Painted marks at El Mèdol quarry near

Tarragona: observations on the logistics and

organisation of a Roman limestone quarry
Maria Serena Vinci
This paper aims to present a set of painted marks and inscriptions on stone blocks from
El Mèdol quarry (Tarragona), the limestone quarry which supplied the stone used to build
most of the public edifices, including what is known as the Provincial Forum, in the nearby
town of Tarraco, capital of Hispania Citerior. There are 16 marks in red paint and 3 written
in charcoal, all linked to the phases of extracting and shaping the stone.
Painted inscriptions on building materials are rarely preserved, normally being lost
with the shaping of the ashlar or due to weathering. Although quarry marks are often
difficult to interpret and their meaning is not easily accessible to us, they are useful for
approaching aspects of the work organisation that would otherwise remain unknown. The
discovery of El Mèdol inscriptions provides basic information for study of the complex
mechanism that was a building site and the organisation of the work from quarry to com-
pleted monument.
The study of the modus operandi and management of a cantiere edilizio1 presents chal-
lenges caused by a lack of clear information regarding some of the working phases. The
ancient sources and epigraphy often provide insufficient information on the logistics and
management of both quarry and building site activities. This is even more true in the case
of the supply of regular building materials such as limestone or pozzolana, by contrast with
the better-known operations relating to prestige materials such as marble.2 In some cases,
however, quarry and masons’ marks3 can be a useful tool for considering the organisa-
tion of the phases of work in the quarry and/or at the building site. Studying their use and
function has potential for adding to our knowledge as they are a kind of tracking code, a
communications system used by the ancients to record the material and working stages.
Graphic, numerical or alphabetic signs used in construction processes are found in
structures as early as the Archaic Greek period,4 but the best-known cases are the Roman
notae lapicidinarum documented on Imperial marble. We are beginning to understand how
they were used on marble in the different working phases of extraction, accounting, trans-
port, storage, control and trade.5 The marks, or in some cases actual inscriptions labelling

1 For a terminological examination of the English equivalent for “cantiere edilizio”, see J. DeLaine,
“Conclusions,” in S. Camporeale, H. Dessales and A. Pizzo (edd.), Arqueología de la construcción
I (Mérida 2008) 324.
2 From as early as the end of the 1st c. A.D. the imperial quarries in the provinces were characterised
by a complex organisational system; it became fully developed from the Hadrianic period.
3 The definition and terminology for signs or marks on stone is complicated. Bibliography for the
Middle Ages has dealt with this issue more systematically, coming up with “signes lapidaires”:
J.-L. Van Belle, “Les signes lapidaires: essai de terminologie,” in id. (ed.), Actes du Colloque int.
de glyptographie de Saragosse, 1982 (Zaragoza 1983) 29-43; id., Pour comprendre les signes lapidaires
(Brussels 2014).
4 For the Greek world, see an earlier summary in R. Martin, Manuel d’architecture grecque, I (Paris
1965) 225-31.
5 The bibliography is vast. It includes J. C. Fant, Cavum antrum Phrygiae (BAR S482; Oxford 1989);
© Journal of Roman Archaeology 32 (2019)
252 M. S. Vinci

the marble, offer different kinds of information, including consular dating, promoters,
administrative indications about the quarry, an exact point of origin within the quarry, and
the workshop in charge of the extraction or processing of the stone. This procedure was
strictly connected to the marble-quarrying districts and the marble trade within the admin-
istrative system of the imperial patrimonium. Following early studies in this field, the first
hypothesis for an interpretative systematisation of these signs was proposed by J. C. Fant,6
according to whom the function of these marks can be broadly divided into two groups:
marks used by the quarry’s administrative staff, which are normally brief and composed
of a few letters; and marks used to manage operations outside the quarry, for example, at
storage or distribution sites.
Much more complex is the question of marks on non-prestige building materials.
Here the challenge is often due to the difficulty of understanding the legal status of the
quarry and the possible hierarchy of workers. Moreover, since such materials were usu-
ally addressed to a limited range of destinations, quarry marks are often internal messages
for the quarry workers and are thus symbols or codes that are difficult for us to interpret.
The system adopted for their study has often incorrectly associated such marks with those
on imperial marbles,7 creating confusion and hindering advances in this field.8 Despite
this, recent studies are shedding light on how this precious evidence can contribute to
our understanding of the organisational system in quarries that produced regular build-
ing materials. A good example is the set of quarry marks found at the Roman dam at
Muel near Caesaraugusta (Zaragoza).9 The numerous marks on the limestone blocks have

P. Pensabene, Le vie del marmo (Rome 1994); id., “Il fenomeno del marmo nella Roma tardo-
repubblicana e imperiale,” in id. and A. Alvarez y Perez (edd.), Marmi antichi 2 (Rome) 333-91;
J. C. Fant, “Quarrying and stoneworking,” in J. P. Oleson (ed.), The Oxford handbook for engineering
and technology in the classical world (Oxford 2008) 121-35; A. M. Hirt, lmperial mines and quarries
in the Roman world (Oxford 2010); P. Pensabene, “Cave di marmo bianco e pavonazzetto in
Frigia. Sulla produzione e sui dati epigrafici,” Marmora 6 (2010) 71-134; id., “Sigle di cava,
amministrazione imperiale, appalti e commercio,” in Camporeale et al. (supra n.1) 41-57; E. Pari-
beni and S. Segenni (edd.), Notae lapicidinarum dalle cave di Carrara (Pisa 2015).
6 J. C. Fant, “Ideology, gift and trade: a distribution model for the Roman imperial marbles,” in
W. V. Harris (ed.), The inscribed economy (JRA Suppl. 6, 1993) 145-70; id., “The Roman imperial
marble trade: a distribution model,” in R. Francovich (ed.), Archeologia delle attività estrattive e
metallurgiche (Florence 1993) 71-96. The mediaeval bibliography treats this subject in a more
specific and systematic way: see Van Belle 1983 (supra n.3) 29-43; N. Reveyron, “Marques
lapidaires: the state of the question,” Gesta 62 (2003) 161-69; Van Belle 2014 (supra n.3) 41-51.
7 See, e.g., the question about El Mèdol, where the hypothesis of imperial management was
justified in particular by the discovery of the mark CAS interpreted as CA(e)S(aris), testifying
to work organised by the provincial government: R. Mar and P. Pensabene. “Finanziamento
dell’edilizia pubblica e calcolo dei costi dei materiali lapidei: il caso del foro superiore di
Tarraco,” in Camporeale et al. (supra n.1) 512-13. Considering the context of the discovery and
the group of some 70 marks in which they were found, among which none refers to a possible
imperial management, this interpretation is certainly questionable. Moreover, regarding the
term caesaris, no reference to workers or tasks linked to the government appears at Tarraco.
8 The legal context and the modalities of quarry management also change in relation to their
importance; they vary according to the number of workers employed, the value of the quarried
stone, and the economy that the whole productive system may have generated. A quarry for
non-prestige stone cannot be considered equal to a quarry for marble: see G. Poma, “Aspetti
giuridici e amministrativi nella gestione delle cave,” in C. Guarnieri (ed.), Il vetro di pietra: il lapis
specularis nel mondo romano dall’estrazione all’uso (Faenza 2014) 37-38.
9 P. Uribe et al., “Nuevas marcas epigráficas procedentes de la presa romana del Muel (Zaragoza),”
 Painted marks at El Mèdol quarry near Tarragona 253

allowed references to different loci at the extraction site to be identified. In some cases they
are associated with the letter P, for which a reading of p(ublico) has been proposed. This
abbreviation suggests that some sectors of the quarry would have been public property,
probably belonging to the colonia of Caesaraugusta. In the same category is the interesting
group of charcoal and minium inscriptions from the Conero quarry near Ancona10 (see
further below).
Within this context, the aim of this paper is to present a group of unpublished quarry
inscriptions found near the main limestone quarry that supplied the stone for Colonia Iulia
Urbis Triumphalis Tarraco. The town, capital of Hispania Citerior from Augustan times, in
the Early Imperial period underwent an intensive building programme with a widespread
use of limestone, culminating in construction of the so-called Provincial Forum. It is to this
complex that the organisational activity described by the paint and charcoal inscriptions
examined here can probably be linked.

