• 1 e.g Yaron 1959; Härder 1975, p. 17-30; Frier 1983, p. 257-295; Bouvier – Wehrli 1983, p. 261-266; K (...)
• 2 Cf. Lenel 1889, p. 1121 n. 2 (an inference from the commentaries of Pomponius, Ulpian and Paul ad S (...)
• 3 Frier 1983.
• 4 E.g. D.18.6.1 (Ulp. 28 ad Sab.); D. 18.6.2 (Gaius 2 Rer. Cott.); D. 18.6.4 (Ulp. 28 ad Sab.); D. 18 (...)

1Wine was one of the three core commercial staple products (alongside grain and olive oil) that dominated Roman trade. In contrast to the other two, the sale of wine has not only generated a vast number of academic publications,1 but its particularities were also a topic of considerable interest for the Roman jurists. As an example, the early jurist Sabinus apparently devoted a section of his commentary on civil law to the sale of wine.2 The reasons for the special consideration given to this product are easy to guess. Wine is a ‘living’ commodity, meaning that it keeps maturing after being bottled and contact with air could expose it to subsequent oxidation and result in it becoming its derivate product, vinegar.3 In that sense, both buyer and seller need to well define the risks when establishing an agreement for the sale of wine, and many cases described in the Digest underline the issues and responsibilities involved regarding time, quality, and quantity.4

• 5 Liou 1987a, p. 74 ; Laubenheimer 2004, p. 153-171 ; Revilla Calvo 2000-2001, p. 209-216; Komar 2020 (...)
• 6 Djaoui – Tran 2014, p. 487-494.

2One can discern the importance of the processes for selling wine from the inscriptions painted on amphorae, which reflect its quality, production, or bottling dates.5 These artefacts, and their epigraphic record, reflect elements of the agreements of which they were a part, and therefore constitute proof of a transactional practice (a contract, an exchange of goods for a price). Thus, comparing different inscriptions regarding wine amphorae will allow us to understand the repetitive elements needed to establish the main features of an agreement, and at the same time, the divergences in some inscriptions will reflect the particularities of the agreement concluded by the parties. In that sense, it would be a mistake not to consider inscriptions on wine containers as evidence revealing first-hand traits of mercantile agreements.6

• 7 See nt. 127.
• 8 Dzierbicka 2005, p. 9-92; 2015, p. 201-208; 2018; Jané 2008.

3This paper is devoted to the study of ten inscriptions written on Italic amphorae that reflect elements of the wine’s harvesting and bottling. Until now, contracts for the sale of wine have attracted deeper consideration by scholars, thanks to the Egyptian papyrological evidence, but the situation in the western Mediterranean is still poorly understood, given the lack of similar evidence for what concerns sale contracts. For the west, scholars have considered several texts from Justinian’s Digest, as well as the few literary sources describing the processes for the production and distribution of wine (e.g. the works of Cato and Columella). In many cases, scholars tend to use the Egyptian evidence to compare and frame the contracts used for the sale of wine in the west. However, Egyptian papyri mostly record one specific kind of sale, the sale of wine on delivery or advanced sale of wine, which does not apply to the evidence studied here (see section 4),7 reason why papyri have not been used as a main source in this contribution. In addition, reading these papyrological studies has made me wonder if we can use Egyptian evidence to understand other parts of the Mediterranean, when features such as calendar dates and distribution patterns slightly differ from the western ones.8

• 9 Sen., Ep., 114.26. acknowledges that fact. Aspice veteraria nostra et plena multorum saeculorum vin (...)
• 10 Pérez 1995, p. 175-186, who establishes the chronology on 140 BC to 50-30 BC, and summarises the fi (...)
• 11 Liou 1987b, p. 131-148.
• 12 Marques Martín et al. 2009, p. 363-371.
• 13 CIL IV, 5512.
• 14 Rigato – Mongardi 2016, p. 101-130.

5Different archaeological contexts have revealed the presence of wine amphorae bearing inscriptions indicating a consular date, ranging from the Roman Republic (second century BC) through the High Empire.9 While the published assemblage of republican amphorae identifies them as Italic containers,10 other examples from later periods indicate that consular dates are also attested on amphorae from Gaul,11 the Iberian Peninsula,12 and Greece.13 Regarding the republican and imperial evidence of consular dates written on amphorae, the most complete compilation is that of Rigato and Mongardi, who provide a very helpful and thorough summary of the materials available.14

• 15 Aguilera Martín 2012, p. 139.
• 16 Aguilera Martín 2007, p. 19-20.
• 17 E.g. Remesal Rodríguez 1986.
• 18 Vera 2006, p. 303-317; Vera 2010, p. 211-227.
• 19 Aguilera Martín 2007, 135-143.
• 20 Rigato – Mongardi 2016, p. 108-125.

6Writing a consular date on an amphorae is a practice that can also be seen in the case of Baetican Dressel 20, starting from mid-second century AD (concretely, during the reign of Antoninus Pius).15 In the case of the Dressel 20, including the consular date was a response to an administrative/bureaucratic need, that is, to indicate the date on which the olive oil was inspected and registered,16 as it was part of the goods included in the annona.17 In contrast, it is a well-known fact that wine was not included in public distributions before Late Antiquity,18 and until now, no amphorae bearing consular dates from late antique contexts has been found. Thus, the most possible answer would be that tituli picti depicting consular dates referred to the needs of private customers. The lack of standardisation on the way of painting inscriptions in the Italic wine amphorae (differently from the case of Dressel 20)19 can be interpreted as a clear sign of the private contexts in which these inscriptions were written.20

• 21 Tchernia 1986, p. 30. Also, Pérez Ballester 1995, p. 180-181, referring to one amphora from Toulous (...)
• 22 Rigato – Mongardi 2016, p. 105. This view was also supported by Peña 2007, p. 51.

