
Conn Herriott
Related Authors
Nigel Goring-Morris
The Hebrew University of Jerusalem
Anna Belfer-Cohen
The Hebrew University of Jerusalem
Clinton Briar
University of Leicester
Anna Hanzelková
Masaryk University
Ofer Bar-Yosef
Harvard University
ESTELLE ORRELLE
University of East London
Leore Grosman
The Hebrew University of Jerusalem
InterestsView All (24)
Uploads
Papers by Conn Herriott
Part 1 of the dissertation comprises the results of the ethnological analysis, relying on quantitative data from the SCCS. Here is presented statistical patterning in the global distribution of eleven traits of religion: three degrees of religious organisation; three types of ritual ceremonies; three ceremonial elements; and two ideological characteristics. Within the constraints of data availability, these religious traits were selected according to their value in prehistoric archaeology and also ethnology. Some of the SCCS variables had to be reformatted in order to tailor them to the current research questions. Also the author corrected certain data, according to source research. The goal in this cross-cultural analysis was to achieve clearer understanding of the social and economic conditions in which each religious trait is most common.
It was found that the selected characteristics of religion generally track with community size and socio-economic complexity. However, each trait follows a different trajectory along this continuum, with distinct and surprising evolutionary inflection points. For instance, the analysis suggests that religious institutionalisation does not correlate strongly with ‘high god’ belief (e.g., Shariff et al. 2011; Peoples and Marlowe 2012; Botero et al. 2014; Watts et al. 2015; Lang et al. 2019). Religious organisation appears to lag behind social and economic developments, and only becomes common in chiefdom-type societies. At the other end of the scale, even a modest ‘shamanic’ degree of religious centralisation is rather rare among the smallest, least variegated societies. Meanwhile, some religious characteristics—e.g., collective ceremonial feasting and reincarnation beliefs—are so widespread that it is difficult to identify the social and economic conditions in which these traits are rare among non-industrial societies. There are also many unexpected religion-lifeway associations or co-occurrences. For example, strong emphasis on male supernatural entities is frequently found where females enjoy relatively high status in economic and religious terms.
In Part 2 of the dissertation these ethnological trends are brought to bear on four case studies spanning the transition between the Epipalaeolithic and Neolithic periods in the southern Levant. Selected on the basis of the quality and extent of published data, the case studies are: the Geometric Kebaran of the Mount Carmel region (16,500–14,500 calBP); the Early Natufian site of Eynan (Ain Mallaha) in the Hula Valley (15,000–13,750 calBP); PPNA Netiv Hagdud in the Jordan Valley (11,600–10,750 calBP); and Late PPNB Basta in the western highlands of modern-day Jordan (9,540–9,040 calBP). For each case, ethnological patterns are used to frame interpretations of direct archaeological evidence for ritual and symbolic behaviour, as well as religious beliefs. Rather than relying on anecdotal use of ethnographic analogy, the current study engages the physical remains through the lens of robust statistical patterns drawn from the SCCS. This approach helps to rank alternative views of physical evidence—e.g., the possible symbolic meanings of burials, or figurines and imagery, or the inter- and intra-site distributions of potentially ritual-related artefacts. Ethnological trends offer some gauge of the likelihood of different religious traits in each of these prehistoric cases—e.g., different ceremonial emphases, human sacrifice, or special structures for religious ritual and symbolic behaviour.
Beyond these specific religious traits and case studies, this dissertation’s ethnology-based analysis of several ‘snapshots’ in the Epipalaeolithic-Neolithic transition may offer some counterbalance for several dominant paradigms in the archaeology of religion. One such is the post-processualist emphasis on cultural particularism, as exemplified by researchers such as Hodder (e.g., 1992, 2003). Another is Cauvin’s (1994) influential révolution des symboles, which has given rise to a dominant narrative around religion as a driver of change. Although such perspectives are well taken, the current research argues that they are often overstated. It is more justified and useful to recognise the social and economic forces acting upon religious developments in human history, and to frame archaeological interpretation of ritual and symbolic evidence through reference to these dynamics, as demonstrated in ethnological patterns.
