Abstract
Building on recent studies of Jesus’s social network, this article seeks to explore how the relational dynamics surrounding Jesus’s life and ministry are depicted differently in the canonical Gospels. Using different perspectives and methods than those usually employed by biblical scholars, the network analyses provide rich illustrations and descriptions of structural dynamics that have not traditionally been the focus of Gospel scholarship. Analyses examine the extent to which the Gospels’ social networks overlap, as well as differences in levels of relational prominence and in relational structures across the Gospels. The results provide a unique window into the relational dynamics portrayed by the Gospels, producing a variety of insights, some which may not surprise biblical scholars but others which hopefully will inspire further consideration.
Several studies have examined Jesus’s relationships and interactions using social network analysis (Duling 1999, 2000; Grandjean; McClure 2016, 2018), an important and growing subfield within social psychology (Felmlee & Faris 2013). My recent depictions of Jesus’s social network have been based on a compilation of the relationships recorded in all four canonical Gospels (McClure 2016, 2018). However, there is ample evidence that significant differences exist among the Gospels and that the Gospels depict Jesus’s life and message uniquely based on their authors’ theological perspectives and audiences (e.g., Powell). For this reason, it is important to understand how the Gospels differ in depicting the relational dynamics surrounding Jesus’s life and ministry. This study develops and analyzes social networks for each Gospel, with the hope of identifying new interpersonally-oriented insights about the Gospels.
Scholarly interest in Jesus’s social network largely began as a “thought experiment” in response to Gerd Theissen’s eminent argument that Jesus and his followers sought to renew Judaism by addressing social, economic, and political divisions, relinquished many earthly possessions, and were able to do so through receiving support from local Galilean communities (Duling 1999: 156). Dennis Duling (1999, 2000) used social network analysis to hypothesize about how geographical routes and interpersonal connections shaped and facilitated what Theissen calls the “Jesus movement” (1).
To understand how spatial, or geographic, networks shaped Jesus’s ministry, much of which took place in Galilee, Duling drew on rich archeological and historical information concerning numerous Galilean cities, towns, and villages to describe their geographies, structures, cultures, and commerce. Particular attention was paid to the following:
Nazareth, a village in southern Galilee were Jesus grew up (Duling 1999:159; Matt 2:23; John 1:45–46);
Capernaum, a fishing community on the Sea of Galilee that served as the primary locus for Jesus’s ministry (Duling 1999: 161–62; Theissen & Merz:166; e.g., Matt 4:13, Mark 1:21, Luke 4:31, John 2:12); and
Bethsaida, another fishing community near the Sea of Galilee which was mentioned in conjunction with many of Jesus’s miracles and from which some of Jesus’s disciples came (Duling 1999: 163–64; e.g., Mark 8:22–26; Luke 9:10–17; John 1:44).
To create the spatial network, Duling drew on archeological research about possible roads and sailing patterns that may have facilitated trade, interpersonal connections, and communication among these communities; this research suggests that Galilee was well populated with robust travel and trade (1999: 165–68; for archeological sources, including some cited by Duling and others more recent, see Dorsey; Tsafrir, Di Segni, & Roll; Meyers; Strange 1997, 2014; and Sawicki). Within this spatial network, Capernaum was a strategic location for Jesus’s ministry because of its connections via roads and sailing to many Galilean villages; it allowed Jesus and his followers access to many communities within a day’s journey and provided a haven for their return (Duling 1999: 170).
Duling largely framed the social dynamics of the Jesus movement through the lens of an ego network (2000: 5), or an analysis of the immediate actors surrounding a person, where the central person is designated as “ego,” and the people to whom he or she is connected are “alters” (Prell: 118). Duling identified three types of alters in Jesus’s ego network (2000: 6–8): the disciples who forsook all and followed him (e.g., Matt 19:27, 29); people who remained in their communities and supported him (e.g., Mark 5:18–20); the anonymous crowds present during some sermons and miracles (e.g., Matt 5–7; Matt 8:13–21). He also examined familial networks (2000: 8–10), including those between: Zachariah, Elizabeth, John the Baptist, Mary the mother of Jesus, and Jesus (Luke 1); Andrew and Simon Peter (Matt 4:18–20); James, John, their father Zebedee, and their mother (Matt 4:21–22, 20:20–21). In addition, some of Jesus’s followers were also connected through shared work, like fishing (Luke 5:2–11). Jesus’s ministry was undergirded by networks of people who followed and provided for him.
Duling’s analyses provide a helpful theoretical framework for understanding the dynamics that allowed the “Jesus movement” to emerge from within first century Palestinian Judaism and to develop into a larger religious movement in the Greco-Roman world (1999, 2000; Theissen); however, they did not intend to provide a complete picture of Jesus’s social network, and they did not include many of the actors mentioned in the Gospels. Building on these studies, I sought to develop a social network that included all the actors whom the Gospels depict as interacting with Jesus and to investigate its dynamics through more quantitative social network analytical techniques, some of which Duling drew on to inform his discussions (1999: 158; 2000: 4–5) but did not employ.
