Saddle querns and rotary mills from
the Iron Age site of La Cervera
(La Font de la Figuera, Valencia, Spain)

Introduction¹

The 1st millennium BC in the western Mediterranean was a period marked by significant social and economic changes. Deriving from a surge of connections within the Mediterranean, these changes are evidenced by new settlement patterns, technological innovations and the emergence of complex social and political organisations. Although there are local specificities among this general trend, agrarian systems underwent a rapid and profound transformation throughout the western Mediterranean from the outset of this millennium (Buxó 2008; Bouby 2014; Pérez-Jordà et al. 2017). Particularly key features were the incorporation of new crops and technologies and the spread of innovations designed to intensify production.

Stone tools serving to grind grain into flour played a key role in the analysis of this process. Research on querns and mills in eastern Iberia has focused for the most part on their typological variability and technological changes, as well as on their link to social organisation (for an overview see Alonso & Anderson 2019 and Alonso & Pérez-Jordà 2014). During the course of the 5th and 4th centuries BC, the saddle quern, a tool in use since the 6th millennium BC, was abandoned in favour of the rotary quern and soon after its larger counterpart,
the Iberian rotary pushing mill (Alonso & Anderson 2019). Although the studies cited above offer valuable information on their economic and social role, other topics of research have attracted less attention. These include notions as to their raw material and provenance, notably their petrology and mineralogy, microsedimentary and use-wear traces (Bofill et al. 2013) and the quarries where they were extracted (Anderson 2016).

The current study therefore offers new evidence as to these materials based on a multidisciplinary analysis of 27 saddle and rotary querns and mills from the site of La Cervera (La Font de la Figuera, Valencia). The first three sections consist of a description of the site and the mills recovered in different contexts as well as a thorough examination of their typology, lithology and their outcrops. The final section contextualises the findings in the framework of the current discussion regarding the adoption of the rotary motion for grinding in the western Mediterranean so as to assess the social, economic and political conditions yielding the change from the “to and fro” to the “rotary” motion. Overall, the aim is to demonstrate the key role of mills in the framework of the broader processes of urbanisation and social change transpiring in the mid-1st millennium BC.

The site of La Cervera

The Iron Age site of La Cervera is located to the west of the present-day town of La Font de la Figuera (Valencia) on a gentle hill overlooking the Cànyoles River Valley (Fig. 1). The site was partially excavated in 2011 and 2012 in the framework of a rescue operation prior to the construction of the A-33 motorway. As it is estimated to stretch over a surface of about seven hectares, only 20% has been explored. The excavation of Area 1 took place in 2011 while that of Area 2, a small tip to the north of the hill, dates to 2012. The saddle querns and rotary mills presented in this paper come from these two areas which in terms of function, reveal no differences (López Serrano et al. 2018) (Fig. 2).

The site’s different structures and layers were moderately affected by erosion and by contemporary agricultural activity. The later finds comprise potsherds dated to the Roman, Medieval and Modern times associated with terraces and vineyards. However, the site’s main occupation dates to the Iron Age, more specifically, between the second half of the 6th and the 4th century BC. This chronological range is based on the pottery, a type of find that has benefitted from many years of study in this area and offers a secure relative dating. The locally produced “grey ware” pottery characteristic of the 6th and 5th centuries BC, as well as the presence of Attic imports, point to specific timeframes between the 5th and 4th centuries BC. The dating thus stems from the associations of finds in the stratigraphic units (SU) and base itself on the most recent materials of each assemblage. Therefore, to be on the safe side, the querns and mills were assigned the latest chronological phases of their corresponding layers.

The site is enclosed by a ditch with a U-section stretching across the southern, most accessible, area of the hill. This feature cutting through the bedrock is about 123 m long, 4.5 m wide and a maximum 2.5 m deep. It was sunk during the earliest phase of the site between the 6th and 5th century BC. It subsequently began to be filled with materials, gravel and soil by the 5th century BC (López Serrano et al. 2013).

The only dwelling identified consists of a structure with at least four rooms and two adjacent spaces that probably served other functions. The majority of the site’s features consist of terraced walls, pebbled areas, and platforms of stones serving a variety of activities including zones linked to combustion. Furnaces and metalworking areas dating to between the late 5th and the first half of the 4th centuries BC are distributed throughout different sectors of the site. Eight are interpreted as smithy hearths (elongated pits 30 to 50 cm long and 20-30 cm wide with burnt walls) linked to the presence of hammer scales (López Serrano et al. 2018). La Cervera reveals clear morphological differences with respect to other contemporary regional sites such as La Bastida de les Alcusses (Moixent), La Mola de Torró and l’Alt del Frare (La Font de la Figuera) as its internal organisation is not articulated along the line of groups of blocks of domestic constructions separated by areas of circulation.

The stratigraphic contexts of the querns and mills

There is no apparent functional pattern linked to the find contexts of the querns and mills. Certain were recovered among the layers of abandonment of dwellings, whereas others come from working areas or pits. Two were reused for other purposes and others are in SUs that offer little information as to their use or function (n°3 in SU 4028; n°4 in SU 2022; n°8 in SU 3148). The contexts of the others remain undetermined.