El Mèdol quarry and the construction of Roman Tarraco

From the very first phases of its monumentalisation, public building at Tarragona
involved the intensive exploitation of local resources thanks to a geology that provided
good-quality building stone. In fact, almost all its monumental public edifices were built
of limestone quarried from sites in the immediate vicinity. The main extraction site was El
Mèdol quarry, located some 9 km northeast of the town11 (fig. 1). It provided a Miocene
bioclastic limestone12 characterised by a yellow-golden colour and a high content of bio-
clasts (especially shells) and mollusc fragments, which are visible even to the naked eye.13
It is quite easy to extract and work, two characteristics that make it excellent for building.
The preservation of this opencast quarry is exceptional; evidence of ancient quarrying,
including tool marks and the negative imprint of blocks extracted, is still clearly visible
(see fig. 12a below). Until a few years ago,14 we had little data regarding the exploitation
of the quarry in antiquity or information about the operation and organisation of the work
in some of its sectors. Six extraction points have been identified to date, of which the main
one is “El Clot” (fig. 2). It was in this sector that an access ramp was identified during the
most recent excavation campaign;15 it is built of stoneworking waste and linked to a con-
tainment wall.16 Also documented was a zone interpreted as a roofed control area (4 x 3 m)

Saldvie 16 (2016) 209-45; M. Navarro Caballero et al., “La presa romana de Muel (Zaragoza,
España): ¿una obra militar?,” in F. Cadiou and M. Navarro Caballero (edd.), La guerre et ses
traces (Bordeaux 2014) 573-98.
10 G. Paci, “Le iscrizioni della cava romana del Conero,” in id. (ed.), Contributi all’epigrafia di età
augustea. XIIIe rencontre franco-italienne (Tivoli 2007) 217-46.
11 El Mèdol quarry produces two types of stone: soldó and El Mèdol. They differ in the amount of
bioclastic components, although their appearance, general porosity and compactness can also
vary significantly: see A Gutiérrez García-M. et al., “Can a fire broaden our understanding of
a Roman quarry? the case of El Mèdol (Tarragona, Spain),” in P. Pensabene and E. Gasparini
(edd.), ASMOSIA X (Rome 2015) 779.
12 A. Gutiérrez García-M., Roman quarries in the northeast of Hispania (modern Catalonia) (Tarragona
2009) 108.
13 Ibid. 105.
14 Important data were obtained in 2010 after a fire burned off the vegetation to reveal unknown
evidence of the quarry’s Roman exploitation: Gutiérrez García-M. et al. (supra n.11) 779-89.
15 The excavation in 2013 was directed by A. Gutiérrez García-M. and J. López Vilar.
16 After exhausting the SE area of the “El Clot” sector, it was filled with the débris of stoneworking.
 254
M. S. Vinci

Fig. 2. El Mèdol quarry: the quarry sectors identified (López Vilar and Gutiérrez García [infra n.17] 178).
 Painted marks at El Mèdol quarry near Tarragona 255

Fig. 1. Location of El Mèdol quarry (Gutiérrez García [supra n.12] 113).

built up against the rock wall17 and associated with the ramp. What small finds there were
in the upper layers were dated to the mid-1st c. A.D.18 The best-known image from the
entire quarry is the “Agulla” (Catalan: “needle”) intentionally left standing in the “El Clot”
sector when all the rock around it was extracted19 (fig. 3). It is 20 m high, which makes it
possible to reconstruct the original level of the rock before quarrying began. Excavations
carried out at the base of this pillar revealed traces of charcoal, whose 14C analysis date it to
the Augustan period, highlighting how an extended area will already have been exploited,
perhaps in the Late Republican period.20
The first monumental phase of Roman Tarraco dates back to the Republican period, but
almost no archaeological evidence linked to this period has been found in the “Part Alta”
(the main historical quarter).21 From as early as the 2nd c. B.C., the colony was developing

It is estimated that about 500 m3 of material was required for the construction of the ramp:
J. López Vilar and A. Gutiérrez García-M., Memòria de l’excavació arqueológica i del seguiment de les
obres de rehabilitació de la pedrera del Mèdol (Tarragona 2014) 14.
17 Traces of a posthole were documented in this area, as well as 7 rectangular holes carved in the
rock wall that were probably used to slot in beams to support a roof of perishable material:
J. López Vilar and A. Gutiérrez García-M., “Intervencions arqueològiques a la Pedrera del Mèdol
(Tarragona),” Tribuna d’Arqueologia 2013-14 (2016) 185.
18 Ibid. 184-85. There were two coins: an illegible bronze coin and a denarius from the mint of
Lugdunum bearing the bust of Tiberius on the obverse and Livia seated as Pax on the reverse
(RIC I, 30). The latter can be dated to A.D. 36/37 and provides a terminus post quem for the
layer. Sherds of thin-walled pottery (Mayet XIX and XXXVII1c; López Mullor LIV) provide an
approximate Julio-Claudian date, which would be in keeping with the only amphora fragment
found, a Tarraconensis Dressel 2/4.
19 There are other examples, the closest being at the quarry of Glanum: J.-Cl. Bessac, “Le travail
de la pierre à Glanum,” in Pierres en Provence (Aix-en-Provence 1987) 79-92; id. and N. Lambert,
“La pierre à Glanum,” Les dossiers d’archeologie 140 (juillet-aôut 1989) 8-13.
20 López Vilar and Gutiérrez García-M. (supra n.17) 188-91.
21 The site’s topography and contours had a clear impact right from the first Roman contact,
256 M. S. Vinci

Fig. 3. El Mèdol quarry: the “Agulla” (needle) standing in the “El Clot” sector (author).

Fig. 4. Tarragona’s setting in the Imperial era (J. M. Macias et al., Planimetria arqueològica de Tarraco
[Tarragona 2007] 29).
 Painted marks at El Mèdol quarry near Tarragona 257

around a twin monumental centre adapted to the natural contours: a highly strategic hill
rising to c.80 m asl (the “Part Alta”) and a flat area reaching down to the coast (the “Part
Baixa”) (fig. 4). The first major exploitation of El Mèdol quarry could have been during
this period when two major projects were undertaken. Between 150 and 125 B.C., the town
walls were extended to enclose both “Parts” as far as the port22 and the civic complex now
known as the Colonial Forum was built.23 It was under Augustus24 that a remarkable period
of building began that continued throughout the 1st c. A.D. The town reached its peak

during the Second Punic War (218 B.C.), when troops led by Scipio established a praesidium
on the summit of Tarraco hill: see J. M. Macias and I. Rodà, “Tarraco, the first capital,” Catalan
Historical Review 8 (2015) 9-28. However, in this area there is little evidence of Republican-era
works. Some traces of works to level out the bedrock have been noted on the upper part of the
hill, but it is not yet possible to put forward any hypothesis on the shape or plan of the town.
22 The first phase of the wall may have been linked to the town’s administrative status changing to
that of provincial capital in 197 B.C. or to Cato’s repressive campaigns in 195. See R. Mar et al.,
Tarraco. Arquitectura y urbanismo de una capital provincial romana, vol. II (Tarragona 2015) 51-52;
J. Menchón, La muralla romana de Tarragona: una aproximació (Barcelona 2009) 48-49. The first
construction phase only has a terminus ante quem from pottery associated with the next building
phase. For an interpretation of the building phases of the walls, see Th. Hauschild, “Torre de
Minerva (San Magín). Ein Turm der römischen Stadtmauer von Tarragona”, MadMitt 16 (1975)
246-62 id., “Die römische Stadtmauer von Tarragona,” MadMitt 20 (1979) 246-62; Arquitectura
romana de Tarragona (Tarragona 1983); “Die römische Tore des 2. Jhs. v.Chr. in der Stadtmauer
von Tarragona,” in T. Schattner and F. Valdes (edd.), Stadttore. Bautyp und Kunstform (Mainz
2006) 153-72.
23 X. Aquilué, “Arquitectura oficial,” in X. Dupré (ed.), Las capitales provinciales de Hispania.
Tarragona, Colonia Iulia Urbs Triumphalis Tarraco (Tarragona 2004) 42-46; R. Mar, J. Ruiz de Arbulo
and D. Vivó, “El foro de la colonia de Tarraco entre la República y el Imperio,” in R. Gon-
zález Villaescusa and J. Ruiz de Arbulo (edd.), Simulacra Romae II (Reims 2010) 39-70. The forum
of the colonia started off, in its Republican construction phase, with a peripteros sine postico
Capitoline temple with three cellas on a small podium. It went through different phases after
which, between 50 and 25 B.C., it was transformed into a peripteros hexastyle temple set on a
high opus quadratum podium. Probably in the Hadrianic period it was transformed again, this
time into a mighty temple with a tripartite plan and without side porticoes: R. Mar et al., Tarraco.
Arquitectura y urbanismo de una capital provincial romana, vol. I (Tarragona 2012) 162-80.
24 At this time major modifications were made to the town’s lower part: the old Republican forum,
as well as the area next to the waterfront, was extensively altered; the theatre was built, and
a new urban layout was established. The road network was re-organised, along with other
parts of the territorial infrastructure. See, e.g., J. M. Gurt and I. Rodà, “El pont del Diable. El
monumento romano dentro de la política territorial augustea,” AEspArq 78 (2005) 147-65. The
urban and architectural evolution of the town’s upper part during this period is not as well
known, but the presence in Tarraco of Augustus between 26 and 25 B.C. (Quint., Inst. 6.3.77)
makes it likely that there was at least one early construction project on the hill; see M. S. Vinci
and A. Ottati, “La monumentalizzazione delle Hispaniae: alcune riflessioni su progettualità
e realizzazione del Foro Provinciale di Tarraco,” in M. Livadiotti et al. (edd.), Theatroeideis.
L’immagine della città, la città delle immagini, vol. II (Thiasos Monografie 11; Rome 2018) 169-82.
Beyond the presence of the altar dedicated to Augustus while the emperor was still alive (Mar
et al. 2012 [supra n.23] 345-48; cf. RPC 1, no. 218 for a representation of the Tiberian period),
two branches of one of two aqueducts, both built in Augustan times, have been documented:
I. Mesas, “Los acueductos romanos de Tarraco: cronología y nuevos tramos,” in J. López Vilar
(ed.), Tarraco biennal. Actes del 2r Congrés int. d’Arqueologia i Món Antic. August i les províncies
occidentals (Tarragona 2015) 249; see J. López and D. Gorostidi, “Aqva Avgvsta a Tarraco?,” ibid.
253, for the inscription A[QV]AM [AVGVS]TA[M] documented on a block. The work on a water
system suggests the existence of a project to make the upper part into a monumental area.
258 M. S. Vinci

Fig. 5. Tarragona, Provincial Forum: section and plan (author).

with the massive exploitation of El Mèdol quarry25 and

construction of the Provincial Forum, which would
completely change the architectural character of the
provincial capital of Hispania Citerior26 (fig. 5).