The excavation team was made up of workers from Bir al-Maksur, Galilee, as well as I.Bransburg and G. Suleimani (field supervisors) and Y. Govrin (archaeological consultation and logistics).
The settlement at Horvat Hermas existed from the late Byzantine period through the Mamluk period, with a peak in the Early Islamic period. During this peak period it included several building complexes, courtyards, an industrial area, and refuse pits. The excavation provides a significant addition to the material culture corpus of the period and the region and supplements the settlement picture in the region during late antiquity.
Part 1 of the dissertation comprises the results of the ethnological analysis, relying on quantitative data from the SCCS. Here is presented statistical patterning in the global distribution of eleven traits of religion: three degrees of religious organisation; three types of ritual ceremonies; three ceremonial elements; and two ideological characteristics. Within the constraints of data availability, these religious traits were selected according to their value in prehistoric archaeology and also ethnology. Some of the SCCS variables had to be reformatted in order to tailor them to the current research questions. Also the author corrected certain data, according to source research. The goal in this cross-cultural analysis was to achieve clearer understanding of the social and economic conditions in which each religious trait is most common.
It was found that the selected characteristics of religion generally track with community size and socio-economic complexity. However, each trait follows a different trajectory along this continuum, with distinct and surprising evolutionary inflection points. For instance, the analysis suggests that religious institutionalisation does not correlate strongly with ‘high god’ belief (e.g., Shariff et al. 2011; Peoples and Marlowe 2012; Botero et al. 2014; Watts et al. 2015; Lang et al. 2019). Religious organisation appears to lag behind social and economic developments, and only becomes common in chiefdom-type societies. At the other end of the scale, even a modest ‘shamanic’ degree of religious centralisation is rather rare among the smallest, least variegated societies. Meanwhile, some religious characteristics—e.g., collective ceremonial feasting and reincarnation beliefs—are so widespread that it is difficult to identify the social and economic conditions in which these traits are rare among non-industrial societies. There are also many unexpected religion-lifeway associations or co-occurrences. For example, strong emphasis on male supernatural entities is frequently found where females enjoy relatively high status in economic and religious terms.
In Part 2 of the dissertation these ethnological trends are brought to bear on four case studies spanning the transition between the Epipalaeolithic and Neolithic periods in the southern Levant. Selected on the basis of the quality and extent of published data, the case studies are: the Geometric Kebaran of the Mount Carmel region (16,500–14,500 calBP); the Early Natufian site of Eynan (Ain Mallaha) in the Hula Valley (15,000–13,750 calBP); PPNA Netiv Hagdud in the Jordan Valley (11,600–10,750 calBP); and Late PPNB Basta in the western highlands of modern-day Jordan (9,540–9,040 calBP). For each case, ethnological patterns are used to frame interpretations of direct archaeological evidence for ritual and symbolic behaviour, as well as religious beliefs. Rather than relying on anecdotal use of ethnographic analogy, the current study engages the physical remains through the lens of robust statistical patterns drawn from the SCCS. This approach helps to rank alternative views of physical evidence—e.g., the possible symbolic meanings of burials, or figurines and imagery, or the inter- and intra-site distributions of potentially ritual-related artefacts. Ethnological trends offer some gauge of the likelihood of different religious traits in each of these prehistoric cases—e.g., different ceremonial emphases, human sacrifice, or special structures for religious ritual and symbolic behaviour.
Beyond these specific religious traits and case studies, this dissertation’s ethnology-based analysis of several ‘snapshots’ in the Epipalaeolithic-Neolithic transition may offer some counterbalance for several dominant paradigms in the archaeology of religion. One such is the post-processualist emphasis on cultural particularism, as exemplified by researchers such as Hodder (e.g., 1992, 2003). Another is Cauvin’s (1994) influential révolution des symboles, which has given rise to a dominant narrative around religion as a driver of change. Although such perspectives are well taken, the current research argues that they are often overstated. It is more justified and useful to recognise the social and economic forces acting upon religious developments in human history, and to frame archaeological interpretation of ritual and symbolic evidence through reference to these dynamics, as demonstrated in ethnological patterns.