Among the querns and mills yielding the most contextual information is n°6 (SU 3164) recovered in the only domestic area identified at the settlement. Another is in a layer of abandonment layer of a storage area (n°11, SU 3154). Certain are in SUs liked to workshop with undetermined combustion structures (n°16, a rotary mill) and metallurgical structures (n°17 and 22-24, saddle querns). SU 2003 in Area 2 is interpreted as a midden containing potsherds, faunal remains and fragments of burnt earth interpreted as the walls of hearths. SU 2046 belongs to a level of abandonment associated with a floor linked to iron and lead slag. Two combustion structures (SUs 2005 and 2045) interpreted as smithy hearths are recorded among these units. Finally, two undetermined saddle querns were recovered in the layer of abandonment of working area (n°5 and 7, SUs 2003 and 2006).

A pebbled area identified in SU 3090 may have served as the base of a structure serving for various undetermined activities. Although there are two rotary mills in this unit, there is little contextual information providing insight into these activities. The fact that an area of burnt earth was identified next to the pebbles might point to a connection with smithy hearths and undetermined metallurgical activities.

Three querns were unearthed in the fills in the ditch cut through the bedrock. The first is an undetermined saddle quern (n°10) dating to the 6th – 5th centuries BC. The other two are fragments of rotary mills (n°14 – 15) dated to the 4th century BC. Five fragments of saddle querns (n°18 – 21 and 26) and one hand-held upper stone (n°27) were recovered in the layers of fill of a large oval pit (44 m²) interpreted as a clay quarry. This structure was backfilled with debris (SU 2010, 2014 and 2035) including potsherds, faunal remains, mudbricks and charcoal.

Two saddle quern fragments (n°9 and 25) were identified respectively in the foundation of construction SU 305 and that of the terraced wall SU 202 associated with stones of diverse sizes in an alignment about 3.5 m long. The poor state of conservation of this feature does not allow any further interpretation as to its function but it may well have been part of the structure serving to delimit a working area.

Two final remarks about the contexts of the mills. Firstly, there is no clear pattern allowing to link each of the two types (saddle and rotary) with spatial functions. Secondly, there is no evidence of the coexistence of the two mill types. The data in fact points to a substitution of the saddle quern for the rotary mill in a timeframe corresponding to the turn of the 5th to the 4th century BC. Saddle querns belong to SUs dated to the 6th and 5th centuries BC although they also appear in the fills of the 4th century BC. Rotary mills, in turn, can be securely dated to the 4th century BC, unearthed in a dwelling and in layers of fill of workshops and terraced areas. The relatively high concentration of saddle querns in Area 2 (Fig. 2) can be explained by the fact that the occupation of the 4th century BC was not well preserved in this sector of the site.

Mill typology and lithology

The database comprises 25 mill fragments and two hand-held rubbers or upper stones (Fig. 3; Table 1). The lot can be further broken down into 10 saddle quern fragments, and nine rotary mill fragments and six undetermined. As none is complete and all were discarded and recovered in secondary position, it is not always possible to distinguish between the rotary mill types.

Saddle querns

The saddle querns serving to grind grain consist of a rounded hand-held upper stones or rubbers, driven with a to and fro motion, on a flat or concave lower stone. This quern type is known in the eastern Iberia since the Neolithic and continued to serve in the mid-1st millennium BC. La Cervera has yielded 10 lower and two upper stones (rubbers) (Fig. 3). The former belongs to a minimum of eight different querns. It has not been possible to measure the total length of any of the saddle querns. Their width, in turn, ranges from 11 to 20 cm and their thickness between 4.54 and 11 cm. Due to the high degree of fragmentation, only two can be identified typologically on the basis of their morphological features. They fall according to the classification advanced by Alonso (1999) into types V-P1 and V-P2. Their forms are ellipsoidal and rectangular, depending on whether their corners are rounded or not.

Various types of rock served to manufacture the saddle querns unearthed at La Cervera, notably seven sedimentary and five igneous types (Fig. 4). The sedimentary rocks comprise bioclastic calcirudites and undetermined sandstones. Calcirudites are characterised by abundant millimetric rounded and angular quartz clasts and very frequent hematoid quartz crystals which also present unusually abundant pores, possibly related to processes of dissolution. The sandstones reveal a variety of textures and lithologies, predominated by angular medium and thick quartz sands. The matrices are light grey and yellowish. Certain among the yellowish cases are micritic and others bear traits related to calcirudites.

The igneous rocks serving for the saddle querns are of two types. Predominant is the light grey (light grey 7 / N Gley 1) type with vacuoles or circular or oval vesicles surpassing 1 mm. Macroscopic observations reveal black, white and transparent crystals that could correspond to phenocrysts of pyroxene and plagioclase of porphyritic texture. The second is a dark grey (dark grey 4 / N Gley 1) type characterised by small vacuoles (less than 1 mm) of a similar texture.

