North-Western Negev Settlements

A decrease in population in other parts of the Empire would create a decrease in the population of the arid regions, as people would move to recently vacated lands, which were safer and more fertile. Thus, such arid regions are considered good indicators of societal change, and so great emphasis was placed on their excavations and subsequent analysis.^[36] One such area where a supposed decline would be visible is the settlements of the north-western Negev desert (Fig. 1).

Fig. 1:

Map of the north-western Negev and its vicinity with the city of Elusa and the settlements of Nessana and Shivta indicated. The map was created by using ArcGIS Pro 2.9.3.

The site at the crux of the theory regarding the decline of this region is Elusa. Originally a Nabatean waystation that was established in the 3^rd c. BCE,^[37] it evolved into the main polis of the Roman Negev a few centuries later. The city prospered during the 4^th to the 6^th c. CE from its extensive agriculture and viticulture industry.^[38] The polis was surrounded by a number of large garbage mounds, and the surveying team conjectured that an abandonment of trash mounds could be an indicator of the breakdown of municipal facilities, societal collapse, and an emigration of people.^[39] Thus, a survey and an examination of the garbage mounds surrounding the city of Elusa was conducted, and their end use was dated through the use of radiocarbon dating and ceramic typology. According to them, the polis and the surrounding settlements started to deteriorate in the mid-6^th c. CE, and so this decline was attributed to the Justinianic plague and the LALIA.^[40] However, the methods of dating used by the NEGEVBYZ project were problematic.

Regarding the radiocarbon dating, only eight botanical samples were dated from an area of more than thirteen hectares. In addition, five of these samples originated from one excavation trench 1.3 m deep.^[41] Moreover, the samples were taken from areas which did not have an extensive representation of pottery sherds designated to later dates, even though there were several areas on the trash mounds with large quantities of such sherds. The radiocarbon dating of organic samples from such areas would have been best suited for understanding the last use of the trash mounds, but this did not occur.^[42] Furthermore, the botanical sample from probe 22 was claimed to indicate that the end use of the mound happened in the mid-6^th c. CE. However, the full date range of this sample is 425 to 580 CE.^[43] The fact that the dating of the botanical sample was rounded down, and the full range of dates was not provided in the article, is not the only problem with this dating.^[44]

When inputting the radiocarbon data into the newly updated OxCal (a program used to calibrate radiocarbon dates),^[45] a version that was unavailable to the NEGEVBYZ project in 2019, the dates for all the samples were pushed forward, with the new dating for probe 22 being 435 to 597 CE at 95.4 % probability.^[46] A further analysis of the dates using the mathematical modelling function in the OxCal program was also implemented.^[47] In this case, the raw dates of the eight samples were entered into a single-phase model to try and calculate the possible end date of the garbage mounds, something the original publication neglected to do. According to the model, the end use of the mound fell within a large range of 476 to 681 CE at 95.4 % probability, or 515 to 599 CE at 68.3 % probability.^[48] It is important to note that the dating and modelling are so imprecise due to the severe lack of dates for such a large area. Moreover, the samples were taken from places where there were no recent ceramic sherds, unlike other areas on the mound where there were such recent sherds dated to the 7^th c. CE. Consequently, it is safe to argue that the radiocarbon dates, be they modelled or unmodelled, and especially when considering their limited number, do not prove that the garbage mounds ceased to be used in 550 CE.

The second dating method used to complement and strengthen the radiocarbon dating of the garbage mounds was ceramic typology, which was entirely based on a 1995 article by Majcherek describing the dating of Gaza jars.^[49] In Elusa, vast numbers of Gaza jar sherds were found, especially from those belonging to Majcherek’s Type 2 and Type 3. According to Majcherek, Type 2 represents the 4^th to the end of the 5^th c. CE, while Type 3 was in production from the end of the 5^th until the beginning of the 7^th c. CE, with a peak in its production after 535 CE.^[50] More updated typologies of this jar, which is also called the Late Roman Amphora 4 (LRA 4), have slightly amended Majcherek’s dating. Pieri divided Majcherek’s Type 2 into two separate typologies, one dated to the 4^th and 5^th c. CE called LRA 4A2,^[51] and the other dated to the second third of the 5^th c. to the first half of the 6^th c. CE, known as LRA 4B1.^[52] Majcherek’s Type 3, reclassified as LRA 4B2, was re-dated to the beginning of the second half of the 6^th c. to the 7^th c. CE.^[53] Sazanov attempted to further improve on this classification and categorised the vessels into smaller sub-categories.^[54]