The excavation team was made up of workers from Bir al-Maksur, Galilee, as well as I.Bransburg and G. Suleimani (field supervisors) and Y. Govrin (archaeological consultation and logistics).
The settlement at Horvat Hermas existed from the late Byzantine period through the Mamluk period, with a peak in the Early Islamic period. During this peak period it included several building complexes, courtyards, an industrial area, and refuse pits. The excavation provides a significant addition to the material culture corpus of the period and the region and supplements the settlement picture in the region during late antiquity.
Part 1 of the dissertation comprises the results of the ethnological analysis, relying on quantitative data from the SCCS. Here is presented statistical patterning in the global distribution of eleven traits of religion: three degrees of religious organisation; three types of ritual ceremonies; three ceremonial elements; and two ideological characteristics. Within the constraints of data availability, these religious traits were selected according to their value in prehistoric archaeology and also ethnology. Some of the SCCS variables had to be reformatted in order to tailor them to the current research questions. Also the author corrected certain data, according to source research. The goal in this cross-cultural analysis was to achieve clearer understanding of the social and economic conditions in which each religious trait is most common.
It was found that the selected characteristics of religion generally track with community size and socio-economic complexity. However, each trait follows a different trajectory along this continuum, with distinct and surprising evolutionary inflection points. For instance, the analysis suggests that religious institutionalisation does not correlate strongly with ‘high god’ belief (e.g., Shariff et al. 2011; Peoples and Marlowe 2012; Botero et al. 2014; Watts et al. 2015; Lang et al. 2019). Religious organisation appears to lag behind social and economic developments, and only becomes common
in chiefdom-type societies. At the other end of the scale, even a modest ‘shamanic’ degree of religious centralisation is rather rare among the smallest, least variegated societies. Meanwhile, some religious characteristics—e.g., collective ceremonial feasting and reincarnation beliefs—are so widespread that it is difficult to identify the social and economic conditions in which these traits are rare among non-industrial societies. There are also many unexpected religion-lifeway associations or co-occurrences. For example, strong emphasis on male supernatural entities is frequently found where females enjoy relatively high status in economic and religious terms.
In Part 2 of the dissertation these ethnological trends are brought to bear on four case studies spanning the transition between the Epipalaeolithic and Neolithic periods in the southern Levant. Selected on the basis of the quality and extent of published data, the case studies are: the Geometric Kebaran of the Mount Carmel region (16,500–14,500 calBP); the Early Natufian site of Eynan (Ain Mallaha) in the Hula Valley (15,000–13,750 calBP); PPNA Netiv Hagdud in the Jordan Valley (11,600–10,750 calBP); and Late PPNB Basta in the western highlands of modern-day Jordan (9,540–9,040 calBP). For each case, ethnological patterns are used to frame interpretations of direct archaeological evidence for ritual and symbolic behaviour, as well as religious beliefs. Rather than relying on anecdotal use of ethnographic analogy, the current study engages the physical remains through the lens of robust statistical patterns drawn from the SCCS. This approach helps to rank alternative views of physical evidence—e.g., the possible symbolic meanings of burials, or figurines and imagery, or the inter- and intra-site distributions of potentially ritual-related artefacts. Ethnological trends offer some gauge of the likelihood of different religious traits in each of these prehistoric cases—e.g., different ceremonial emphases, human sacrifice, or special structures for religious ritual and symbolic behaviour.
Beyond these specific religious traits and case studies, this dissertation’s ethnology-based analysis of several ‘snapshots’ in the Epipalaeolithic-Neolithic transition may offer some counterbalance for several dominant paradigms in the archaeology of religion. One such is the post-processualist emphasis on cultural particularism, as exemplified by researchers such as Hodder (e.g., 1992, 2003). Another is Cauvin’s (1994) influential révolution des symboles, which has given rise to a dominant narrative around religion as a driver of change. Although such perspectives are well taken, the current research argues that they are often overstated. It is more justified and useful to recognise the social and economic forces acting upon religious developments in human history, and to frame archaeological interpretation of ritual and symbolic evidence through reference to these dynamics, as demonstrated in ethnological patterns.