In spite of the lack of mineralogical or petrological analyses, the two rock types can be classified respectively as volcanic or subvolcanic. The texture of the first group is most often porphyritic, vacuolar or vesicular resulting from the cooling of magma at low pressure allowing the expansion of the volatile substances. The subvolcanic rocks, in turn, although occasionally revealing a porphyritic texture, are mainly ophitic or diabasic. Although less common, vacuoles can likewise develop in disturbed rocks, for example through hydrothermal processes giving rise to the formation of secondary minerals.

Rotary mills

Rotary mills consist of two superimposed circular stones: a stationary or passive lower stone mounted by a mobile or active upper stone. This smaller rotary quern model, usually not surpassing 50 cm in diameter, was driven by hand, presumably in a sitting position, and generally associated with domestic tasks. The larger Iberian pushing mill, by contrast, was driven from a standing position (often on a podium) (Alonso-Frankel 2017). Yet for the La Cervera corpus of six upper and three lower highly fragmented stones with a diameter estimated at about 50 cm it is often not possible to determine whether they correspond to rotary or Iberian pushing mills (Table 1; Fig. 3)

In any case, calcirudite is the rock which exclusively served to manufacture the rotary models. Its properties are similar to those used to make saddle querns, notably an intrinsic coarseness assuring a “bite” even after extended use, and not requiring any type of grinding surface dressing.

Metrics, in particular diameter and thickness, as noted above, can serve as criteria to characterise these mills (Fig. 5). A basic criterion to differentiate rotary mill types is whether they surpass 50 cm in diameter (Longepierre 2012) as it is difficult to envision the larger models to be operated by a single person. The thickness of rotary upper stones also has a major bearing on performance. None among the La Cervera group (ranging between 9.9 and 13.5 cm in width) attain the dimensions designated for the larger mills. This factor thus appears to differentiate the La Cervera rotary mills from others in the area such as those of La Bastida de les Alcusses and L’Alt del Punxó (Bonet & Vives-Ferrándiz 2011; Espí et al. 2009).

Another difference corresponds to the inclination of the stone’s working surface, respectively less or greater than 20º according to if they line up with the smaller and larger models. Concerning diameter, half of the assemblage reveal diameters within the average range of 39 to 44 cm known for rotary querns of the Iron Age (Alonso & Pérez-Jordà 2014). Although the remaining reveal diameters that surpass this average, only two (n°1 and 8) can be classified as larger Iberian pushing mills (Table 1; Fig. 3 : 5).

The upper surface of the upper stones reveals a slight inclination towards the central hole. Moreover, three have a border or rim serving to contain the grains. Two types can be differentiated according to the inclination of the lower surface: three cases ranging between 22º and 24º and two between 46º and 47º. The upper stone grinding surfaces reveal an angle between 13º and 18º.

Two of the rotary mills are characterised by perforated handle holes through protruding lugs (n°6 and 15). The second of the two, in fact, corresponds as evidenced by a new perforation near the rim to a repair following the break of the original hole. Perforated handle holes, rims and ear-shaped lugs are upper stone features of both rotary querns and the larger Iberian pushing mills typical of sites in the Cànyoles River Valley such as La Bastida de les Alcusses (Bonet et al. 2011) and Alt del Frare. These morphological elements are likewise recognised at sites in adjoining valleys such as Cerro Lucena (Enguera, Valencia) and L’Alt del Punxó (Espí et al. 2009) appearing to reflect a regional tradition of manufacture.

The lower stones are completely perforated by a central eye. Certain reveal an irregular base suggesting they were set in a hollow at floor level or on podia no longer preserved raised with stones and earth.

Rock outcrops

Various types of sedimentary and igneous rocks were exploited for the manufacture of saddle querns and rotary mills. As there is no outcrop at the site itself, their identification is relevant to the understanding of the relationship of the site and control of the surrounding area and its resources.

The calcirudites serving to manufacture both the saddle and rotary mills (Fig. 6) bear unique features that allow a relatively clear identification of their source. Bioclastic calcirudites are calcareous sandstones characterised by a matrix of calcium carbonate and thick detritic elements (frequently > 2 mm in diameter), terrigenous inclusions (mainly quartz), and marine, fossil and intraclastic organic remains. They stem from coastal environments and are thus found in various phases of the Mesocenozoic associated with transgressive marine phases.

This different rock outcrops around La Cervera (Fig. 7) originated in the framework of an intra-Neogene transgressive phase. Its dating ranges between the Langhian and the Tortonian, with progressively younger values to the north/northwest and west (Martínez del Olmo & Benzaquen 1975a and b; Almela et al. 1975; Lendínez & de Tena-Dávila 1981; Goy et al. 1980, 1982).

