (as argued earlier) in our view this text is not an epitaph tout court, but an honorific memorial accorded to the dead person by others who were fellow members of a private association. It is an indication of Iulia Prokla’s rank that she is named before the widow who lacks Roman citizenship; the ordering of the relatives’ names there may remind us that a woman with the same name was listed before her older step-brother on the gravestone for their mother at Rome (§III.B, above). Her prior position reflects her superior rank.

Iulia Prokla lived her short marriage at Thessaloniki, and buried her husband there. By virtue of that marriage she acquired some relatives on her husband’s side. We know of one sister-in-law, at least, Aelia Naevia Artemidora mentioned on the sarcophagus in that city; and in view of the word that Iulia uses to describe her link with Tiberius Kathekon in the new inscription, his wife Epigone was Iulia’s relative by marriage. Other relatives by marriage can be adduced: Edson provides a stemma of the family of Iulia Prokla’s husband as part of his commentary on IG X 2.1.204. We can now augment this somewhat with details of Iulia’s family in our revision of that stemma (§VI, below). In summary, in view of the common provenance of the two inscriptions as Thessaloniki or nearby, there is a strong case for identifying Iulia Prokla in the new inscription with the homonymous woman in IG X, 2.1.558. That she is to be identified with the woman of that name in IGUR 2.2.732 is not provable, but well worth con­sideration for the reasons already given.

Hitherto, we have considered only Greek epigraphic attestations relevant to the new text. However, the name Iulia Procula is attested in the following eleven Latin inscriptions from Rome in CIL VI. Although no dates are proposed in CIL for any of these texts, we infer that most (if not all) are of Imperial date, and from the first two centuries of the present era. The type of stone is not always indicated, though marble predominates. All these texts are quite modest memorials, with brief and fairly formulaic wording: where complete, they have DM, name of deceased, name(s) of the memorialiser(s) and relation­ship to the deceased, and sometimes the latter’s age.

Fasc. 2.10176 (marble), on the back of an epitaph erected by his parents for a two-year-old child, Iulia Procula is memorialised by her husband, a murmillo veteranus.

Fasc. 3.19986 (marble), C. Iulius C. f. Faustinus and Iulia C. f. Procula memorialise their son.

20561 (marble), on a double tombstone, Iulia Procula, aged 26, is memorialised by her husband Iulius Atimetus and a slave(?) Myrtus.

20640, Iulia Procla provides a tomb for her father C. Iulius Agathopus, for herself and for her liberti and libertae and their descendants.

20642 (marble), tombstone of Iulia Procula and Iulius Glycerus.

20643, Sariolena Aug<<e>> memorialises her daughter Iulia Procula, aged 27.

Fasc. 4, 1.35604, a woman (unnamed) provides a tombstone for her daughter Iulia Procula and her husband Iulius Fidelis; the order of the names suggest that the latter is the son-in-law of the mother.

Fasc. 4, 2.37239 (marble fragment), funerary monument for a soldier erected by (his wife?) Iulia Procla.

37811 (marble), M. Caecilius erects a tomb for his wife Magna Iulia Procula, aged 22.

38205 (marble), Claudia Thalia, daughter of Tiberius, and Iulia Procula, daughter of Gaius, provide together the memorial for Claudia Zosime their mother and grand­mother, respectively.

38796 (marble), Tiberius Claudius provides a tombstone for his wife Iulia Procula, daughter of Gaius.

The majority of these CIL epitaphs can readily be ruled out (and two on more than one ground) as referring to the Iulia Prokla of IGUR 2.2.732. In that epitaph:

i. Her mother is Laevia Ithake, so we may exclude 20643, 38205.

