Masters of mudbrick: Geoarchaeological analysis of Iron Age earthen public buildings at Ashdod‐Yam (Israel)

Excavations at Ashdod‐Yam exposed a fortification system that features a massive mudbrick wall with large earthen ramparts laid on either side. This fortified horseshoe‐shaped enclosure once surrounded what was likely a human‐made harbor and an adjacent acropolis with complex earthen architecture, constructed and active during Iron Age IIB–C (eighth–seventh centuries B.C.E.). These Iron Age public structures are at the center of the current research. In this paper, we present the geoarchaeological analyses of Ashdod‐Yam's earthen architecture. We applied a multidisciplinary methodology to new evidence for mudbrick manufacture with the goal of understanding the relationship between governing bodies and craftsmen. The analyses combine X‐ray fluorescence, loss on ignition, environmental scanning electron microscopy, and thin‐section petrography to investigate raw material procurement, manufacturing choices, and labor organization at Ashdod‐Yam during Iron IIB–C. Construction techniques and the standardization of the mudbrick recipe point to a local enterprise regarding the site's public earthen architecture. Furthermore, the degree of labor organization must have been closely observed and supervised by a central political power. Thus, it is argued here that construction and maintenance of the site was carried out by the kingdom of Ashdod, either as a part of its own local initiative or on behalf of the Neo‐Assyrian empire.

. Its study has notably progressed in the last few decades illustrating how earthen architecture is a source of social as well as technological and environmental information, especially when combining new approaches through the application of chemical and physical characterization techniques (Devolder & Lorenzon, 2019;Lorenzon & Iacovou, 2019;Mateu et al., 2022).Thus, geoarchaeological and architectural analyses of earthen architectural features can provide valuable information regarding manufacturing processes and technological choices as well as economic status and social stratification at these settlements (Homsher, 2012;Lorenzon et al., 2020;Nodarou et al., 2008).One such site is the Iron IIB-C enclosure at Ashdod-Yam, which features one of the largest earthen Iron Age construction projects in the southern Levant.
The analysis of mudbricks has a long history in West Asian archaeology within the prehistoric periods (Goldberg, 1979;Love, 2012;Rosen, 1986;Rosenberg et al., 2020).However, only a few studies have concentrated on these aspects during later periods in the southern Levant.Furthermore, much of the work undertaken on Levantine earthen architecture has stressed the technological aspects of production and the complexity of destruction rather than the anthropological and sociocultural developments involved in the construction process (Forget et al., 2015;Homsher, 2012;Ramírez et al., 2020).
Our research examines the relation of craftsmanship to the built and natural environments at Ashdod-Yam within the framework of earthen architectural construction.The study's main aim is to understand how people engage with public earthen architecture, especially during its construction and maintenance.This research focuses on an often-overlooked form of material culture, specifically that of mudbrick and other earthen building materials, to understand the mechanisms that influence craftsmanship (for discussion on earthen architecture craftsmanship see Fodde [2009] and Devolder and Lorenzon [2019]).Such mechanisms may include labor organization, diachronic/synchronic skill transfer between groups, and hierarchical or heterarchical control over the construction process (Lorenzon & Iacovou, 2019).To study these aspects, two lines of evidence were collected: (1) macroscopic data which provide information on building techniques and macro changes, and (2) microscopic records that relate craftsmen to raw source procurement, manufacturing techniques, and adaptability within the natural environment.

| Archaeological context
Ashdod-Yam (Ashdod by the Sea) is located on the coast of Israel, ca. 5 km northwest of Tel Ashdod, which served as the main settlement in the region from the Bronze Age until the late Iron Age (Dothan, 1971;Dothan & Freedman, 1967;Dothan & Porath, 1982, 1993) (Figure 1).The fate of Ashdod-Yam was always connected to the capital city of Ashdod, although the region's center of gravity shifted from Ashdod to Ashdod-Yam in the classical periods and, perhaps, already during Iron Age IIB-C (Fantalkin, 2014).
Ashdod-Yam is a large site, spanning at least 2 km from north to south and ca.1.5 km from east to west.It consists of several clearly definable areas, representing different periods of its history.A unique feature of the site is a large human-made enclosure at its southern end, which contains a massive fortification system and an associated acropolis, dating to Iron IIB-C (Figure 2).Additionally, a small site dating to the Late Bronze Age was uncovered ca. 1 km to the south of the enclosure (Nahshoni, 2013).
The Iron Age fortifications were partially excavated by Kaplan (1969).Ten cross-sections were dug along the edges of the rampart and glacis and exposed segments of the city wall.The fortification system consists of a 3.5-4.5-m-thickmudbrick wall, which served as a core for a large earthen rampart and an outer glacis laid on both sides (Kaplan, 1969, pp. 141-143).According to Kaplan, an additional segment of the wall, which would have fully enclosed the compound, stood at the western edge of the mound and was destroyed by erosion (Kaplan, 1953(Kaplan, , p. 32, 1969, p. 138), p. 138).Since Iron Age pottery was found during a survey of the site, beyond the ramparts, it is suspected that the fortified enclosure was part of a larger settlement, which may be buried under the accumulation of deposits and anthropic levels from the classical periods.The pottery retrieved from the rampart, outer glacis, and nearby locations was dated to Iron IIB (Kaplan, 1969, pp. 144-147).
Renewed excavations at the Iron Age compound of Ashdod-Yam were initiated in 2013 under the directorship of A. Fantalkin on behalf of the Institute of Archeology at Tel Aviv University.Four excavation seasons uncovered three main phases of occupation, which date to Iron IIB (Area B), Iron IIC (Areas C and D), and the Hellenistic period (Areas A, A1, and D) (Ashkenazi & Fantalkin, 2019;Fantalkin, 2014Fantalkin, , 2018;;Fantalkin et al., 2016) (Figure 2).According to Fantalkin's investigation, the fortification system was designed from the beginning in a crescent-shaped defensive form that extended over an area of more than 15 acres and featured a wide opening to the sea (contra Kaplan, 1969).Thus, the entire enclosure was constructed during the Iron IIB to protect a human-made harbor at Ashdod-Yam.
The construction of the site's Iron Age enclosure and its relation to the settlement at Tel Ashdod have been the subject of much debate over the past two decades (see summaries in Fantalkin [2001Fantalkin [ , 2014Fantalkin [ , 2018] ] and Itkin [2022]).While Kaplan (1969) argued that the enclosure was built on a local initiative as a response to a Neo-Assyrian threat, others interpret it as a Neo-Assyrian imperial enterprise (Finkelstein & Singer-Avitz, 2001;Na'aman, 2001).The question of the site's builders is essential to our understanding of the relations between the governing bodies and the craftsmen as well as the mechanisms that influenced craftsmanship at Ashdod-Yam.Thus, one of the major aims of the renewed excavations at the site is to understand the agency behind the establishment of the fortification system at Ashdod-Yam, be it on behalf of the Neo-Assyrian ruling regime or on behalf of the kingdom of Ashdod, which was later incorporated into the Assyrian realm.
In what follows, the architectural remains associated with the Iron Age sequence at Ashdod-Yam are discussed.These primarily include the fortification wall (Wall 2002, Area B) and a large public structure (Building 5175, Area D) uncovered on the acropolis (Figure 3).Additionally, the Hellenistic remains at the site are also mentioned (Areas A, A1, and D) for comparative purposes.