Calcirudites comprise diverse facies whose characteristics vary according to the sedimentary environment that formed them and their supply basin. The facies range from conglomerates with quartzite edges to biolytic limestones formed in situ by organisms, through molasses of very fine white or yellow sands in all cases with fossil remains of marine origin. This is, for example, the case of the highly fossiliferous sandy limestones and the compact limestones of the Mugrón formation which occupy the entire relief bearing the same name in Almansa 33 km from La Cervera (Calvo et al. 1974), the fine-grained biocalcarenites and bioclastic limestone of El Perdigón, also in Almansa 25 km away (Lendínez & de Tena-Dávila 1980), and the fine white sandstones of Vallada and Alcoi (Martínez del Olmo & Benzaquen 1975a and b).

Of the 12 outcrops surveyed in a radius of 50 km from La Cervera, only those of Port dels Alforins (17 km) and Montesa (30 km) reveal similar features. A true extractive millstone quarry has in fact been identified at La Mola in the town of Montesa (Fig. 8). Cylinders at this quarry measuring between 50 cm and 1.20 m were scored directly from bedrock on multiple horizontal planes yielding a high vertical tubular quarry face or in the form of clusters in benches at different levels. This extraction presumably dates to the Middle Ages and later. It nevertheless cannot be ruled out that certain were Roman (Anderson 2014: 161; Anderson 2016: 106, 594).

Igneous rock outcrops are relatively frequent in the Valencian region and its surroundings (Fig. 7) with a predominance of subvolcanic diabases. They are traditionally called ophites because they appear associated with extrusions of Triassic materials in areas with deep fault lines, although their texture and composition differ from those of true ophites or tholeiitic dolerites (Gallart & Lago 1988). The diabases outcrop on a small hill in Villena (22 km), at Tossal Negre de Barxeta (50 km) and at Cerro Negro de Quesa (55 km). Noteworthy are also the extensive group of quarries respectively at La Alcoraya de Alicante and Albatera (more than 70 km away) and Orxeta, Finestrat, Altea, Calp, etc. (more than 100 km away). The outcrops in the case of Bartxeta and Quesa, however, do not present either light grey rocks or vacuoles. The first outcrops of diabases comprise spilites while the second, with a very similar mineralogical and geochemical composition (exploited in prehistory for polished stone tools) comprise pyroxenes of varying size and plagioclases that include pyroxenes with diabasic texture as well as amphiboles (Gallart & Lago 1988, 236). In Villena and its surroundings stands out a small promontory (Soler García 1987) with a few diabase outcrops. Although these materials were exploited since the Neolithic to produce polished tools, their use appears to have significantly decreased in the Bronze Age.

Volcanic outcrops are less frequent with the nearest are at Cofrentes and Picassent respectively 65 and 85 km away. Other volcanic districts are recorded further away at Cabo de Gata, Murcia, Campo de Calatrava and Girona (Cebriá 2002). The rocks of Cofrentes (Castillo, Cerro de Agras and Fraile), originally primary magmas, are limburgites consisting of olivine nephelinites. Also frequent are dunites, harzburgites, lherzolites and werhlites, all olivine-rich ultramafic rocks altered to form iddingsite. The olivine basalts of Picassent form part of a small outcrop (max. 10 m thick) described as trachybasaltic and basanite/tephrite (Ancochea et al. 1984; Ancochea & Huertas 2002). The features of the latter may match those described for the igneous rocks serving to manufacture the querns of La Cervera, which do not appear to contain olivine although may have a porphyritic texture. However, it is not possible to confirm this due to the subsequent exploitations and transformations at Picassent that have significantly reduced the outcrops.

Finally, the undetermined sandstone saddle querns might represent singular facies of the outcrops cited above. The others bear lithological features that resemble other types of facies, of similar origin, dating from the Cretaceous to the Miocene. Obviously, these rocks are frequently present in the immediate surroundings of La Cervera (see references above).

Discussion

Quern and mill studies yield valuable information on ancient societies and their economies. If a Europe without querns is unthinkable, as Peacock and Williams have rightly argued (2011, ix), then an ancient Mediterranean without querns is likewise unthinkable. As the main tools serving to transform grains (a staple throughout most of its history) into flour for consumption, these mechanisms were vital to human subsistence. They likewise shed light on issues relative to social relations, the organisation of daily tasks, patterns of exchange and distribution, technological innovations, and the use of raw materials.

Querns and rotary mills in their historical context

The first point of discussion regards the change from the saddle to the rotary quern and the social and economic conditions surrounding the new mill type. The rotary motion is a technological innovation that took place in the western Mediterranean around the 5th century BC (Alonso 1997). Seen through a Mediterranean lens, this is not an exception or a one-off event. It coincided with other milling innovations in other parts of the region. These include the Olynthus style mill in Greece or the Morgantina type in central Mediterranean (Alonso & Frankel 2017). The earliest rotary querns appear in contexts dated to around 500-450 BC in northeastern Iberia at sites such as Els Vilars (Arbeca, Lleida), Alorda Park (Calafell, Tarragona), Turó de Ca n’Olivé (Cerdanyola, Barcelona), and further north at Pech Maho (Sigean, Aude, France) to the north of the Pyrenees. Although they were not imported from distant regions, there is evidence of exchanges operating in local networks. The earliest rotary quern from the inland site of Els Vilars, dated to the middle of the 5th century BC, is thought to be an import from the Catalan coast (Alonso 1999, 253).