Conversely, the team at Elusa incorrectly assigned Type 3 to Levantine archaeology’s Middle Byzantine period (450 to 550 CE).^[55] Additionally, as stated by Majcherek, vessels are often used long after their production has ceased,^[56] as usable jars and vessels were not discarded simply because they were no longer produced. Therefore, it is safe to assume that Type 3 was still used extensively in the first quarter of the 7^th c. CE and later, until the Islamic conquest of 636, and possibly even a few decades after. Hence, Type 3 represents the norm during the end of the Middle and all of the Late Byzantine archaeological periods in the Levant. Accordingly, the ceramic record suggests a longer and prosperous habitation of the region, rather than a deterioration in the mid-6^th c. CE.^[57]

This is supported by the large concentration of pottery sherds from the Late Byzantine i.e., from post 550 CE,^[58] which were found in squares VII–IX of the survey that was conducted on the garbage mounds of Elusa.^[59] The ceramic concentration in these squares was larger than any other found in the other survey squares, and was highly unusual. Thus, these squares may have been the last areas where garbage disposal was practiced, and it strongly suggests that this continued well into the 7^th c. CE,^[60] and indicates that the cessation in municipal services occurred only in the 7^th c. CE.^[61]

This lends weight to discoveries that show a shift in the ratio of pottery typologies in the Negev, showing a drastic increase in the number of sherds, and especially the percentage of bag-shaped jars, compared to a decrease in the percentage of Gaza jar sherds, especially in the assemblages of Shivta (Sobota) and Nessana. According to the NEGEVBYZ project, Gaza jars were used only for trade, and were especially built to be transported on camels and ships, while bag-shaped jars were used for storage.^[62] This, alongside a change in the ratio between grain seeds and grape pips in different contexts, had brought them to say that the decrease in Gaza jar sherds correlates with a change in agricultural produce.^[63] However, bag-shaped jars had a dual purpose and were also used for trade.^[64] Their sherds can be found in large quantities throughout the Empire, similar to the dissemination of Gaza jars (Fig. 2).^[65]

Fig. 2:

A distribution map of LRA 4 amphorae (in red, top) and LRA 5 amphorae (in blue, bottom). The data for this distribution map was taken from: Lloyd 1977; Kingsley 1999; Pieri 2005.

Moreover, there is a stark increase in the total number of sherds found after 550 CE, possibly indicating an increase in the industrial capacity and prosperity of the region. This is especially noticeable in Area A in Nessana, which was dated to 550 to 700 CE, where a total of 16,148 sherds were found, a larger number than all the other areas and contexts from all the sites combined.^[66] Such a number can be used to support a population peak in the early 7^th c. CE, or it can be characterised as untrustworthy and possibly anomalous.

In terms of yield change, grape pips were found in higher numbers from contexts dated to after the mid-6^th c. CE than in previous periods. Thus, it is definitely not a reliable indication of a decline in the viticulture industry, nor sufficient evidence for a decline of the region.^[67] It is also worth noting that the number of botanical remains, such as pips and seeds, is small, so changes in their ratio could be incidental. Furthermore, most of the botanical finds were dated to post-550 CE,^[68] which can indicate that the peak in agricultural production occurred in the late 6^th or the 7^th c. CE, or possibly accidental. Additionally, the dating of the data from Shivta, Elusa and the other Negev sites, needs to be amended. As elaborated previously, what was dated to 450 to 550 CE should be at least redated to 475 to 636 CE or 550–650 CE, if not later. During this period, the grape pip percentage of the cereal and grape assemblages reported in the middens of Shivta and Elusa sites was 42 % and 43 %, respectively, making it the phase with the largest percentage of grape pips. As a result, it implies the highest period of viticulture output was at least until, and in the first few decades following, the Islamic conquest.

The shift in the pottery and botanical assemblage ratios can be explained by the Roman-Persian wars, which severed trade routes, curtailed the sale of commodities, and constricted the market. It may also have compelled a conversion from Gaza jars to bag-shaped jars in the hope that the produce could be stored, and the conflict would finish shortly so that the product could be sold. Bag-shaped jars were chosen because they were easier to store than Gaza jars but could still be used for transportation. The Islamic Conquest prolonged and exacerbated the degradation of trade and markets, potentially driving many farmers to bankruptcy. It is plausible to infer that conflicts and trade restrictions also encouraged many farmers to switch from growing grapes to growing grain, as grain, despite being less profitable, was easier to store for longer durations. This transition may have caused a decrease in the area’s prosperity, as well as the deterioration and eventual abandonment of the sites in the following decades.