Blocks with painted quarry marks: contextualisation

and extraction
The use of painted marks as a tool for the organisa-
tional and administrative system of the extraction and
construction activities of a building site (“cantiere”) is
well known. Their use must have been widespread,
but their ephemeral nature, both with regards to their immediate function and how they
were written, means that finds are extremely rare. Where painted traces survive, they are
often very poorly preserved, especially on building materials. The painted marks from
Tarraco owe their survival to the particular conditions in which they are found. They were
discovered in a mound of blocks that had been piled up in ancient times, a circumstance
that ensured their preservation and limited the degree of deterioration. This accumulation
of blocks was found in an area right opposite the entrance to El Mèdol quarry, outside
the actual extraction site. Today it is virtually impossible to reconstruct the ancient land-
scape as the area has been much altered by modern road construction27 (fig. 6). It is not

25 Gutiérrez García-M. (supra n.12) 151-52. The archaeological data point to a significant decline in
quarrying from the 3rd c. on.
26 Two terraces were built to create an area divided into three enormous spaces: the Circus on
the lowest terrace; the “plaza de representación”, a wide intermediate terrace with political
and administrative functions for the province; and, at the top of the hill, the “recinto de culto”,
the religious area with the temple of Augustus enclosed by a portico which acted as a temenos.
In spite of the debate on the exact chronology, a series of changes during construction of the
Provincial Forum between the Julio-Claudian and Vespasianic periods has been identified:
J. Sánchez Real, “Exploración arqueológica en el jardín de la Catedral de Tarragona,” MadMitt
10 (1969) 281; M. S. Vinci et al., “El subsuelo de la Torre del Pretorio: substructiones de tradición
helenística bajo la sede del Concilium Provinciae Hispaniae Citerioris (Tarraco),” Arqueología
de la Arquitectura 11 (2014), at http://arqarqt.revistas.csic.es/index.php/arqarqt/article/
viewArticle/165/187); J. López Vilar and L. Piñol, Terracotes arquitectòniques romanes. Les troballes
de la Plaça de la Font (Tarragona) (Tarragona 2008). For epigraphic data, see G. Alföldy, Flamines
provinciae Hispaniae Citerioris (Anejos de ArchEspArq 6, 1973). For architectural decoration, see
P. Pensabene, “La decorazione architettonica dei monumenti provinciali di Tarraco,” in R. Mar
(ed.), Els monuments provincials de Tàrraco. Noves aportacions al seu coneixement (Tarragona 1993);
id. and R. Mar, “Dos frisos marmóreos en el acrópolis de Tarraco, el templo de Augusto y el
complejo provincial de culto imperial,” in J. Ruiz de Arbulo (ed.), Simulacra Romae. Roma y las
capitales provinciales del Occidente europeo (Tarragona 2004) 73-88; P. Pensabene and R. Mar, “Il
tempio di Augusto a Tarraco: gigantismo e marmo lunense nei luoghi di culto imperiale in
Hispania e Gallia,” ArchCl 61 n. 11 (2010) 258-59.
27 These are the AP-7 motorway and the A-7 dual carriageway. The N-340 main road, which
follows the route of the Via Augusta, is also near the site.
 Painted marks at El Mèdol quarry near Tarragona 259

Fig. 6. El Mèdol quarry, showing the current position of most of the blocks coming from the mound.
possible to establish whether
this was part of the work area
of the Roman quarry. It was
during roadworks that the
accumulation of blocks was
discovered and partially doc-
umented.28 A mound of some
6,000 blocks rose to a height
of c.10 m above the bedrock
(fig. 7). A total of 77 quarry
marks and inscriptions have
been found on some of the
blocks: 58 are engraved, while
19 are written (16 of these are
painted in red, 3 traced in
charcoal).
Unfortunately, virtually
no archaeological finds were
made among the blocks, but
some fragments of N Afri-
can terra sigillata and a coin
of the Tiberian period29 were
documented, although they
cannot be linked to a particu-
lar stratum due to the absence
of an excavation report. The
mound does not appear to Fig. 7. (a) accumulation of blocks at El Mèdol during the excavations of
2008-9 (CODEX); (b) current situation of part of the blocks (author).

28 The excavations were carried out between 2007 and 2009. Unfortunately, the changeover from
one archaeology cooperative to another resulted in the loss of a large amount of information,
and the report was never deposited.
29 Another denarius from the time of Tiberius was found during excavation of the access ramp to
the “El Clot” sector.
260 M. S. Vinci

have been created all at the

same time, since at least three
different layers of compacting
were identified in which frag-
ments and stone splinters were
distinguished.30
The blocks are not wholly
in the parallelepiped form in
which they were extracted from
the quarry,31 but are broken
(fig. 8). This implies they were
discarded or abandoned, prob-
ably because they were faulty.
This hypothesis is supported Fig. 8. Some broken
blocks with red paint
by the fact that some blocks
(author).
have natural fractures, a clue

Fig. 9. Blocks with natural fractures (author).

Fig. 10a. Blocks abandoned at El Mèdol during extraction (author); fig. 10b. evidence of a groove made by a
pick (author).

30 J. F. Roig Pérez et al., “El dipòsit de carreus del Mèdol (Tarragona). Resultats preliminars,”
Tribuna d’Arqueologia 2009-10 (2011) 403.
31 The standard block shapes and sizes (1.6 x 0.8 x 0.7 m) can be deduced from the negative traces
still visible on the quarry faces.
 Painted marks at El Mèdol quarry near Tarragona 261

that corroborates the

idea that this was the
reason for their rejec-
tion (fig. 9). The blocks
with painted inscrip-
tions are squared and
have regular faces as
a result of the process
of detachment from
the rockface using Fig. 10c. A stonecutter opening trenches (J.-Cl. Bessac, La pierre en Gaule narbon-
trenches and picks. naise et les carrières du Bois des Lens (Nîmes) (JRA Suppl. 16, 1996) 211, figs. 131-32).
Usually at least one
of the faces is very
irregular and should
correspond to the
detachment surface.
The blocks were quar-
ried from vertical
walls (fig. 12a) with
the ensuing forma-
tion of terraces that
are still visible where
the working face was Fig. 11. El Mèdol quarry: negative traces of rectangular wedge sockets (author).
abandoned before it had become exhausted.32 The demarcation of each block by opening
2-3 trenches is attested both by the evidence at the quarry and by grooves on the blocks
from the mound33 (fig. 10). These grooves were made by a pick when the trenches for
detaching the upper piece from the quarry front were defined prior to its extraction. They
would have been removed once the roughing-out process had started. This evidence on
the blocks examined confirms that this phase of the working had not yet been completed
or perhaps had not even started. The definitive detachment from the quarry face would
have been achieved by inserting rectangular (10-14 cm) wedge sockets, traces of which can
be documented (fig. 11, no. 12).
Unlike those just described , an initial phase of work is attested on a few blocks:
(a) some completely irregular faces34 have one or two of the edges worked with a flat
chisel. Once the chiselling of all the edges had been finished, the piece would have
been left bossaged, or the quadratarius would have eliminated the excess material and
smoothed the surface35 (fig. 12c = no. 6);

32 Gutiérrez García-M. (supra n.12) 153.

33 The trenches are square in section and 10-13 cm wide. On the blocks examined, the thickness of
the point of the tool is 2 cm.
34 Each block would have presented at least two irregular faces: a lower one resulting from the
detachment of the block from the bedrock, and an upper one resulting from the detachment of
the previously extracted block above it.
35 One can see on the upper terrace of the Provincial Forum how the square perimeter is built
with blocks with a strongly emphasised bossage (an excess of material that has not been
eliminated): M. S. Vinci, El Foro Provincial de Tarraco: documentación y análisis de técnicas y procesos
de construcción (Ph.D. diss., Univ. Rovira i Virgili, Tarragona 2014) 268 and 273.
262 M. S. Vinci

Fig. 12. El Mèdol quarry: (a) evidence of tool marks and

the negative imprint of extracted blocks on the quarry face;
(b) block with flat chisel marks;
(c) block with one edge worked with a flat chisel.