7The meaning of these consular dates is still a subject of debate. While Tchernia and Salviat believed that the consular dates corresponded to the date of the vintage,21 Rigato and Mongardi, carrying out a close study of consular inscriptions in Italic amphorae, reached the conclusion that the consular indication referred to the bottling date of the wine.22 According to Mongardi, interpreting these as indicating the production date is difficult, because the consuls documented on tituli picti were in office during different periods of the year, and these do not always coincide with the cycles of wine production as attested in literary sources such as Cato. In that sense, one may argue that if we think of the references to the ordinary and suffect consuls as indicating the vintage’s year, then the buyer would know when the grapes had been placed in dolia for fermentation. Instead, if one thinks about the consular date as referring to the time of bottling, the buyer may not know for how long the wine had been fermenting in the dolia before being poured in amphorae. However, as I suggest that the wine calendar was part of the intrinsic knowledge of any person involved in the wine trade, and therefore by indicating the date of the bottling the buyer may have been able to guess the dates of the wine cycle behind it. In addition, the inclusion of the term diffusio indicates that the wines contained in these amphorae were special, either because they have aged for some years (e.g. CIL IV, 2551), or because they belonged to one concrete dolia that contained a specific must (e.g. CIL IV, 5519), among other reasons. The latter bears witness of the diversity of wines available in the market, as well as of the specific features considered when selling these special products.

• 24 CIL XV, 4582, Sex. Papinio./ et. Plautio. co(n)̣(ulibus) (n.36 p. Chr.)/ mul(sum) fact(um) [Kal. ve (...)

9That assertion is proven by cases such as the Dressel 6 amphora recorded in CIL XV, 4582, in which the consular date specifically indicates the year when the wine was stored in the amphora for aging (between the 17^th October and the 16^th November of the year AD 36).24 The inscriptions included in this paper generally follow the principle of the bottling date (an additional reason of why this assemblage was selected), but cases such as the one quoted above evidence the exceptions to that rule, as well as other particularities that reflect the diversity of practices carried out by the merchants.

• 25 Berni Millet 2021, p. 23.
• 26 Harris 1989, p. 8.
• 27 Harris 1995, p. 19-27, and later, Woolf 2009, p. 56-58.
• 28 The well-known negotiatotres vinarii, e.g. CIL XIII, 1911; CIL XIII, 1921= AE 1974, 422; CIL XIII, (...)
• 29 e.g. Cat., Agr., 23; 25; 113; 148.
• 30 e.g. Varr., Rust., 1.44.
• 31 e.g. Plin., H.N., 18.315-316.
• 32 e.g. Col., Rust., 11.2.60.
• 33 For a summary on vintage sources, see Thurmond 2017, p. 137, and on fermentation, p. 178.
• 34 The Vinalia Urbana (Varro., Ling., 6.16.20; Ov., Fast., 4.863-869; Plin., H.N., 18.287; Fast. Praen (...)
• 35 The feast of anthesteria in Attica took place on 11-13 Anthesterion, which is the end of February. (...)

10The formulary and abbreviated character of inscriptions painted on commercial amphorae is a sign that these were meant to be read by specialists working in a specific business, and not by the wide public.25 In that sense, I should recall the concept of “craftsman’s literacy”, first mentioned by Harris in 1989,26 and his later designation “specialist literacy” to characterise the decoding skills required for those who manufactured, transported, and traded the portable objects known as instrumentum domesticum.27 Following that, it would be important for these persons to know the date when the wine was bottled, because combining that with their general knowledge regarding the dates of the wine production cycle would give them an idea of the product’s features. In that sense, I would like to think that the phases and dates of the wine calendar from a specific region may have been quite straightforward and known both to specialists working in the wine business and also to the buyers.28 The latter can be inferred from the sources available for Italy (Cato,29 Varro,30 Pliny the Elder,31 and Columella).32 It is interesting that all of them seem to say very little about a specific “calendar”, possibly because they treated the months in which the respective stages of winemaking took place as something obvious. They only seem to address the issue of the duration of certain phases, such as for how long certain kinds of wine should be stored after the vintage, where and how. They indicate the number of days or months but mentions of specific dates or even the names of months are rare.33 The fluid nature of the wine calendar in the Mediterranean is well illustrated by the attested dates for the tapping of the vintages,34 as well as for the opening of jars of new wine.35

• 36 See, for example, BGU XII 2209 (Herm. 614) for an example of a complete contract of this type.
• 37 Eg, BGU XII (n. 2176, 2207, 2209); CPR VIII 60 (p. 155) and Pringsheim 1950, p. 494-503.
• 38 Dzierbicka 2018, p. 180-194.
• 39 Kruit 1992, p. 273.
• 40 Dzierbicka 2011, p. 3, it was the low level of wine production that prevented the wine from being s (...)

11A similar specialist knowledge can be assumed for the Egyptian wine calendar when reading documentary papyri. However, the situation there was slightly different, because of the different climate cycles, which can be appreciated thanks to the attestations from contracts (especially loans) that required specific dates.36 We know from several texts that the period for picking the grapes was between the end of July and the middle of August.37 Immediately after gathering the grapes, they were pressed so the fermentation process began, and the wine remained in the vats until the wine was made available for consumption.38 Then the wine was drawn off from the vats in the months Choiak, Tybi, Mecheir or Phamenoth, i.e. in the period December-March.39 These dates may have been known by the locals who were trading and acquiring the wine, as it is well-known that Egyptian wine was not exported on a large scale.40

• 41 Cato., Agr., 147-148. These dates are indirectly considered by Ulpian in the Digest’ texts D. 18.6. (...)
• 42 Aubert 2004, p. 132, uses a similar reflection concerning agricultural production cycles in general (...)
• 43 Gaius drew a distinction between the wine’s first seller and a “merchant who is accustomed to buy a (...)