| The fortification wall (Area B)
During the renewed excavations of the Iron Age compound, one of Kaplan's sections was reopened and enlarged (section 2) (Kaplan, 1969, p. 139, fig.2).This section was originally created by Kaplan with mechanical tools and, at that time, the upper segment of the fortification line (including the mudbrick wall) was removed.During our excavations, the inner revetment was cleaned down to the foundation of the preserved section of the fortification wall (ca. 4 m wide and ca. 5 m high) and the massive outer glacis was partially excavated (Fantalkin, 2014) (Figure 4).The fortification wall (Wall 2002) was built directly on the natural kurkar (fossilized dune stone).
The top of the eastern end of the wall is well preserved and a series of rectangular mudbricks are visible.The bricks were tightly attached to one another, with very thin joints, while at the same time, some gaps caused by a defective arrangement of the layers were filled with mud material to ensure uniformity.The measurements of these bricks broadly correspond to the measurements of bricks reported by Kaplan from his Trench No. 1 (ca. 55 × 35 × 15 cm) (Kaplan, 1969, pp. 140-141).The central and western portions of the top of the designed with engineering in mind: the surfaces removed rainwater from the base of the wall once it was covered by the rampart, revetment, and glacis.Furthermore, the inner face of the wall was built at an inclination of some 15°.A similar angle was detected by Kaplan (1969, p. 140, fig. 3) in section 1. Engineering is also indicated by the mud plaster detected on the entire exterior of the mudbrick wall's upper courses.Several layers of mud plaster on the wall suggest that the construction process was rather long and demanded the wall's constant maintenance and replastering before its final covering.
The ceramic assemblage uncovered on the surfaces, from the fills between and from the favissa, as well as C 14 results from the organic materials discovered on the upper surface (Locus 2010), corroborate Kaplan's dating of the fortification system to the Iron IIB (eighth-midseventh century B.C.E.).

| Building 5175 (Area D)
Area D is located in the southwestern sector of the mound, on the acropolis.Excavations in Area D were carried out for two seasons, in 2017 and 2019, during which time the remains of a large public structure (Building 5175) were uncovered in the northeastern part of the area.This is an open courtyard structure, built of mudbrick walls and stone orthostats (Fantalkin, 2018).Although the building was only partially excavated, it appears to be a rectangular structure, bounded by four mudbrick walls (Walls 5061,5104,5162,and 5166) measuring ca.12.30 × 9.00 m.The western end of the structure as well as its northwestern and southeastern corners were almost completely destroyed due to later digging activities.The damaged portions of the structure were reconstructed in line with the remaining wall segments and the contours of several mudbricks visible in pits (Figure 5).The structure is divided into three spaces, two of which (Room 5172 and Room 5173) were clearly defined.The third unit (Square EA22), located just to the east of Room 5173, is divided into two narrow rectangular spaces which are delineated by mudbrick walls.These might have served as partition walls or as the foundation of a floor that was poorly preserved and is represented only by small segments throughout the unit (Loci 5053,5073,5091).
The entire structure was destroyed by fire; evidence consists of thick ash deposits on the floors of the building, charred organic materials (including a collapsed door lying at the entrance to Room 5172, Locus 5170), small finds, and restorable vessels.Among the latter were the remains of several almost complete hole-mouth storage jars, which were concentrated mainly on the plaster floor of Room 5172.Furthermore, over a 100 rim fragments of the same vessel type were uncovered throughout the structure, testifying to the building's administrative nature.Other elements within the ceramic assemblage, including several fragments of well-dated East Greek pottery, indicate an Iron IIC (late seventh-early sixth century B.C.E.) date for the destruction.Such a date is also corroborated by paleomagnetic results taken from the floors of the complex (Vaknin et al., 2022).
An additional surface (Locus 5112) was uncovered during the 2017 season ca.20 m southwest of Building 5175, in Area D West, with traces of destruction (e.g., burnt organic material, complete vessels).Although only a small portion of this surface was excavated, it was dated to Iron IIB (i.e., earlier than Building 5175), based on its lower elevation and the ceramic material on top of it.