Another option is that the Gaza jars were less prevalent in the Negev after the wars reached the area, as the vessels arrived, most likely filled, from the coastal region.^[69] As the Negev settlements’ ability to import products decreased, so did the quantity of imported vessels and the ability to reuse them for export. Consequently, the Negev villages began to increase the local manufacturing of bag-shaped jars.^[70] This theory could be tested further by examining the colour of the sherds and determining their origins.

The decline of the area drove traders and farmers into bankruptcy. Since pigeons were raised for their faeces, which was used as fertilizer, the abandonment of pigeon towers in the Negev could be an indication for that decline.^[71] In a recent article of the NEGEVBYZ project, pigeon bones from several towers were radiocarbon dated, providing a range of dates in the second half of the 6^th c. and the beginning of the 7^th c. CE, suggesting that the decline occurred in the 7^th c. CE.^[72] Like the case of Elusa, a meagre number of tests were done, and so these dates are not enough to conclusively prove and provide exact dates. For example, only three bones were examined in Horvat Saadon, with date ranges of 430–585 CE, 430–600 CE and 440–640 CE at 95.4 % probability.^[73] According to these results, it is possible that the use of these facilities in Horvat Saadon continued into the late 6^th c. and even the 7^th c. CE, such as in other places. The finds in other towers in Shivta and Beer Sheba provided later dates, which can indicate a decline in the 7^th c. CE, whose causes are most likely connected to the general deterioration of commerce and wealth due to the Persian war and the Islamic conquest, as the NEGEBYZ project pointed out itself.^[74]

The presented data, including a reinterpretation of the radiocarbon dates and the ceramic typology of the north-western Negev settlements, indicate that the decline, economic deterioration, and partial abandonment of the region most probably occurred from the 7^th c. CE. This is corroborated by the presence of certain ceramic typologies originating from the east, and Judaea/Palaestina in particular, throughout the Mediterranean. These vessels, namely Gaza jars (LRA 4) and bag-shaped jars (LRA 5), were widely found in multiple sites, such as Marseille, Naples, Carthage, eastern Spain and Alexandria, from contexts dating to the end of the 6^th and the first decades of the 7^th c. CE.^[75] According to the archaeological record, there was a collapse in the presence of these vessels from the early to the mid-7^th c. CE. This implies that trade in the eastern Roman Empire reached a height in the late 6^th and beginning of the 7^th c. CE. The further implications are that both the settlements in Judaea/Palaestina, as well as in the rest of the Roman east, were at the pinnacle of their industrial capabilities, and provided commodities to the entire Mediterranean basin during this period.

The communities of the north-western Negev also relied on the export and contents of these vessels (LRA 4 and LRA 5), as is evident by the number of their sherds found throughout the region, especially in archaeological contexts dated to post-550 CE. This, as well as other evidence, suggests there was a peak in the prosperity and welfare in the sites of the Negev in the late 6^th and early 7^th c. CE. This is substantiated by the corpus of papyri and inscriptions from Nessana, most of which are dated to before the Islamic conquest. Their contents show that Nessana and its surroundings reached a maximum in size and population in the late 6^th and early 7^th c. CE.^[76] Other evidence for the region’s later prosperity comes from the dedicatory inscriptions for buildings and new mosaics from within Negev settlements, such as the mosaic floor in the church in Shivta from the year 607 CE.^[77] Both the construction of new buildings and the instalment of new mosaics are considered as major indicators of wealth and economic prosperity.

To summarise, there is no clear evidence for a decline in the north-western Negev region in the 6^th c. CE, whether from Elusa’s garbage mounds, the region’s pigeon towers, or other archaeological data. Rather, the evidence suggests that there was continuity in everyday life and that the decline occurred from the 7^th c. CE. This suggests that neither climate nor the Justinianic Plague played a role in the region’s collapse. Conversely, the region’s decline was caused by a long period of warfare between the Romans and the Persians, and later the Arabs, which destroyed the trade routes to and from the north-western Negev desert.