12These different groups of documents only give us ballpark dates on when wine was considered ready to drink, as this was – technically speaking – not their main concern, but rather when it would potentially start to spoil. Following Cato, the grapes may have been placed in the vats in January, and then stored for fermentation until spring (March-April), and could be poured out until October.41 Aditionally, the individual vigneron will have made an exact determination as to the specific day that the vintage will commence, just as today. So it seems that even though the calendar was roughly the same throughout the Mediterranean, it differed locally for a number of reasons and the producers had to “play it by ear”. In any way, what seems clear is that people involved in the wine world, either producing or selling, may have had some knowledge about the wine chronology and phases,42 what would have provided some information in addition to the consular date.43

• 44 Zimmerman 1996, p. 284-285.
• 45 Rossiter 1981, p. 346-353.
• 46 Van Oyen 2020, p. 51-52. Indeed, the layout and organisation of the storeroom promoted long (er)-te (...)
• 47 Ancient authors describe how the residue from the press-beds could be mixed with water to provide a (...)
• 48 Cf. the Phoenician-Punic tradition of having wine ferment in amphorae or transport containers, Peña (...)
• 49 Such as at the Catalan site of Baetulo, see: Peña Cervantes 2010, p. 164-165.
• 50 Peña Cervantes 2010, p. 32-33; Tchernia 1986, p. 28-36; Tchernia – Brun 1999.
• 51 Ath. I, 26c–27b; Hor., Od., 3.8; 9-12; 21.1; 23.5-8; epod., 13.6.; Gal., Antid., 14.15 (on Surrenti (...)
• 52 E.g. AE 2009, 709; AE 1988, 862b; CIL IV, 5526c; Pais 1076, 10.
• 53 E.g. CIL XI, 6697, 3; CIL XV, 4532.
• 54 CIL XV, 4539 = ILS, 8580.

13However, that specialist knowledge would not have been that helpful for wine stored in dolia before being bottled unless some indication was provided, as the buyer may only know the date of bottling through the consular mention. When talking about wine, some modern authors have indicated that wine from dolia was usually new wine (even referring to the inefficient sealing of the dolia),44 but indeed that was not always the case, as wine could also be stored in dolia for quite long periods. The material transformation of wine from grapes followed a segmented production process that could be interrupted at various stages to deliver products of varying quality.45 Thus, storage fed into this process of qualitative differentiation.46 The cheapest, low-quality wines were probably never stored at all,47 but used or exported directly upon pressing,48 or after a first, short fermentation period in an open basin or lacus.49 The longer and more segmented the production process – including fermentation through storage – the higher the quality and value of the resulting product.50 All good vintages must, therefore age. Several ancient authors mentioned their preferences in terms of wine aging, generally agreeing on an optimal age of between five and twenty-five years.51 Many inscriptions on amphorae specify the age of the wine in some way, from the simplest, which only say old wine,52 or wine so many years old,53 up to the most elaborate, which indicate not only the consular date of the harvest but also that of the placing in an amphora, after a period of aging in a dolium that could last up to five years.54 This is the case with the ten inscriptions studied here, which should be considered as examples of this practice of selling high-quality wines. As I will show in the next section, the assemblage confirms that the consular date corresponds to the bottling date, but also that the added information regarding the diffusio among other details recorded in the inscriptions supply extra information to supplement the customer’s intrinsic knowledge about wine cycles.

• 55 The Egyptian documents rarely provide information on this process, barely referred to in P. Oxy. XI (...)
• 56 D.18.6.1.2 (Ulp., 28, ad Sab.).
• 57 Billiard 1913, p. 487-489; Thurmond 2017, p. 208.
• 58 Plin., HN., 18.232; Cato., Agr., 120; Varr., Rust., 65 (referring to mustum).
• 59 For more details on the tituli, see Rigato and Mongardi 2016, p. 108-125.
• 60 Again, following Cato., Agr., 184 (which does not mean that this chronology can be subject to local (...)
• 61 Thurmond 2017, p. 208.

14Diffusio or tapping was a procedure performed when the wine was poured out of jars containing wine which were sealed or corked.55 This action was related to the tasting of wine, because it helped preserve the quality of the liquid and prevent substitution.56 Tapping can be linked to three different moments in the wine cycle; first, the spring tapping of bulk wine not previously bottled, and preparation for shipment;57 second, it could be bottled for further aging inside amphora;58 and finally, it could be bottled after having fermented for a while inside the dolia. Within the context of the ten tituli picti analysed in the contribution, it seems appropriate to make a distinction between those that report the consular date referring to bottling, as well as an additional consular date indicating the vintage year (CIL XV, 4573; 4539; 4571; 4588; 4610 and IV, 2551), and the four specimens (CIL IV, 2561; 5519 and XV, 4531; 4576) which apparently present only the indication of the diffusio.59 In the event that the absence of the mention to diffusio is actually not simply a gap in the inscriptions, this would in fact be the usual indication probably also underlying the remaining tituli on Italic wine amphorae bearing a single consular date, usually without specification of the term diffusio, which in many cases indicated a special vintage. In addition, that double-dating practice makes special sense in cases where the consuls quoted were active over a long period (e.g. one whole year), because in absence of other indications a single consular date provides only a loose idea about the date when the wine was poured out (from April and until October).60 The latter reaffirms my previous argument on the knowledge that wine merchants would have had about the wine production cycles. For the case of wine aged inside amphorae, they only needed to consider that the liquid would have been poured inside the dolia during the spring of the year attested by the consular date.61

• 62 Also mentioned by Dressel 1879, p. 50-51.
• 63 Cfr.BCom 1879, p. 36-112; BCom 1980 82-117; Lugli 1940, p. 375; 1977, III G, 13-14, a e b, via Mont (...)
• 64 Antico Gallina 2011, p. 189; Rizzo et al. 2022, p. 234, 243-250.

15The first titulus is CIL XV, 4573 (fig. 1),62 written on a Campanian amphora Dressel 2-4, found in the fossa aggeris at the castra Praetoria in Rome. This structure consists of a huge amphora deposit, used to fill a pit located in an area from the regio VI (Alta Semita), between the ancient Porta Collina and the Porta Viminalis, outside the servian walls.63 Regarding this archaelogical context, it is important to indicate the recent correction proposed to the initial dating of the site by Dressel, who limited its chronology from the first century BC until half of the first century AD. Instead, a thorough study of the context and all its materials has allowed to enlarge that chronological frame until the end of the first century AD.64

• 65 Many Christian authors refer to these consuls, as AD 29 is the typical date indicated for the cruci (...)
• 66 Tchernia 1986, p. 167-168, 260, 338, 348-349.
• 67 Tchernia 1986, p. 55.