The rotary mill was adopted relatively quickly in the lower part of the Ebro River  Valley and further south, along the Valencian inland and coastal areas at the sites of Los Villares/Kelin (Caudete de las Fuentes, Valencia) in the 5th and 4th centuries BC (Mata 2019: 88, 147) and in the 4th century BC at El Taratrato (Alcañiz, Teruel), El Tos Pelat (Moncada, Valencia) as well as further south in the Meseta at Alarcos (Ciudad Real) (Alonso & Pérez-Jordà 2014). However, this process was by no means uniform. As in the case of many other technologies throughout history, rotary mill adoption was politically driven.

Rotary querns and mills in the upper Cànyoles River Valley dated to 400-330 BC are recorded at La Bastida de les Alcusses (Bonet & Vives-Ferrándiz 2011). The finds from La Cervera therefore confirm this pattern of change placing this type of mill in the 4th century BC. Moreover, there are arguments that this mill did not coexist with the saddle quern. However, it remains uncertain whether saddle querns fell into complete disuse or whether the two mill types coexisted for a certain time at other sites.

Certain saddle and rotary mills from La Cervera were unearthed in metallurgical workshops and smithy contexts. The assumption is that they were served to grind the grain consumed by the workshop workers. A hypothetical use of the saddle querns at some stage in the processing of metal for crushing or as anvils cannot be confirmed without functional and use-wear analyses.

The social implications of the adoption of rotary mills are compelling. Given that they were technological innovations, and that technology is not politically neutral (Foxhall 2003, 75), it is necessary to contextualise this mill within other major social and economic changes in the 5th and 4th centuries BC, notably the processes of urbanisation, the intensification of relations in the western Mediterranean, and a new framework of social relations based on empowered households. In other words, it is imperative to delve into the interplay between the political economy, knowledge and technology.

There is a common consensus that rotary querns replaced saddle querns due to their greater efficiency (Alonso & Pérez-Jordà 2014). There is no evidence suggesting that the change of mill type correlates with the grain types as the cereals grown remained largely unchanged. Hulled barley and naked wheat continued being the staple in the framework of the earliest rotary querns as it had throughout the previous millennia (Pérez-Jordà 2013). This bolsters the notion that the innovation resulted in a more efficient and greater yield of flour while decreasing the amount of labour and investment of time per person.

It is possible to rule out the hypothesis that the earliest rotary querns were the larger pushing mills serving the whole community. Contextual data from eastern Iberia reveals that the earliest models were smaller models in domestic contexts measuring between 35-45 cm in diameter (Alonso & Pérez Jordà 2014, 242, fig. 2) thus serving the household. This implies that the technological change was driven by decisions inside households and did not stem from a top-down phenomenon imposed by a broader political entity. However, the larger pushing mills soon coexisted alongside the smaller counterparts because by the 4th century BC millstones surpassing 50 cm in diameter can be linked to collective milling structures at La Bastida de les Alcusses and at L’Alt del Punxó probably controlled by empowered households (Espí et al. 2009; Bonet & Vives-Ferrándiz 2011; Alonso & Pérez-Jordà 2014).

All in all, labour saving for households was as crucial as increasing production evidenced by the simultaneous mobilisation of other productive technologies. Iron for example did not become widespread productive technology until the rise of the demand for surplus, a circumstance that took place roughly from the 5th century BC onwards (Vives-Ferrándiz & Mata 2020). It appears that new concepts and techniques were catalysed in the 5th century BC in the framework of population growth and aggregation, the founding of new settlements and the creation of new social relations. This study now delves deeper into the particular local features of this situation.

Querns and mills, outcrops and territory

From the outset of the 4th century BC (and probably already during the second half of the preceding century) the settlement pattern in the upper Cànyoles River Valley was dramatically altered with the creation of new sites on hilltops or on the slopes of the mountains surrounding the valley such as La Bastida de les Alcusses, La Mola de Torró, Alt del Frare and Castellaret de Baix (Fig. 1). Certain stood out as they were heavily defended as evidenced by walls. Tombs with sculptures at the cemetery of Corral de Saus (Moixent) next to the main route following the valley exemplified the power of certain families (Izquierdo 2000). Cemeteries were more often than not part of the processes of urbanisation (Riva 2010, 177) and Corral de Saus might have drawn together the tombs of the elite from several settlements such as La Bastida de les Alcusses. All of these sites reveal violent levels of destruction dating to the late 4th century BC. This aspect is yet another argument revealing the relation between these sites in the landscape.