Further, on the basis of the hypothesis that all three of the Greek inscriptions of approximately similar period on which the name occurs may point to the same indi­vidual,

ii. if she were married at Rome prior to marriage to a provincial family of consider­able local standing, a marriage at Rome to a gladiator or to a soldier is highly unlikely, ruling out 10176 and 37239. And

iii. given her marriage to C. Iulius Rufinianus Artemidorus at Thessaloniki where she buried him prior to the birth of any children, epitaphs for Iulia Procula at Rome are excluded: 10176, 20561, 20642, 20643, 35604, 37811, 38796.

This analysis of the eleven CIL VI attestations of the name at Rome, leaves two possibilities for an identification with the woman mentioned in the new inscription: nos. 19986 and 20640. In the first of these, Iulia Procula and her husband (also a Iulius) bury their son. This fits well with the suggestion already raised that at Thessaloniki Iulia marries into a provincial family of considerable local visibility for whom the same gentilicium is visible for three generations on the maternal side through her mother-in-law. So we would have to infer that Iulia contracted a first marriage with a member of the Julian family at Rome and bore him a son who died in infancy. If that epitaph is to be linked with our Iulia Prokla, then her husband must have died (and of that we have no surviving trace) in view of her marriage to Rufinianus at Thessaloniki. There are too many imponderables raised by the attempt to link that CIL epitaph with the Iulia who is our focus, and we therefore conclude that it should be treated as highly unlikely.

Concerning the remaining CIL text, no. 20640, caution is required not to over­emphasise the spelling of her name; but the decision to memorialise her mother in the IGUR text in Greek with the orthography Prokla (as against the latinised Procula) allows us to note two instances of the graecised spelling in Latin at nos. 20640 and 37239. The latter is too fragmentary to permit us to infer more than that she married a soldier. But, in the former we see that her father was a Iulius, and a libertus with an originally Greek servile name: C. Iulius Agathopus.

In this epitaph, a Iulia Procla is the sole person involved in memorialising her father. His name and formal rank would make him an appropriate ‘fit’ as husband of the liberta Laevia Ithake, and also as father of a freeborn daughter carrying the Julian gentilicium. To the gradual building of such a circumstantial case, we may now recall the com­paratively ornate (hence, expensive) gravestone for her mother (pl. 4, fig. 6). Whether her father died before Iulia’s mother Laevia Ithake cannot be settled with certainty, though we have raised that likelihood earlier (in §IV) due to the absence of his name from Ithake’s tombstone. In that case, it would be odd that this man’s wife is not named in CIL no. 20640, were she still alive.

It is tantalising that a homonymous man at Thessaloniki in AD I is attested in IG X, 2.1.259, col. 2.21 as a member of a private association whose religious interest focused not only on Zeus and Dionysos, but also on at least one Egyptian deity, Sarapis. With others, C. Iulius Agathopous contributes to a gift to Zeus Dionysos Gongylos. [36] But the epitaph above from Rome implies that his daughter Iulia Procla was the sole heir in view of the plans she had already conceived for further burials in that tomb; and a certain (though precisely indeterminable) level of wealth is suggested by the wording of that text. She acts alone, so we should infer that she was not yet married. This Iulia Procla is clearly independent as she is described as having liberti. However, were there acceptance of the AD I–II date suggested above for the quite unornamented memorial reprinted here from CIL, and AD I for the mystai donative from Thessaloniki, then a link with the second- or third-century Iulia Prokla who buried her mother at Rome fades from view. [37]

Regarding these Latin attestations of the name Iulia Proc(u)la at Rome, therefore, we conclude that none of them (not even the least unlikely candidate, CIL VI, 3.20640) is to be connected with Iulia Prokla in any of the three Greek inscriptions which mention a woman with that name.

We reproduce here Edson’s stemma for the family from Thessaloniki (from his IG X, 2.1 p.86), slightly revised and with supplements, contingent upon the hypothesis advanced above, providing details about Iulia Prokla’s relatives at Rome (see Appen­dix).

Edson’s reconstruction of the family generations is not the only possible one — some other options can be raised, due to the increasingly fluid naming patterns of the third century. [38] Our revision of Edson’s stemma reflects a number of proposed alter­ations to the familial connections as Edson conceived them.