My examinations of Jesus’s social network have focused on three main groups of people that Jesus interacted with throughout his life and ministry: his family and followers, the civil and religious authorities, and stigmatized people (McClure 2016, 2018). Key actors in the first group include his parents, his brothers, John the Baptist, his twelve disciples, additional women followers, and others (Bauckham 1990, 2002; Duling 2000: 9; McClure 2016: 4–5; McClure 2018: 5). Combining Jesus’s family with his followers may appear to be problematic due to arguments, for example, that Mark depicts “opposition between Jesus and his relatives” (Crossan: 113); however, scholars question the extent of the opposition (Lambrecht: 256) and even frame the posited opposition as an “overreading” of the text (Malbon: 219). My analyses indicate that Jesus had many positive relationships with his family and followers (McClure 2016: 11).
This assertion may be counterintuitive because the Gospels record a number of instances in which Jesus rebukes his family and followers for their unbelief, resistance to his ministry, and lack of support (Matthew 8:23–27, 16:21–23, 19:13–15; Mark 4:13, 14:32–42; John 2:1–5, 7:1–9). The ties are overwhelmingly positive, however, because whether a tie is positive or negative depends on whether there were more positive or negative interactions between a pair of actors. Although Jesus had negative interactions with his family and followers, the positive interactions largely outweighed the negative interactions.
Jesus’s family and followers were the most central, or prominent, category of actors in his social network, and the Gospels mostly presented Jesus’s relationships with them as supportive (McClure 2016: 12–14).
According to the Gospels, Jesus had mostly negative ties with the civil and religious authorities (McClure 2016: 11), and actors in this group included Herod Antipas, Pontius Pilate, Pharisees, Sadducees, scribes, elders, chief priests, and other Jewish religious leaders (Riches; Bond; Jeffers; Jensen; McClure 2016: 5–6; McClure 2018: 5–6).
However, there are different perspectives on whether these interactions were negative. Debates were normative when Jewish religious leaders were discussing religious and theological questions (Schwartz 2012: xi, xv), and it is difficult to tell if these interactions were actually negative or just dialogue among leaders with different perspectives. [My] study, however, [considered] them to be negative because the Gospel writers often frame the religious leaders as being antagonistic toward Jesus.
This second category of actors had average centrality, or prominence, in Jesus’s social network, and Jesus’s interactions with them were depicted as conflictual (McClure 2016: 12–14).
The final category of actors in Jesus’s social network included people who were stigmatized due to sickness, deformity, demon-possession, sinfulness, criminal activity, ethnicity, or nationality (Goffman; Riches; Knoppers; McClure 2016: 6; McClure 2018: 6). Actors in this category had the lowest centrality, or prominence, in Jesus’s social network; however, Jesus’s interactions with them were compassionate (McClure 2016: 12–14).
Additional analyses in which I examined the structure of Jesus’s social network in more depth provided three key insights. First, they underscored the importance of Jesus’s ties with the stigmatized people in his network, many of whom had very few ties to other actors, and his work in integrating them into the broader social network (McClure 2018: 14–16). They also described how many of the authorities who were antagonistic toward Jesus were also antagonistic toward stigma-tized people, indicating that a key source of conflict between Jesus and the religious authorities, in particular, involved some religious authorities’ opposition to Jesus’s compassionate interactions with stigmatized people (McClure 2018: 18; for examples, see Matt 9:10–13, Mark 2:15–17, Luke 5:29–32, and John 9:24–34). In addition, they analyzed patterns of interpersonal tension in Jesus’s social network, much of which occurred in relation to the climax and major crisis of the Gospels: Jesus’s betrayal, crucifixion, and burial (McClure 2018: 19; see also Rutledge: 41, 43; Binz: 9).
Because biblical scholars (e.g., Powell) have identified many literary, textual, and theological differences among the Gospels, it is important to investigate whether the ways in which the Gospels depict Jesus’s social network might differ, as well. Doing so will elucidate whether certain characteristics of Jesus’s social network are, for example, shared among the synoptic Gospels or unique to a specific Gospel. This analysis differs from the textually- and theologically-oriented approaches of biblical scholars by comparing relational aspects of the Gospels. Before proceeding to the research questions, it is important to acknowledge some preliminary work by Martin Grandjean, who developed networks for each Gospel and described the number of characters and ties in each Gospel, as well as patterns related to the density of ties and fragmentation in each Gospel. His approach differs from mine, however, by focusing explicitly on narrative properties, not on interpersonal relationships. He states:
There is no ‘social network’ in our study, because our goal is to avoid these ‘social’ approaches by analyzing the simultaneous presence of characters in the scenes of the story…. It is therefore not a mapping of social interactions but a comprehensive and interpretative vision of the facts that are told, a focus on the content of the narrative and not its form.
There are other differences between Grandjean’s study and my studies: Grandjean includes non-human actors, like angels, in his analyses (3), while I do not (McClure 2016: 7); Grand-jean includes actors who do not interact with Jesus, like the parents of the blind man in John 9 (5), but I do not (McClure 2016: 8); Grandjean sometimes examines subsets of the social network (2, 7–8), while I have focused on the whole social network (McClure 2016, 2018). Grandjean’s analyses provide valuable insights but are not inherently social or interpersonal, thus necessitating a study that compares and contrasts how the Gospels depict Jesus’s social network.
