¹ I would also like to thank Helmut Koester and HTR for permitting us this forum for what I hope is a productive scholarly exchange on interpretation of archaeological data from the world of early Judaism and Christianity.

² Runesson, Because (Anders Runesson, “The Oldest Original Synagogue Building in the Diaspora: A Response to L. Michael White,” HTR 92 [1999] 409, n. 1)Google Scholar comments on the different versions of my own discussion in various publications, I should simply state that the HTR article ( White, L. Michael. “Synagogue and Society in Imperial Ostia: Archaeological and Epigraphic Evidence,” HTR 90 [1997] 23–58CrossRefGoogle Scholar ) was the most complete, despite the fact that it appeared in print before that in Donfried, Karl P. and Richardson, Peter, eds., Judaism and Christianity in First Century Rome (Grand Rapids: Eerdmans, 1988).Google Scholar The reason is simple: the latter article was submitted in draft form first, and some important changes and corrections that I later incorporated into the HTR version could not be included. For the sake of the present discussion, I consider the HTR article and the summary in my work, The Social Origins of Christian Architecture, vol. 2: The Christian Domus Ecclesiae in its Environment (Valley Forge, PA: Trinity Press International, 1997) 379–97Google Scholar to be the clearest articulations of my argument up till now.

³ Here I refer primarily to the following articles of Maria Floriana Squarciapino, which have by far been the most accessible and the source for most of the secondhand reporting of the site: “La sinogaga di Ostia: secondo campagna di scavo,” in Atti di VI Congresso internazionale diarcheologia cristiana, 1962 (Studi di Antichita Cristiana 26; Rome: Pontifical Press, 1965) 299–315Google Scholar , and “The Synagogue at Ostia,” Archaeology 16 (1963) 194–203Google Scholar.

⁴ Idem, “The Synagogue at Ostia,” 203. Here, note Runesson's statements regarding the discovery of the benches in D (“Oldest Original Synagogue Building,” 421, and n. 44) because I had clearly used the discussion just noted in the Archeology article. It was my mistake not to include the above citiation in my n. 27; however, pace Runesson, their presence was already documented in the second campaign.

⁵ See idem, “The Synagogue at Ostia,” 203, where these three steps are spelled out. Compare idem, “La sinogaga di Ostia: secondo campagna,” 306.

⁶ Runesson basically accepts Squarciapino's identification and dating of the original building as a de novo synagogue edifice, as well as her three-stage process of renovation. As I will indicate later, I have some serious reservations on all three points.

⁷ Here I think that Runesson (“Oldest Original Synagogue Building,” 412, n. 8; 423, n. 54) has understated the reservations expressed long ago by Kraabel, A. Thomas in his article, “The Diaspora Synagogue: Archaeological and Epigraphic Evidence since Sukenik,” ANRW 2.19.1 (1979) 477–510Google Scholar , esp. 498-500. Even though Kraabel originally followed Squarciapino in identifying the first stage as a synagogue, based on the curved wall of D (p. 498); he questioned the “monumentality” of the design and the ‘temple-like” colonnade, certainly for the first stage. He suggests that a “simpler design would be more likely” (p. 499). He also questioned Squarciapino's stages of renovation, especially as they relate to the torah shrine (p. 500). Runesson cites these views in passing only at the end of his notes (n. 101).

⁸ , White, “Synagogue and Society,” 23–58.Google Scholar

⁹ , Runesson, “Oldest Original Synagogue Building,” 413.Google Scholar

¹⁰ Reprinted as The Social Origins of Christian Architecture, vol. 1: Building God's House in the Roman World: Architectural Adaptation among Pagans, Jews, and Christians (Valley Forge, PA: Trinity Press International, 1996) 62 (cited in Runesson, “Oldest Original Synagogue Building,” 413, n. 16).

¹¹ White, Social Origins of Christian Architecture, 1. 69, where I also cite the excavator's view that it was a de novo synagogoue building.

¹² In my later discussions (“Synagogue and Society” 17; and Social Origins of Christian Architecture, 2. 391), I was more cautious in calling it an “insula complex.”

¹³ , Runesson, “Oldest Original Synagogue Building,” 420–21.Google Scholar

¹⁴ So notice “Synagogue and Society,” 27, n. 15, where I use the term insula this way; it comes from the standard nomenclature (as used in all reference works on Ostian archaeology) for numbering site locations in Ostia by region (I-V), followed by insula (meaning “block”), then by edifice. For an exemplary study of one such insula block see Boersma, Johannes, Amoenissima Civitas: Block V.ii at Ostia: description and analysis of its visible remains (Assen: Van Gorcum, 1985).Google Scholar In its final form this insula contained 11 different edifices, including two warehouse buildings (nos. 2, 14), one apartment building (an insula of the other type, no. 3), four smaller houses (nos. 9-13), some assorted shops (nos. 1, 9, and miscellaneous other rooms), and three large buildings: the House of the Fortuna Annonaria (a collegial hall, no. 8), the Baths of the Philosopher (probably another collegial hall, nos. 6–7), and the House of the Porch (a large domus, nos. 4–5). Originally, there might have been as many as 14 different edifices, and almost all of these individual buildings went through significant stages of renovation to reach their final forms. This fact is indicated in the edifice numbers in which two earlier buildings were combined to form a new, larger edifice.