The points of difference in the stemma are:

a. C. Iulia Artemidora may have had a prior marriage to a Iulius, which would account for the nomen Iulius in her son, C. Iulius Rufinianus Artemidoros (see §IV, above).

b. The gentilicium of P. Aelia Naevia Artemidora indicates that she is the daughter of P. Aelius Rufinianus Proclus (and C. Iulia Artemidora); consequently she is a step-sister of C. Iulius Rufinianus Artemidoros (see §IV, above).

c. We suggest that P. Aelius Rufinianus Nikostratos, registered as the boys’ pent­athlon victor in 252/3, is more likely to be identified with P. Aelius Rufinianus Proklos, also called Nikostratos, the second(?) husband of C. Iulia Artemidora, than that he is a third son of that marriage.

d. As Edson made no link between the Iulia Prokla known from Thessaloniki both in the sarcophagus inscription for her husband and in the new inscription, and the Roman family connections of Iulia Prokla in the IGUR epitaph, that is an addition to his stemma which we have proposed as a hypothesis (see §IV first paragraph, above).

On the basis of our hypothesis, Iulia Prokla, whose mother must have married into the Julian gens at Rome after her manumission, herself marries a Iulius whose mother and maternal grandfather were also members of that gens. Her sister-in-law married a Iulius, and the son of that union carried the nomen as well. So the Julian gentilicium, still strongly attested in AD II, was a feature of this local aristocratic family in provincial Thessaloniki, though Aelii are also prominent. This suggests that, even in a period of onomastic flux, at least some provincial elite families preferred certain gentilicia via the choice of marriage partners for their children. It may or may not be coincidental that Iulia Prokla’s father-in law, P. Aelius Rufinianus Proklos (also called Nikostratos), carries the same cognomen as she. There is too little basis on which to extrapolate any conclusion from that name being shared by them.

As for the matter of fluid naming patterns, an instance reflective of the practice, and relating directly to this family, is to hand. Iulia Prokla’s sister-in-law is named on the epitaph for Iulia’s husband (IG X, 2.1.558, printed above at §III.C) as N(a)evia Arte­midora, whereas on no. 205 — the epitaph for her own husband — she names herself as P(ublia) Aelia N(a)evia.

i. A definite provenance of Thessaloniki or its environs for the new inscription has been settled, thanks to Professor Nigdelis’ discovery of a copy of the text made perhaps a century ago at a now closed museum in the city.

ii. We regard this new text as a posthumous honorific, not simply an epitaph, on the ground of its dimensions, the different position within the text of the standard funerary formula, and the lesser role accorded to the immediate family of the deceased. Its size (and notably, thinness) is suggestive of a plaque affixed to a wall in a room or building where meetings occurred of the association of which Tiberius Kathekon was a member. The dimensions and the position of the funerary cliché of IG X, 2.1.288 (§III.A above) provide a pertinent analogy, except that no family members are included.

iii. The Iulia Prokla in the new text from Thessaloniki or its environs is almost certainly to be identified with the woman of that name who is responsible for providing the inscribed sarcophagus at Thessaloniki for her husband. Whether she may also be the same person at a younger age in the monument from Rome is a possibility we have advanced as a hypothesis. This linking of the three texts is proposed on the basis of their approximately similar date, and the fact that they are the only inscriptions attesting this name which happen to be in Greek. In consequence, we may cautiously draw some inferences about her mobility (both physical and social) and independence. In particular, as the child of a liberta (maybe even of liberti) she has married into an elite provincial family, her husband having been a bouleutes and her brother-in-law, T. Iulius Lysimakhos, a bouleutes ἐκ προγόνων (see §III and §IV, above).

Appendix

Tafel 1

Tafel 2

Tafel 3

Tafel 4

Tafel 5

Tafel 6