Social network analysis is distinguished from other perspectives and methodologies because of its focus on relationships and the larger social structures that they form, as well as its ability to depict relational dynamics with engaging visualizations (Felmlee and Faris 2013:440–441). Drawing on these emphases, this study investigates similarities and differences in the relational dynamics of the four Gospels, focusing more on structural factors than individual actors, even though individual actors are used to illustrate the larger structural insights. The structural orientation is intentional in contrast to the approaches of other scholars who have analyzed specific characters in the Gospels (Bauckham 1990, 2002; Malbon; Farelly; Hunt, Tolmie, and Zimmermann); their work provides many valuable insights which this study seeks not to duplicate but to complement with a more structural lens. The three main research questions are as follows:
To what extent do the Gospels’ social networks overlap? Is overlap higher among the synoptic Gospels?
To what extent do the Gospels differ in depicting certain actors as more or less central or prominent?
To what extent do the relational structures of the Gospels differ?
Data and Methods
Data
This study examines Jesus’s social network, as depicted in the four Gospels of the New Testament. This study often refers to the Gospels by their shorter name (i.e., “Mark” instead of “the Gospel of Mark”), dropping “the Gospel of” to improve flow. Using the shorter name does not, however, indicate that this study espouses traditional authorship of the Gospels. When saying something about a certain Gospel (e.g., Mark depicts…), the title should be interpreted as referring to the Gospel, not the traditional author. There were slight changes in the data between my previous studies (Mc-Clure 2016, 2018), and this study uses the data from my 2018 study.
The actors, or nodes, in the study include Jesus and the human actors, both the individuals and the concrete groups, with whom he interacted; across the four Gospels, there are 121 actors.
Individuals (Jesus, Simon Peter, etc.) and concrete groups of people (the Pharisees, Sadducees, disciples of Jesus, etc.) were included as actors. Some groups were treated as actors, while individual members were also treated as distinct actors. The most common instance includes coding the group of Jesus’s disciples as a node and then coding each of the individual disciples as nodes. References to larger conglomerates like crowds and multitudes were not coded (e.g., Matthew 14:14; Luke 6:17).
In addition, “actors who had multiple names, like Simon Peter or Nathaniel/Bartholomew, were coded as one actor, based on traditional understandings of the names (Glover 1939:44–45)” (McClure 2016:8), and John son of Zebedee was coded as the Beloved Disciple, despite scholarly debates about the Beloved Disciple’s identity (Powell:117; Charlesworth; Bauckham 2007). These decisions, despite their shortcomings, were made in the initial coding of the data. I maintain this approach in order to provide consistency across my studies. I do not point out these coding decisions to be antagonistic toward critical methods and insights, which I value, but to provide transparency about the data. Seventy-nine percent of the actors in Jesus’s social network are men, and 21% are women. Groups, like the Pharisees and the disciples of Jesus, tended to be coded as male. Twenty-eight percent of actors with whom Jesus interacted (N=120) are his family and followers, 20% are civil and religious authorities, 43% are stigmatized people, and 8% are others (McClure 2016:9). “Actors who fit into multiple categories were placed into the category that best fit their role in the narrative. For example, Pontius Pilate, who was a Gentile (i.e., stigmatized) and a civil authority, was coded with the civil and religious authorities, not stigmatized people” (McClure 2016:8). Actors are listed by category in Appendix Ap.
There are relational ties present between actors: if the actors interacted; if an individual was a member of a group; if an actor responded to another actor, even if that actor was not present in the situation; if an actor was mentioned in relation to another actor. Of the ties in Jesus’s social network (N=400), 89% of the ties are positive, and 11% are negative (McClure 2018:8). “Whether ties were positive or negative was determined by coding each [connection] between each pair of actors as either positive or negative. If there were more positive than negative [connections], the tie was coded as positive; on the other hand, if there were more negative than positive [connections], the tie was coded as negative” (Mc-Clure 2016:8). The ties are coded to reflect the interactions and relationships between the actors explicitly stated in the text. This is particularly important to note in relation to John, where the religious leaders are often called “the Jews” and not differentiated into more specific types (Beutler: 146–48). I have “[referred] to this group as the ‘Jewish leaders’ instead of ‘the Jews’ because the term ‘Jew(s)’ can have a derogative connotation” (McClure 2016:6). In coding the ties, interactions and relationships that involve the Jewish leaders and that do not mention more specific leaders, like scribes, elders, Sadducees, and others, are only coded for the Jewish leaders and are not coded for more specific types of leaders that are not explicitly presented in the text. More information about this data collection is presented in my earlier studies (McClure 2016:7–8, 2018:28–30).
The main difference between the data used in my earlier studies (McClure 2016, 2018) and the data used in this study is that the overall dataset for Jesus’s social network has been separated into four datasets, one for each Gospel. The data were coded based on what is canonically accepted in each Gospel, including passages that are absent in some early manuscripts; for example, the additional endings in Mark (Coogan et al.: 1824–25) and the narrative about a woman caught in adultery in John (Coogan et al.: 1895–96) are included in Mark’s and John’s social networks, respectively. Each dataset only contains the actors and the relational ties present in that Gospel. In developing the Gospel-specific social networks, I started with the entire social network from my 2018 study and then excluded actors missing from that Gospel and all of their ties. Next, I examined each of the remaining ties to see which ties were present in that Gospel. Lastly, I analyzed the interpersonal connections between each pair of actors in that Gospel in order to determine if there were more positive or negative interpersonal connections. Like in my earlier studies (McClure 2016:8, 2018), if there were more positive connections, the tie was positive; if there were more negative connections, the tie was negative. I checked the coding for each Gospel-specific social network three times.