Access to long-distance exchange and the control and redistribution of agricultural surpluses bolstered the power of the elites at these sites. The region as early as the 6thand 5th centuries BC formed part of a broader world. This is evidenced by finds of Greek imports at Castellaret de Baix (Moixent) and La Cervera itself. Although the density of the occupation, the architecture and the internal organisation of these sites are fairly similar, La Bastida de les Alcusses stands out due to the extension of its domestic area, its fortification type, its economic infrastructures (a granary and roads for carts) and its long-distance connections. La Cervera with its smithies devoted to post-reduction activities and other copper and lead metalworks thus blended into the territory. These metallurgical workshops did not operate at a domestic scale as was the case in contemporary settlements such as La Bastida de les Alcusses (Pérez-Jordà et al. 2011). Thus, the presence of certain upscale craft activities in specialised settlements fits into a complex organisation of the territory aiming at the exploitation of resources, their transformation and distribution.

Pursuing this line of thought leads to delving into the significant changes linked to the material used to produce the querns and mills. While production of saddle querns of the 6th and 5th centuries BC resorted to a variety of rocks, notably calcirudite, sandstone and igneous rocks, the rotary mills of the 4th century BC were all hewn from calcirudites probably procured locally at Montesa. At this point, sandstone and igneous rocks were completely abandoned. This pattern of specialisation in the exploitation of an outcrop for rotary querns and mills is confirmed by other nearby sites such as La Bastida de les Alcusses where all rotary querns are calcirudites (Fig. 9). Thus, from the 4th century BC onwards the manufacture of querns as a centralised activity appears to have gone hand in hand with the adoption of the new type. At Els Vilars, a similar change in the type of stone is recorded. Saddle querns were fashioned from a variety of materials, often of igneous origin, whereas most rotary querns and mills from the 4th century BC were hewn from nearby porous limestone outcrops (Alonso et al. 2011, 57, fig. 6).

Little is known about the mechanisms of manufacture, exchange and distribution of saddle querns. In any case, their heterogeneity points to complex connections in the region and different spheres of exchange. This situation changed in the 4th century BC when resorting to calcirudite to fashion rotary mills became widespread. There are potentially several reasons behind the choice of specific outcrop. These are in the first place functional aspects deriving from the properties of this rock and its availability. The igneous outcrops were small when compared to those of the calcirudites. Hence a need to increase the rate of manufacture might have provoked the change of rock. It is necessary to recall that calcirudite had earlier served for saddle querns so the inhabitants of the territory were familiar with its properties. It is characterised by abundant pores which give it an abrasive quality, a property that would have been very useful for grinding cereals in circular mills. Also, although relatively simple to carve, its hardness and density required regular dressing. Be this as it may, the widespread use of calcirudite for rotary querns must have been connected to their morphology, speed of rotation or grinding properties. Only comparative mechanical testing and petrographic analyses can offer further insight into these questions.

But there may also have been a political reason for this choice linked to the creation of a structured territory. It is clear that the source of calcirudite was local (probably at La Mola in Montesa). A quarry serving different sites fits well into the framework of an organisation of specialised settlements in a territory, much like the extra-domestic iron working at La Cervera. The production and distribution of querns and mills must thus have been local. But what remains unknown due to the lack of a systematic survey is whether mills of this rock type attained other regions. Other questions also remain unsolved. Was rotary mill manufacture standardised? How was the exploitation of the quarry organised? There are likewise issues of ownership, the mobilisation of labour and other logistical aspects. Only the identification and archaeological study of the quarry itself can shed light on certain of these questions.

As yet, there is no evidence – not even a single cylindrical roughout – that rotary mills were manufactured inside the settlement. The only case of this type of work recorded so far in the framework of the Mediterranean Iron Age is in a layer dating to the 4th century BC at the Fortress of Els Vilars in the Catalan hinterland. The original source serving to procure this rock is an outcrop of porous limestone in the site’s vicinity (2 km away) (Alonso et al. 2011). The organisation in the upper part of the Cànyoles River Valley, in turn, may have differed. For the moment is it only possible to state that the resources were mobilised in the quarry and that the finished products were distributed among different settlements.

Conclusions and future research

This case study of saddle and rotary querns and mills from La Cervera, a site in southeastern Iberia dated between the 6th and 4th centuries BC, contributes to the archaeological study of mills and milling in the Mediterranean Iron Age. The mills were recovered in a variety of contexts, most often that of metallurgical workshops dedicated to post-reduction processes. Those dating to the 6th and 5th centuries BC were saddle querns of diverse typology. Various outcrops were exploited simultaneously for their manufacture, notably sandstones, calcirudites and igneous rocks of uncertain origin.

Major changes took place in the early 4th century BC with the introduction of the rotary motion. The fact that all these mills are of calcirudite points to the emergence of new pattern of outcrop exploitation. The promontory of La Mola at Montesa, 30 km from La Cervera, may have served as the source of this stone. The changes transpired in the framework of population growth and urbanisation stemming from political integration into new forms of social relationships. Although already underway in the 5th century BC, the changes developed even more in the 4th century BC. In this scenario, rotary querns and mills represented an innovation that overcame the main obstacle to increasing production – the shortage of labour – alongside other technological innovations.