¹⁵ See Packer, James E., “Housing and Population in Imperial Ostia and Rome,” JRomS 57 (1967) 8–95Google Scholar and his monumental study, The Insulae of Imperial Ostia (Memoirs of the American Academy in Rome 31; Rome: American Academy, 1971).Google Scholar See also Hermansen, Gustav, Ostia: Aspects of Roman City Life (Edmonton: University of Alberta Press, 1981) 17–54Google Scholar.

¹⁶ As Runesson also grants (“Oldest Original Synagogue Building, 414); however, he seems to want to segregate K physically, archaeologically, and socially from the synagogue edifice.

¹⁷ And this is precisely how I phrased it in my article “Synagogue and Society,” 33.

¹⁸ As shown by Boersma's careful study of the buildings in block V. 2, based on detailed masonry analysis (see n. 7 above).

¹⁹ Runesson, “Oldest Original Synagogue Building,” 417, fig.2.

²⁰ Ibid., 414.

²¹ Here Runesson's discussion is essentially correct. It must be remembered, however, that my plan was the first to assign letters and numbers to these areas for clarity of reference in discussion. No such referencing system appears in Squarciapino's earlier articles, and the lack of one makes disciphering the different areas through distinct layers more difficult. Also, I have now had a chance to see under the podium to confirm the wall spur and note that it was an integral construction with the south wall of D. The wall spur also makes it clear that this partition wall between D and C₃ was only cut down at the time of the construction of the aedicula, since it was used as part of the podium base. Also, a portion of the wall, bonded into the south wall of D, was preserved to the height of the aedicula's pediment. It was cut back and the tufa quoins apparently redressed in order to be used as the front edge of the curved niche wall on the south side. This feature was already reflected in my earlier plan of the edifice (= Runesson, “Oldest Original Synagogue Building,” 415, fig. 1).

²² See White, “Synagagoue and Society,” 58, fig. 5, also reflected in the masonry profile in fig. 4.

²³ It is ca. 1.37 m wide (or approximately 4.5 Roman feet); for comparison, the doorway in the original reticulate wall of F/G measures only ca. 1.18 m (approximately 4 Roman feet).

²⁴ See my discussion in “Synagagoue and Society,” 35 and n. 34.

²⁵ Ibid., 35 and my photo of the synagogue's north wall of room D (fig. 5).

²⁶ Ibid., also marked in my fig. 5.

²⁷ Also there is a dissociated fragment of either a door or window frame still stacked among wall fragments in room D. Its original position is not noted in any report of which I am aware, although it may be catalogued among the excavation's logs, which I have not had the privilege of seeing.

²⁸ For further discussion of Runesson's fig. 2, see the Appendix at the end of this article.

²⁹ Beyond this point on each end its line cannot be confirmed, because the walls are not extant to this height; however, from the perspective of the masonry type, one would expect it to continue.

³⁰ It was on the basis of this brick band, by analogy with the position of the windows on the north wall, that 1 had suggested that there could have been windows at this lower level on the south wall (“Synagagoue and Society,” 35, n. 34; pace Runesson, “Oldest Original Synagogue Building,” 416, n. 28). But again this proposal is merely a structural possibility suggested by the nature of the masonry; there is no direct evidence.

³¹ This question cannot be answered at present, but I note that there is no extant portion of the wall from ground to full height in the restoration on site. The extant upper wall section surrounding the window has been installed above a modern restoration section. Also, this section of restoration work has failed to include the lower band of reinfocing brick at a lower level (ca. 1.5 m above the bedding course), preserved in situ on the south wall and in the corners of the west and east walls. If further analysis were to suggest that the extant portion of the original wall containing the windows is not in its proper vertical position, then all these reconstructions will need to be revised. In any case, future masonry studies of the building will need to give careful consideration to the presence of two bands of brick coursing in these walls.

³² Runesson, “Oldest Original Synagogue Building,” 416.

³³ It was this observation that earlier made me start looking for the “missing” stairs, which would have disappeared in any case as a result of a later renovation that removed the second floor to create a higher interior elevation.

³⁴ From a purely architectural perspective, I am not sure that this logic is valid, but we shall focus only on the issue of the columns.

³⁵ , Runesson, “Oldest Original Synagogue Building,” 414.Google Scholar

³⁶ Another flaw in Runesson's logic enters here. He argues that the position of the columns on their foundation blocks “proves” that they were there in the original building, since they had been rearranged at some time. He then wants to equate their rearrangement with the intermediate phase of renovation, when area B was given its tripartite configuration. While I accept the latter point–that the columns were “arranged” in conjunction with the partitioning of B–that does not prove that these same columns were in the earlier position. They could just as easily have been introduced at this same point in time and set in their final position. In my view, this is more likely, as I shall discuss below in conjunction with the dating of the capitals (see n. 42 below). On archaeological grounds, therefore, there is no evidence to prove that these four column drums were themselves original to the building. Kraabel (“Diaspora Synagogue,” 498-99) also argued that they look like a later embellishment.