The networks that are presented and analyzed in this study include all relational ties among actors who interact with Jesus, for each Gospel. To look at larger structural dynamics and to maintain consistency with my previous studies (McClure 2016, 2018), the analyses do not look at subsets of the network or vignettes for a passage, and they do not look at the time order or development of ties (e.g., in John 12, Greeks ask Philip to see Jesus, Philip tells Andrew, and Andrew and Philip together tell Jesus). Table 1 describes the composition of each Gospel’s social network. The actors present in each Gospel are listed by category in Appendix A.
Composition of the Gospels' Social Networks
Notes: Some percentages do not add up to 100% due to rounding error.
Methods
This study uses three social network techniques, one for each research question. The method used to examine overlap among the social networks of the four Gospels is QAP correlation. The Quadratic Assignment Procedure (QAP) can be used to determine interrelations among matrices (i.e., social networks) (Prell: 202; see Felmlee & Faris 2016 for an example of QAP analysis). QAP correlations examine, for example, whether a tie in one social network is also present in another social network, for all ties. A positive correlation indicates that a tie in one social network is likely to exist in another social network; a negative correlation indicates that a tie in one social network is not likely to exist in the other social network (Prell: 202). For the present study, QAP correlations are used to determine if there are similar relational patterns when comparing the Gospels’ social networks. In order to correctly calculate the QAP correlations, all actors, even if they are not present in every Gospel, need to be present in every Gospel’s social network. For the QAP analyses only, actors who are not present in a Gospel (e.g., there are no Sadducees in John) are included as isolates, or as actors with no ties to other actors (Wasserman & Faust: 100), in that social network.
When analyzing the prominence of actors within a social network, centrality measures are typically used (Wasserman & Faust: 169). Network centrality focuses on the extent to which actors are “well-connected” (Scott: 84) in a social network, and social network analysts have developed many measures which quantify different aspects of network prominence (Wasserman & Faust). Centrality measures are typically designed to be compared within, not between, social networks; when comparing centrality measures across social networks of contrasting sizes, the centrality measures need to be adjusted, or normalized (e.g., Prell: 99–100). This study uses two normalized measures. The first is degree centrality, or the number of relational ties per node (Borgatti, Everett, & Johnson: 165), whose normalized form can be interpreted as the percentage of other actors in a social network with which an actor has ties (Prell: 100). The second is eigenvector centrality, a modified form of degree centrality that measures popularity within a social network by gauging not only how many ties an actor has but also the number of ties that its alters (the other nodes to which it is tied) have (Borgatti et al.: 28, 168). More specifically, eigenvector centrality “is the sum of an actor’s connections to other actors [alters], weighted by [the alters’, or other actors’] degree centrality” (Prell: 101). With eigenvector centrality, an actor “with small degree could have a higher score than [an actor] with high degree if the first [actor’s] friends are very popular while the second [actor’s] friends are not” (Borgatti et al.: 168). The numerical values for eigenvector centrality, both in its unnormalized and normalized forms, do not have as intuitive interpretations as degree centrality. Using normalized measures in this study is beneficial because the Gospels’ social networks vary by size and doing so reduces the likelihood that observed differences in centrality are due to differences in network size (Prell: 99–100). Both measures are used to increase the likelihood that observed differences are meaningful and not just statistical artifacts of one centrality measure.
This study explores the relational structure of each Gospel’s social network by identifying subgroups, or smaller, cohesive, and interconnected clusters of actors (Borgatti et al.: 181). Within social network analysis, there are many ways to identify subgroups (Wasserman & Faust), and this study uses the Clauset-Newman-Moore algorithm within NodeXL. This algorithm divides a social network into mutually exclusive subgroups (i.e., actors can only be in one subgroup) in such a way that it maximizes ties within clusters and minimizes ties between clusters (Clauset, Newman, & Moore: 1).
My previous analyses of Jesus’s social network have analyzed patterns of positive and negative relational ties (Mc-Clure 2016, 2018). The analyses for this study focus on whether ties exist, not whether they are positive or negative. Figures of the Gospels’ social networks, however, depict whether ties are positive or negative.
Results
Network Overlap
The first research question concerns similarities in relational patterns among the Gospels’ social networks. The QAP correlations for this analysis indicate the extent to which the Gospels’ social networks overlap, and they are presented in Table 2. All of the QAP correlations in Table 2 are positive, and positive correlations indicate that, for all relational ties, a tie in one social network is likely to be present the other social network (Prell: 202). The QAP correlations are stronger among the three Synoptic Gospels (ranging from 0.685 to 0.868), with Matthew and Mark having the highest level of overlap; they are weaker between each of the Synoptic Gospels and John (ranging from 0.347 to 0.408), indicating that there is more similarity in the relational patterns of the synoptic Gospels than there is between each of the synoptic Gospels and John. QAP correlational analyses produce results that are consistent with the levels of overlap that biblical scholars have observed among the Gospels (e.g., Aland; Nickle: 89–90).