16The item was inscribed by two hands, the first part located in campana written in rubrum reading Geminis. II(duobus) co(n)s(ulibus) / Ha[tr]anian(um?) / diffu(sum) [--] vel diffu[--] trimum. The other part of the titulus (also located in campana) is overwritten in atramentum, following the previous one, and reads A/ Sex. Papinio./ Q. Plautio co(n)s(ulibus). Regarding the first part of the inscription, the two consuls Gemini, what allows us to date this first inscription to the year AD 29.65 Tchernia doubts the reading Hatrianum (that indeed should read Hadrianum) describing the wine, since this supposedly corresponds to wine bottled in the areas of Hadria (Hadriaticum from Greece), and generally bottled in Dressel 6 from the Adriatic area.66 However, as no good location of wine producing amphorae from the republican period has been found in the Adriatic, another most possible option is that the inscription actually concerns its provenance from Atri/Hadria, in southern Picenum.67

• 68 CIL XV, 4610 and CIL IV, 2551 (= 5520).
• 69 Peña 2007, p. 61-70; Thurmond 2017, p. 227; Pecci 2017, p. 515-521.

17The last bit of this first part of the inscription refers to the diffusio, which took place three years (trimum) after the wine was placed inside one dolium. Other inscriptions included in the assemblage studied in this paper also use the adjectives trimum or bimum,68 which I think will support my views here. In this case, I think that we need to think of the wine as originally dating to the year AD 26, then having aged inside the dolia for three years, and subsequently being bottled in the year AD 29. For the second part of the writing, there is an additional consular date that corresponds to AD 36, and which I think corresponds to a reuse of the container, which was usual for wine amphorae.69 In that case, the consular date would refer to the bottling of the second wine in the vessel.

• 70 A densely populated area, see: Simelius 2022, p. 217, 219.
• 71 Dressel refers to the consul suffectus Anicius Cerialis, quoted in Tac., Ann.,15.74.
• 72 Aging the wine during two years in amphorae is mentioned by Hor., Carm., 3.8.9 ff.
• 73 See also, CIL IV, 9316; CIL XV, 04619; CIL XV, 4585, and see also Lanciani 1874, p. 215.
• 74 Gasperini 1971, p. 138-139.
• 75 e.g. AE 1936, 90; AE 1912, 242; Bloch 1947, p. 10, n. 15.
• 76 E.g. Djaoui – Tran 2014, p. 1-2, referring to the producer Valerius Proculus.
• 77 Tchernia 1986, (Falernum), Marzano 2007, p. 21, 30 (caecubum and fundanum), see also Plin., H.N., 3 (...)

18The titulus recorded in CIL IV, 2551=5520 is a fragment of an unknown amphora type found in Pompeii in July 1867, at the peristile of the Via Stabiana.70 The piece provides additional arguments for the interpretation of the previous inscription. In this case, it is written by one single hand, in album, and reads: C. Pomponio. C. Anicio co(n)s(ulibus)/ ex fund(o). Badiano/ diff(usum). Id(ibus). Aug(ustis). bimum.71 So the consular and botling year was AD 65, although the wine had been aged for two years, as indicated by the adjective bimum,72 so constituting a wine from the vintage of AD 63. In addition, the wine was bottled on the 13^th of August (idibus augustis), when the consuls were still in their roles. An additional detail in this inscription is that it refers to the fundus Badianus, where the wine was produced,73 one detail that also appears written on some wine samples,74 and otherwise, on inscriptions written on tegulae.75 Mentioning the estate where the wine was produced was a way to indicate the good quality of the wine by acknowledging its origin, as was also the case when the producing state or producer was identified in relation to a wine,76 thus indicating the products’ origin in a different fashion.77

• 78 On the tabularium, see LTUR V, p. 17-20.
• 79 For the consular dates, see Phillips 1997, p. 107-108.
• 80 As indicate Rigato – Mongardi 2016, p. 107.

19The inscription CIL XV, 4588 (fig. 2) corresponds to a Dressel 4 similis that was found at the Tabularium in Rome,78 written in rubrum, apparently by the same hand, in two different parts of the amphora. The first bit, written on the neck (in collo), states: Pr(---) vel [---]pr(---)/d(iffusum). Pasien(o), while the second bit, on the belly (in ventre), reads: (vinum allatum est) Non(is). F(ebruariis). ex/ Ros(ciano). n(ostro vel novo). C(aesare). XII. co(n)s(ule)/ d(iffusum). N(onis) Iul(iis). Pasien(o).79 Like the previous case, this inscription mentions the name of the estate or fundus where the wine was produced (Rosciano). This inscription has two peculiar details: first that the consular name referring to the year of the diffusio (4 BC) appears written twice; and the second is that the additional date, supposedly indicating the year when the wine was produced. Regarding the first occurrence, I would say that this could be simply something that can be attributed to the person writing the inscription, who perhaps wanted to clearly indicate the bottling date in both parts of the vessel, thus highlighting the importance of this procedure for the wine production cycle. As for the second date (fifth of February from the year 5 BC), if this corresponds to the production date of the wine, it seems a bit off from the alleged traditional placement of the grapes in vats in January.80 However, as I have indicated previously, these dates could vary from one vineyard to another, and secondly, the date does not really seem to be that far from the end of January, when the grapes, according to Cato, should be placed inside the vats for fermenting.

• 81 Rigato – Mongardi 2016, p. 105, indicating the uncertainty of the modalities of communicating the c (...)