³⁷ See , FaustoZevi, “La sinagogadi Ostia,” Rassegna mensile di Israel 38 (1972) 140Google Scholar , cited by , Runesson, “Oldest Original Synagogue Building,” 420, n. 37. Unfortunately, I have not been able to see Zevi's articleGoogle Scholar.

³⁸ , Runesson, “Oldest Original Synagogue Building,” 420.Google Scholar A propos of this suggestion, Runesson does not cite Squarciapino's proposal that the capitals are from a “later restoration” (“si protrebbe pensare ad un restauro posteriore”) found in “La sinagoga recentemente scoperta ad Ostia,” Rendiconti della Pontificia Accademia Romana di Archeologia 34 (1961-1962) 129Google Scholar ; and cited in Patrizio Pensabene's discussion (see the following note). Squarciapino, however, thought that the columns were repositioned at the time of the fourthcentury renovation.

³⁹ Pensabene, Patrizio, Scavi di Ostia VII, I Capitelli (Rome: Libreria dello stato, 1973) 61–62Google Scholar (= catalogue nos. 232–34 and n. 1); 217-18. While Runesson cites Zevi's 1972 article “La sinagoga di ostia,” he does not consult Pensabene's final publication on the capitals (1973). It is my understanding from sources at Ostia, though not fully confirmed, that Pensabene had also done the preliminary analysis of the capitals on which Zevi's article was based.

⁴⁰ Pavolini, Carlo, Ostia: Guide archeologiche Laterza (3rd ed.; Rome and Bari: Laterza and Figli, 1989) 176Google Scholar : “un accesso monumentale sottolineato da quattro colonne corinzie disposte a rettangolo, databili nel II sec. d.C. ma qui riutilizzate nell'ambito della ricostruzione tardoantica.” While this statement makes it clear that Pavolini assumes they were not original to the building, it leaves unclear whether the term “reutilized” (riutilizate) should be taken to mean that they had been added to this building in an intermediate stage or were placed there only at the time of the later (fourth-century) reconstruction.

⁴¹ I made this point in my earlier article, with special mention of the fact that one of the columns was broken and repaired in antiquity. The drums are described only briefly by Pensabene (see n. 39 above). On this point, I have now identified three other column drums of similar marble and proportions: two were used in primary construction of a building of the second century (Hadrianic); one more was reused in the entryway to the fourth century bath complex across the Via Severiana from the synagogue itself (and is shown in my fig. 2). Its presumed mate is not extant. Not only is the size and type of the drum similar to that in the synagogue, but also the base. Another indication that these are spoils is the fact that the bases (which are in a white marble with a distinctive profile), the drums, and the columns are not matched in size and have been altered to fit in thier present space. On the morphology, see also Pensabene, Scavi di Ostia VII, Plate LXXXI.

⁴² , Runesson, “Oldest Original Synagogue Building,” 419.Google Scholar

⁴³ As he notes (Runesson, “Oldest Original Synagogue Building,” 419, n. 40), Squarciapino dated the repositioning of the columns to the major renovation of the early fourth century. Here I am inclined to keep in mind Squarciapino's chronological parameters for reasons of archaeological caution: on the basis of present evidence the terminus ante quern for the repositioning of the columns cannot be definitively set earlier than the construction of the aedicula (first phase) which abutted one of the columns. Of course, it means that the column had already been moved into this position at some earlier point, but how much earlier cannot be determined on this evidence. If, as Squarciapino argued, the installation of the aedicula occurred as a secondary project after the monumental renovation of the early fourth century, then from an archaeological perspective, the repositioning of the columns could logically have occurred at the same time as this renovation, unless direct archaeological evidence shows otherwise. Runesson's only suggestion in this regard is that the walls used to partition area B, which he also dates to the renovation of the second to the third century, fit with this realignment. Here I think Runesson is partially correct but needs further archaeological evidence to support and refine his conjecture. I will return to this point later.

⁴⁴ , Runesson, “Oldest Original Synagogue Building,” 417, fig. 2, and 419,Google Scholar including notes citing Squarciapino. “Secondo campagna,” 313 and “The Synagogue at Ostia,” 203. See also my earlier comments on the benches in D, n. 3 above.

⁴⁵ , Squarciapino, “Secondo campagna,” 312Google Scholar : “Pavimentazione in cocciopesto, in due strati sovrapposti si e trovata immediatamente sotto al mosaico nella parte centrale dell'aula sinagogale (fig. 12) [note that this figure shows the area B₂]. Tale pavimentazione poggia direttamente sulla sabbia” [my emphasis]. It is not clear from the sentence whether the two strata of cocciopesto apply to all areas of the floor from B-G through D. Also, Squarciapino clearly treats the two strata of cocciopesto as if they were one floor.