A thorough survey of all the Iron Age querns and mills both in the Cànyoles River Valley and the region is still pending. For now, it is only possible to draw attention to the lithological and typological characteristics of these implements and point out several potential avenues of research. Future projects should focus on the mills themselves through lithological and use-wear analysis. Geological analyses can potentially shed light on the different properties of the igneous and sedimentary rocks. Landscape studies can glean information on the distribution of mills and exchange patterns. Finally, the identification of quarries is paramount although it is difficult to offer precise chronological references. Overall, this molinological research has proven to be fruitful and offers a promising framework for future archaeological studies.

Bibliography

• Almela, A., Gómez, E., Quintero, I. and Mansilla, H. (1975): Mapa geológico de la Hoja nº821 (Alcoy). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Alonso, N. (1997): “Origen y expansión del molino rotativo bajo en el Mediterráneo occidental”, in: Meeks, D. G. D. (ed.), Techniques et économie antiques et médiévales. Le temps de l’innovation, Aix-en-Provence, 15-19.
• Alonso, N. (1999): De la llavor a la farina. Els processos agrícoles protohistòrics a la Catalunya Occidental, Lattes, UMR 154, CNRS.
• Alonso, N. and Anderson, T.J. (2019): “A brief overview of archaeological research on ancient mills and milling in Spain”, in: Anderson, T.J. and Alonso, N. (eds), Tilting at Mills: Colloquium on the Archaeology and Geology of Mills and Milling (Almería, marzo 5-8, 2014), Revista d’Arqueologia de Ponent, Extra 4,13-31.
• Alonso, N. and Frankel, R. (2017): “A survey of ancient grain milling systems in the Mediterranean”, Revue Archéologique de l’Est Suppl. 43, 461-478.
• Alonso, N. and Pérez-Jordà, G. (2014): “Molins rotatius de petit format, de gran format i espais de producció en la cultura Ibèrica de l’est peninsular”, Revista d’Arqueologia de Ponent, 24, 239-256.
• Alonso, N., Aulinas, M., Garcia, M. T., Martín, F., Prats, G. and Vila, S. (2011): “Manufacturing rotary querns in the 4th century BC fortified settlement of Els Vilars (Arbeca, Catalonia, Spain)”, in: Williams, D.  and Peacock, D. (eds), Bread for the People: The Archaeology of Mills and Milling, Proceedings of a colloquium held in the British School at Rome, 4th-7th November 2009, BAR Int. Ser. 2274, Series in Archaeology, 3, 55-65.
• Ancochea, E. and Huertas, M. J. (2002): “Nuevos datos geocronológicos y geoquímicos de las manifestaciones volcánicas de Picasent y Cofrentes (Valencia)”, Geogaceta, 32, 31-34.
• Ancochea, E., Muñoz and M. Sagredo, J. (1984): “Las manifestaciones volcánicas de Cofrentes y Picasent, provincia de Valencia”, in: Congreso Español de Geología. Segovia 1, t. 2, 1-13.
• Anderson: T.J. (2014): “Moleras en la Península Ibérica: una primera clasificación de las canteras de molinos”, Revista d’Arqueologia de Ponent, 24, 157-74.
• Anderson, T.J. (2016): Turning Stone to Bread. A Diachronic Study of Millstone Making in Southern Spain, Southampton Monographs in Archaeology, New Series, 5.
• Bofill, M., Procopiou, H., Vargiolu, R. and Zahouani, H. (2013): “Use-wear analysis of near eastern prehistoric grinding stones”, Regards croisés sur les outils liés au travail des végétaux, Antibes, APDCA,225-242.
• Bonet Rosado, H. and Vives-Ferrándiz, J. (2011): La Bastida de les Alcusses. 1928-2010, Valencia, Diputación de Valencia.
• Bonet, H., Soria L. and Vives-Ferrándiz, J. (2011): “La vida en las casas: Producción doméstica, alimentación, enseres y ocupantes”, in: Bonet, H. and Vives-Ferrándiz J. (eds), La Bastida de les Alcusses. 1928-2010, Valencia, Diputación de Valencia, 139-175.
• Bouby, L. (2014): L’agriculture dans le Bassin du Rhône du Bronze Final à l’Antiquité. Toulouse.
• Buxó, R. (2008): “The agricultural consequences of colonial contacts on the Iberian Peninsula in the first millennium”, Vegetation History and Archaeobotany, 17(1), 145-154.
• Calvo, J.P., Ordóñez, S. and Usera, J. (1974): “Estudio del Terciario marino de la sierra del Mugrón (Prov. Albacete y Valencia)”, Acta Geológica Hispánica, IX, 5, 174-178.
• Cebriá, J.M. (2002): Geoquímica de las rocas basálticas y leucititas de la región volcánica de Campo de Calatrava, España, Doctoral Thesis. Universidad Complutense de Madrid. Facultad de Ciencias Geológicas. Departamento de Petrología y Geoquímica.
