1 Prior to Ruyt's, C. De Macellum: marché alimentaire des Romains (Louvain-la-Neuve 1983)Google Scholar, macella were the subject of some brief accounts. See especially Crema, L., L'architettura romana (Enciclopedia Classica, sez. 3.12.1; 1959) 171-73, 286-87 and 515–21Google Scholar; Boëthius, A. and Ward-Perkins, J. B., Etruscan and early Roman architecture (Harmondsworth 1970) 211, 294-95, 298, 372, 468-69 and 482-83Google Scholar; Nabers, N., “The architectural variations of the macellum,” OpuscRom 9 (1973) 173–76Google Scholar, which, as an abstract of an unpublished 1967 Princeton Ph.D. dissertation (Macella: a study in Roman archaeology), was the first attempt at an architectural classification.

2 “Il peut sembler étrange de consacrer un chapitre à une notion qui évoque plutôt, dans notre langue, une fonction économique plutôt qu'une structure bâtie”: Gros, P., L'architecture romaine du début du 3ème s. av. J.-C. au Haut-Empire. I. Les monuments publics (Paris 1996) 450 Google Scholar.

3 The aediles or agoranomoi were responsible for the annona macelli: De Ruyt (supra n.1) 356-58; Holleran, C., Shopping in ancient Rome: the retail trade in the Late Republic and the Principate (Oxford 2012) 175 CrossRefGoogle Scholar.

4 Rare seafish, especially, was praised and sold in public auction: Holleran ibid. 176-77.

5 For this list, cf. De Ruyt (supra n.1) 341-50. Holleran (supra n.3) 172 underscores the dearth of sources concerning the sale in macella of non- or less-perishable food items such as bread, oil, wine, garum or cheese.

6 Ruyt, C. De, “Les produits vendus au macellum,” Food & History 5.1 (2007) 135–50CrossRefGoogle Scholar.

7 It has long been assumed that the functional specialization and visual uniformity characterizing the Late Classical/Early Hellenistic Ionian agora inspired the architecture and function of early Roman macella, but Ionia's pioneering rôle in the evolution of the Greek agora is increasingly being re-appraised in favour of mainland Greece and Magna Graecia: see, e.g., Hellmann, M. C., L'architecture grecque. 3. Habitat, urbanisme et fortifications (Paris 2010) 260–67Google Scholar. Smaller peristyle courtyards functioning as markets (e.g., at Kassope [Epirus], dated to the late 3rd c. B.C.) are now considered better prototypes: Gros (supra n.2) 451; Hellmann ibid. 260-62. On Thasos, the recently-excavated Imperial macellum took over the shape and function of a Hellenistic commercial square: Marc, J. Y., “Un macellum d'époque hellénistique à Thasos,” in Cavalier, L., Descat, R. and Courtils, J. Des (edd.), Basiliques et agoras de Grèce et d'Asie Mineure (Bordeaux 2012) 225–39Google Scholar. The Roman macellum should thus no longer be considered the result of the clear-cut and unilateral influence of Greek agorai, but rather as the product of a long and flexible process of cross-fertilization between different types of courtyard-shaped commercial structures built in Greece and Italy in the Early Hellenistic period.

8 De Ruyt (supra n.1) 236-52, based on Plautus and later authors. Gros (supra n.2) 450 favours a dating in the late 3rd or early 2nd c. B.C. The earlier dating was still defended by De Ruyt, “Exigences fonctionnelles et variété des interprétations dans l'architecture des macella du monde romain,” in Cascio, E. Lo (ed.), Mercati permanenti e mercati periodici nel mondo romano (Bari 2000) 178 Google Scholar, but in 2007 (supra n.6, p. 135) she refers to the late 3rd c. Gaggiotti, B.C. M. (“Macellum e magalia: ricezione di elementi ‘culturali’ di origine punica in ambiente romano-repubblicano,” in Mastino, A. [ed.], L'Africa romana VII [Sassari 1990] 773-82)Google Scholar argued that the advent of separate food markets and use of the term macellum occurred between the First and Second Punic Wars (c.240-220 B.C.). Holleran (supra n.3) 163 dates the first macellum before 210 B.C. Cf. also Sartorio, G. Pisani, “Macellum, Macellum Liviae, Macellum Magnum,” in LTUR III (1996) 201–2Google Scholar and Morcillo, M. García, “El Macellum Magnum y la Roma de Nero,” Iberia 3 (2000) 266 Google Scholar, both neutral on the issue.