(b) two blocks, in addition to being the only ones with two-line inscriptions (nos. 15-16),
are well squared and, unlike the previous type, have flat chisel-marks on the whole face
with the painted inscriptions. The distribution of the chisel-marks allows us to recon-
struct the movements made by the stonecutter with this tool (fig. 12b = no. 16).
In line with the above, the area selected to hold these blocks would have lain outside
the extraction area. These blocks for various reasons did not meet the optimal conditions
for sending to the construction site, and were abandoned. The rejection phase may have
occurred immediately after the blocks had been extracted or during the early stages of
their working, after the quadratarius had begun the roughing-out phase but then realised
that there were structural problems. This area would also have been a kind of initial pro-
cessing area, a hypothesis suggested by blocks with the shaping process barely started, as
well as by stone fragments and splinters from the working found within the whole deposit.

Catalogue
All the pieces came from a rescue excavation carried out between 2007 and 2009.
The references to the sides of the blocks (front, back, upper, lower, etc.) refer to their current posi-
tions after it was decided to call the side with the inscription the front.
1 (fig. 13 in colour). Block of bioclastic limestone, h. 0.56 / w. 0.69 / th. 0.70. Now in Catalan Gov-
ernment store in Tarragona. Parallelepiped block, partially preserved. Two of its faces have a fairly
regular surface on which traces of pick marks can be made out. On the front, on which the inscription
is written, there are pick marks in different directions, the result of the stonecutter taking different
positions as he worked. The other preserved faces have entirely irregular surfaces. One of the block’s
lower corners is worked in a kind of double step. What is now the rear is the breakage surface.
Inscription in charcoal, poorly preserved, Latin capital letters 0.20-0.22 m in size.
V K(alendas) [---?]
The inscription is written on a single line and is difficult to read. The suggestion for the letter K after
the numeral V is mainly derived from the context of the other inscriptions found at the site. After the
letters V and K, more written in charcoal are still visible, so the presence of further letters is assumed.
 Painted marks at El Mèdol quarry near Tarragona 263

2 (fig. 14 in colour). Block of bioclastic limestone, h. 0.73 / w. 0.87 / th. 1.35. Now in Catalan Govern-
ment store in Tarragona. Parallelepiped block, partially preserved. Two of its faces (front and upper)
have fairly regular surfaces on which traces of pick marks can be made out. The front side also bears
marks of at least 3 blows from a tool with a rectangular tip (the marks have a width of c.0.02 m). The
right side is preserved and has a completely irregular surface. The remaining sides are the breakage
surfaces.
Inscription in charcoal, poorly preserved. Latin capital letters 0.11 / 0.28 m in size.
XIIII K(alendas) [---?]
The inscription is on a single line and is composed of at least 6 letters, of which the first is an X and is
notably larger than the others. The last recognisable letter is a K, on which the two legs are separated
from the vertical stroke and the lower leg is shorter than the upper one. The presence of further let-
ters cannot be completely ruled out.
3. Block of bioclastic limestone, h. 0.79 / w. 0.73 / th. 0.80. Now in Catalan Government store in Tarra-
gona. Parallelepiped block, partially preserved. Two of its faces (upper and front) have fairly regular
surfaces with traces of pick marks. On the front side with the painted inscription the pick marks are
oblique and run parallel to one another. The other sides are completely irregular. The back is the
breakage surface.
Inscription in charcoal, fair preservation. Latin capital letters, 0.16 / 0.19 m in size.
IIIX Ḳ(alendas) [---?]
The inscription is on a single line and difficult to read. At least 5 letters can be identified. The numeral
IIIX can be seen, although of the first character only a short part of the upper vertical stroke remains.
The number seven is expressed by subtracting three from the numeral ten. K for kalendas is proposed
for the last letter. Only the lower part of the vertical stroke and the lower leg are preserved. The pres-
ence of further letters cannot be completely ruled out.
4. Block of bioclastic limestone, h. 0.70 / w. 0.92 / th. 0.80. Now in Catalan Government store in Tar-
ragona. Parallelepiped block, partially preserved. Two of its faces (of which one is the front) have
fairly regular surfaces and pick marks can be made out. On the front side with the inscription the
pick marks are oblique and run parallel to one another; a fracture runs vertically across the whole
face. The upper face is irregular and slightly concave, which suggests that it may be the surface that
was detached from the bedrock. The rear of the block is the breakage surface. The remaining sides
are completely irregular.
Inscription painted in red, fair preservation. Latin capital letters, 0.04 / 0.07 m in size.
IIX K(alendas) [---?]
The inscription is on a single line. It is difficult to read as the state of conservation of the text is only
fair. At least four letters can be identified. The number eight is expressed by subtracting two from the
numeral ten. After the numeral, the letter K can be identified. It has the two legs that start from the
vertical stroke with the lower leg finishing with a small trait towards the top. The presence of further
letters cannot be completely ruled out.
5. Block of bioclastic limestone, h. 0.80 / w. 0.89 / th. 1.67/. Now in National Archaeological Museum
of Tarragona store. Large parallelepiped block, mostly preserved. Pick marks can be made out on
almost all its sides, which have regular surfaces except for the upper face, which has an irregular,
concave surface, probably because it is the surface that was detached from the bedrock. The rear of
the block is the breakage surface; a natural fracture crosses the whole block obliquely. The right side
also has a fracture crossing the whole block vertically.
Inscription painted in red, poor preservation. Latin capital letters, 0.08 m in size.
V [---]
The inscription is written on a single line. In its current state of conservation only one letter has been
identified, but there are traces of red paint so there would certainly have been other letters.
6. Block of bioclastic limestone, h. 0.85 / w. 0.78 / th. 1.00, now in Catalan Government store in Tar-
ragona. Parallelepiped block, partially preserved in an early working phase. The surface with the
inscription is regular; on its lower part there is a carved moulding that was still being worked on. On
the same surface, corresponding to the right edge, there are two grooves (0.03 m wide) that could be
the marks of the pick used to open the trench to extract the overlaid block or to facilitate its detach-
ment, the tool being used as a sort of lever. The block’s right-side surface is irregular, but the upper
264 M. S. Vinci

Fig. 13. Inscription no.

1 in charcoal (author).

Fig. 14. Inscription no. 2 in

charcoal (author).

Fig. 15. Inscription no. 8 in red paint (author).

Fig. 16. Inscription no.

9 in red paint (author).

Fig. 17.
Inscription
no. 10 in
red paint
(author).

Fig. 18. Inscription no. 11 in red paint (author).

 Painted marks at El Mèdol quarry near Tarragona 265

edge is chiselled, probably because it was being bossaged (fig. 12c). The left side was probably the
breakage surface.
Inscription painted in red, poor preservation. Latin capital letters, 0.05 / 0.07 m in size.
Reading 1: Reading 2:
IIIX [---?] III K(alendas) [---?]
I [---] I [---]
The inscription is on 2 lines. On the first, there are three vertical strokes and a letter that could be an
X or a K. On the second, a single vertical stroke can be made out, although the traces of red paint
suggest there were once other letters.
7. Block of bioclastic limestone, h. 0.82 / w. 0.89 / th. 0.76, now in Catalan Government store in
Tarragona. Parallelepiped block, partially preserved. Block is squared. On the front, which has the
inscription, pick marks (between 0.05 and 0.07 m in length) in different directions can be made out.
Clearly preserved on the upper side is the groove left by the pick for opening the trench to extract
the overlaid block (fig. 10b).
Inscription painted in red, poor preservation. Latin capital letters, 0.04 / 0.06 m in size.
II [---?] Ḳ [---?]
The inscription is on a single line. At least 2 vertical strokes and 1 alphabetic character can be iden-
tified. The latter is probably a K (of which only the lower part of the vertical stroke and the leg are
preserved). The space between this letter and the numeral suggests there was another letter that is
no longer preserved.
8 (fig. 15 in colour). Block of bioclastic limestone, measurements unknown, current location unknown.
Parallelepiped block, almost completely preserved. The surface with the inscription is regular and
bears pick marks (only the lower right-hand corner has been broken off). The other sides are also
regular and smooth and show pick marks. The top, the only side that is not regular, is probably
the detachment surface, but on this same side an early working phase to smooth the surface can be
detected on one of the edges.
Inscription painted in red, good preservation. Latin capital letters, measurements not available.
IX K(alendas) [---?]
The block has been studied from photographs and excavation notes (cf. Roig Pérez et al. [supra n.30],
no reading proposed there). The inscription is on a single line. Following the numeral IX there is the
letter K written with very short legs that are slightly detached from the vertical stroke. It is not pos-
sible to determine whether there were other letters.
9 (fig. 16 in colour). Block of bioclastic limestone, measurements unknown, current location unknown.
Parallelepiped block, almost completely preserved. Almost all its sides are squared and pick marks
can be made out on them. Only the upper side is completely irregular and this was probably the
detachment surface.
Inscription painted in red, good preservation. Latin capital letters, measurements not available.
IIX ID(us) [---?]
The block was analysed from photographs and excavation notes. The inscription, on a single line,
is composed of at least 5 letters, 1 numeral and 2 alphabetic letters. On the first numeric letter it can
be seen how the brush, which had just been dipped in the paint, was placed on the block to make a
vertical stroke from top to bottom. The last letter, a D, was painted in a single stroke.
10 (fig. 17 in colour). Block of bioclastic limestone, h. 0.61 / w. 0.40 / th. 0.57, now in Catalan Govern-
ment store in Tarragona. Parallelepiped block, only a small part is preserved. The preserved sides
have smooth, even surfaces. On the right edge of the front side, which bears the inscription, concave
traces at regular intervals (width 0.03 m) can be interpreted as the negative marks of wedges used to
detach the block from the bedrock.
Inscription painted in red, good preservation. Latin capital letters, 0.05 / 0.15 m.
IV [---?] ‘APR’(iles)
The inscription is on a single line. Two numerical characters can be documented in the first part of
the inscription, followed by a gap that could be completed by the term kalendas. In the last part of the
inscription there are 3 letters in ligature for which a reading of APR(iles) is proposed. See Roig Pérez
et al. (supra n.30) (no proposed reading).
266 M. S. Vinci