Cities in Judaea/Palaestina

Further examples of demographic and economic decline dated to the 7^th c. CE can be seen in archaeological sites from the centre of Judaea/Palaestina. An example for this can be found in Scythopolis, where the public buildings were so neglected during the 7^th c. CE that they were dismantled by the public for use as building materials.^[78] This is also visible in other archaeological sites, such as the cities of Yavne-Yam, Yavne, and Emmaus, and various other agricultural settlements in their vicinity.^[79] The causes of this societal and economic collapse can be explained by the deterioration of the trade routes and economy as a result of the region’s political control changing multiple times in a relatively short time period. In addition, the new rulers of the region were in an almost constant state of war with the late Eastern Roman Empire, and the local population possibly fled to what was left under the rule of the Emperor in Constantinople. Therefore, due to extensive migration and the instability of the economy and trade, the settlements and cities of the eastern Mediterranean deteriorated.^[80]

On the other hand, the city of Jerusalem remained relatively stable in the 7^th c. CE as the Persian and Islamic conquests of the city did not destroy the infrastructure of the city.^[81] This may have been due to the importance of the city as a seat of power in the area and an important administrative centre.^[82] However, there are discoveries of mass graves in Jerusalem dated to the Persian conquest, which are also mentioned in ancient texts. These texts show that the Christians were marched out of the city to be killed, and so there are few visible signs of destruction in the city itself.^[83]

Shipwreck Analysis

The first type of information that can provide a broader picture is an analysis of shipwreck data, based on two well-known databases: the shipwreck database of Harvard University,^[84] and the Oxford Roman Economy Project (OXREP) database,^[85] which is partially based on the Harvard database.^[86] These databases aggregated data on shipwrecks from antiquity, including their dates, site/shipwreck name, GPS location, and cargo. However, the biggest impediment to easily interpreting this data is a lack of information on the body of water from which the ship came. Consequently, this information was completed, with the combined dataset including the body of water in which each shipwreck was found. This enabled the easy exclusion of ships that were not discovered in the Mediterranean Sea.

But why do shipwrecks matter? The use of this type of data implements a method that has recently been applied in different studies, such as Justin Leidwanger’s latest book.^[87] This method assumes that the number of shipwrecks has statistical significance, and greater amounts of maritime traffic are reflected in higher numbers of shipwrecks in certain periods. As the number of ships in use increases, so does the probability that some of them will sink due to storms and other calamities. Generally, the comparison of the number of shipwrecks between half-centuries and centuries is considered acceptable and common, and is an important tool in understanding the volume of sea-borne trade. The assumption at the foundation of this system is that unique catastrophic events leading to numerous ships sinking simultaneously are short-lived and rare, and a comparison between long periods neutralises the distinctive effect of such calamities.^[88]

The main obstacle in employing shipwrecks is the difficulty in exactly establishing the date of the ship’s sinking. As a result, in most cases, only broad estimates are available, ranging from a few decades to many centuries, depending on the type and extent of investigation undertaken on the shipwreck by surveyors and/or excavators, as well as the artefacts uncovered. To compensate for this difficulty, the full date range of each shipwreck was inputted into the graph. Thus, a ship can appear in multiple columns. This method is not entirely accurate, but it provides a valuable and universal picture, although with some deviation.^[89] The outcome of this method is that the sum of all the columns would vastly outnumber the number of ships appearing in the database. Leidwanger presented similar raw figures in his book. However, the number of ships he examined was less than what will be presented here.^[90] Henceforth, the data in this article will be the most comprehensive and up-to-date data available, despite the fact that it is only a partial study of a continually evolving dataset.^[91]

As can be appraised from the data showing the number of shipwrecks in the whole Mediterranean (Fig. 3), there was an increase in the number of shipwrecks in the Late Republic. As Rome gradually united the entire Mediterranean Sea, and the Mediterranean truly became a mare nostrum, naval traffic and trade became safer and easier, leading to a surge in commerce during the relatively peaceful 1^st c. CE.^[92] In addition, the thrill of suddenly available exotic commodities boosted this peak in naval commerce.^[93] With the possible novelty of the exported goods wearing off, there is a decline in the number of shipwrecks in the 2^nd c. CE.

Fig. 3:

Graph showing the number of shipwrecks per period in the Mediterranean according to the Harvard and OXREP databases.

However, from the 2^nd c. to the end of the 5^th c. CE, the number of shipwrecks remained consistent and near to what was seen at the end of the Republic, implying that this was a regular level of naval trade during full Roman control of the entire Mediterranean. Then, at the very end of the 5^th c. CE, there is a sharp decline of almost fifty percent in shipwreck numbers. The reason for such a severe reduction was most probably due to the fall of the Western Roman Empire in the late 5^th c. CE.^[94] The fall of the west also symbolised the decline of the city of Rome and other western trade cities and their hinterlands, and their subsequent reduction in population. Later, stability in the number of shipwrecks can be seen in the 6^th c. CE, followed by a decline from the 7^th c. CE onwards. This decline was most probably an outcome of the Persian war, and the Islamic conquest shortly after, which deprived Constantinople of most of the territories that were previously under the rule of the Eastern Roman Empire. These events damaged the trade networks between the former regions of the Empire, as well as between the eastern and western sides of the Mediterranean.