20CIL XV, 4610 (fig. 3) corresponds to an unidentified amphora fragment found in Rome, written by a single hand in rubrum, reading: Sisen[na Stati]lio / [S]cribonio Libone co(n)s(ulibus) / [dif]fusum bim[um 3] / Idus Maias. The consular date provided corresponds to AD 16, but again, bearing in mind the adjective bimum, the vintage of the wine needs to be chronologically located in AD 14. In this case, the writing indicates that the wine was concretely bottled on May 15, a bit of a problematic date, because the fasti consularis indicate that Scribonius Libo was in charge only until April 12. While one option could be that this indicates that a new consul had not yet been appointed, other could be that the person writing the titulus did not know about the new person in charge and still thought that Scribonius was the active consul.81 However, the most probable explanation for this is that they remained in their roles for the entire semester, what would make sense in terms of the indicated bottling date.

• 82 Also mentioned in Pérez Ballester 1995, p. 185.
• 83 Lanciani 1874, p. 40-41.

21The titulus from CIL XV, 453982 (fig. 4) constitutes one of the most singular inscriptions relating to wine that has come down to us, according to Lanciani.83 The inscription was found in the Esquilinum neighbourhood from Rome, in the Atrium Effossae (viale principessa Margherita with Ager Servianus), and was written in rubrum by a single hand. The inscription reads: ti. claudio. p. Quinctilio. co(n)s(ulibus)/ a(nte). d(iem). XIII. K(alendas). iun(ias). vinum/ diffusum. quod. natum. est/ duobus Lentulis. co(n)s(ulibus)/ Autocr(atum).

• 84 Tchernia 1986, p. 30.
• 85 In other cases, referred by writing vindemia on the amphora, e.g. AE 1997, 1221b-c; Piccottini 1997 (...)
• 86 In the CIL, Dressel indicates that autocr(atum) appears in Plin., HN., 15.78, in Galen.,10.832 and (...)

22As can be read, the inscription includes two consular dates, the first one corresponding to 13 BC (ti. Claudius Nero i; P. Quinctilius Varo), and the second dating to 18 BC (P. Cornelius Lentulus Marcellinus; cn. Cornelius Lentulus). The first consular reference provides the date when the wine was bottled (around the 20^th of May), while the second indicates the year of the vintage, 84expressed by quod natum est.85 Concerning the mention of Autocratum, the meaning of the inscription is unknown, and it does not appear on other amphorae nor inscriptions.86 Instead, it seems appropriate to recall the passage from Athenaeus (I 32F), already mentioned by Dessau (in ILS 8580) in which the adjective is used in reference to one of the typologies of Chios wine.

• 87 Bodel 1995, p. 289.
• 88 CIL XV, 4646; CIL XV, 4784.
• 89 CIL XV, 4559.

23Another inscription (CIL XV, 4571 (= ILS, 8581)) displayed in fig. 5, uses natum to describe the wine’s features. It is an amphora Dressel 2-4, bearing an inscription written in rubrum by one hand, which reads: (vinum) nat(um). Gavian(o fundo). m(ilia). p(assum) XLV./ Imp. Caesare. et. Silvano co(n)s(ulibus)/ diff(usum). M. Servilio. L. Aelio Lamia. co(n)s(ulibus)./ suff(ectis). P. Silio. L. Volusio. Saturnin(o). In this case, the details not only include the year of the vintage (2 BC), thanks to the consular dating,87 but also the location where the grapes were produced, that is, 45 miles from the fundus Gavianus. The wine was poured into the amphora on AD 3, after aging for five years inside a dolia. It seems appropriate to point out that the chronological indication of the diffusio, unless one assumes that the specific date was included either in the final part of line 2 or in a fifth line that has not been preserved, is very generic, including not only the ordinary consuls but also of those suffects for AD 3. This amphora was found in the fossa aggeris in Rome, where other three amphorae were discovered bearing the same consular dating of AD 3,88 at least one of which referred to a high-quality wine such as Falernum.89

• 90 It should be noted that, despite the identification in the CIL as Schöne-Mau X (= Dressel 10), it i (...)
• 91 For the consular couple mentioned in CIL IV, 2561 (p. 14), i.e. the ordinary consuls in the first h (...)
• 92 Tchernia 1986, p. 45, 176, 202 and esp. p. 344-347; and for a complete list of the sources in which (...)
• 93 CIL IV, 2555; CIL IV, 2556; CIL IV, 5514; CIL IV, 5521; CIL IV, 5522; CIL IV, 5525; CIL IV, 5560; C (...)
• 94 Hor., Sat., 2.4.5.
• 95 Strab., 5.4.2.
• 96 D. 33.6.16 (Proc. 3 ex post. Lab.) Qui vinum Surrentinum in urnalibus habebat diffusum, is tibi vin (...)
• 97 It should be noted that, despite the identification in the CIL as Schöne XIV (= Dressel 12), it is (...)
• 98 AE 1972, 117, and Gasperini 1971, p. 138-139.

24The titulus recorded in CIL IV, 2561 is written in album by a single hand on a Dressel 2-4 found in Pompeii,90 bearing the inscription: Surr[entinum?) vel Suff(usum?)]/Crii[---]bi[----]/[dif]f(usum) pr(idie) Idus/ Iun(ias)/ Ve(r)ginio et Re(gulo) co(n)s(ulibus)].91 First, it is necesssary to acknowledge that this amphora contained a high-quality wine, the Surrentinum,92 from the Pompei-Sorrento area, mentioned in other amphorae from the area,93 and acknowledged by several ancient authors such as Horatius,94 Strabo,95 and even the imperial jurist Labeo.96 In this case, the wine was bottled in AD 63, on the day before the 13^th of June. Another Pompeian titulus corresponds to inscription recorded in CIL IV, 5519, and dates to the year AD 63, also identified as a Dressel 2-4.97 The inscription, written by a single hand in album, reads: Dol(ium). I./ dif(fusum) est Id(ibus). Iu[niis vel Iuliis] /L. Verginio C. Regulo/co(n)s(ulibus). Thus, the wine was bottled before June 13 and, in this case, mentions the dolia (number 1), probably where the wine was stored. The latter also appears written in two written samples from Taranto, inscription that mentions the kind of wine, then the fundus of origin, and finally the dolia in which the wines were kept.98

• 99 Tchernia 1986, p. 65, 103, 159, 160, 184; and for a complete list of the sources in which this wine (...)
• 100 This wine was also mentioned on a jar found in Arles, which was meant to serve as a sample for a po (...)
• 101 Berni Millet 2021, p. 25.
• 102 D. 8.3.6pr (Paul 15, ad Plaut.); D. 33.7.7 (Scaev., 22, Dig.); Di Porto 1984a, p. 72-82, and Manaco (...)