9 In Alatri (Latium): CIL 12 1529; De Ruyt (supra n.1) 225.

10 The earliest Italian macella attested archaeologically predate the Social War. The macellum at Morgantina (second half of the 2nd c. B.C.) is the earliest documented: Nabers 1973 (supra n.1) 173-74; De Ruyt (supra n.1) 109-14.

11 The new Macellum Liviae built by Augustus on the Esquiline and the Macellum Magnum of Nero possibly on the Caelian demonstrate the importance of large specialized market buildings for the supply of the city: Nabers 1973 (supra n.1) 175; De Ruyt (supra n.1) 163-84; Pisani Sartorio (supra n.8) 203-6; García Morcillo (supra n.8) 265-86; Holleran (supra n.3) 162-67.

12 Even if many macella are known in the West, only a few were comprehensively excavated and published. See especially Lugdunum Convenarum, of the early 1stc.A.D. (G.Fabreand J. L. Paillet, Saint-Bertrand-de-Comminges IV. Le macellum [Pessac 2009]), and Baelo Claudia, of late 1st/early 2nd c. A.D. ( Didierjean, F., Ney, C. and Paillet, J. L., Belo III. Le macellum [Madrid 1986])Google Scholar.

13 On macella in Asia Minor, see the brief but confusing contribution by S. Atik, “The ancient delicatessen ‘macellum’. Some thoughts about the macella in Anatolia,” in Menozzi, O., Marzio, M. L. Di and Fossataro, D. (edd.), SOMA 2005. Proc. IX Symposium on Mediterranean Archaeology, Chieti 2005 (BAR S1739; Oxford 2008) 325–30Google Scholar.

14 It possibly derived from the Semitic root miklā, bearing the same meaning. Another Semitic root was proposed as an alternative: ʿkl meaning ‘eating,’ based on which the word *máʾkal or ‘eating place’ was formed: Gaggiotti (supra n.8). De Ruyt (supra n.6) 137 now considers this hypothesis more likely.

15 De Ruyt (supra n.1) 351-55.

16 This sum may have covered only the erection of the colonnades. Indeed, an unpublished statue base for an Eros that was found in the macellum indicates that the city paid for ‘a section of pavement” and a "pyramid" (the conical roof of the tholos?).

17 Wiegand, T., Didyma. Die Inschriften von Albert Rehm (Berlin 1958) 321 Google Scholar.

18 On the meaning of the term διπλη̑ στοά, see Coulton, J. J., “διπλη̑ στοά,” AJA 75 (1971) 183-84CrossRefGoogle Scholar; id., The architectural development of the Greek stoa (Oxford 1976) 4; Gros (supra n.2) 95-97. We do not necessarily follow Özdizbay, A., Perge'nin M.S. 1.-2. yüzyıllardaki gelişimi (Antalya 2012) 70–71 Google Scholar, according to whom the διπλη̑ στοά would refer to the external portico flanking the market on the W side. It could just as well be applied to the internal single-storey colonnades with a row of shops in the rear nave.

19 The amount of statues erected is unclear in SEG 43. 865 Google Scholar.

20 CIG 3705: ἔστρωσεν ἐκ τῶν ἰδίων τὴν πλατεīαν ἀπὸ τοῦ ζυγοστασίου μέχρι τῆς ὑποχωρήσεως. Note that Robert, L. (“Les inscriptions,” in Gagniers, J. Des et al. [edd.], Laodicée du Lycos. Le nymphée [Québec 1969] 259)Google Scholar defines ζυγο στάσιον as “public weights”.