QAP Correlations among the Gospels’ Social Networks
All correlations are statistically significant at p<0.001.
Differences in Actor Centrality
In order to address the second research question about whether specific Gospels depict actors as more or less prominent, this study compares normalized centrality scores from the Gospels’ social networks. Degree centrality is based on each actor’s number of relational ties (Borgatti et al.: 165), and eigenvector centrality measures the extent to actors are connected to other actors “that are themselves well connected” in the social network (Borgatti et al.: 168); these measures were normalized to allow comparison across the Gospels’ social networks, which differ in size (Prell: 99–100). Table 3 presents median normalized centrality scores for the three main categories of actors in Jesus’s social network, as well as for women; median, not average, values are presented because some of measures are skewed. While there are many differences among the centrality scores in this table, four stand out: (1) Jesus’s family and followers are less prominent in Luke; (2) the civil and religious authorities are more prominent in John; (3) stigmatized people are more prominent in John; (4) women are more prominent in John.
Normalized Centrality Scores by Category
Jesus’s family and followers are less prominent in Luke than in the other Gospels, both in terms of the percentage of actors and the prominence of the actors with which they have relationships. Understanding why involves differentiating between whether Jesus’s family and followers were present in Luke and at least one other Gospel or only in Luke. For Jesus’s family and followers, actors who are present in Luke and at least one other Gospel have similar levels of centrality in Luke as in other Gospels; for example, Simon Peter and John, son of Zebedee, are not substantially less prominent in Luke than in the other Gospels. For Jesus’s family and followers that are present in Luke and at least one other Gospel, the median degree centrality is 0.146, and the median eigenvector centrality is 29.889. However, Jesus’s family members and followers that are depicted only in Luke (e.g., Anna, Joanna, Susanna, Symeon, etc.) have substantially lower levels of centrality: their median degree centrality is 0.034, and their median eigenvector centrality is 5.192. Additionally, Luke has many unique actors among Jesus’s family and followers. Unique actors make up 4.2% of Jesus’s family and followers in Matthew, 0% in Mark, 20% in Luke, and 9.5% in John. Jesus’s family and followers are less prominent in Luke because they include a larger number of less prominent actors that are unique to Luke.
The civil and religious authorities are more prominent in John. While there are authorities who are present in all of the Gospels and who are substantially more prominent in John, like the chief priests and the Pharisees, the best example for demonstrating this group’s prominence involves an actor unique to the Gospel of John—“the Jews.” John uses this term to refer to the Jewish religious leaders without differentiating between more specific types of leaders (Powell: 121; Beutler: 146–48; e.g., John 5, 7, 8, 10). As I have stated earlier, I have “[referred] to this group as the ‘Jewish leaders’ instead of ‘the Jews’ because the term ‘Jew(s)’ can have a derogative connotation” (McClure 2016: 6). The Jewish leaders are the most prominent of the civil and religious authorities in John. In addition, they are substantially more prominent than the most central authorities in other Gospels, as is demonstrated in Table 4, which presents the normalized centrality scores for the most prominent civil and religious authorities in each Gospel. The normalized centrality scores indicate that the Jewish leaders are relationally tied with a higher percentage of actors and also with more prominent actors in John, while the most central civil or religious authorities in synoptic Gospels are tied with lower percentages of actors and less prominent actors. The Jewish leaders’ prominence in John reflects their extensive relational ties to a variety of actors, including Jesus, his disciples, John the Baptist, Lazarus and his sisters, other civil and religious authorities, and a variety of stigmatized people.
Normalized Centrality Scores for Prominent Civil and Religious Authorities
John also depicts stigmatized people more prominently. Across the Gospels, stigmatized people have low levels of centrality, and John does not depict them as prominently as it depicts Jesus’s family and followers or the civil and religious authorities. However, compared to the synoptic Gospels, stigmatized people in John have ties with a greater percentage of its social network and with more prominent actors. In the synoptic Gospels, many stigmatized people only have relationships with Jesus and maybe one other person. The percentage of stigmatized people who have ties only with Jesus is 30% in Matthew, 57% in Mark, 36% in Luke, and 0% in John. The percentage of stigmatized people who at most have ties with Jesus and one other person is 78% in Matthew, 76% in Mark, 48% in Luke, and 45% in John. When stigmatized people have relational ties with actors other than Jesus in the synoptic Gospels, these ties are typically with other stigmatized people; however, almost all of John’s stigmatized actors have ties with actors who are not stigmatized. In contrast, the percentage of stigmatized actors who only have ties to Jesus and, if they have other ties, to other stigmatized people is 61% in Matthew, 67% in Mark, 70% in Luke, and 9% in John. Stigmatized people in John are more prominent because they have relationships with a greater percentage of actors and are more likely to have relationships with actors that are more central, like Jesus’s family and followers and the civil and religious authorities.