25Back in Rome, the inscription from CIL XV, 4531, which we can see in fig. 6, is from a Dressel 2-4, and was written in gilvum by a single hand. The text is incomplete but reads: Alb(anum)/ Gaetulico et. Calvisio [co(n)]s(ulibus)/ diff(usum) Gamus l(ibertus) [--] vel L[--]. The wine is identified as an Albanus, which was a well-known99 high-quality wine from the Alban Mountains in Latium.100 In this case, the bottling was done in AD 26 by someone named Gamus, who Dressel identified with a freedman. The latter can be interpreted as a sign that the vineyard had freedmen as workers and mentioning them could be a way of documenting their labor, as well as acknowledging their liability for the action of bottling the wine. The practice of indicating the names of freedmen in instrumentum domesticum to document their work is known and can be seen for example in inscriptions from Dressel 20. However, in the case of the Dressel 20, these liberti were mentioned in stamps or graffiti that were meant to be read by the personnel working in the kilns, who would then know the amount of work done by them.101 Instead, this inscription is a painted titulus, what makes me wonder if it could mean that the freedman was not contracted by the wine estate, but was sort of working on commission or in specifically contracted work, which may be the reason why he was mentioned in an inscription meant to be read by the customer.102 Otherwise, this could simply be linked to the issue of acknowledging his liability in relation to the delicate task of pouring wine into amphorae.

• 103 CIL XIV, 2112 = FIRA III, 35.

26The final inscription appears written in rubrum in an amphora Dressel 2-4 (CIL XV, 4576). The writing is quite faded, but when reading it, Dressel interpreted: Lani(vinum)/ Cn. Domitio. Ahenoba(rbo)/ Camillo Sc[ribo]niano/ VIII Idus. d(iffusum). [per?] Epaphum l(ibertum)[---]. The first bit was reconstructed as Lanivinum, thus originally from Lanuvium in Latium. The place is perhaps not that well known for its wine production, but it was mentioned in an inscription indicating that the members of their undertakers’ guild used to drink good wine offered by the slave members of the collegium on the day when they were freed.103

• 104 D. 14.3.16 (Paul 30, ad Ed.); Di Porto 1984b, p. 3256; Manacorda 1985, p. 142.

27The other part of the inscription refers to the year when the wine was bottled (AD 32), being either on the eighth of May or the fifth of June (VIII ante idus), since L. Arruntius camillus scribonianus was ordinary consul from January until June of that year. Finally, the titulus indicates the name of the freedman who carried out the bottling. As I have indicated for the previous inscription, this mention may have related to some sort of task carried out by the freedman in the vineyard, evidencing the existence of integrated enterprises in the wine trade. In this context, freedmen participated in the transformation of the produce of the fundus, the production of containers for its preservation, and in their commercialisation.104

• 105 See also: Mataix Ferrándiz 2019, p. 177-205; 2022, p. 131-157.

28In sum, that the producers decided to include several extra details in addition to the bottling date (referred by the consular mention) also indicates that these wines had a specific quality that made them special (for example, belonging to a specific estate). In addition, I have already mentioned that the double-dating practice was especially helpful when the consuls quoted were active over a long period, to provide a better idea about the actual date when the wine was poured out. The latter also indicated a special concern by the producer to make clear to the customer the actual dating of the wine, indicating that this was a quality product. These inscriptions had a value for both merchants and customers and were considered when establishing their agreements, and therefore they should be read considering the particularities of the sale of wine.105

• 106 D.32.60.2 (Alf. 2, a Paul epit.); 36.6.6 (Proc. 5, epist.); D.33.6.14 (Pomp. 6, ad Sab.); D.33.6.15 (...)
• 107 Peña 2007, p. 47-49.

29The significance of these inscriptions would have been understood by the subjects involved in wine trade, as a practice that acquired its meaning through its use by merchants in their transactions. These practices, as many others used by merchants across the Mediterranean, were justified by daily uses and needs, trust and mutual understanding. Why do I say so? Because when one looks at the diverse texts from the Digest mentioning diffusio, these are merely concerned with the character of wine containers as being part of the goods included in an inheritance.106 The Digest’s texts differentiate among dolia or vasa, which were fixed facilities (included in the instrumentum fundi) that served for the temporary storage of wine prior to its sale, from amphorae or cadi, which served for its later and longer storage and distribution.107 In sum, there is no mention to diffusio in The Digest in relation to the sale of wine.

• 108 E.g. D.11.6.5.2 (Ulp. 24, ad Ed.); D.18.1.35.5 (Gai. 10, ad Ed. Prov.); D.18.1.71 (Pap. 1, Const.); (...)
• 109 E.g. Honoré 2010, p. 79-109; Liebs 2021, p. 261-262.
• 110 Babusiaux 2016, p. 177-186.
• 111 D.13.6.18.3 (Gai. 9, ad ed. Prov.); D.18.6.1pr. (Ulp. 28, ad Sab.); D.18.6.1.3 (Ulp. 28, ad Sab.); (...)
• 112 I will refer further to these issues in the next section.
• 113 Even though sealing is a well-attested practice in both the archaeological and textual evidence. Se (...)
• 114 See nt. 107. This difference between the world of books and actual commercial practice is not uniqu (...)