Fig. 19. Inscription no. 12 in red paint (author).

Fig. 20. Inscription no. 13 in red paint (author).

Fig. 21. Inscription no. 14 in red paint (author).

Fig. 22. Inscription no. 15 in red paint (author).

Fig. 23. Inscription no. 16 in red paint (author).

 Painted marks at El Mèdol quarry near Tarragona 267

11 (fig. 18 in colour). Block of bioclastic limestone, h. 0.74 / w. 1.30 / th. 0.80, now in the Catalan
Government store in Tarragona. Parallelepiped block, of which only half is preserved. The front is
concave and has a regular surface. The other sides have irregular surfaces.
Inscription painted in red, fair preservation. Latin capitals letters, 0.10 / 0.14 m in size.
Reading 1: Reading 2:
Θ X ṾII KA(lendas) Θ X ṾII K(alendas) A(priles)
The inscription. written on a single line, is composed of at least 6 characters (4 numerical and 2 alpha-
betic). The characters are not the same height and are not perfectly aligned. What is interpreted as the
numeral X is made in an oblique direction with respect to the other letters. At the start of the inscrip-
tion there is a symbol/character, a kind of Greek theta, whose meaning is not clear. On the letter K the
upper leg is shorter than the lower one. Two possible readings are proposed for the interpretation of
the last two letters: the abbreviation KA for the term kalendas, or the initial letters K and A for kalendas
apriles (in the latter case the date would correspond to March 16). See Roig Pérez et al. (supra n.30)
(no proposed reading).
12 (fig. 19 in colour). Block of bioclastic limestone, h. 0.92 / w. 1.11 / th. 1, now in the Catalan Gov-
ernment store in Tarragona. Parallelepiped block, partially preserved (most of its length is no longer
preserved). The preserved sides have pick marks. Almost all the sides have well-squared regular
surfaces, with the exception of the upper and probably the lower ones. The current upper side is
probably the detachment surface; on the left edge there are 4 rectangular negative traces interpreted
as having been made by wedges used to detach the block from the bedrock (wedge sizes 0.9 m; 0.14
m; 0.13 m; 0.10 m; the distance between the wedges is 0.10 m: fig. 11a). On the left edge of the lower
side a cut in the stone appears to correspond to marks made with a pick (with a point size of 0.02 m).
The cut could be interpreted as a trace of the trench made to detach the overlaid block, which had
previously been extracted. If so, the original position of the block would be upside-down. The right
and back side are the breakage surfaces.
Inscription painted in red, fair preservation. Latin capital letters, 0.09 / 0.12 m in size.
[---?] XVI Ḳ (alendas) · MẠ(ias)
The inscription, on a single line, is composed of at least 6 characters (3 numerical and 3 alpha-
betic). The reading is difficult due to the state of preservation. The first character, whose meaning
is unknown, is a kind of semicircle that appears to surround the inscription almost entirely. After
the numerals and the letter K comes a punctuation mark. Finally, the letter M is followed by a single
oblique stroke which could correspond to the letter A and could represent the abbreviation for Maias.
Hence the interpretation of April 16. See Roig Pérez et al. (supra n.30) (no reading proposed).
13 (fig. 20 in colour). Block of bioclastic limestone, h. 0.86 / w. 1.30 / th. 0.69, now in the Catalan
Government store in Tarragona. Parallelepiped block, only a small part preserved. All the sides are
broken with the exception of the front, which is quite regular but not smooth.
Inscription painted in red, fair preservation. Latin capital letters 0.06 / 0.07 m.
V K(alendas) ‘ẠPṚ’ (iles)
The inscription is on a single line. In its state of preservation it is difficult to propose a reading, espe-
cially for the last part. The first letter has been painted without lifting the brush from the stone, in a
single movement that starts at the top left of the letter. The first brush contact with the stone is clearly
visible from an accumulation of paint at that point. In the digitally processed photographs, a drop of
paint is visible right here where the brush tip came into contact with the block. The letter K is made
with the two legs inserted into the vertical stroke. The interpretation of the last letters is very dif-
ficult. The proposal, which takes into account the other inscriptions, is of 3 joined letters interpreted
as APR(iles).
14 (fig. 21 in colour). Block of bioclastic limestone, h. 0.75 / w. 1.00 / th. 0.80, now in the Catalan Gov-
ernment store in Tarragona. Parallelepiped block, only a small part preserved. The block’s upper
surface is regular; the front with the painted inscription is regular and smooth. The left and rear sides
are the breakage surfaces.
Inscription painted in red, good preservation. Latin capital letters 0.07 / 0.12 m.
PR(idie) NON(as) ‘ẠPR’(iles)
The inscription, on a single line, is composed of at least 8 letters. They are not the same height and
are not perfectly aligned. On the first part of the inscription the letters are particularly close together
268 M. S. Vinci

(P is overlapped by R; the vertical stroke of the first N overlaps the leg of the R). Both Rs have a small
bow. It is possible to observe how the brush had just been dipped into the paint when it came into
contact with the stone (see, e.g., the upper part of the vertical stroke on the letter P, or the first letter
N). The reading of the last part of the inscription is not easy. The proposal is of 3 joined-up letters
interpreted as APR(iles). Only the left stroke of the A is preserved. See Roig Pérez et al. 2010 (no read-
ing proposed).
15 (fig. 22 in colour). Block of bioclastic limestone, h. 0.64 / w. 1.28 / th. 0.78, now in the National
Archaeological Museum of Tarragona’s store. Parallelepiped block, almost entirely preserved. Well-
squared, all the preserved sides are regular with pick marks. The only exception is the surface with
the inscription, where the marks of a flat chisel can be documented (the point of the tool is between
0.03 and 0.04 m). These marks run in different directions following the changing position of the
stonecutter as he worked.
Inscription painted in red, fair preservation. Latin capital letters 0.09 / 0.10 m.
XIII K(alendas) APṚ(iles)
VER(e)CUND(us) or VER(e)CUND(i)
The inscription is on 2 lines with letters that are not the same height and not perfectly aligned. The
first line has been interpreted as a date: after the numeral composed of 4 letters comes a letter K, the
two legs traced with a single wavy trait. The letter R preserves the vertical stroke, part of the leg and
a few traces of paint from the bow. The second line has been interpreted as a cognomen. The letters
are not as well preserved but the horizontal strokes of the letter E are inclined upwards and the letter
R has a well-rounded bow.
16 (fig. 23 in colour). Block of bioclastic limestone, h. 0.66 / w. 0.57 / th. 0.82, now in the National
Archaeological Museum of Tarragona’s store. Parallelepiped block, most of its length is no longer
preserved. Well-squared, all its preserved sides are regular and show pick marks. On the side with
the inscription the marks of a flat chisel can clearly be seen. The tool marks (in the middle they are
c.07 m in size and on the left between 0.02 and 0.04 m in size) run in different directions according to
the positions adopted by the stonecutter during the roughing-out phase.
Inscription painted in red, good preservation. Latin capital letters 0.05 / 0.15 m.
Reading 1: Reading 2:
ṾỊ ID(us) · ‘IAṆ’(uarias) [---?] ṾỊ ID(us) · ‘IAṆ’(uarias) [---?]
L(ucius) · SILVÎNỊ L(ucius) · SILVANỊ
The inscription, on 2 lines, presents the same typological scheme as no. 15. The first 2 letters have
been read as the number VI followed by ID(us). After the letter D there is a punctuation mark. After
that, it reads IAN with the A and the N joined up. The traces of paint suggest there were probably
other letters. On l.2 there is a punctuation mark after the first letter L. The proposed reading of l.2 is
a praenomen and a cognomen. See Roig Pérez et al. (supra n.30) (no reading proposed).
17. Block of bioclastic limestone, h. 0.71 / w. 1.00 / th. 0.72, now in the Catalan Government store in
Tarragona. Parallelepiped block, most of its length is no longer preserved. Well squared, all its pre-
served sides are regular and show pick marks. On the front, with the inscription there are pick and
axe marks which run in different directions according to the positions adopted by the stonecutter as
he worked.
Inscription painted in red, very poor preservation. Latin capital letters 0.05 / 0.08 m.
There are traces of paint but they cannot be linked to any letter.
18. Block of bioclastic limestone, h. 0.73 / w. 0.85 / th. 0.93, now in the National Archaeological
Museum of Tarragona’s store. Parallelepiped block, partially preserved. Its preserved sides, well-
squared with regular surfaces, show pick marks. There are also axe marks on the face with the
inscription. The upper and lower sides, which have entirely irregular surfaces, could be the detach-
ment surfaces from the rock.
Inscription painted in red, very poor preservation.
There are traces of paint but they cannot be linked to any letter.
19. Block of bioclastic limestone, h. 0.71 / w. 1.17 / th. 0.75, now in the National Archaeological
Museum of Tarragona’s store. Parallelepiped block, only a small part of which is preserved. Except
for the front and the back, the sides are completely irregular. On the front, pick marks can be seen, all
 Painted marks at El Mèdol quarry near Tarragona 269