⁴⁶ Unfortunately Squarciapino never reported any altimetric readings or stratigraphic measurements for these floors. Here is what I have found from in situ remains (please note that the following measurements are as accurate as possible using three-dimensional levels but should not be considered exact; also they depend on the precision of the restorations): (1) the cocciopesto floor surface in area G is 64 cm below the top of the bedding course of bricks (56.5 cm from the bottom), which serves as the foot of the reticulate masonry; (2) the mosaic floor in G is 39 cm from the top of the reticulate's brick bedding course; (3) the front line of the stibadium is preserved to a height of 17 cm below the mosaic (with ca. 8 cm more of bedding to the surface of the mosaic); (4) the surface of the mosaic floor in B₂ is 41 cm below the top reticulate's bedding course (measured at the SE corner, at the door to A). Assuming the same amount of bedding, this means that the cocciopesto floor in B₂ might be as much as 17-20 cm higher than that in G; (5) the surface of the mosaic floor in B, (measured at the NW corner, adjacent to C,) is only 21 cm below the top bedding course of the reticulate; and (6) the small area of mosaic surface in D (on the north wall adjacent to C,) is just 3 cm below the top of the bedding course of the reticulate (roughly the same level as the opus sectile pavement). Unfortunately I was not able to measure the level of the cocciopesto floor in D.

⁴⁷ The only exceptions are the two rough partition walls of B, which seem clearly to cut through the cocciopesto; they are thus later than the cocciopesto floors.

⁴⁸ There are rough and irregularly laid foundation walls below this bedding course of brick, but these were typically filled in with packed earth as the base for flooring.

⁴⁹ See also the appendix at the end of this article for further comments regarding these doors. Runesson fails to consider the three original doors in G, in his discussion of its form and function in the first phase.

⁵⁰ These three doors are almost identical in size and construction. They are 1.18 m wide (or 4 Roman feet) framed in tufa blocks. The northern door from A originally faced directly opposite the door from F, but was later partially blocked (width 83 cm). The reticulate wall between A/G shows extensive rebuilding at a later stage on the face in A, continuing north to the later doorway cut in at B₃; this door frame was rebuilt both inside and out, thus obscuring elements of the earlier phases of masonry. See Appendix at locus 13.

⁵¹ Here we should note the work of Dunbabbin, Katherine M. D. (“Triclinium and Stibadium,” in Slater, William J., ed.. Dining in a Classical Context, [Ann Arbor: University of Michigan Press, 1991] 121–48, esp. 128-31)Google Scholar for the types and chronology of stibadia; see also her “lit Graeco More Biberetur: Greeks and Romans on the Dining Couch,” in Nielsen, Inge and Nielsen, Hanne Sigismund, eds., Meals in a Social Context: Aspects of the Communal Meal in the Hellenistic and Roman World, (Aarhus Studies in Mediterranean Antiquity 1; Aarhus: Aarhus University Press, 1998) 81–101, esp. 92-97.Google Scholar Dunbabbin argues that the stibadium form of dining room was typical of Roman influence on communal dining practices; a variation of the triclinium, it came to be used in private domestic entertaining and especially by collegia. It did not arise, as some had argued, from bringing outdoor dining forms indoors. At 1.5 m or wider the benches were still designed for reclining while dining. She also discusses the several forms or types of stibadium bench configurations: all are set in rectangular rooms with entrance from one of the short sides where the bench is left open. The simplest form is what she calls the U-shaped, in which the corners are all 90-degree angles (not curved); next is the U+T configuration, in which the top of the T is set at the open end of the U and all angles are still 90 degrees. Finally, there is a lunate form of the U-shaped bench, which in the later empire gave rise to the sigma-couch design, seen in many later floor mosaics. Thus the lunate stibadium of G-B₃, even though a plaster bench rather than couch construction, may suggest a later form. The shift to mosaic floors with layout for couches is also typical of stibadia or triclinia of the later second to third century CE (and continuing into the later empire), and may suggest an appropriate form for the character of the mosaic panel in the south end of G from the later period, as suggested in the following discussion.

⁵² See my plan in Runesson, “Oldest Original Synagogue Building,” 415, fig. 1, and in the Appendix of this article, at locus 5.

⁵³ As I shall suggest below, it was probably resealed when area B was given its tripartite configuration.

⁵⁴ Squarciapino “Secondo campagna,” 312. Squarciapino calls the upper floor mosaico, while she designates the lower floor as musivo (“tesselated”). She does not explain the difference, and I have not seen photos of the lower floor in published reports.

⁵⁵ Ibid.

⁵⁶ Ibid. She also reports a wooden screen across the lower mosaic, thus indicating traffic/use patterns in this intermediate phase. The new door at B₃ would not have provided access to F, only to G.

⁵⁷ This is based on empirical observation alone and assumes that the restoration work reflects accurately what was excavated.