30However, that Roman jurists used some terms to refer to concrete procedures, as for example happens in the case of mensura connected with the sale of wine,108 and disregarded others, does not mean that these procedures were not employed by commercial practitioners. On the one hand (and belonging to a different discourse), there is the issue of the scattered number of sources compiled in the Digest, what may have implied that several discourses may have been left out when selecting the sources to include in it.109 On the other hand, one needs to bear in mind that jurists tend to comment on specific cases, and thus used certain kinds of narratives to convey their messages to the readers.110 In that sense, mensura and the other procedure of effundere (to spill)111 were more problematic for wine sold from dolia, as these determined the price and marked the moment until the wine sold in advance could be collected,112 and thus could imply more issues related to contractual matters than for bottled, sealed wine.113 Finally, there are many elements of the contracts employed in commerce described by Roman jurists that can be connected to the diffusion, such as the need to taste the product before acquiring it (degustatio) for the sale of wine, attesting the importance of having in mind when the wine was actually poured out of the vat. In that sense, one must recognise that this mention to diffusio in the context of wine sales was part of the world of commercial practice, so mundane to be even mentioned by the jurists in their texts, as happens as well with the samples, another ignored element in the jurist’s works, but quite used in commercial practices.114 In sum, even though the term diffusio does not seem to be considered by the Roman jurists except in cases of issues regarding inheritance, its mention on inscribed containers indicates a specific commercial practice to sell high-quality wines.

• 115 See bibliography in nt.2.
• 116 Or even Britannia, which deserves to be studied as its own case study, see e.g.: Brown et al. 2001, (...)
• 117 Sale was binding upon consent before delivery of the goods or payment of the price, see e.g.: Yaron (...)
• 118 In addition, the sale contract was governed by the ius gentium (law of the peoples), so that all su (...)
• 119 Arangio Ruiz 1978, p. 86-87, indicates the variety of documents that we can find reflecting sales i (...)
• 120 Jakab 2015, p. 214.
• 121 A sale that was only valid after having performed the tasting of the wine.
• 122 A sale that was only valid after having measured out the wine from the dolia.
• 123 Jakab 1997, p. 318-324.
• 124 E.g. Frier 1983, p. 276 n.76 finds five different models in the sale of wine. Olszak 1990, p. 361-3 (...)

31Although wine has been a topic of interest for many scholars, there is unfortunately no material evidence attesting a contract of sale as such for the western part of the Mediterranean. For the eastern Mediterranean, studies rely on papyri which, given the absence of other sources in the west, have been used as paradigm to understand how the contract of sale worked.115 However, Egyptian wine’s distribution, as well as its agricultural cycles had some particular qualities and cycles that differed from other areas such as Gaul, Italy, or Hispania.116 Notwithstanding the area differences, these procedures could be assumed to have operated under the umbrella of the contract of sale, which constitutes a consensual exchange117 of an object for a price.118 It was a contract that was very adaptable to circumstances, so the parties involved could add conditions to the deal consisting of inducements for the acceptance of the sale in each case.119 The Digest of Justinian deals with sale transactions (emptio venditio) in detail, as the decisions and opinions of Roman jurists are compiled in eight chapters. Of these, three chapters are written from a rather theoretical point of view (D. 18.1, 18.2 and D. 19.12) while six mirror contractual practice (see D. 18.2-73). The latter indicates that the Roman law of sale was a matter of routine and experience, and that the rules for concluding a contract, specifying the liability of the parties, and allocating special risks seem deeply rooted in every day legal practice (law in action).120 To understand these practices, Roman law scholars have long attempted to explain the wine passages in D. 18.6 according to different agreement models (emptio ad degustationem,121 emptio ad mensuram122 and their variants),123 whereby the starting point is always the compatibility with the so-called theory of perfection (regarding the moment when the contract is valid and openly displays all of its effects).124 That mindset, however, leads to contradictions that can often only be resolved by attempts at interpolation or dogmatic compromises, which in addition did not consider in any case material sources such as the ones presented in this contribution.

• 125 Jakab 1999a, p. 71-111.
• 126 See also, Bagnall 2011, p. 85-96; Gallimore 2012, p.151-165; Kruit 1992, p. 167-184; Thür 1995; 199 (...)
• 127 Jakab 1999a, p. 71; 79-80; 84-85.
• 128 Jakab 1999b, p. 38.
• 129 To build up her case, Jakab used several documents such as BGU XII 2209 or SB XVI 12486. For a comp (...)

32Contrasting with these views, Jakab studied the sale of wine focusing on Egyptian papyri, since the large number of surviving documents allows a “statistical” examination of the usual contract clauses, which reveals the contours of certain typical agreement models employed in Egypt.125 Even though there are other kinds, Jakab focuses her attention on a concrete contract, such as the sale of wine on delivery, whereby a party acknowledges receipt of money as the price for goods to be delivered later, which is the most known and widely discussed type of transaction thanks to the papyrological evidence.126 However, the article also provides useful insights that generally apply to the wine trade overall, as she recognizes that the interplay of the complicated methods of winemaking and individual economic interests necessarily led to a colorful variety of business types.127 Notwithstanding these limitations, she indicates that the most accurate way to divide the documents would be into two major groups: with and without guarantee .128 In the first one, the seller would be liable until the wine is delivered to the buyer, while in the second case, the purchase was the fresh must, which was immediately handed over by the wine press.129 Although Jakab proposed these two sorts of possibilities thinking of the contract of wine sold on delivery, I think that the idea that some wines may have been sold under certain warranties should not be completely disregarded for other case studies.

• 130 Tchernia 2011, p. 201-209.

33To understand our ten amphorae, one needs to combine the legal expertise of Jakab with the knowledge of ancient wine from André Tchernia, who proposed the existence of three kinds of wine sale. First, the producer makes and looks after his wine, and then the trader comes to the cellar door to make the purchase and take delivery after the vinalia in April, or later. Second, the producer sells the standing crop; and third, the producers undertake the transport of their own wine from the estate.130 For the case of these ten vessels, the only available options seem to be the first and the second types, because all the cases imply the sale of wines that have fermented and aged in the seller’s dolia.