21 Scherrer, P. and Trinkl, E., Ephesos XIII/2: Die Tetragonos Agora in Ephesos (Wien 2006) 11 and 42–43 CrossRefGoogle Scholar. The fact that the commercial agora of Ephesos may have been called macellum will be considered below.

22 Rozenfeld, B. T. and Menirav, J., Markets and marketing in Roman Palestine (Leiden 2005) 26–29 CrossRefGoogle Scholar.

23 τὸ ἐκεῖ τὰ χοίρειακρέα κόπτεσθαι μόνον: Chron. 12.7.71. The chronology of the Makellon is disputed, even if Malalas’ account is sufficient to say that it was near the Forum of Valens: Downey, G., A history of Antioch in Syria from Seleucus to the Arab conquest (Princeton, NJ 1974) 406–7Google Scholar.

24 De Ruyt (supra n.6) 143-44.

25 The βιῶτικὴ ἀγορά of Oenoanda had three, one single-storey and two double-storey. See also Atık (supra n.13) 329, who does not fully discuss their possible function as markets for the sale of foodstuffs.

26 Transcription and translation by W. Clarysse, publication in preparation.

27 Contra Holleran (supra n.3); see above n.5.

28 Taeuber, H., “Inschriften,” in Pülz, A. (ed.), Ephesos IV/4. Das sog. Lukasgrab in Ephesos (Wien 2010) 348 Google Scholar, cat. Lk Epi 31-32 (no dates given; Pülz ibid. 93 simply underlines their value for demonstrating the commercial function of the complex). At Didyma, the topos inscription (Wiegand [supra n.17] no. 522) also refers to a baker; A. Rehm attributes it to the Markthalle without further information.

29 IK Perge II, nos. 300-2.

30 Richard, J. and Waelkens, M., “Le Macellum de Sagalassos (Turquie): un marché ‘romain’ dans les montagnes du Taurus? Compte-rendu préliminaire des fouilles archéologiques menées depuis 2005,” in Chankowski, V. and Karvonis, P. (edd.), Tout vendre, tout acheter. Structures et équipements des marchés antiques (Bordeaux 2012) 83–104 Google Scholar.

31 The larger dimensions (c.45 by 40 m) given ibid. 88 were estimated when significant parts of the complex were still unexcavated. The S side of the courtyard was probably open towards the lower city through a simple portico, and lacked shops.

32 Mansel, A. M., “Ausgrabungen in Pamphylien 1957-1972,” AA 90 (1975) 49–96 Google Scholar; De Ruyt (supra n.1) 129-33; Özdizbay (supra n.18).

33 Mansel, A. M., Die Ruinen von Side (Berlin 1963) 97–107 CrossRefGoogle Scholar.

34 Curiously, De Ruyt invariably refers to “l'agora de Side” but includes the complex of Perge in her catalogue of macella. Mansel, A. M. (Die Agora von Side und die benachbarten Bauten. Bericht über die Ausgrabungen im Jahre 1948 [Ankara 1956] 31 Google Scholar, and id. [supra n.33] 100-1) refers to a Peristylen Agora and to an Agora in the case of Side; he also establishes a parallel with the Marktanlagen of Perge, Sagalassos, and Selge. In Side. 1947-1966 yılları kazıları ve araştırmalarının sonuçları (Ankara 1978) 167, Mansel makes a parallel between Side and other macella in the region. When describing the square at Perge (supra n.32, p. 76), he refers to the Römische Agora, before acknowledging on p. 83 that it presented the characteristics of a macellum. He then compares it to the Marktanlagen of Side and Sagalassos, referring to the latter’s dedicatory inscription mentioning the word makellon. More recently, Özdizbay (supra n.18) refers neutrally to macellum/agora.

35 The number of gates on the E side, not fully documented, is an estimate: Mansel 1956 (supra n.34) 29; id. 1978 (supra n.34) 150.

36 Pülz, A., “Das sog. Lukasgrab in Ephesos. Vorbericht der Nachuntersuchungen 1997-2000,” Mitteilungen zur Christlischen Archäologie 7 (2001) 9–25 Google Scholar.