When comparing the trends of shipwrecks in the entire Mediterranean to the eastern Mediterranean, and when contrasting the two sides of this sea, this image becomes clearer, and the contrasts between the eastern and western Mediterranean become more pronounced (Fig. 4 and 5).^[95] This was done to eliminate the numerical bias of the shipwreck data, as more research was conducted in the western Mediterranean, and the original data is focused mainly on this area.

In the eastern Mediterranean, there was an increase in maritime traffic in the last two centuries BCE. This increase reached a peak in the 1^st c. CE, similar to what is visible in the Mediterranean as a whole. However, contrary to the general trend, after a slight decrease there was stability in the 2^nd and 3^rd c. CE, followed by a continued increase in the number of shipwrecks which reached a peak in the late 5^th c. CE, with this trend continuing until the beginning of the 7^th c. CE. This peak was even greater than the one in the 1^st c. CE, demonstrating the Eastern Roman Empire’s stability as well as the potential that trade was even more robust in the region during Late Antiquity. The peak in eastern Mediterranean trade continued at least until 600 CE, and so implies that there was no decline in the mid-6^th c. CE.^[96]

Fig. 4:

Graph showing the number of shipwrecks per period in the eastern Mediterranean according to the Harvard and OXREP databases.

Fig. 5:

Graph comparing the percentage of shipwrecks in each half century, out of the total number of shipwrecks in that half of the Mediterranean, according to the Harvard and OXREP databases. Orange represents the western Mediterranean, blue represents the eastern Mediterranean.

This stability in maritime trade in both the Mediterranean as a whole, and specifically the eastern Mediterranean during the 6^th c. CE, diminishes the arguments of different researchers who have claimed that the Justinianic plague and the LALIA were major factors in the breakdown of the Eastern Roman Empire. In addition, following the 6^th c. CE, there is a fall in the number of shipwrecks during the 7^th c. CE in both the Mediterranean as a whole, and the eastern Mediterranean in particular. Such a trend highlights that the deterioration of naval commerce occurred in the 7^th c. CE and was most probably connected to the Roman-Persian War of 602–628,^[97] and the Islamic conquests that started in 636.^[98] The Arab wars with the empire further deteriorated the situation and hampered commerce well after the 7^th c. CE.^[99]

Since the shipwreck data is based on dating, which provides a range rather than precise years, it is crucial to note and demonstrate that this information and conclusions are consistent with other evidence. These include the quantitative analysis of settlement trends in Judaea/Palaestina, a comparison with other regions, and other previously discussed and soon to be analysed findings.

Archaeological Survey of Israel

The Archaeological Survey of Israel is the most thorough archaeological survey conducted in a country. The entire country was carefully divided into survey grids of 100 km² (10x10 km) in the early 1960s, with each grid being published as a survey map,^[100] of which 70 % have been surveyed and published since the 1960s.^[101] Each map was published after a team of archaeologists systematically surveyed the area on foot, and methodically inspected every square metre of the grid where possible. Individual teams plotted, marked, and noted any site with signs of anthropogenic activity, as well as dated the site where possible using indicative sherds or other artefacts. Upon the completion of a survey, the team would include a summary and analysis of the various periods identified in the survey grid, as well as their conclusions concerning changes in settlement patterns, the date of these changes, and their possible causes.

Other important surveys conducted in Israel outside of the general Israel survey is a recent salvage survey in Beth Shemesh,^[102] as well as emergency surveys in Judaea and Samaria in the West-Bank between the years 1967 and 1968.^[103] Moreover, there are the Manasseh Hill Country Surveys that were conducted by Adam Zertal and encompass most of the territories of Samaria.^[104] All these surveys were conducted in a similar manner to the Israel survey, and each map and survey were published individually, electronically and/or in hard copy. Some of them were more organised and straightforward, while others were less so. The information from all the mentioned surveys was collected, and a quantitative analysis of the number of settlements from the Hellenistic to the Early Islamic period was conducted.^[105]

Fig. 6:

Graph indicating the number of sites per period in modern-day Israel.