• 131 Cato., Agr., 154.
• 132 Almagro and Vilar Sancho 1968), p. 323-336; Hesnard – Gianfrotta 1989, p. 393-441.
• 133 D. 18.6.4pr. (Ulp. 18, ad  Sab.); D.18.6.16 (Gaius 2, Rer. Cott.).
• 134 D. 18.6.4.1 (Ulp. 18, ad Sab.) difficile autem est, ut quisquam sic emat, ut ne degustet.
• 135 Tchernia 1986, p. 31.

34Thus, I think that, except for the amphorae compiled in CIL XV, 4531 and 4576, the rest of the inscriptions were bought from the cellar, and the seller took care of all the processes of fermentation and aging of the grapes.131 Then the liability for the wine would be transferred to the trader, who in many cases (especially for Dressel 1) marked their names on the amphora stoppers.132 This agreement may have included a condition of tasting,133 which seems to have been more common for wine bought in bulk, but as Ulpian notes: “it is hard to believe that anyone would buy wine without a proviso that it is to be tasted”.134 Since all of these wines were poured into amphorae a long time after the grapes fermented, it would have been common sense that the buyer tested the wine before the diffusio, especially considering that the high quality of these wines (Surrentinum, Albanum) and their age (bimum, trimum) would also have had an impact on their price.135

• 136 AE 1972, 117 and Gasperini 1971, p. 138-139.
• 137 D. 18.9.2 (Ulp. 28, ad Sab.).

35One peculiar case is that of CIL IV, 5519, in which the inscription indicates that the wine was came from one concrete dolia from the cellar. As I have mentioned previously, two samples from Taranto also indicate the dolia from where the wine was poured.136 I am not saying that this case must necessarily have implied that the wine was tasted before being acquired by the buyer by using a sample, but that could have been a possible option. In any case, I think that here the buyer may have had some knowledge of the kind of wines offered by the seller, and concretely wanted the produce of that specific dolia. The inscription in this case will imply the liability of the seller that the wine sold was that specific one, and no other from his cellar, which implies a specificity of the object regarded in the contract that could invalidate it.137

• 138 Rigato – Mongardi 2016, p. 111 n. 37.
• 139 See the inscription CIL XV, 4571 (=ILS, 8581) studied previously in this paper, as well as CIL XV, (...)
• 140 See CIL XV, 4653 for a similar numeral inscription.
• 141 Zevi 1996, p. 216-219; Tchernia 1986, p. 132.

36Regarding that first kind of sale, I would like to underline another concrete example that diverges from those in this contribution, but that allows us to see the many different options for wine sales that were available to practitioners. This is the case of CIL XV, 4559, which displays two inscriptions written in rubrum by two different hands. The first bit, inscribed on the upper neck, reads: Fal(ernum) fund(i) Aniani. / M. App(uleio). P. Sil(io) co(n)s(ulibus), thus referring to two consuls dating to the year 20 BC,138 while the second part, written in the neck indicates: L. Saturnini./ CCX[-]IX, thus indicating the name of a consul in charge in the year AD 3.139 In addition, the second inscription includes a numeral that should be the number of amphorae from the lot to which this one belonged.140 Here we could be either facing a reuse, in which the second date would correspond to the bottling date of the new wine, or (as I think), this is a case of a very special wine, aged for 23 years inside the amphora. Why indicate the name of the consul, when the buyer could calculate the wine’s age thanks to the first consular date? I think that in this case the inscription may refer to some kind of promemoria for the internal use of the cellar, which would have been one of the labels from the amphorae intended for sale.141

• 142 See n. 92.
• 143 D.14.3.4 (Paul 30, ad Ed.) cum interdum etiam ad homines honestos adferant merces et ibi vendant. N (...)
• 144 Lenel 1989, p. 1024.
• 145 D.14.3.16 (Paul 29, ad Ed.) Si cum vilico alicuius contractum sit, non datur in dominum actio, quia (...)
• 146 Tchernia 2011, p. 206-207.
• 147 D.14.3.1 (Ulp. 28, ad Ed.).

37Finally, concerning our inscriptions CIL XV, 4531 and CIL XV, 4576, the inscription mentioning the freedmen could act as a note to the producer, who may have put them in charge of bottling and shipping the wine from the estate.142 The Digest includes several texts referring to the responsibilities of slaves and freedmen working in estates, such as the case of D. 14.3.4 (Paul 30, ad Ed.),143 thanks to which it can be affirmed that the goods conveyed were probably agricultural produce, as the previous comment on the edict was focused on the role of the vilicus in a farm,144 see D. 14.3.16 (Paul 29, ad Ed.).145 As indicated previously, both of the wines bottled in these jars came from Latium (Albanum and Lanivinum), and the amphorae were found in Rome, which can suggest that these were sold to nearby buyers,146 who could complain in case of trouble, having registered in writing who was in charge of the pouring.147

• 148 Jakab 1999a, p. 84-85.
• 149 Johnston 1999, p. 9, “to write history using the legal sources alone is inadvisable, whenever possi (...)

39The study of these amphorae in relation to legal and literary texts underlines the need to combine sources to get the closest idea of how such transactions worked. As suggested in this paper, many elements of a sale were customary and belonged to the world of commercial practice, which is the reason why the Digest’s texts, the literary sources, and the papyri do not provide the economic content of different trading practices at first glance. For example, the Digest constituted a scattered compilation of juridical texts that were written by jurists for a specific audience, which would vary depending on the kind of book to which the fragments belonged. However, the continued use of papyri in which scribes notarised wine purchases led to a certain standardisation, formulating a somewhat schematic document, in which the wording has hardly changed over the centuries.148 Although I have not used papyri in this contribution, the latter needs to be considered to underline the incompleteness of the available sources, and therefore the need to combine their study with material evidence to study commercial practices.149 Indeed, the analysis of different documents concerning wine and viticulture indicated that many things, such as the wine calendar, were part of the expertise of people involved in commerce, as was the writing and understanding of the different tituli written on amphorae, which would have likely been indecipherable to people not involved in these specific professional circles. The latter again illustrates the undeniable need to combine different sources in the study of commerce.