⁵⁸ ln area G: F1 = original building (not extant); F2 = cocciopesto floor/stibadium (lower than F1); F3 = mosaic (first phase); F4 = mosaic (reworked); F5 = earthen floor (coterminous with oven installation). In area B₃: F1 = original building (not extant); F2a = cocciopesto, lower layer; F2b = cocciopesto, upper layer/stibadium (lower than Fl and coterminous with F2 in G); F3 = lower mosaic (coterminous with partitioning of B); F4 = upper mosaic (coterminous with fourth-century constructions in opus vittatum). I am not yet prepared to make a proposal regarding how this picture fits with the floor levels in D, other than to suggest that it allows for the cocciopesto floor and benches to be a secondary renovation. I also wonder whether the opus sectile floor might be a secondary stage of the fourth-century project (that is, at the time of the erection of the aedicula), and/or a replacement for an intervening mosaic floor (as preserved in the NE corner of D). Finally, the mosaic floor of K, must now be examined more closely for similarities to those in G, B, C, and D.

⁵⁹ I am assuming that the position and level of the mosaics may come from the earlier phase (when B was partitioned), but that the mosaics could have been reworked later, perhaps at the time of the fourth-century renovation, initial phase.

⁶⁰ See n. 38 above.

⁶¹ Runesson, “Oldest Original Synagogue Building,” 420, 427. Note that Squarciapino would place her intermediate stage near the beginning of the third century. I think Squarciapino is correct, as I will argue below.

⁶² Ibid., 419.

⁶³ Ibid., 426.

⁶⁴ The brickstamps are reported and analyzed by Zappa, Giulia Garfalo, “Nuovi Bolli Laterizi di Ostia,” in Barbieri, G., et al., eds., Terza Miscellanea Grecae Romana, (Studi pubblicati dall'istituto italiano per la storia antica 21; Rome: Istituto italiano per la storia antica, 1971) 283–85.Google Scholar They were found in a deposit of cement wall core, where they had been reused. This wall core was found deposited on the floor of the fourth century, not as implied by Runesson, with the bricks simply lying scattered on the later floor. Hence, they seem to come from brick spoils reused in making the cement core of walls, either of the fourth- century project or from walls of an intermediate stage ( , Zappa, “Nuovi Bolli,” 285)Google Scholar.

⁶⁵ , Runesson, “Oldest Original Synagogue Building,” 426.Google Scholar We should be more precise. Out of 28 datable stamps, only two came from the “first century” (period unspecified), but this can hardly constitute evidence for the original construction of our building. , Zappa (“Nuovi Bolli,” 283)Google Scholar says explicitly that the stamps are not assigned to the first building; they are spoils. The remaining 26 stamps come from the second century as follows: six date to the reign of Trajan, eighteen to the reign of Hadrian, and two to the reign of Antoninus Pius. As Zappa discusses, the Hadrianic corpus is the most significant, since it also yields 14 from the single year 123 CE and one more from 124 CE. In my view this corpus correlates best with the construction of building K-E-H, but it may point to other interior renovations in areas D-C-B-G-A.

⁶⁶ Ibid. Also note that Runesson says that the paintings on these partition walls date to “the beginning of the second century”; however, I find no such reference to “the beginning of the [century]” in Zappa's article. Of the paintings, she says only: “di due muretti costruiti in opera cementizia, che tagliano il pavimento della prima fase e che, in base soprattutto ad alcune pitture che li decorano, sono databili all incirca al II secolo.” (p. 285). He cites only Squarciapino, “Secondo campagna,” 212-13, which does not date the paintings at all.

⁶⁷ They might also be spoils used in composition of the opus vittatum wall core of the fourthcentury renovation (Zappa,"Nuovi Bolli,” 285). What Runesson is apparently doing is saying that the whole bricks came from his “first renovation,” and they were later used in the fourth-century renovation. This seems doubtful to me. Instead, the bricks may well come from some part of the construction associated with area K; if so, then they suggest to me another, even earlier renovation than Runesson's “first.” See discussion below.

⁶⁸ To be discussed below.

⁶ The inscription reflects two stages, given that an earlier text was erased and recut with the name of Mindius Faustus. As Runesson notes (“Oldest Original Synagogue Building,” 430, n. 96), this earlier text can only be dated in general terms to the later second century CE, following the analysis of Noy, David, Jewish Inscriptions of Western Europe (2 vols.; Cambridge: Cambridge University Press, 1993-1995) 1. 22–26Google Scholar (= no. 13). The basis for this dating are the letter forms and the introductory Latin formula Pro salute Aug. Noy prefers the later Antonine period as the more probable, based on the letter forms, but notes that the same formula continued to be used under the early Severans (especially Septimius Severus and Caracalla). In my article I had also cited the Severan examples of this formula in assigning a late second to early third-century date to the first inscription (“Synagogue and Society, “ 41). Noy also dates the reinscribed lines to the later third century, based on letter forms. More recently Maria Letizia Lazzarini has suggested a similar dating on orthographic grounds “L'incremento del partrimonio epigrafico grecoostiensedopo ‘Roman Ostia’, “in Zevi, AnnaGallina and Claridge, Amanda, eds., ‘Roman Ostia’ Revisited: Archaeological and Historical Papers in Memory of Russel Meiggs, [London: British School at Rome, 1996] 240–41.)Google Scholar She dates the first redaction to the second century but the second redaction to the end of the third century. In general, 1 am happy to accept a late Antonine to early Severan date (ca. 160-200 CE) for the first redaction and to correlate it with the renovations in B, C, and G suggested for this same period. Such dating is consisitent with my suggestions in my earlier article. Runesson's discussion of this inscription (“Oldest Original Synagogue Building,” 430) would seem to leave it out of the picture for his “first renovation,” but he does not say clearly how the installation of the keiboton described would fit his overall picture.