Figure 6 represents the overall quantitative analysis of the survey results,^[106] and a significant increase in the number of sites with anthropogenic activity is visible during the Early Roman period (red-brown), defined as the mid-1^st c. BCE until the beginning of the 4^th c. CE. In this period, the number of sites more than tripled compared to the Hellenistic period (column in grey representing the end of the 4^th c. BCE until the mid-1^st c. BCE). More significant is the continuous increase in the number of sites from the Early Roman to the Late Roman period (blue column which represents the early 4^th c. to the early 7^th c. CE). This 61.77 % increase over the preceding period establishes the Late Roman period as a high point in the number of sites in Judaea/Palaestina, which was only surpassed in the 20^th c. CE. The substantial increase points to a well-documented population explosion that happened throughout the Eastern Roman Empire. However, this settlement boom ended in the 7^th c. CE, followed by a protracted period of decline in the number of sites and population.

Thus, when did the number of settlements peak between the 4^th to the beginning of the 7^th c. CE? There was a severe contraction, and the number of active sites during Early Islamic period was fewer than a third of what had previously been active under the reign of the Eastern Roman Empire. If the peak arrived in the mid-6^th c. CE, the decline in the number of sites can be tied to the LALIA and the Justinianic plague. On the other hand, a decline in the 7^th c. CE would indicate that climate and plague most probably did not majorly contribute, and the main reasons for this deterioration was manmade – the Roman-Persian wars and the Islamic conquest.

Fortunately, the surveyors left plenty of information that can help determine the timing of the fall. Numerous surveyors indicated that it is possible that there were more Early Islamic settlements than what was indicated in the maps and surveys, as the pottery found was of the types that may have continued several decades into the Early Islamic period. Hence, it seems that the Byzantine (Late Roman) ceramic, and many of the identified Byzantine (Late Roman) sites can be dated to the Late Byzantine period (550–636 CE). These observations prompted researchers, such as Gideon Avni, to define settlements with such assemblages as Byzantine-Early Islamic in the surveys they conducted,^[107] defining hundreds of sites as such.^[108] Furthermore, many surveyors saw the cause of the said decline as the Islamic conquest, and they believed that this decline continued throughout the first few decades of Early Islamic rule.^[109] In fact, just one surveyor suggested that the decline could be linked to the Justinianic plague.^[110] Other surveyors either did not deal with the issue of decline, or attributed the decline entirely to the Islamic conquest.^[111]

The reliability of the survey, its finds, and the dating of the Byzantine period in particular, were in the midst of several academic discussions and major analyses. Today, the accepted opinion is that in most of the Byzantine sites, the finds are from the later period,^[112] and their use continued into the first decades of the Early Islamic period.^[113] Moreover, a re-examination of some of the survey squares in the Galilee was conducted, alongside a new survey, excavations and dozens of test pits and probes. This examination showed that the original survey finds stand the test of time and a more thorough examination, and the reliability of this data is high.^[114] It is important to note that the survey finds for a single site may be inaccurate, but the survey as a whole is accurate in detecting general changes and trends in large regions – the type of research and debate that the current article wishes to conduct.

To conclude, the ceramic typologies that were found in the different sites caused the different researchers and surveyors to assume that settlement decline occurred in the 7^th c. CE and occurred due to the wars and geopolitical changes which transpired during this period. This hypothesis that claims the effects of the Justinianic plague and the LALIA were limited, and minimal, and anthropogenic activities and war were the main causes, is also supported by other evidence. For example, dated inscriptions provide evidence of a vibrant daily life and the construction of numerous buildings even after the 6^th c. CE in modern-day Israel and Jordan. While a decrease in the number of such finds occurred briefly in the fifth decade of the 6^th c. CE, it was immediately followed by a full recovery, inferring that the effects of the Justinianic plague or the LALIA were short and minimal in the Eastern Roman Empire. On the other hand, a sharper decrease in these dated inscriptions can be found in the 7^th c. CE due to the Persian war. Even though there was a slight recuperation the following decade after 628 CE, when Rome recovered the area, an even sharper decline was visible after the Islamic conquest of 636 CE, after which there was no recovery.^[115] The trend of stability in terms of the number of inscriptions in the 6^th c. CE can also be seen when looking at the eastern Mediterranean basin as a whole.^[116] Thus, the decline and abandonment of settlements in the Late Eastern Roman Empire most probably needs to be dated to the 7^th c. CE and not to the 6^th c. CE.