37 Pülz (supra n.28) 97.

38 Naumann, R. and Naumann, F., Der Rundbau in Aezani mit dem Preisedikt des Diokletian und das Gebäude mit dem Edikt in Stratonikeia (IstMitt Beih. 10, 1973)Google Scholar; De Ruyt (supra n.1) 22-25.

39 Naumann and Naumann ibid. 28-67; Crawford, M. H. and Reynolds, J., “The Aezani copy of the Prices Edict,” ZPE 26 (1977) 125–51Google Scholar.

40 The limited extent of the excavations did not allow a full documentation of the layout of the porticoes. Bielfeldt, R., “Das Macellum von Pompeiopolis: ein neuer Markt für Kleinasien. Vorbericht zu den Grabungskampagnen 2008 und 2009,” in Summerer, L. (ed.), Pompeiopolis I. Eine Zwischenbilanz aus der Metropole Paphlagoniens nach fünf Kampagnen (2006-2010) (Langenweißbach 2011) 49–62 Google Scholar.

41 This dating is hypothetical: Bielfeldt ibid. 10.

42 Morselli, C., “Osservazioni sulle strutture di età romana,” in Schneider, E. Equini (ed.), Elaiussa Sebaste I (Rome 1999) 263–68Google Scholar.

43 Bueno, M. Martín, “Notes préliminaires sur le macellum de Gerasa,” Syria 66 (1989) 177–99CrossRefGoogle Scholar; Uscatescu, A. and Bueno, M. Martín, “The macellum of Gerasa (Jerash, Jordan): from a market place to an industrial area,” BASOR 307 (1997) 67–88 Google Scholar.

44 Uscatescu and Martín Bueno ibid. 72-73.

45 The latter is a supposition based on symmetry (the area has not been excavated): ibid. 75.

46 Dentzer, J. M. et al., “Le développement urbain de Bosra de l’époque nabatéenne à l’époque byzantine: bilan des recherches françaises 1981-2002,” Syria 79 (2002) 75–154 Google Scholar; Dentzer-Feydy, J. et al., Bosra. Aux portes de l’Arabie (Beirut 2007) 262–64CrossRefGoogle Scholar.

47 The inscription in Wilberg, W., Ephesos III: Die Agora (Wien 1923) no. 15Google Scholar, referring to the makellon of Ephesos, cannot be attributed to the Lukasgrab. This wall architrave was in fact found at the South Gate of the Tetragonos Agora and may suggest that the name makellon was in use to designate that complex in the early 3rd c. On this question, see De Ruyt (supra n.1) 71 and now Pülz (supra n.28) 94, n.478. On the Tetragonos Agora, see Scherrer and Trinkl (supra n.21).

48 As underscored by De Ruyt (supra n.1) 284, c.30 different ground-plans of macella are known throughout the Empire, which makes a detailed classification hazardous.

49 Cf. the example of Rome: Holleran (supra n.3) 96-97.

50 Van Andringa, W., “Du sanctuaire au macellum: sacrifices, commerce et consommation de la viande à Pompéi,” Food & History 5.1 (2007) 47–72 CrossRefGoogle Scholar.

51 One could enter the main vestibule from the portico along the cardo, while on the opposite side of the courtyard a secondary vestibule had two entrances. Two pairs of interrelated tabernae in the middle of the N and S aisles were used as subsidiary entrances. See Uscatescu and Martín Bueno (supra n.43) 70.

52 See the cross-section of the macellum at Gerasa in Martín-Bueno (supra n.43) 181.

53 The inscription containing the expression macellu[m …. cum … et pi]scario from the early 1st c. A.D. macellum at Corinth is considered by De Ruyt (supra n.1) p. 313 and (supra n.6) p. 144 as evidence for the use of some basins as fish ponds. For the complete inscription, see West, A. B., Corinth VIII.2. Latin inscriptions 1896-1926 (Princeton, NJ 1931) nos. 124-25Google Scholar, with an update based on new fragments by Kent, J. H., Corinth VIII.3. The inscriptions 1926-1950 (Princeton, NJ 1966) no. 321Google Scholar.