of them running in the same direction. On the lower part of the right side, a rectangular trace (width
0.06 m) could correspond to a wedge used to detach the block from the bedrock. On the top of the
rear side, a natural fracture runs horizontally across the whole piece (fig. 9b).
Inscription painted in red, very poor preservation.
There are traces of paint that cannot be linked to any letter.

Painted quarry marks: description and analysis

Analysis of the red paint established that it is red ochre,36 a mineral pigment made with
iron oxide (also known as sanguine or red ochre, it is a commonly available element) that
can be used dry or diluted in water. The condition of the marks hinders complete readings
as there are many gaps in the painted traces. In order to improve their visibility, the images
of all the inscriptions were processed with imaging software37 that alters the colours and
contrasts to obtain a remarkable improvement in their readability. This allows us to note
the constant presence of capital letters, often of variable sizes, in the same inscription and
of writing that is never horizontal. The ductus is thus irregular, written quickly without
any special care. This reflects the ephemeral nature of the texts, which are all brief except
for two that are written on two lines. As they are functional inscriptions referring to the
extraction/construction processes, they lack any official form. Moreover, they were prob-
ably written by poorly-educated or semi-illiterate people.
From an initial analysis, it was possible to identify certain common elements: the pres-
ence of the letter K and, in two cases, the binomial ID. These are interpreted as dates in the
Roman calendar, counting the days until the next monthly festivity (K = kalendae, NON =
nonae, ID = idus). The kalendae were established on the first of the month, the nonae on the
fifth except for March, May, July and October when it was the seventh day. In those same
months, the idus were on the 15th instead of the 13th.38 This means that some dates are only
possible in the months mentioned above.
It was not possible to propose a reading for a few inscriptions that are visible only as
traces of red paint (nos. 17-19), and of the others some do not include or preserve mention
of the month. Most of this group of red-painted inscriptions contain reference to the kalen-
dae (nos. 4-8) of an indefinite month; this is the case with no. 8 (fig. 15), where the ninth
before the kalendae of a month is indicated.39 The same is true for the three inscriptions writ-
ten in charcoal (nos. 1-3; figs. 13-14) where the numeral is always associated with the letter
K. In only one of the red-painted inscriptions in which the indication of the month does not
appear is the numeral not followed by the reference to the kalendae, but to the idus (no. 9; fig.
16). For the remaining inscriptions (nos. 10-16), it was possible to propose a reading of the
month and thus an exact day, even though the year is lacking. Although there is frequent
reference to the month of March, the lack of any mention of the year makes it impossible
to establish a chronological sequence (fig. 24); nor do we have enough data to arrive at

36 The painting was analysed using Fibre Optic Reflectance Spectroscopy (FORS) and a scanning
electron microscope coupled with energy-dispersive X-rays. I wish to thank Aurélie Mounier
(IRAMAT-CRP2A, Michel de Montaigne Université Bordeaux) who carried out the analysis.
37 DStrecht software, an extremely useful tool for bringing out faint pictographs that are otherwise
invisible to the human eye.
38 A. Invernizzi, Il calendario (Rome 1994).
39 This could correspond to the 24th in a month of 31 days (Ianuarius, Martius, Maius, Quintilis,
Iulius, Augustus, October), the 23rd in the case of a 30-day month (Aprilis, Iunius, September,
November), or the 21st in February.
270 M. S. Vinci

an absolute chronology.
The difficulty in reading
some of the characters
still leaves some doubts
about their interpreta-
tion, as is the case of no.
11 (fig. 18), correspond-
ing to the date March
16, in which it was not
possible to identify the
meaning of the symbol
or character (a kind of
theta) at the start of the
text. One hypothesis is
that it could simply be
a symbol indicating a
particular operation or
check of the block. The
same could be said for
inscription no. 12 (fig.
19), which includes the
date April 16 and where
the first character seems
to be a sort of oval that
encloses the painted
text.
Only one of the in-
scriptions includes a ref-
erence to nonae (no. 14;
fig. 21), the pridie nonas
of April (= April 4).
This is one of the best-
preserved inscriptions
Fig. 24. Painted dates from El Mèdol.
in which it is possible to
observe the strokes of the paint brush, which was quite thick (2 cm). On the first letter one
can see clearly how the brush, just dipped in the paint, goes from top to bottom, executing
the vertical stroke of a P, then the R, in three different traits (vertical stroke, bow and leg),
overlapping the bow of the P. In no. 13 (fig. 20) the tip of the brush, just removed from the
paint, is also distinguishable as it releases drops of excess paint on the block.
Mention should be made of the two red-painted inscriptions on two lines. The state
of preservation and the fact that they appear to have been written quickly, without much
care, make the reading difficult. Both seem to present the same pattern: on the first line a
calendar date, on the second a cognomen. In no. 15 (fig. 22), after the reference to the date
of March 20, the reading Verecundus/i40 can be proposed on l.2. Inscription no. 16 (fig. 23)

40 In The Latin cognomina (Rome 1982) 68, I. Kajanto mentions it among cognomina that refer to
mental abilities evoking ideas of modesty, humility and discretion (among which Verecundus
 Painted marks at El Mèdol quarry near Tarragona 271

also has a date on the first line, in this case a reference to 6 days before the idus of January
( = January 8). Judging by a few preserved traces, it seems conceivable that there were once
other characters. A reading of L. Silvini or L. Silvani41 is proposed for the second line.

Interpretation and comparisons

These are inscriptions made in haste in a common, informal script with no attention
paid to their epigraphic aspects. The hypothesis proposed for the interpretation of the
painted inscriptions is that of a system used to note the date of the stonecutter’s working
day and the amount of stone extracted, calculating the correct remuneration. In this way
the stonecutter (or, more likely, the supervisor responsible for each team of quarrymen/
quadratarii) would have recorded the work done by the end of a specific period, presum-
ably the end of each day or week.42 The omission of the month in some cases may have
been to avoid giving superfluous information, as the corresponding pay would have been
credited with reference to the current month, although we cannot completely rule out the
possibility that the original indication of the month has simply not been preserved.
It is also possible to link the quarry mark to the extraction activity and not the process-
ing phase, as the absence or very initial working phase of the blocks confirms. In fact, the
roughing-out appears not yet to have started or in some cases is only in its initial stages.
Following the detachment of the block or its initial working phase, the stonecutter may
have encountered natural fractures in the stone, which then would have caused the block
to break. Such a scenario would not affect the interpretation of the inscriptions, as the
workers would still have wanted to keep track of all the stone and thus of the work they
had carried out. Each of the marked blocks would thus represent a group of blocks, which
would avoid having to report each of the elements extracted.
We also cannot rule out that the inscriptions were made on the quarry walls at the end
of the extraction activity. In this case the marks would refer to the material that had pre-
viously been extracted. In this regard, a cogent comparison are the painted inscriptions
from Docimium where on one of the quarry walls or extraction faces there is a sequence
of painted dates, based on which M. Bruno hypothesised a chronology of the extractive
phases.43
Despite the scarcity of painted inscriptions on building materials, available compari-
sons help us to advance an hypothesis for this site. First there are the Baths of Trajan on
Rome’s Oppian Hill. On the walls of the underground tunnel that runs along one of the
edges of the supporting terracing a veritable “building journal” was found. It consists of

is the most used). This cognomen appears 387 times in the CIL: for 261 men, of whom 26 are
freedmen or slaves, and for 87 women, of whom 13 are freedmen and slaves (ibid. 264). For
the Iberian peninsula, J. M. Abascal Palazón, Los nombres personales en las inscripciones latinas de
Hispania (Murcia 1994) 542) reviews 19 cases of Verecundus/a and Verecundinus/a, of which three
are found at Tarragona (CIL II 4161, 4175 and 4209).
41 Or Silvanianus: cf. Abascal ibid. 512; Kajanto ibid. 213-14; B. Lőrincz, Onomasticon Provinciarum
Europae Latinarum (Vienna 2002).
42 Also in the Mediaeval period it has been noted how the methods of paying workers could vary
and be either daily or weekly: R. Dionigi, “I segni dei lapicidi: evidenze europee,” in I magistri
commacini: mito e realtà del Medioevo lombardo. XIX Congresso (Spoleto 2009) 375.
43 M. Bruno, “Tituli picti su due fronti di cava nel distretto di Bacakale a Docimium (Iscehisar,
Afyonkarahisar),” JRA 30 (2017) 469-89.
272 M. S. Vinci