⁷⁰ Squarciapino was most likely using the taxonomy developed by Blake, Marion Elizabeth, Ancient Roman Construction in Italy (3 vols; Washington: American Philosophical Society, 1947-1973), esp. 2. 63–66 for examples, such as the Horreum of Hortensius (V. 12. 1) and the Sanctuary of Bona Dea just outside the Porta Marina (IV. 8. 3), both of which would date to at least the time of Claudius if not slightly laterGoogle Scholar.

⁷¹ , Runesson, “Oldest Original Synagogue Building,” 425.Google Scholar

⁷² Here Runesson is also following the work of Heres, Theodora Leonore (Paries: A Proposal for a Dating System of Late-Antique Masonry Structures in Rome and Ostia [Amsterdam: Rodopi, 1982])Google Scholar , who says that bricks began to replace the tufa quoins “at the end of the first century” (p. 29) and further qualifies this statement by saying that such replacement did not begin “before Claudius” (p. 38, n. 40). However, I would further note that my own analysis of the masonry follows that of Boersma (Amoenissima Civitas, 302, and elsewhere). Boersma has developed an even more sophisticated taxonomy for the masonry. For example, he identifies four distinct types of opus mixtum (designated A-D) and two distinct types of opus vittatum (designated A and B, and he does not always call it vittatum mixtum, even though it clearly is). I would add that Heres's work, Paries, was originally written as her doctoral dissertation under Boersma, and she is credited with the masonry analysis in Amoenissima Civitas, 11-71. Hence, I take this later work as a refinement of her own system proposed in Paries. Unfortunately, Runesson does not seem to consult Boersma's work.

⁷³ See White, “Synagogue and Society,” 29, and n. 19; notwithstanding Runesson's characterization (“Oldest Original Synagogue Building,” 425 and n. 62), which I think is misleading.

⁷⁴ Here I must take issue with Runesson's handling of the Plotius Fortunatus inscription. I agree that, for the sake of completeness, I should have included it in my discussion, even though it has little bearing on the building (not even in the same way that the Julius Justus inscription does). But my difficulty with Runesson's interpretation is the date, because he wants to assign it arbitrarily to about the year 100 CE, without any sort of archaeological or epigraphic evidence (Ibid., 424, n. 60, and 427-28). As Noy (Jewish Inscriptions of Western Europe, 1. 26-28, and n. 14) notes, the inscription can only be dated in the most general terms to the first or second centuries. On the basis of the letter forms it could easily be nearly as late as the first recension of the Mindius Faustus inscription. Runesson cites Noy's dating but gives no rationale for assigning it arbitrarily to the year 100. It does not constitute, therefore, the kind of direct, datable archaeological evidence that he seems to claim.

⁷⁵ Runesson presents no such archaeological evidence, with the exception of a suggestion of an early renovation in the time of Hadrian. However, I have already shown that his evidence for this renovation and the elements of the building that he assigns to it are problematic. In the final analysis, his only evidence for a Claudian date is circumstantial and speculative, at best, because we have no other direct archaeological evidence of Jewish activitiy at Ostia in that period and no clear epigraphic evidence for an established community in the first century.

⁷⁶ My survey of comparanda is still in an early stage and is thus incomplete; however, I have checked all of the dated walls analyzed by Boersma in insula V. 2, and other edifices or walls in reticulate typically dated to the middle of the first century and later.

⁷⁷ I would note that, as far as 1 have been able to observe, the building at Ostia that has the greatest similarity in masonry to that of the particular configuration of opus mixtum a in the synagogue complex is the triple apsidal nymphaeum (II. 8. 3) located in the area of the Four Republican Temples (II. 8. 2). While the adjacent Shrine of Jupiter (II. 8. 4) is usually dated to the early principate because of its opus reticulatum masonry, the nymphaeum – which blends tufa facing and quoins with brick reinforcing bands very much like the synagogue building–is dated by Pavolini to the Hadrianic period; see Ostia: Guide archeologiche, 3. 72.

⁷⁸ This does not include some of the incidental small structures of the latest phase, such as the small appurtenances in F.

⁷⁹ This point will need to be reevaluated in the light of some new observations made above. For example, if the tripartite partitioning of B occurred in the renovation of the late second century, at the same time that columns were introduced (or reset) in their present position, then it is conceivable that the aedicula was part of the design of the fourth century renovation from the start, rather than an afterthought.

⁸⁰ This masonry is characterized by reticulate facing with two bands of lateral brickwork reinforcing and interlocking brickwork quoins. The walls are generally ca. 45 cm thick, except the external wall on the west side (made thicker by an outer face of opus latericium) and the thinner partition walls in the rooms under the stairs (ca. 37 cm thick).