The final difference in centrality among the Gospels involves women, who are more prominent in John. There are a number of women who are more prominent in John than in the synoptic Gospels, like Mary Magdalene and Mary and Martha of Bethany. However, the best example of women’s higher prominence in John is Mary, the mother of Jesus. In the synoptic Gospels, Mary interacts mainly with family members—Jesus, his brothers, and Joseph—and others from Jesus’s infancy narratives, including wise men (Matt 2:1–12), shepherds (Luke 2: 8–20), and Symeon (Luke 2:25–35). As a point of clarification, it is important to note that I have coded Mary, the mother of Jesus, and Mary, the mother of James and Joses, as different actors (McClure 2016:19; Bauckham 2002: 298–99). In John, however, Mary interacts with a larger number of actors, including Jesus, his disciples, his brothers, a group of women present at Jesus’s crucifixion, and the disciple whom Jesus loved (John 2:1–2, 12; John 19:25–27). The centrality scores reflect Mary’s increased prominence in John. Her normalized degree centrality scores are 0.103 in Matthew, 0.078 in Mark, 0.090 in Luke, and 0.217 in John. Her normalized eigenvector centrality scores are 6.093 in Matthew, 5.335 in Mark, 7.100 in Luke, and 27.154 in John. Compared to the synoptic Gospels, John depicts Mary as having relationships with a greater percentage of its actors and with more prominent actors.
Differences in Network Structure
The final analysis investigates differences in the relational structures of the Gospels by comparing subgroups, or clusters of actors, across the Gospels (Clauset et al.: 1). The social networks for Matthew and Mark, which are presented in Figures 1 and 2, have similar relational structures with five subgroups and similar actors in each subgroup. The subgroups for all figures are numbeed, starting with Jesus’s subgroup and then moving counterclockwise. The first subgroup includes Jesus, John the Baptist, many stigmatized people, some civil and religious authorities (e.g., Herod Antipas, Pharisees, and Sadducees), and people present at Jesus’s crucifixion (two criminals, Simon of Cyrene, soldiers, etc.). The second subgroup includes the disciples, when coded as a group, and actors that had ties with them, while the third subgroup includes eleven individual disciples and actors that had ties with these disciples specifically. The fourth subgroup consists of people who were involved in Jesus’s arrest and trial, including Judas Iscariot, the high priest, chief priests, elders, scribes, and Pontius Pilate (Matt 26–27; Mark 14–15). The fifth subgroup includes Jesus’s family members, including Mary, Joseph (in Matthew), and his brothers, as well as people who interacted with them during Jesus’s infancy, like the wise men (Matt 2:1–12). This alignment, together with the QAP correlation between Matthew’s and Mark’s social networks, indicates significant levels of similarity between Matthew’s and Mark’s social networks.
Subgroups in Matthew’s Social Network
Subgroups in Mark’s Social Network
Legend for All Figures
Main categories
Circle: Jesus’ family and followers
Triangle: Civil and religious authorities
Square: Stigmatized people
Diamond: Others
Gender
Outlined Shapes: Men
Solid Shapes: Women
Tie Type
Lighter solid line: Mostly positive connections
Darker dotted line: Mostly negative connections
Node Size: Based on number of ties
Luke’s relational structure, which is illustrated in Figure 3, has five subgroups. The first includes Jesus, John the Baptist, stigmatized people, people related to Jesus’s crucifixion (criminals, Simon of Cyrene, centurion, soldiers, etc.), and almost all of the civil and religious authorities. The second subgroup includes the disciples when coded as a group and those who interacted with them, while the third subgroup includes the twelve individual disciples and people who interacted with specific disciples. The fourth subgroup includes ten lepers whom Jesus healed (Luke 17:11–19), and the fifth subgroup is mainly Jesus’s family members, like Mary, Joseph, and his brothers, as well as people who interacted with them during Jesus’s infancy, like the shepherds and Symeon (Luke 2:8–20, 25–35). The three synoptic Gospels’ social networks are similar in: clustering Jesus, John the Baptist, stigmatized people, and people related to his crucifixion into the same subgroup; categorizing individual disciples into a different subgroup than the group of Jesus’s disciples; having a subgroup mainly comprised of Jesus’s family.
Subgroups in Luke’ Social Network
There are two key differences in relational structure between Luke and the other synoptic Gospels. First, in Luke, the civil and religious authorities involved in Jesus’s arrest and trial are grouped with Jesus, John the Baptist, some other civil and religious authorities, stigmatized people, and people related to Jesus’s crucifixion, not in their own subgroup. A potential explanation for this grouping involves Herod Antipas, the ruler in Galilee during much of Jesus’s life (Jensen 2014). All three synoptic Gospels mention Herod Antipas beheading John the Baptist (Matt 14:–12; Mark 6:14–29; Luke 9:7–9). However, only Luke depicts Herod Antipas as participating in Jesus’s trial (23:6–12). This difference is reflected in Herod Antipas being moderately more prominent in Luke than in Matthew and Mark. His normalized degree centrality scores are 0.029 in Matthew, 0.031 in Mark, and 0.067 in Luke. His normalized eigenvector centrality scores are 4.286 in Matthew, 4.113 in Mark, and 7.266 in Luke. In Matthew’s and Mark’s social networks, Herod Antipas is placed in the first cluster due to his relational ties with Jesus and John the Baptist, while the civil and religious authorities involved in Jesus’s arrest and trial are placed in a separate cluster. However, in Luke, the civil and religious authorities involved in Jesus’s arrest and trial are placed in the same cluster as Herod Antipas and Jesus, and this difference is likely due to Herod’s role in and interactions during Jesus’s trial.