Fig. 25. (a) Painted inscriptions from the Baths of Trajan (Volpe [infra
n.44] 384); (b) quarry marks at Conero, Ancona (Paci [supra n.10] 227
and 239).
painted dates referring to kalendae, nonae or idus, identifying the organisation of the build-
ing work for various parts of the gallery.44 In 2009, another 60 red-painted inscriptions
were documented on the walls of the SW exedra of the baths. The comparison of the vari-
ous inscriptions showed how several teams worked in parallel on the same structure, as
well as the timeline of the work.45 Some of the labels include the full information on the
calendar date, but there are numerous inscriptions with no mention of the month (fig. 25a).
However, their positioning, which respects their original ordering on the walls, made it
possible to trace them back to the month of reference even in cases where only the day is
mentioned. The dates can be interpreted as annotations affixed at the end of the working
day as a report on the activity carried out, probably for calculating the payment due.46
The use of painted marks in the construction process must have been widespread, as
is shown also by another example at Rome’s Crypta Balbi. A block in the colonnade foun-
dation has the inscription V K (the fifth day before the kalendae of an undefined month),
interpreted as the date it was placed there.47
Even more cogent comparisons with El Mèdol are found at extraction sites. In the
quarry at Kriemhildenstuhl, three inscriptions on the quarry front indicate dates every

44 R. Volpe, “Un antico giornale di cantiere delle terme di Traiano,” RömMitt 109 (2002) 377-94.
45 R. Volpe and F. M. Rossi, “Nuovi dati sull’esedra sud-ovest delle Terme di Traiano sul Colle
Oppio: percorsi, iscrizioni dipinte e tempi di costruzione,” in S. Camporeale, H. Dessales and
A. Pizzo (edd.), Arqueología de la construcción III (Anejos de AEspArq 64, 2012) 69-81; R. Volpe,
“Le giornate di lavoro nelle iscrizioni dipinte dalle Terme di Traiano,” in M. L. Caldelli,
G. Gregori and S. Orlandi (edd.), Epigrafia 2006. Atti XVIe rencontre sur l’épigraphie in onore di
Silvio Panciera (Tituli 9; 2008) 453-66; ead., “Organizzazione e tempi di lavoro nel cantiere
delle Terme di Traiano sul Colle Oppio,” in S. Camporeale, H. Dessales and A. Pizzo (edd.),
Arqueología de la construcción II (Anejos de AEspArq 57, 2010) 81-91.
46 However, a use related at the construction process is not excluded. They could be a useful way
of recording the execution phases of the structural parts, requiring particular processes to be
carried out according to pre-established times and methods in order to strengthen the wall:
Volpe (supra n.44) 383-93.
47 M. Cante, “La cosiddetta crypta del Teatro di Balbo: tipologia e struttura architettonica,” Palladio
33 (2004) 8.
 Painted marks at El Mèdol quarry near Tarragona 273

three months.48 They have been interpreted as recoding the arrival dates of new groups
of workers alternating at the site. Further evidence from an extraction site are the quarry
marks documented in the quarry of Conero (Ancona), where minium and charcoal inscrip-
tions alternate, both on the extracted blocks and directly on the quarry walls. Among the
written information, which includes counts or references to the size of the quarry area to
be exploited,49 there are calendar dates50 (fig. 25b). Here too it seems clear that the quarry
marks can be linked to the management of the quarrying tasks. Worthy of note in two of
the red-painted inscriptions is the mention of two different cognomina, one of whom is
believed to have been a quaestor at the colonia of Ancona.51 The fact that the two cognomina
appear in pairs leads to the belief that they were the duoviri there, whose charge in this case
would not have been explicit due to the extemporaneous character of these inscriptions.
The mention of these two individuals could provide information on the ownership of the
quarry, referring to a quarry area granted in locatio by the colony to private citizens.52 The
two cognomina (nos. 15-16) at El Mèdol might seem to be a parallel, but the epigraphy from
the town does not provide any elements to support this notion.53 One possible hypothesis
is that they refer to a person in charge of a quarry sector54 or a person responsible for one
of the teams of workers, connected to a date recording when the work was carried out.
Another possibility is that the inscriptions were related to orders or deliveries, accompa-
nied by the name of the individual who placed the order. We may compare a pedestal of
local marble preserved in the Roman baths of Caldes de Montbui that bears a dedication
by L. Vibius to Apollo (IRC I 35)55 and whose rear face carries a graffito with the cognomen
VIBI, probably added in the quarry to indicate the ownership of the block that this indi-
vidual had chosen and paid for.

From quarry to monument: organisational aspects of the construction of the Provincial

Forum
Fifty-eight engraved marks have also been documented in the accumulation of blocks
found next to El Mèdol quarry. They are alphabetic or numerical and in groups of 1, 2
or 3 characters. They are therefore short marks that in some cases could be interpreted
as simple numerals (including A, H, +, D, M), but in others could refer to the use of an

48 The dates are: III Idus Feb(ruarias) (Feb. 11); III Idus Mai(as) (May 13); VI Id(us) Aug(ustas) (Aug.
8). A fourth date is missing, referring to a few days before the idus of November; according to
the author, it would not be due to a gap in the preserved marks but to the interruption of work
during the winter: F. Sprater, Limburg und Kriemhildenstuhl (Speyer-am-Rhein 1948) 45-46.
49 Paci (supra n.10). See respectively inscriptions nos. 1, 2 and 3 written in charcoal and no. 9 in
minium.
50 Paci ibid. 227-29, inscription nos. 4, 6 and 7.
51 This identification is based on documentation of the quaestor at Ancona (Paci ibid. 237-38).
52 The hypothesis is reinforced by the abbreviations LOC and ME in the inscription, interpreted
by the author as locare and metallum (Paci ibid. 233-37).
53 A L(ucio) Aemilio Clemen[ti] Silvani[ano] has been attested, but he was curator capitoli in the
Hadrianic period (CIL 02-14, 02305; CIL 02-14, 01201; RIT 00922).
54 See a painted quarry-mark on one of the travertine blocks in the Colosseum. It refers to a man
by name of Seleucus, probably the head of a quarry sector where the travertine was extracted:
C. Conti and S. Orlandi, “Sui travertini del Colosseo tra restauro ed epigrafia,” RendPontAc 85
(2013) 82-83.
55 A. Álvarez Pérez et al., El marmor de Tarraco: explotació, utilització i comercialització de la pedra de
Santa Tecla en época romana (Tarragona 2009) 54-55.
274 M. S. Vinci

Fig. 26. Engraved quarry marks from El Mèdol.

 Painted marks at El Mèdol quarry near Tarragona 275

abbreviated terminology, probably anthroponyms (including TIR, FLA, BVCOLI, CLONI,

CIL: fig. 26 in colour). One example could be the quarry mark TIR, which Th. Hauschild
proposes reading as the cognomen Tiro.56
The same types of marks have been documented in the Provincial Forum at Tarragona.
A total of 27 engraved marks has been attested,57 mainly in the area of the “Praetorium
Tower”. This is in the SE sector of the “Representation Plaza”, at the junction of the build-
ing with the entrance to the middle terrace from the south.58 The marks have the same
typology59 — both simple numerals and more complex abbreviations. Moreover, some of
these marks are the same type as those found at the quarry, for example, TIR, VR, M, D, V
(fig. 27 in colour). The mark TIR is documented at the quarry and it is attested 7 times in
the monument. This helps link the work at the quarry with the construction site. Construc-
tion of the Provincial Forum involved a massive exploitation of El Mèdol quarry, whose
limestone was the main material used for the opus quadratum of this monument, which was
the largest in the colonia. It is no coincidence that the Early Imperial era was when exploita-
tion of the quarry reached its peak.60 Since they bear no explicit chronological indicators,
the pile of blocks found at the quarry could be quarrying rejects discarded steadily over
a period of time between the Late Republican and Early Imperial periods,61 but we prefer
to link the presence of quarry marks from two different contexts to the same extractive/
constructive process – i.e., for building the Provincial Forum. The dates of the beginning
and end of construction works for the Provincial Forum are not precisely known, but in all
likelihood the activity extended, if not continuously, over at least a century. The volume of
stone extracted for what became the symbolic monument of the imperial town would have
been remarkable,62 as must have been the organisation needed to manage it.