⁸¹ Hadrian's Villa at Tivoli employs an opus mixtum a similar to that of the synagogue (reticulate face with tufa quoins and brick bands), alongside the more typical Hadrianic masonry of opus mixtum b (reticulate face with brick quoins and brick bands).

⁸² This is one of the key features that will need to be treated through systematic survey.

⁸³ Caution is still required in interpreting the relationship between the two parts of the complex, since the precise phasing of the architectural components of K-E/J-H has not been clarified.

⁸⁴ Runesson ignores the areas in A that are in the same masonry as the original building, which are, in my view, part of the original complex.

⁸⁵ It appears that a passageway existed immediately to the west of building K-E-H, leading from the Via Severiana toward the beach. Just to the west of it is an unexcavated area that apparently contains rubble and architectural remains, the nature of which is so far unknown.

⁸⁶ These walls and wall segments are not shown correctly in Runesson's plan (fig. 2), and my plan needs to be corrected and supplemented. See Appendix below.

⁸⁷ The nymphaeum (length 4.9 m) faces east. These areas are not shown in Runesson's plan (fig. 2), and his walls are drawn incorrectly for the rear (south) portions of room 5. See locus 23 in the Appendix. From the masonry, it appears that the nymphaeum was secondary to the walls of K, although how much later cannot be determined at present.

⁸⁸ This is probably a distinct type of mixtum a (though possibly an example of mixtum c), because it contains sections of tufa block (but not reticulate) intersected by courses of brick. It appears to be coterminous with the original building.

⁸⁹ Runesson (“Oldest Original Synagogue Building,” 416) criticizes my plan in relation to the front wall of K that extends toward the Via Severiana, saying that I had extended it too far. He is partially correct. The wall is not extant beyond the middle of the west wall of D; however, the masonry shows that the wall was not terminated at this point (as Runesson assumes). It must have continued to the north beyond this point. I should have drawn it as dashed lines, since the actual termination is not clear.

⁹⁰ In addition to the main constructions of this period, which are in opus vittatum a or b, there are also some small sections of opus vittatum simplex (to which I had at one point made some reference; see Runesson, “Oldest Original Synagogue Building,” 409, n. 1).

⁹¹ For example, the two flanking doorways from A into B, and B₃ respectively were clearly later additions cut into the original wall, as indicated by their brickwork framing; however, it is possible that they came from the late second-century renovation, rather than that of the fourth century, at the same time that area B was partitioned with the early “rough walls” (as Squarciapino called them). I think that the entry to A from the street probably had two phases of construction, both later than the original building.

⁹² It does not conform to a typical house plan, at least not in the remains now known. It extended farther to the south and east, but these areas now lie under the modern highway.

⁹³ Something like this may have been at work in areas E and F, and perhaps elsewhere, including G and A.

⁹⁴ See Runesson, “Oldest Original Synagogue Building,” 418 and n. 32. My observation that the main door from A into B₂ might have been widened at some point is based on the width of the tufa quoining that frames this door, in comparison with other doors in the edifice, since it is only three blocks rather than four to five. There is other evidence that they were able to cut back such masonry and reface it in a fashion similar to the original. If so, such work might have modified the width of this door by up to 50–75 cm. Nonetheless, Runesson is correct in stating that this door was the largest door in the complex (even if narrower by this much). I only suggest the narrowing as a possiblity based on the extant archaeological evidence.

⁹⁵ If area A is not part of the edifice, then my argument that there are contemporaneous buildings that must be considered in the larger complex is strengthened further.

⁹⁶ See locus 7 on the annotated plan below.

⁹⁷ Annotated plan, locus 8.

⁹⁸ Annotated plan, loci 4 and 5, respectively. Runesson previously read the door at locus 4 to be original to the building; but, based on the masonry, such a suggestion is not possible. The door had to have been cut in later, if it functioned as a door at all.

⁹⁹ I had previously assumed that these doors were sealed at the time of the fourth-century renovation; however, I now wonder, in the case of the door in B, at least, if such sealing might have occurred at the time of the partitioning of B, when the new doors were cut through from A. Thus, if the secondary door of B, correlates with the installation of the stibadium, it further supports the idea that the stibadium was introduced in a secondary phase of renovation after the original construction of the building. If the door of Bj was then resealed at the time of the partitioningof B, then this door only functioned in my phase II of the building. See below.

¹⁰⁰ White, “Synagogue and Society,” 35. Runesson, “Oldest Original Synagogue Building,” 423.

¹⁰¹ , Runesson, “Oldest Original Synagogue Building,” 416.Google Scholar

¹⁰² Ibid., 424, 431-32.