| Hellenistic remains
Hellenistic remains discovered on the acropolis consist of a monumental stone-built citadel destroyed in the second half of the second century B.C.E.(Area A1).Numerous artifacts, including pottery, coins, weaponry, and weights, were unearthed in association with this building.Adjacent structures from the same period (Areas A and D) were constructed of mudbricks that either stood on stone foundations made of local beachrock or were laid directly on the sand.These structures were abandoned shortly before the destruction of the citadel and were finally destroyed by an earthquake.Several pieces of Hellenistic military equipment were unearthed in relation to the monumental structure in Area A1.It is possible that this building and the other auxiliary buildings in the Area should be viewed within the framework of Seleucid military activity.The Hellenistic occupation of the site perhaps represents a mercenary garrison stationed at Ashdod-Yam in the service of the empire.The monumental building, it seems, was destroyed in the late second century B.C.E. as a result of the Hasmonean expansion (Ashkenazi and Fantalkin, 2019;Fantalkin, 2014) (Figure 6).While numerous Hellenistic mudbricks were studied macroscopically, in this study, which focuses on Iron Age architecture, we decided to include only one Hellenistic sample in the microscopic analysis to showcase the diversity of earthen architecture in later periods.

| Geological background
Ashdod-Yam is situated in the littoral area south of the Lachish Stream.This area is characterized by eolian sediments that create low hills of red sandy soil (hamra), aeolianites ridges, coastline cliffs (kurkar), and sand dunes (Tsoar & Cohen-Zada, 2020, p. 1).Within this dynamic environment, Ashdod-Yam was built directly on geological deposits near the sea (Tsoar & Cohen-Zada, 2020).This location involved close human-environment interaction, considering the development of a purpose-built harbor and exploitation of the natural creek (Fantalkin, 2014) (Figure 7).
The geological formation of the site was produced during the Oligocene (33.9 ± 0.1 to 23.03 ± 0.05 Ma), when tectonic movements in Israel and low sea levels generated the deposition of clastic material in the Ashdod submarine canyon (Buchbinder et al., 2005).
These materials are mainly Oligocene sandstones rich in quartz and conglomerate, arranged in "lower clastics," "middle marls," and "upper clastics" (Buchbinder et al., 2005).This stratigraphy is the basis for understanding the frequency of microfossils present in each geological layer.The geomorphological evolution of the area includes sediments from the development of the kurkar ridge during the last 65 ka (Buchbinder et al., 2005), and movements in the dune systems, (e.g., the formation of the large dune of Ashdod) (Tsoar & Cohen-Zada, 2020).During Iron II, there was an episode of progressive intensification of aridity which, compared to more mesic conditions during the Late Bronze Age (Olsvig-Whittaker et al., 2015), correlates with the driest phases detected in the Levant (Cheng et al., 2015;Kaniewski et al., 2017Kaniewski et al., , 2010)).
Changes in the sea level and sand deposition on the coast and on the seabed have affected the location of the shoreline during different periods (Sivan et al., 2004;Toker et al., 2019Toker et al., , 2011)).These changes had a direct impact on the location of port facilities in relation to the coastline.Sea level in the eastern Mediterranean gradually rose from the Late Bronze until the Hellenistic period, at which point it reached the elevation that it is today (Sivan et al., 2001).During this period, the major finds from Ashdod-Yam are located in the southern sector of the site.Based on information obtained from ancient coastal wells, it has been suggested that following the Roman period sea level varied.
While a slightly higher sea level was recorded in the Byzantine  (Vunsh et al., 2018).The major finds from these periods at Ashdod-Yam are located in the northern sector of the site.
Following the excavations of the Iron Age compound, Kaplan (1969, p. 143, n. 8) observed a layer of deposits that characterize sand-choked estuaries or lake bottoms at the outer side of the glacis.British Mandate aerial photographs as well as modern LIDAR images suggest that the traces of a former branch of the Lachish River can be observed from both sides of the Iron Age compound.A major estuary of this branch seems to have extended towards the sea and was located just south of the enclosure.This would suggest the Late Bronze site was established to the south of the estuary, and the sediments laid during the construction of the Iron Age compound became runoff, infilling the estuary, and changing the site's topography.
F I G U R E 7 Geological map of the area (drawing by Maija Holappa after Sneh and Rosensaft [2004]).