Archaeological Surveys and Excavations in Turkey, Cyprus, and Greece

Evidence for this trend can also be seen in surveys and excavations in some regions of Asia Minor (modern-day Turkey).^[117] During the Late Roman period (325–636 CE), for example, the Maeander Valley saw a rise in agricultural acreage, population, and settlement size.^[118] Similar trends were also seen in Phrygia,^[119] Miletus,^[120] Sagalassos,^[121] and the Troad.^[122] Moreover, several recent works have emphasised and demonstrated that an increase in population occurred in many regions of the Anatolian countryside in the 4^th c. CE, and that this pattern persisted until the mid-7^th c. CE in several phases.^[123]

The first phase, from the 4^th to the early 6^th c. CE, was devoid of any climatic change, be it an increase or decrease in temperature or precipitation. Yet, during the second phase of this trend, from the mid-6^th to the 7^th c. CE, there was an increase in population density that was only matched in modern times and is comparable to the demographic fluorescence in Judaea/Palaestina and other Mediterranean basin regions. This second phase of growth overlapped with a colder period and higher precipitation, which coincides with the LALIA. While the LALIA is commonly thought to be between 550 and 660 CE at most, this colder and wetter period lasted until the mid-8^th century CE. Such a climate was reputedly ideal, which may have contributed to the expansion in population and agricultural industry witnessed in other sections of the Mediterranean. Even though there was no change in climate in the mid-7^th c. CE, the number of sites and the regional pollen record decreased.^[124] This decline, like in other regions, was driven by anthropogenic elements. War ravaged the land, decimating trade, which was especially destructive in Anatolia, as it became the principal border zone between the Byzantine Empire, which arose from the ashes of the Eastern Roman Empire, and the Islamic Caliphates.^[125]

Another notable example is the archaeological excavations and surveys carried out in Cyprus. From them, it is clear that a peak in the number of settlements, and in many cases in their size, occurred between the 5^th and 7^th c. CE. The scholars emphasised in numerous places that the peak must be dated to the end of the 6^th c. or the beginning of the 7^th c. CE, and that the decline began only in the mid-7^th c. CE in most regions, and lasted for a very long time after.^[126]

The same can be seen in Greece, with many areas declining only in the 7^th c. CE. This is evident from the archaeological surveys conducted in Greece, which can generally be split into two main groups. The first is surveys where there is a peak in habitation during the Classical and Hellenistic period, a decrease between the Hellenistic and Early Roman followed by centuries of stability until the end of the Late Roman period. The second is a gradual increase in habitation which reached a peak in the Late Roman period. An unequivocal trend, visible in almost all the surveys, is a stark decline in the 7^th and/or 8^th c. CE, as can be seen in Laconia,^[127] Methana,^[128] Boetica,^[129] Argolid,^[130] and the Asea valley.^[131] The deterioration that can be seen in the 7^th and 8^th c. CE cannot merely be associated with an unfamiliarity with the ceramic assemblage, as the material culture of the 6^th to 8^th c. CE in most of the Greek sub-regions is well understood.^[132] Thus, this decrease is due to an abandonment of territories and sites. However, there is a chance that some of the sites that were considered abandoned were not necessarily so.^[133]

The decline in population can be seen not only through the abandonment of sites, but also through other finds such as the decrease in the size of the Basilica at Perissa in Santorini in the 8^th c. CE.^[134] It is true that the number of sites that continued after this period is greater than what was presumed in the past, yet the severe decline in the size of the population is undoubtable.^[135] In general, the 7^th c. CE saw the newly transformed Byzantine Empire reorganise their holdings in the Aegean region, and heavily invest in fortifications throughout these islands. While there was a decline in population in several areas of Greece, Gortyn in Crete saw an increase, visible in the division of larger buildings to create new public and private spaces, as well as the construction of a new church.^[136]

While it is relatively established that a decline in settlement size and number occurred in the 7^th or 8^th c. CE in Turkey, Cyprus and Greece, this region saw numerous events in the 6^th c. CE which should and could have caused a deterioration of the area. One of the most concrete examples is the tsunamigenic earthquakes in the middle of the 6^th c. CE.^[137] However, instead there was rapid renovation and rejuvenation of the region, especially in the port of Kos in Greece,^[138] and Beirut in Lebanon.^[139] Such rapid recovery and rejuvenation of cities after devastating natural disasters is substantial evidence of an economically resilient and demographically stable Empire, which was possibly at its strongest.