• Espí, I., Grau, I., López, E. and Torregrosa Giménez, P. (2009): “La aldea Ibérica de l’Alt del Punxó: producción agrícola y asentamiento campesino”, Lucentum, 28, 23-50.
• Foxhall, L. (2003): “Cultures, Landscapes and Identities in the Mediterranean World”, Mediterranean Historical Review, 18(2), 75-92.
• Gallart, M.D. and Lago, M. (1988): “Procedencia de las hachas pulidas del poblado de la Ereta del Pedregal (Navarrés, Valencia)”, Archivo de Prehistoria Levantina, 18, 233-237.
• Lendínez, A. and de Tena-Dávila, M. (1980): Mapa geológico de la Hoja nº793 (Almansa). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición, IGME.
• Lendínez, A. and de Tena-Dávila, M. (1981): Mapa geológico de la Hoja nº818 (Caudete). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Martínez del Olmo W. and Benzaquen, M. (1975a): Mapa geológico de la Hoja nº794 (Canals). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Martínez del Olmo W. and Benzaquen, M. (1975b): Mapa geológico de la Hoja nº820 (Onteniente). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Goy, J. L., Zazo, C., Ríos, L. M., Beltrán, F. J. and Zapatero, M.A. (1980): Mapa geológico de la Hoja nº769 (Navarrés). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Goy, J. L., Zazo, C., Ríos, L. M., Beltrán, F. J. and Zapatero, M.A. (1982): Mapa geológico de la Hoja nº770 (Alcira). Mapa Geológico de España, E. 1:50.000. Segunda Serie (MAGNA), Primera edición. IGME.
• Izquierdo, I. (2000): Monumentos funerarios ibéricos: los pilares-estela, Diputación de Valencia, Serie de Trabajos Varios del SIP 98, Valencia.
• Longepierre, S. (2012): Meules, moulins et meulières en Gaule méridionale du II^e s. av. J.-C. au VII^e s. ap. J.-C., Montagnac.
• López Serrano, D., Valero Climent, A., García Borja, P., Rodríguez Traver, J. A. and Vives-Ferrándiz, J. (2013): “El foso ibérico de La Cervera (La Font de la Figuera, València)”, in: García Borja, P., Revert Francés, E., Ribera i Gomes, A. and Biosca Cirujeda, V. (eds), El Naixement d’un poble. Història i arqueologia de la Font de la Figuera, Ajuntament Font de la Figuera, 93-103.
• López Serrano, D., Valero Climent, A., García Borja, P., Rodríguez Traver, J. A. and Vives-Ferrándiz, J. (2018): “Excavaciones arqueológicas en el yacimiento ibérico de La Cervera (La Font de la Figuera, Valencia)”, in: Actas de las Jornadas de Arqueología de la Comunidad Valenciana 2013-2015, Generalitat Valenciana, 41-53.
• Mata, C. (2019): De Kelin a Los Villares (Caudete de las Fuentes, Valencia). Nacimiento y decadencia de una ciudad ibera, Diputación de Valencia, Serie de Trabajos Varios del SIP 122, Valencia.
• Pérez-Jordà, G. (2013): La agricultura en el País Valenciano entre el VI y el I milenio a.C., Universitat de València, Doctoral thesis.
• Pérez-Jordà, G., Peña-Chocarro, L., García Fernández, M. and Vera Rodríguez, J. C. (2017): “The beginnings of fruit tree cultivation in the Iberian Peninsula: plant remains from the city of Huelva (southern Spain)”, Vegetation History and Archaeobotany, 26, 527-538.
• Riva, C. (2010): The urbanisation of Etruria. Funerary practices and social change, 700-600 BC, Cambridge, Cambridge University Press.
• Soler García, J. M. (1987): Excavaciones arqueológicas en el Cabezo Redondo (Villena, Alicante). Alicante.
• Vives-Ferrándiz, J. and Mata, C. (2020): “Iron metallurgy, political economy and social change during the first millennium BC in eastern Iberia”, in: Belarte, M.C., Rovira, M.C. and Sanmartí, J. (eds), Iron Metallurgy and the Formation of Complex Societies in the Western Mediterranean (1st Millennium BC), Proceedings of the 8th International Archaeological Meeting of Calafell (2016), Barcelona, Universitat de Barcelona, ICAC (Arqueo Mediterrània 15), 141-160.
• Williams, D. and Peacock, D. (2011): Bread for the People: The Archaeology of Mills and Milling, Proceedings of a colloquium held in the British School at Rome, 4th-7th November 2009, BAR Int. Ser. 2274, Series in Archaeology, 3.

Notes

1. The current study forms part of the HAR2014-54414-P (METIPO: Metales, tierra y poder durante la Edad del Hierro en el sureste de la península ibérica) and FRUITCOM “Un Nuevo modelo agrícola: frutales, hierro y comercio durante el I milenio A.E.” (CIDEGENT/2019/003) research projects.