| Materials
The earthen architecture was initially studied in situ to determine bricklaying techniques and the typologies of earthen building materials used.These include mud plaster, mortar, and floors.
Likewise, earthen building materials, specifically mudbricks, which comprise the majority of earthen building materials used in these constructions, were then analyzed at the macroscale to determine issues of standardization, manufacturing practices, and kinesthetic movements made by the brick makers.Geoarchaeological analyses were conducted to shed light on mudbrick composition and manufacturing choices implemented during production.An integrated geoarchaeological approach was designed to include chemical, mineralogical, and petrographic analyses to determine raw material sources, human-induced tempering, and the skill level of manufacturers.
The sampling rationale was to collect fragments ranging from the Iron IIB-C to comparative materials from earthen constructions of later periods.The selected samples came from public buildings (the fortification wall in Area B; Building 5175 in Area D), various earthen materials from walls (e.g., mud mortar, mudbrick, and mud plaster), and typological features (including both exterior and interior walls) that were uncovered during the renewed excavations (Figure 8).In addition, a Hellenistic mudbrick sample from a structure in Area A of the acropolis was also analyzed for comparative purposes.Sampling took place in 2019 and was constrained (to an extent) to wall preservation; 47 samples were taken from the aforementioned contexts at Ashdod-Yam.
Building 5175 is represented by AY-1-19, the fortification wall (Wall 2002) and its associated floors (Loci 2014(Loci , 2029(Loci , and 2031) by AY-20-41 and AY-50-52 (Table 1).An additional Iron Age sample (AY-53) was taken from a floor (Locus 5112) in Area D West.AY-54 is a soil sample from Area B. The Hellenistic sample (AY-42), belonging to one of the auxiliary buildings that surrounded the Hellenistic citadel, was not taken in situ.The geological study conducted by Buchbinder et al. (2005) functions as the baseline and comparison for the material collected, especially as the cores allow for a comprehensive and uncontaminated geological history of the area.

| Macroscopic analysis
Macroscopic observations were carried out in the field and included analysis of the construction techniques, such as bricklaying methods (from which samples were taken), and the identification of the typology of earthen building materials.The analysis also included individual sample characterization, such as the determination of size, color, and macroscopic fabric (Table 1).
The macroscopic fabric was identified as coarse or fine based on the quantity and consistency of inclusions bigger than 2 cm (for methodology see Devolder and Lorenzon [2019], Lorenzon and Iacovou [2019]).

| Microscopic analysis
To determine the mineralogical and geological characteristics of each sample, a series of integrated geochemical analyses were conducted.
The pXRF analysis was selected because of its nondestructive nature and the potential for statistical analyses based on the quantitative data collected, such as principal component analysis (PCA) and bivariate scattergram.These allowed for the identification of patterns of raw source procurement.The elemental analysis was carried out using a Bruker S1 TITAN handheld XRF used in the mining mode.This specific instrument has a 4 W, 50 kV tantalum anode X-ray tube and a graphene high-performance silicon drift detector, a resolution of 145 eV (Mo-Kα), covered by a 20 mm detector window.
LOI was employed to quantify the percentage of organic matter and calcium carbonate equivalent content (CaCO 3 ) (Davies, 1974;Heiri et al., 2001;Love, 2017;Stein, 1984).The samples were initially predried at 105°C and weighed.Then they were placed in the furnace for 4 h at 550°C, cooled, and weighed again to measure organic loss.Subsequently, the same samples were inserted in the furnace for an additional 2 h at 950°C, to remove carbon dioxide (CO 2 ).After the cooling period, the samples were weighed to measure the percentage of CaCO 3, which was arrived at by dividing the CO 2 content by 0.44 (Love, 2013;Stein, 1984) (Supporting Information: Table S2).
A limited number of subsamples were also examined with an environmental SEM (ESEM-FEI Quanta 200FEG) in high vacuum mode and an Everhart-Thornley secondary electron detector at the Faculty of Engineering at Tel Aviv University.The SEM-EDS was calibrated with standard samples from the manufacturer and provided measurements with a first approximation error of 1% (Ashkenazi & Fantalkin, 2019).Bullock, 1985;Nicosia & Stoops, 2017;Quinn, 2013;Whitbread, 1989Whitbread, , 1995)).

| Building techniques
The Plastic earthen material (PEM), which is an earthen material similar to mud mortar, but created with the same composition as mudbrick can easily be employed as fillers and leveling layers (see Devolder & Lorenzon, 2019).
PEM tends to average 1-2 cm but can measure up to 4 cm thick when used as a leveling layer.Likewise, PEM is used inside the walls of Building 5175 in a thick 5-10 cm filler between the opposing headers and stretchers.This was initiated because the wall was wider than the brick length, and thus irregular bricklaying with the PEM provided a cohesive link between the components within each course.
Macroscopic observation of the mudbricks at Ashdod-Yam shows a prominent and homogeneous fine macrofabric (i.e., macroscopic fabric composition) during the Iron IIB-C.These mudbricks are usually reddish in color (7.5YR 5/3, 5/4, and 5/6) and were produced with a screened sediment (pebbles max.20 mm), showcased limited quantities of vegetal temper, and contained minimal inclusions.
A portion of the Hellenistic mudbrick wall (Wall 117) was discovered on top of the Iron Age fortifications in Area A. The wall was preserved to a height of five courses, with bricks laid in a running bond technique.The wall was erected directly on sand, without stone foundations (Figure 11a).The mudbricks measured approximately 39 × 39 × 10 cm and presented a coarse macrofabric with many broken seashells and small pottery sherds, as opposed to the fine macrofabric of the Iron Age mudbricks (Figure 11b).The collapsed Hellenistic mudbricks analyzed are light in color (generally 10YR 6/3).The mud mortar and PEM intermixed between the courses of the mudbricks present a composition similar to the bricks, while the mud plaster samples show a variety of colors but limited vegetal inclusions (Table 1).The latter is quite atypical for plaster material.No adjoining floors or surfaces were detected in the field except for the remains of an impressive collapse adjacent to the wall (Locus 118).The Hellenistic sample (AY-42) was taken from this location.