Archaeological Surveys in North Africa

North Africa is divided into two distinct regions. Egypt lies on one side, with modern-day Tunisia, Libya, Algeria, and Morocco on the other. Egypt is typically differentiated because it is greatly influenced by the Nile, whose sources depend on the monsoons of the Indian Ocean.^[140] During the late 2^nd to the 5^th c. CE in southern Egypt (Upper Egypt), there is a well-documented fall in the number of sites and population.^[141] Conversely, the northern part of Egypt (Lower Egypt), comprising the Nile Delta and Alexandria, had a boom in population and settlements from the 4^th to the 7^th c. CE.^[142] This corresponds and is verified by what can be seen in other countries of the Mediterranean basin, and supports the notion that a population fall did not occur in the 6^th c. CE due to the Justinianic plague, but rather in the 7^th c. CE due to warfare, and the deterioration of trade.

The rest of North Africa is another important region for understanding the changes occurring in the mid-6^th to the early 8^th c. CE due to its unique history.^[143] Compared to the climate of Egypt, the other regions of Northern Africa are part of the Mediterranean climate zone. This region saw frequent changes in control from the early 5^th until the 8^th c. CE, with the Vandal invasion ending the Late Roman period in 429 CE.^[144] Their Kingdom lasted over a century before being deposed by the Eastern Roman Expeditionary force led by Belisarius, who was dispatched by Emperor Justinian I. Following a year of warfare, the conflict concluded with the Vandals’ ultimate surrender in 534 CE.^[145] These changes in governance corresponded to the settlement of people in the newly conquered regions, who carried their own distinct material culture with them, synthesising with the local one. As a result, these changes are readily visible in the archaeological record, and so distinct and relatively short periods of time may be recognized using the ceramic assemblage.

The lands of North Africa were ruled by the Late Eastern Roman Empire from 534 CE until the Islamic conquest in the late 7^th c. CE, a period known in archaeology as the Byzantine period.^[146] As a result, the Islamic period is usually defined only from 689 CE.^[147] Therefore, due to the Byzantine archaeological period coinciding with the beginning of the LALIA and the Justinianic plague, an examination of this region’s archaeology will reveal whether these calamities had a significant impact on the Eastern Roman Empire.

Although some surveys in North Africa show a decline in the number of sites in the 5^th or 6^th c. CE, this was attributed to a change in the fabric of the society before the LALIA and Justinianic plague or the reconquest conducted by the Eastern Roman (Byzantine) Empire, with a movement from an extensive agricultural hinterland to nucleated communities in urban centres. Yet in some cases, these conclusions may be derived from problems in the dating of the ceramic assemblages.^[148] Nevertheless, under Eastern Roman/Byzantine control, when the LALIA and Justinianic plagues reportedly plagued the Empire, there was no general decline in population size in the region. Even those attempting to downplay the significance of the Islamic Conquest in the deterioration of societal structures in North Africa are forced to admit that according to the archaeological record, the main decline occurred in the following century and a half after the Islamic Conquest, i.e. in the late 7^th and throughout the 8^th c. CE.^[149] The best way to examine this period, and overcome the problems of dating visible in older surveys, is to examine one of the latest surveys. The Thugga survey was conducted by the University of Trento (Italy) and the Institut National du Patrimoine of Tunisia, and an area of 660 km² was surveyed in the Medjerda Valley in the Tunisian High Tell.^[150] In this survey, 643 anthropogenic features were detected and documented. There were 141 separate characteristics and locations dated between 31 BCE and 250 CE, and this number climbed to 153 sites between 250 CE and the Vandal conquest of 430 CE. The number of sites decreased to 147 during the Vandal period, which lasted from 430 to 533 CE. By contrast, the number of sites climbed to 176 during the rule of the Eastern Roman Empire (534–690 CE). However, after the Islamic conquest, the number of sites decreased, with only 21 sites and features attributed to the 8^th–13^th c. CE (Fig. 7a and b).^[151]

Fig. 7a and b:

Graph indicating the number of settlements per period in the Thugga region, modern-day Tunisia. Data taken from: http://rusafricum.org/en/thuggasurvey/datazioni_siti/, last accessed 04.08.2023.

According to this survey, while an area was ruled by the Eastern Roman Empire and was not influenced by wars, invasions, raids, or brief conquests, the region thrived. When the LALIA and the Justinianic plague were the only supposedly apocalyptic occurrences that a region experienced, there was no decline, while in some places there was a peak, suggesting that they were not nearly as devastating in that region. Since the Justinianic plague’s outcome is death, and its consequences could never be described as positive and conducive to population growth, the LALIA with its small fall in temperature could be considered beneficial, similar to what Bruins suggested,^[152] as it possibly encouraged and supported the prosperity under Eastern Roman rule. This is true not only for Vandal North Africa, that was conquered by the Eastern Romans, but also for other Vandal territories which were conquered by the Empire, such as the islands of Corsica,^[153] and Sicily.