The second difference involves Luke including more unique actors than the other synoptic Gospels do, and it includes, in particular, the most unique stigma-tized actors and women. To compare, Matthew has eight unique actors, which comprise 11.6% of its social network. Mark has two unique actors, which comprise 3.1% of its social network. Luke has 26 unique actors, which comprise 28.9% of its social network. Of Matthew’s stigmatized actors, 26% (6 of 23) are unique. Of Mark’s stigmatized actors, 10% (2 of 21) are unique. Of Luke’s stigmatized actors, 48% (16 of 33) are unique. Of Matthew’s female actors, 9% (1 of 11) are unique. Of Mark’s female actors, 0% (0 of 12) are unique. Of Luke’s female actors, 38% (6 of 16) are unique. Many of these actors are grouped with Jesus in the first subgroup, and ten lepers, whom Jesus healed (Luke 17:11–19), form their own subgroup.
The relational structure of John, which is depicted in Figure 4, differs from those of the synoptic Gospels. John has four subgroups, and key actors in the first subgroup include Jesus, some stigmatized people, and civil authorities and others involved in Jesus’s crucifixion (Pontius Pilate, soldiers, two criminals, etc.). The second subgroup includes Jesus’s disciples when mentioned as a collective, some individual disciples, the women who were present at Jesus’s crucifixion and burial, and Jesus’s mother and brothers. Key actors in the third subgroup include the religious authorities, stigmatized people who had negative interactions with them, and Judas Iscariot. The fourth subgroup includes two of Jesus’s disciples, Andrew and Philip, and people whose sole interactions with Jesus’s disciples involved one or both of them, like John the Baptist (John 1:35–40), a boy with five loaves and two fishes (John 6:8–9), and Greeks (John 12:20–22). The main similarity among the relational structures of the four Gospels’ social networks involves the placement of Jesus with stigma-tized people and with people related to his crucifixion. However, there are a variety of differences between the relational structure of John and those of the synoptic Gospels.
Subgroups in John’s Social Network
When compared to the relational structures of the synoptic Gospels, there are four key differences in John’s relational structure, two of which concern Jesus’s family and followers. The first involves clustering Jesus’s disciples, women followers, and family in the same subgroup. This clustering reflects a greater number of relational ties among these actors in John; examples include: Jesus, his disciples, his mother, and his brothers traveling to Capernaum after the wedding during which Jesus turned water into wine (John 2:1–12); the presence of the disciple whom Jesus loved, Jesus’s mother, and other women at Jesus’s crucifixion (John 19:25–27). Second, individual disciples (other than Judas Iscariot) are placed into two separate subgroups. Many of the disciples are clustered with the group of disciples, the women who followed Jesus, and Jesus’s family members. However, Andrew and Philip are in a different subgroup because they serve as intermediaries between Jesus and other actors in three situations: (1) at the beginning of Jesus’s ministry, John the Baptist points Andrew, one of his disciples, toward Jesus (John 1:35–40); (2) Andrew brings to Jesus the boy with five loaves and two fishes, which Jesus miraculously uses to feed thousands of people (John 6:8–9); (3) when some Greeks desire to see Jesus, they approach Philip, who then, with Andrew, tells Jesus (John 12:20–22).
There are two additional differences related to stigma-tized people within the relational structures of John and the synoptic Gospels. The first difference concerns the types of subgroups into which stigmatized people are placed. In the synoptic Gospels, stigmatized people are typically in subgroups where they are predominant. In Matthew and Mark, most (83% and 81%, respectively) of the stigmatized actors are in the one subgroup where they are predominant. In Luke, 88% of stigmatized actors are in the two subgroups where they are predominant. However, in John, a significant number of stigmatized actors are in a subgroup where the religious authorities are predominant. In John, 55% of the stigmatized actors are in a subgroup where they are predominant, but another 36% of stigmatized actors are in a subgroup where the civil and religious authorities, and the religious authorities in particular, are predominant. This difference reflects how a greater percentage of stigmatized actors have ties, and negative ties in particular, to the civil and religious authorities in John than in the synoptic Gospels. The percentage of stigmatized actors who have ties to at least one civil or religious authority is 22% in Matthew, 24% in Mark, 18% in Luke, and 64% in John. Almost all of these ties are negative.
Second, in contrast to Matthew and Mark but similarly to Luke, 30% of John’s social network is comprised of unique actors, and many of John’s unique actors are stigmatized people and women. Of John’s 14 unique actors, seven are stigma-tized, and four are women. Of John’s stigmatized actors, 64% (7 of 11) are unique. Of John’s female actors, 50% (4 of 8) are unique. Most of the stigmatized people who are unique to John are depicted in its “signs” narratives, which portray Jesus’s miracles as signs of his divine identity (Smith 1995:107; Powell: 115). The stigmatized actors unique to John that are present in “signs” narratives include the nobleman’s sick son, whom Jesus healed in John 4; the man with an infirmity, whom Jesus healed in John 5; the man blind from birth, whom Jesus healed in John 9; Lazarus, whom Jesus raised from the dead in John 11 (Powell: 115). The four unique women in John include: the Samaritan woman in John 4 and the woman caught in adultery in John 8 (who is depicted in a narrative that is absent in many ancient manuscripts; see Coogan et al.: 1896), both of whom were stigmatized but not involved in John’s “signs” narratives; Mary the wife of Clopas and the sister of Mary the mother of Jesus, who were present with Mary the mother of Jesus and Mary Magdalene at Jesus’s crucifixion (John 19:25).