| Geochemical analysis
The PCA focused on 12 indicative variables (Al 2 O 3 , SiO 2 , K 2 O, CaO, Ti, Fe, Mn, Cu, Zn, Rb, Zr, Sr), determined on the basis of their reliability in pXRF analysis and limited error margin (Goodale et al., 2011;Hunt & Speakman, 2015).The two main principal components (PC1 and PC2) explain 72.9% of the total variability.The resulting PCA graph shows a clear overlap in raw material sources among the Ashdod-Yam samples, including among different types of earthen building materials from Areas A, B, and D (Figure 12).This reflects the use of common sources of raw material procurement over the Iron IIB-C, and possibly during the Hellenistic period.
T A B L E 1 (Continued)  of the mudbricks, and in this sample group, there is also a preference for phyllosilicates, which were likely used to make the material more plastic in nature (Figure 13).
LOI was used to quantify the percentage of organic temper and CaCO 3 in individual samples, which in turn allows for an assessment of quality and quantity in human-induced tempering.The percentage clearly shows that there is no variation in organic temper between different typologies of earthen architecture across the analyzed structures and soil (Figure 14a).As the percentage of vegetal temper within mudbrick recipes is quite apparent, the lack of distinction and the relatively small amount of organic matter present seems to indicate that vegetal temper was not one of the main human-induced tempers of earthen building materials in the Ashdod region but may have occurred naturally in the soil used in the manufacturing (Figure 14b).There could be multiple reasons for this, such as climate conditions and/or the preference of other tempers in the mudbrick recipes.preparation.The characteristics of the groundmass and the aplastic inclusions point to one common raw material catchment area, compatible with the sedimentary deposits surrounding the site.
Ashdod-Yam stands out for its high homogeneity after comparing a large number of samples from different structures and phases.These results follow the same trend as the chemical data.| 53 added temper is doubtful as the voids are not numerous and the identification of plant tissue remains in some samples (Figure 18g) could occur naturally (AY-22, AY-24, AY-32, AY-33, and AY-37).The five samples in which these remains have been identified are from mudbricks in the fortification wall (Wall 2002).This could point to the raw source material difference between the wall and the earthen materials from the acropolis (Area D).However, according to the chemical results, these same samples fit with the organic percentage identified in the assemblage, suggesting their natural origin.Lastly, another interesting feature is the limited but varied presence of microfossils, such as planktonic and benthic foraminifera (Figure 18c,i) embedded in the matrix as well as very small shell fragments (e.g., in AY-14, AY-26, AY-29, AY-35, and in the Hellenistic sample AY-42).The Hellenistic sample (AY-42) presents some isolated shell fragments (Figure 18f) that could point to the existence of a subfabric in which shells may have been added as tempering.
This is corroborated by macroscopic observations (Figure 11b).F I G U R E 16 Bivariate scatterplot of CaCO 3 equivalent and organic matter percentage illustrating typology of earthen building materials (EBM) and original location.(Credit: Marta Lorenzon).(Singer, 2007).This composition matches the natural geological deposits surrounding Ashdod-Yam.Specifically, the Holocene formations, characterized by eolian sediments, make up the sand dunes, a key feature of the current landscape (Tsoar & Cohen-Zada, 2020).
However, there are also abundant traces of calcrete or caliche, which indicate the presence of elements from the kurkar formations where quartzitic sand occurs together with carbonate-cemented sands.
Such characteristics are found in deposits to the south and northeast of the site (Figure 7).The similarity between Ashdod-Yam's earthen materials and the Oligocene sand levels identified in the cores of the Ashdod canyon further confirm the local provenance of raw materials (Buchbinder et al., 2005, p. 76, fig. 15c).Thus, it can be concluded that, during Iron IIB-C, the raw source procurement occurred within a 2 km radius from the center of the settlement.The Hellenistic sample presented a similar geochemical pattern, but it is impossible to outline raw source procurement areas with the same precision due to the limited number of Hellenistic samples.The single analyzed sample, however, seems to attest to the implementation of different adaptive techniques to the environment.This is indicated by the addition of shell fragments to the brick mixture as a clear tempering (Figure 19), which is quite distinctive from the Iron IIB-C production.
The choice of similar catchment areas during the Iron IIB-C may have been opportunistic, as craftsmen took advantage of nearby soils.Craftsmen may also have used the soil dug out during the construction of the fortification wall, the structures on the acropolis, and other urban architecture at the site.However, basic energetics calculations indicate that the quantity of soil would not have been sufficient to produce the mudbricks necessary for the fortification wall and the structures on the acropolis and thus additional soil was required.The quantity of raw material sources needed for such monumental work presumably had a negative impact on the local landscape.Furthermore, since the results indicate the use of similar soils in Iron IIB-C construction efforts, there may have been a known exploitation area that was located further away from the site.This area functioned as a sediment quarry and may have been revisited later, as the data from the Hellenistic mudbricks seems to suggest.
The preparation of materials probably occurred nearby so that they could easily be transported to the settlement.