56 At “Signa lapicidarum,” CIL II2 14, fasc. IV, 972-76, Th. Hauschild ruled out the cognomina
Tironianus or Tironillianus as they are mentioned only once by Kajanto (supra n.40) 320;
likewise the Greek cognomen Tiridatis (cf. H. Solin, Die griechischen Personennamen in Rom I-III,
CIL Auctarium n.s. 2 [Berlin 2003] 243).
57 Hauschild ibid.; M. S. Vinci, “Notae lapicidinarum: preliminary considerations about the
quarry marks from the Provincial Forum of Tarraco,” in D. Matetić Poljak and K. Marasović
(edd.), ASMOSIA XI (Split 2018a) 699-710.
58 Other quarry marks have been documented in the NE sector of the circus. They are different
blocks with engraved letters (A or H, XXXII) forming part of the podium wall: M. Díaz, Ll. Piñol
and I. Teixell, “Materials i tècniques constructives emprades en l’edificació del circ de Tàrraco,”
in J. López (ed.), Tarraco biennal. Actes 3r Congrés int. d’Arqueologia i Món Antic. La glòria del
circ. Curses de carros i competicións circenses (Tarragona 2017) 257. During the intervention for a
museum in the E part of the circus, occasional blocks were also found bearing engraved marks
in the form of a single character: M. Díaz García and J. F. Roig Pérez, Memòria d’intervenció
arqueològica al C/ del Trinquet Vell, dins del “Projecte bàsic i d’execució per a l’arranjament
del C/ Trinquet Vell” (Tarragona 2014) 185. Since both the Praetorium Tower and the NE sector
of the circus are among the best-preserved parts of the imperial complex, there may well have
been marks in other parts that remain hidden or are no longer preserved.
59 We know of other engraved quarry/construction marks on (probably El Mèdol) limestone on
structures such as the town walls, but their typology is certainly different from those found
at the quarry. For a recent summary, see M. S. Vinci, “Marche di cava e sigle di costruzione:
nota preliminare sul materiale epigrafico proveniente dall’area di Tarraco (Hispania Citerior),”
Aquitania 34 (2018) 143-68.
60 Gutiérrez García-M. et al. (supra n.12) 151-52.
61 Ibid. 152. Judging by the archaeological finds, the extraction activity began to slow down from
the 3rd c. A.D.
62 The estimated volume of extracted stone for the total period of the quarry activity is 150,000 m3:
276 M. S. Vinci

Fig. 27. Engraved quarry marks from the Praetorium Tower in

Tarragona’s Provincial Forum.
 Painted marks at El Mèdol quarry near Tarragona 277

Conclusions
This kind of marking system certainly deserves to be viewed differently than that used
for marble. First, the material often comes from sources located close to the site where they
are to be used. This means that the distribution and trade networks would have been much
narrower in scope than those for marbles. Second, aspects of quarry ownership and man-
agement undoubtedly differed from those in operation for marbles.
The use of inscriptions in red paint or charcoal to organise and manage the quarrying
and/or building tasks must have been common. There are numerous other cases in which
painted or charcoal marks were used at quarries for different kinds of accounting63 but our
knowledge is limited by the fragility of the inscriptions as well as by their nature: extem-
poraneous messages (indeed, ephemeral annotations not intended or expected to survive)
and serving only for the single purpose for which they were written. Yet, paradoxically,
in some cases they have survived, and by chance help us to understand the work pro-
cesses. At El Mèdol, the quarry workers would probably have been paid not long after the
inscriptions were written. This is probably the main reason why they did not have to be
particularly durable and would probably have been removed during the processing phase
of the block. Yet we cannot rule out the hypothesis that the annotations of the daily work
had additional purposes. If, on the one hand, they were used to account for and pay for
work carried out, on the other they would have been a useful system for controlling the
progress of the work with respect to the schedule demanded, as well as for calculating the
time elapsed since the stone had been extracted and, consequently, the time it had been left
standing next to the quarry.
There does not seem to be any significance to whether red paint or charcoal was used:
perhaps it was whatever was available. The marks would have been made either immedi-
ately after the extraction or after the first roughing-out phase, and in fact it was often the
same stonecutters who carried out both the extraction and the initial working of the block.
The inscriptions on two blocks on which squaring had already begun (nos. 15-16) would
appear to confirm that the annotation may have taken place at two different points in the
process. Normally, the first roughing-out phase took place next to the quarry face but in
such a way so as to not impede the quarrying tasks. This meant that less work had to be
carried out at the building site, and it also considerably reduced the weight of stone to be
transported.
The hypothesis that the area in which the accumulation of discarded blocks was found
was also the area in which the blocks were first worked is consistent with the likely trans-
port routes leading to the town. Near the present-day main road, and so next to the
piled-up blocks, was the Via Augusta, a main route to the colony. Yet we cannot rule out a

Gutiérrez García-M. et al. (supra n.11) 779.

63 See, e.g., the evidence from Saint Boil quarry (Saône-et-Loire), where a series of inscriptions
including a list of numbers has been found on the quarry faces and interpreted as referring to
the accounting of the stone extracted, as well as containing a probable reference to the colony
of Cabilonnum and perhaps to an individual in charge of the quarrying: G. Monthel and
M. Pinette, “La carrière gallo-romaine de Saint-Boil,” RAEst 107-8 (1977) 37-56; G. Monthel and
P.-Y. Lambert, “La carrière gallo-romaine de Saint-Boil (Saône-et-Loire),” Gallia 59 (2002) 110-
11. At the Conero Roman quarry, there is similar evidence with inscriptions that include a series
of numbers written in charcoal: Paci (supra n.10) 220-26 and 229-31.
278 M. S. Vinci

sea route, as the quarry was only 300 m from the coast.64 In that case the building materi-
als would have been taken to one of the ports or landing sites on Tarraco’s coast65 and then
transported to the top of the hill.
Perhaps now is the time for scholars to re-evaluate the importance of the information
that the study of quarry marks can provide about the activities carried out at quarries for
less noble materials, and to use such testimonies for investigating further aspects of the
Roman building economy.
maria-serena.vinci@u-bordeaux-montaigne.fr Université Bordeaux Montaigne-AUSO
NIUS UMR 5607, Maison de l’Archéologie

Acknowledgements
This study has been financed by a postdoctoral contract within the “Initiatives d’excellence IDEX”
programme of Bordeaux University. The research was carried out with the collaboration and support
of M. Navarro Caballero (CNRS—Université Bordeaux Montaigne) and D. Gorostidi (ICAC—
Universitat Rovira i Virgili) and is included in the project financed by the Spanish Government
“Economía, Industria y Competitividad” I+D+i (HAR 2015-65319-p): “Officinae Lapidariae Tarraco-
nenses. Canteras, talleres y producciones artísticas en piedra de la Provincia Tarraconensis” directed
by D. Gorostidi.
I would like to thank the AUSONIUS Institute at Bordeaux Montaigne University, where the research
has been carried out thanks to funding from IdEx Bordeaux University, and in particular M. Navarro
Caballero who has supported and encouraged me, enriching this study with teaching, advice and
precious food for thought.
My thanks also to the institutions in Tarragona: the Catalan Institute of Classical Archaeology,
the Catalan Government and the National Archaeological Museum of Tarragona. In particular I
would like to thank D. Gorostidi for having encouraged this study in its early stages, for the initial
documentation and epigraphic consultations; J. López for the surveys of the Roman quarry and
A. Gutiérrez for the consultation on the extractive phases of El Mèdol quarry. I thank J. Menchón,
M. Miró Alaix and M. Adserias Sans for granting me access and permission to study the material
preserved in the Catalan Government store in Tarragona. I would like to thank F. Tarrats and J. A.
Remolà regarding the finds preserved at the National Archaeological Museum of Tarragona. Finally
I thank A. Ottati for operational and scientific support lent to this study.

64 Gutiérrez García-M. et al. (supra n.11) 784-87. The landing point would have been the beach
known as Platja de Calabecs where a ramp 4 m wide carved into the rock that descended into
the sea has been identified. Abundant pottery dating back to the second third of the 1st c. A.D.
has been documented by A. Gutiérrez and J. López c.150 m from this site.
65 It is unlikely that the building material would have been taken to what is known as the town
harbour, for that would have involved a longer sea route, as well as unloading in the lower
part of the town. It would then have had to have been transported across the whole town to the
upper part. For the construction site of the Provincial Forum, it is better to suppose the use of
alternative landing sites that will have existed along the coast. One suggestion is what is known
as Miracle Beach, located at the foot of the hill on which the imperial complex was built. There
is no evidence of port installations in this area, but finds from the sea (e.g., granite columns: see
W. Pérez, Troballes arqueològics al litoral tarragoní: 12 anys d’arqueologia subaquàtica (1968-1980)
[Valls 2007]) leave the possibility open. At Tarraco (Barcelona 1948) 23, A. Schulten claimed that
the port of Tarraco would have been located in Punta del Milagro bay, which offers greater
protection from the east wind.
Reproduced with permission of copyright owner. Further
reproduction prohibited without permission.