54 Martín-Bueno (supra n.43) 187; Uscatescu and Martín-Bueno (supra n.43) 67.

55 Pülz (supra n.28) 60-72.

56 The SE tabernae of the Roman macellum at Thasos were also paved with terracotta plaques: Marc (supra n.7) 228.

57 Sat. Men., Bimarcus XXIII (J. P. Cèbe, Varron, Satires Ménippées [Rome 1974] vol. 2)Google Scholar.

58 The term is applied to the round monuments at Ephesos and Aizanoi.

59 De Ruyt (supra n.1) 299-300.

60 Naumann and Naumann (supra n.38) 23.

61 Pülz 2010 (supra n.28) 97.

62 For Side: Mansel 1956 (supra n.34) 34; 1963 (supra n.33) 107; 1978 (supra n.34) 164-67 (analogy based on coins). For Perge: Mansel (supra n.33) 80 (analogy based on a comparison with Side). This hypothesis was summarized by Özdizbay (supra n.18) 68.

63 Holleran (supra n.3) 176.

64 At Side, Perge and Sagalassos, beam-holes reveal the presence of a second storey or mezzanine above the shops, useful for additional storage or living space.

65 Van Andringa (supra n.50).

66 De Ruyt (supra n.1) 271-73; Van Andringa ibid. 48-49.

67 The physical link between the Hellenistic and Roman market building and the altar in the middle of the agora at Thasos should also be noted: Marc (supra n.7).

68 Spaces between the macellum of Pompeii and two surrounding temples were possibly used as ‘transitional workshops’ for processing sacrificial meat sold in the macellum: van Andringa (supra n.50) 65. On the macellum, see Dobbins, J. J., “Problems of chronology, decoration, and urban design in the Forum at Pompeii,” AJA 98 (1994) 629–94CrossRefGoogle Scholar.

69 For instance, in the SE room of the macellum at Pompeii: De Ruyt (supra n.1) 317; Van Andringa (supra n.50) 59-60.

70 De Ruyt (supra n.1) 318.

71 In the macellum of Djemila: ibid. 320. A measuring table found in a village near Lugdunum Convenarum is hypothetically attributed to the macellum: Fabre and Paillet (supra n.12) 259-61.

72 De Ruyt (supra n.1) 321-22.

73 Özdizbay (supra n.18) 70.

74 Martín-Bueno (supra n.43) 187-88; Uscatescu and Martín-Bueno (supra n.43) 73.

75 Grinding stones, fragments of stone mortars and pestles, knives, hooks, chains and small balance trays were found sporadically in the collapse, but they cannot be specifically related to commercial activities.

76 Built in rubble masonry covered with bricks, the structure was 0.78 m high. Shallow benches along the rear wall in at least three other shops may have fulfilled similar functions. The benches, made of stone masonry covered with a layer of bricks, were 0.80-1.0 m high, a height which would allow people to sit there or merchandise to be displayed.

77 At Perge, the façades of the shops are also equipped with windows but their height makes any direct communication with the inside of the shop impossible.

78 Uscatescu and Martín-Bueno (supra n.43) 75-82.

79 Richard and Waelkens (supra n.30) 95-99.

80 Similar evidence was found in the macellum of Thasos: Marc (supra n.7) 228.

81 A. Galik et al., “Archäozoologischer Befund: die hellenistischen und spätantiken bis frühbyzantinischen tierischen Abfallvergesellschaftungen aus den Grabungen im Bereich des sog. Lukasgrabes,” in Pülz (supra n.28) 359-91.

82 Revell, Paraphrasing L., Roman imperialism and local identities (Cambridge 2009) 1 Google Scholar.

83 For the Lukasgrab at Ephesos, see Galik et al. (supra n.81). The faunal remains from the macella of Sagalassos and Thasos are currently under study. On the latter, see Marc (supra n.7), referring also to recent archeozoological research in several Western macella.

84 De Ruyt (supra n.6) 149.