| Technological choices in production
While chemical, mineralogical, and petrographic analyses indicate that local soils were used for earthen building production, they also suggest that the craftsmen augmented them.After soil selection, mudbrick manufacture involved various technological choices, such as the brick-making recipe, the finish, and the mold size.
In terms of mudbrick production, the first step is to procure the necessary raw materials in addition to the sediment; this includes water and vegetal temper.In the case of Ashdod-Yam, the latter was not added on purpose, but the small quantity present in the mudbricks likely occurred in the soils naturally.This is quite dissimilar for instance to mudbrick production in the Late Bronze Age Tel Lachish in which the use of chaff as vegetal temper was well attested (Goldberg, 1979) The lack of extensive vegetal temper in the Ashdod-Yam recipe during Iron IIB-C could be viewed in light of ancient environmental reconstructions that show that Iron IIB was a drier period (Langgut et al., 2015;Rambeau & Black, 2011).Furthermore, anthropogenic processes, such as the destructive impact of warfare on the availability of resources, should also be taken into account (Itkin, 2022).The selection of soils rich in natural inclusions, such as the quartz and conglomerates identified in the Ashdod-Yam mudbrick recipe, may have compensated for this impact.This may be indicated by both ESEM-EDS results and petrographic analysis, as the bimodal distribution is dominated by fine and medium coarse sand associated with coastal eolian sediments.Similar high sand ratios with limited amounts of vegetal temper were recognized at Amarna, Egypt (French, 1984) and Palaepaphos, Cyprus (Lorenzon & Iacovou, 2019).
The homogeneity of the Ashdod-Yam samples indicates not only a single one raw source procurement site but also the implementation of one standard mudbrick recipe for Iron IIB-C public construction works.This also seems to point to a short production period in which standardized production and construction, as apparent in brick size and bricklaying, were employed in the creation of public structures.
On the other hand, the consistently high presence of sand and similar soil sources could be used to suggest the lack of a highly skilled workforce in mudbrick production.The Ashdod-Yam recipe was likely common knowledge, having been passed down through generations among specialists and nonspecialists alike.Thus, it represents, to a certain degree, a skilled and semi-skilled workforce.Furthermore, the amount of labor needed for the construction of the fortification wall and the structures on the acropolis required all groups, including the skilled, semi-skilled, and unskilled workforce, to aid in manufacturing Similar mudbrick dimensions (ca.54 × 36 × 12, 60 × 40 × 15 cm) and bricklaying techniques were observed in the fortifications system and dwellings at Tel Ashdod during the Iron IIB (Bachi & Ben-Dov, 1971, p. 89;Dothan & Porath, 1982, pp. 19, 29).This is in contrast to the

F
I G U R E 1 Map of the southern Levant with the location of Ashdod-Yam and other sites noted in the text.(Credit: Itamar Ben-Ezra.)F I G U R E 2 View to the south showing the excavation areas at Ashdod-Yam, Area A in which the Hellenistic structures were found, Area B includes part of the fortification wall, and Area D encompasses the acropolis.(Credit: Pascal Partouche, Skyview Photography.)F I G U R E 3 Map of excavation areas in Ashdod Yam.(Credit: Slava Pirskiy.) exposed wall were destroyed by erosion.Within the upper part of the outer glacis, close to the wall, a favissa with discarded cultic objects was discovered.The retaining rampart on the inner side of the wall consisted of several layers of mud and crushed kurkar.Here, a clay surface (Loci 2010, 2029) was discovered abutting the wall.Four mudbrick courses separated it from a lower surface, which lay at the wall's foundation (Locus 2031).The upper surface (Locus 2010) extends some 3.5 m southwest of the wall, sloping down from the base of the wall toward the terminus.It seems that these two surfaces were created during the construction of the fortification system.They were constructed to allow access for the workers constructing the wall.The system was F I G U R E 4 Section of the fortification wall, Iron Age IIB with an explanation of the glacis structure; (a) section of the fortification wall, revetment and glacis; (b) top view of the fortification wall and adjacent surfaces, and (c) the inner face of the fortification wall, a view to the east.(Credits: Pascal Partouche, Skyview Photography, and Philip Sapirstein.)

F
I G U R E 5 Plan of Area D with Iron Age and Hellenistic remains.(Credit: Slava Pirskiy and Eli Itkin.)F I G U R E 6 Hellenistic remains in Area A, showing the mudbricks still in situ and the mudbricks that have collapsed from the upper portion of the wall.(Credit: Pascal Partouche, Skyview Photography.)era, it became lower in the early Islamic period.During the Fatimid/Ayyubid and Crusader periods (11th-13th centuries C.E.), the relative sea level (RSL) reached its lowest elevations (about −0.5 m).The RSL rose again in the 18th-19th centuries C.E.