¹⁰³ Runesson's argument (“Oldest Original Synagogue Building,” 432) for monumentality also assumes that the Ostia synagogue was much larger than the other Diaspora cases that clearly were renovated from private houses. Again, precise comparisons yield a different picture: The area of B-C-D at Ostia is slightly smaller than the original house that was renovated to become the Dura synagogue, and the entire synagogue complex at Ostia, in its final form, is smaller than the final complex at Dura (created by annexing a second house and “monumentalizing” the first). Similarly, the house converted to a synagogue on Delos is larger than the complex at Ostia, and the assembly room before partitioning was larger than room D at Ostia. The Delos synagogue, however, was converted from a domestic complex but was never transformed into a “monumental” style of architecture, even though it bears a strong resemblance to local collegial halls (such as the House of the Poseidoniasts). Finally, one should compare the scale of the villa of Tiberius Claudius Polycharmos, out of which the synagogue at Stobi was built, since it is considerably larger than the Ostian complex, even though the synagogue hall itself (which was only a part of the house complex) was about the same area as room D. So, 1 question what is meant by “monumental,” and especially as it refers to the Ostia synagogue complex prior to the final stage of renovation. (For a synoptic plan of these synagogues that shows their relative size and orientation, see Kraabel, “Diaspora Synagogue,” fig. 1, and adapted in White, Social Origins of Christian Architecture, 1. 63.) In this vein, Runesson's argument regarding monumentality in the collegial halls of Ostia is also problematic (“Oldest Original Synagogue Building,” 432, and n. 107), particularly if one compares the two large collegial buildings renovated from houses at III. 2. 2 (Fabri Novales) and IV. 5. 15 (Schola Traiani).

¹⁰⁴ , Runesson, “Oldest Original Synagogue Building,” 424.Google Scholar

¹⁰⁵ Ibid., 426-27. Pavolini,"Ostia – Saggi lungo la Via Severiana,” Notizie degli Scavi di Antichita 8. 35 (1981) 115-43.

¹⁰⁶ , White, “Synagogue and Society,” 28, and fig. 2.Google Scholar

¹⁰⁷ My translation of Pavolini, “Ostia – Saggi lungo la Via Severiana,” 141: “… urbanizzazione del quartiere gia nel I sec. d.C, e in particolare nella seconda meta, come dimostra la Sinagoga piu antica (p. 15), e agli inizi del II (muro22delsaggioB).” I take Pavolini's phrase “Sinagoga piu antica” to refer only to the earlier of the phases of the synagogue edifice per se, not to its dating relative to other buildings.

¹⁰⁸ As , Runesson (“Oldest Original Synagogue Building,” 427)Google Scholar also notes, and I take this to be completely in keeping with my dating of the building, as discussed earlier. Therefore, here I must take exception to Runesson's comment (Ibid., n. 81), because Pavolini (who was Squarciapino's chief archaeologist and field supervisor in the Via Severiana project of the late 1970s), as expected, cites Squarciapino's work. Moreover, in the brief summary comment in his La vita quotidiana a Ostia (1978; 3d ed.; Gius: Editori Laterza, 1991) 163Google Scholar , the statement (as cited by Runesson) has not changed since the 1978 edition, which was published before his 1981 article in Notizie degli Scavi. Thus, I take the later work as the more authoritative statement of his own view in light of the excavations along the Via Severiana from the late 1970s to 1981. It is also noteworthy that K was not fully excavated at the time of Pavolini's 1981 article.

¹⁰⁹ Pavolini, “Ostia – Saggi lungo la Via Severiana,” 124-26 (plus fig. 7 on p. 121), 141. Note that the floor level of wall 22 is considerably lower than that of the Antonine construction and the Via Severiana itself. The walls of the Antonine/Severan period shops (wall 28) actually used wall 22 as a foundation. Pavolini (127) notes that the wall 28 is also in a form of opus mixtum datable to the second century, but (following Gismondi and Meiggs, Roman Ostia, 2. 539–40) most likely comes from the time of Antoninus Pius. Thus, in the area immediately across the Via Severiana, we have phases of construction and types of masonry that parallel those in the synagogue complex.

¹¹⁰ Ibid., 122.

¹¹¹ Ibid., 141-42; alluded toby Runesson, “OldestOriginal Synagogue Building,” 427, and n. 83.

¹¹² Becatti, Giovanni, in Calza, Guido, et al., Scavi di Ostia I: Topografia generate (Rome: La libreria dello stato, 1953) 113Google Scholar ; Becatti, Giovanni, Scavi di Ostia VI: Edificio con opus sectile fuori Porta Marina (Rome: La libreria dello stato, 1969) 51.Google Scholar (Note that the latter was published well after the discovery of the synagogue. Runesson wants to ignore this road, because it does not fit his picture of an isolated synagogue. He ignores the fact that the Plotius Fortunatus inscription, which also comes from the first or second century (discussed above), probably came from a tomb along this same road a little farther to the south.

¹¹³ Kraabel, “Diaspora Synagogue,” 500. See also my earlier suggestions in “Synagogue and Society,” 36, and above nn. 79 and 43. The chronology outlined in Table 1 marks a revision over my earlier one, even though the basic steps are comparable. For example, what I previously described as Phase Mb is now Phase Ilia, but the dates are the same. It is merely an ongoing process of refinement in our understanding.

¹¹⁴ White, “Synagogue and Society,” 58, fig. 5.