F
I G U R E 8 Samples' location in (a) Area B and (b) Area D. (Credit: Slava Pirskiy and Eli Itkin.)LORENZON ET AL. | 45 T A B L E 1 List of samples and macroanalysis contexts.a macroscopic analysis of Ashdod-Yam's mudbricks indicates that the bricks from the fortification wall (55/50 × 35 × 11/13 cm) share a standard size with those from Building 5175 on the acropolis (50 × 35 × 12 cm; with height variations at 11-13 cm and length variations at 50-55 cm).The bricks presented indications of lateral striation that reveal the use of a wooden mold during manufacture (Figure 9).The fortification wall (Wall 2002) presents consistent English bond bricklaying techniques, that is, an alternation of headers and stretchers.The mudbrick joints were filled with mud mortar measuring 2-4 cm.The wall featured a mud-plaster finish with at least five plastering layers visible in the field, indicating multiple replastering events (Figure 10).In Building 5175, the use of external and partitioning mudbrick walls indicates a cohesive building plan.The mudbrick courses in the building were also laid with English bond techniques with visible joints created by mud mortar.No traces of mud plaster were recovered from this area.
quantitative analysis as indicated by confidence ellipses (95% confidence) shows two manufacturing events that separate Iron IIB (Area B) and IIC (Area D), although some similarities are clearly visible in the overlapping areas.The results of the PCA thus highlight (1) similar soil procurement sources for the construction of the acropolis (Iron IIC) and the fortification wall in Iron IIB, and likely for the Hellenistic structure as suggested by results from sample AY-42, and (2) two separate manufacturing events, as indicated by the ellipses, which clearly reveal human-induced adjustments of local raw sources to produce mudbricks.The triangular scattergram SiO 2 -Al 2 O 3 -CaO illustrates the use of non-calcareous clays in mudbrick production at Ashdod-Yam.The quartzes indicate a reliable presence of sandy sediments in the matrix When visualizing the percentage of CaCO 3 , the results present a different pattern (Figure 15) in the form of two groupings.One group shows a lower percentage of CaCO 3 (average < 3%) in mud mortar, mud plaster, PEM, and soil, while mudbricks present more variability having an equivalent of CaCO 3 ranging between 2% and 4%.Earthen floors show a high percentage of CaCO 3 (average 5%-10%).The results match macroscopic observations that these Iron Age earthen materials present minimal carbonate inclusions, and that the builder's floor (Loci 2010, 2029) features a silty, compact appearance, often characterized by a thin lime plaster finish or was often embedded with shell fragments.The slightly elevated carbonate percentage in the soil sample may reflect lime plasters and wood ashes that decayed around the structure, which in turn affected soil composition.The bivariate scattergram, representing the percentage of organic temper and CaCO 3 , highlights the similarities between the Iron IIB-C materials (Figure 16).Most of the mudbrick samples from the acropolis (Building 5175) have a scant organic percentage.This indicates that mudbrick manufacturers may have adapted their recipes based on the available soil sources.While the minimal amount of vegetal temper may produce bricks that have less compression and tensile strength, sandy soils can still be molded into functional mudbricks (French, 1984; Lorenzon & Iacovou, 2019).AY-53, a possible Iron IIB floor in Area D west (Locus 5112) appears as an outlier, presenting a higher percentage of CaCO 3 than the other samples.This may have been affected by postdepositional factors, such as the progressive deterioration and decay of the lime plaster decorating the walls.The SEM-EDS analysis indicates that Ashdod-Yam mudbricks have a cohesive structure, with small pores.The micrographs show a particle size distribution in the Iron Age samples that has a high sand percentage in comparison to fine fraction, that is, clay and silt (Figure 17).EDS analysis further reinforces the conclusion that sand was either high in the selected sediment or was added as an example of human-induced tempering.The presence of medium and coarse grains corresponds to eolian sand, likely the local soils of the coastal area.Only one sample (AY-41) shows clear evidence of voids that can be associated with chaff, highlighting the limited use of vegetal temper.These data are also supported by the results of the petrographic study, in which voids associated with vegetal temper are uncommon.The Hellenistic sample (AY-42) shows a similar geochemical composition compared to Iron Age samples, but a greater fine fraction.Illite flakes are visible in the micrographs, constituting the main clay mineral recognized in the sample.4.2.2 | Petrographic analysis Thin-section analysis under the polarizing microscope revealed a homogeneous assemblage in geological composition and technological F I G U R E 9 Bricks in situ in Area B with red dotted lines indicating the lateral striations.(Credit: Owen Chesnut and Marta Lorenzon).

Fabric 1 :
Eolian quartz sand, calcrete, vegetal temper, and microfossils Fabric 1 is highly homogeneous in terms of petrographic composition and distribution parameters (c:f:v 50:40:10).It represents 45 samples from Ashdod-Yam, including from mudbricks and floors.The main characteristic of this fabric is a bimodal distribution of inclusions, with the widespread presence of quartz and the common occurrence of cemented conglomerate clasts, specifically calcrete.The groundmass has a brownish-orange color and the aplastic inclusion represents very fine (>0.05 mm) to coarse (<1 mm) sands that follow a poorly sorted distribution.The space distribution is measured within the single to double-spaced frame.More abundant inclusions are represented by monocrystalline quartz of heterogeneous size (≤1 mm), with a sphericity grade from subangular to rounded (the majority present a rounded shape) F I G U R E 10 (a) Bricklaying of the fortification wall in Area B with inserts of the central area.(b) The image showcases the mud plaster covering the mudbricks and the glacis.(c) Highlight of the mud floor and lower levels of Wall 2002.(Credit: Slava Pirsky, Pavel Shrago and Owen Chesnut).(Figure 18a-c).Calcrete inclusions are subrounded and can reach large sizes (≤4.5 mm).Other less frequent inclusions are calcite, clay pellets, and alkali feldspars (both plagioclase and microcline), with very rare pyroxenes and tourmaline.The recognition of calcrete in this group (Figure 18d,e) is particularly interesting, as this has been observed in at least 17 of the 45 samples (AY-2, AY-3, AY-4, AY-10, AY-11, AY-13, AY-14, AY-16, AY-17, AY-21, AY-29, AY-30, AY-35, AY-39, AY-52, AY-53, AY-54).This could indicate the presence of two subgroups, according to the presence or absence of this aplastic inclusion.However, the distribution of these individuals in the chemical subgroups and homogeneity led to the conclusion that its random presence is a natural feature.Fabric 1 is also characterized by long stretched voids which appear in the form of channels and vughs.Their presence is irregular and does not have an oriented distribution.However, some voids are clearly associated with vegetal temper.The use of this material as an F I G U R E 11 (a) Bricklaying and Hellenistic wall in Area A. (b) Close-up of the Hellenistic mudbrick, AY 42, showing the characteristic macrofabric embedded with shell fragments.(Credit: Pavel Shrago and Sasha Flit).F I G U R E 12 Results of PCA considering 12 variables (Al 2 O 3 , SiO 2 , K 2 O, CaO, Ti, Fe, Mn, Cu, Zn, Rb, Zr, Sr) for the 47 mudbricks samples collected in Ashdod-Yam.(Credit: Marta Lorenzon.)PCA, principal component analysis.F I G U R E 13 Triangular scattergram SiO 2 -Al 2 O 3 -CaO.(Credit: Marta Lorenzon).LORENZON ET AL.

|
Raw sources, catchment areas, and environmental impact Macroscopic and microscopic analyses indicate that Ashdod-Yam's mudbricks present a high homogeneity over time.The results point to small differentiations in production during Iron IIB-C, likely due to craftsmen adapting to local resources.But overall, the data reveal a general commonality in practice not only in the manufacture of mudbricks but also in other earthen building materials (e.g., mud plaster, mud mortar, etc.).This then supports the hypothesis of a high degree of standardization within public constructions at Ashdod-Yam.This understanding is essential in assessing the impact of monumental public construction on resource procurement and human-environment interactions.The samples from Building 5175, the fortification wall (Wall 2002), and Area D West (Locus 5112) are F I G U R E 14 Percentage of organic matter in different (a) earthen building materials (EBM) and (b) areas.(Credit: Marta Lorenzon).highly representative (n = 47) and demonstrate that raw materials were locally collected for all earthen building materials.The samples reveal the endemic presence of eolian sandy sediments-mainly characterized by round monocrystalline quartz and phyllosilicates-and the limited amount of clay, identified as illite and smectite as characteristic of the local hamra soils F I G U R E 15 Percentage of CaCO 3 in different earthen building materials and soil.(Credit: Marta Lorenzon).

F
I G U R E 18 Photomicrographs of Fabric 1 (Ashdod Yam): mudbricks from the acropolis ([a] AY-10; [b] AY-14; [c] AY-15; [d] AY-2, XPL only), and mudbricks from the wall ([e] AY-29; [f] AY-42; [g] AY-32); floors from the wall structure ([h] AY-51; [i] AY-53).Plane polarized light and crossed polars.(Credit: Benjamín Cutillas-Victoria.)and construction(Devolder & Lorenzon, 2019;Marchand, 2009).This construction process, which involved various social segments of society, provides evidence of standardization and administrative oversight, at least regarding public buildings.Mudbrick micromorphology points to specific kinesthetic actions at play during manufacture; once the raw sources were collected, they were mixed and kneaded for multiple days.Usually, this was done by stomping on the mud mixture and turning it by hand (Lorenzon personal observation at the site of Tell Timai, Egypt, andMarchand, 2015).In the case of the Hellenistic production, an additional step in the chaîne opératoire included the incorporation of crushed seashells (as in the case of AY-42).Petrographic and micromorphological analyses provide specific evidence for a rotational feature with semi-circular arrangements of subrounded clasts (Figure20).This suggests repetitive kinesthetic movements of the craftsman during manufacture.Ethnographic observations indicate that after daily kneading, the mudbrick makers pressed the mixture into the mold, first by pushing it down with the palm of the hand and folding it on itself, and then by pressing it into the mold (Lorenzon personal observation at Tell Timai).This latter movement is a key step in mudbrick production and causes recognizable rotatory deformation structures visible in the petrographic micrographs.In addition to the geoarchaeological data, over 100 measurements of mudbricks that were taken at Ashdod-Yam indicate the use of the same wooden mold size.The rigorous size standardization of Iron IIB-C mudbricks indicates that all manufacturing teams employed similar tools and a shared recipe.This supports the idea that the production and labor organization at Ashdod-Yam was coordinated and overseen by a governing power.5.3 | Designing earthen architecture under political controlThe results of the analyses from both the fortification wall and structures on the acropolis highlight their homogeneity.It is argued that the acquisition of similar raw materials, use of standardized mold size and consistency in bricklaying techniques point to skill transfer during Iron IIB-C, on a local scale.There are two key elements that support this argument: (1) the presence of a singular recipe, which had minimal human-induced tempering; the Hellenistic manufacturing practices differ, as the macrofabric of the sample shows intensive human-induced tempering; (2) the short time span of construction of Ashdod-Yam's Iron Age monumental public architecture (i.e., the fortification wall).
Neo-Assyrian governor's residence uncovered at Ashdod ad-Halom, where different mudbrick sizes were recorded (38 × 38 × 10 cm) (Kogan-Zehavi, 2008, p. 1573).The Mediterranean coastline of the southern part of Israel has almost no natural sheltered location for building and F I G U R E 19 Comparison between Iron Age and Hellenistic mudbrick macrofabrics.(Photo: Benjamín Cutillas-Victoria.) AY-54 is a soil sample from Area B, the only area in which we reached the virgin level and could sample uncontaminated soil.Geochemical and LOI results of the earthen building material and soil samples.
a T A B L E 2