Discussion and Conclusion
There are many similarities and differences in how the Gospels depict Jesus’s social network, which reflect and even deepen the observations and insights of biblical scholars. The first finding is important, albeit obvious. The social networks of the synoptic Gospels are very similar, whereas John’s social network is distinctive. This study, however, supplements this well-established observation (e.g., Aland; Nickle: 89–90) with unique relational and structural insights. The synoptic Gospels have the highest level of relational overlap, and the levels of prominence with which they depict groups within Jesus’s social network (e.g., Jesus’s family and followers, the civil and religious authorities, stigmatized people, and women) are mostly comparable. They also have similar structures and patterns of clusters. However, there is significantly less relational overlap between John’s social network and those of the synoptic Gospels. John depicts key groups within Jesus’s social network, including the civil and religious authorities, stigmatized people, and women, as more prominent, and it also has a unique relational structure that is not analogous to the synoptic Gospels’.
In addition, the social network analyses corroborate some biblical scholars’ observations related to actors and groups that are more prominent in one Gospel than in others. First, Luke’s social network includes a significant number of stigmatized actors and women that are not present in other Gospels. These actors reflect how Luke strongly emphasizes the stigmatized people in society who were disadvantaged and sometimes excluded, and it includes many narratives about the poor, lepers, blind people, others with sicknesses and disabilities, tax collectors and other “sinners,” as well as Gentiles and Samaritans, who were often seen as outsiders in Jewish society (O’Toole: 109–10, 118–26, 129; Tannehill: 103, 132; Riches: 16–18; Knoppers). Furthermore, “Luke’s concern for women is a legitimate example of [its] overall concern for the oppressed…. Women [were] inevitably placed at a disadvantage in [Luke’s] patriarchal world” (Powell: 94). Second, the Jewish leaders are highly central and well-connected in John. John often refers to the Jewish religious leaders as a collective, typically failing to differentiate between more specific types of Jewish religious leaders (Powell: 121; Beutler: 146–48), and this lack of differentiation depicted the Jewish leaders as having more extensive relational ties than the more differentiated religious authorities in the synoptic Gospels. Third, women and particularly Jesus’s mother are relationally more well-connected in John than in the synoptic Gospels, and this corresponds with biblical scholars’ observations about the importance of women, and of Jesus’s mother in particular, in the narrative and text of John (Smith 2001: 210–11; Karakolis; Conway 2018: 221). A key difference, however, is that, while some biblical scholars argue that in John “women disciples are presented as being of equal standing with men” (Beirne: 220) or that the Gospel of John presents women as “clearly superior to the men” (Conway 1999: 205), this study’s centrality analyses indicate that, although women are more relationally prominent in John than in the synoptic Gospels, women are not more relationally prominent than men in John.
The social network analyses also provide an intriguing window into a cluster of Johannine relational dynamics best illustrated in its “signs” narratives (Smith 1995: 107). The theological purpose of these narratives is to reveal Jesus’s divine and messianic identity and to encourage people to believe in him (Guthrie: 72; Smith 1995: 107; Lee: 266). Scholars have noted that a key relational aspect of some, but not all, of the “signs” narratives involves opposition and antagonism toward Jesus by the Jewish religious leaders (Guthrie: 75–77; Smith 1995: 110–11; Lee: 267–68; e.g., John 5:16, 9:16, 11:45–53). However, the social network analyses illustrate that these narratives also depict tension and sometimes hostility between the religious authorities and the stigmatized people to whom Jesus was ministering (John 5:10, 9:24–34, 12:9–11). The stigmatized actors depicted in “signs” narratives that have negative interactions with religious authorities include the following: the man with an infirmity, whom Jesus healed in John 5; the man blind from birth, whom Jesus healed in John 9; and Lazarus, whom Jesus raised from the dead in John 11 (Powell: 115). These Johannine dynamics, which can be found in but are less prominent in the synoptic Gospels (e.g., Matt 9:10–13; Mark 2:15–17, Luke 7:36–50), are reflected in John’s social network by both groups’ higher levels of prominence and by the clustering of most of the civil and religious authorities and a significant number of stigma-tized people into the same subgroup. These dynamics more closely integrate stigmatized people into the narrative and relational structure of John and tie compassion toward stigma-tized people directly to Jesus’s identity.
Lastly, these social network analyses pinpoint particular actors whose narrative and historical roles have been discussed by biblical scholars but whose roles in the Gospels’ relational structures have needed further examination. One example involves Herod Antipas, who has a relatively obscure role in Matthew and Mark but who is depicted as participating in Jesus’ trial in Luke 23. This “minor [scene]” (Fitzmyer: 1480), whose historicity scholars have debated (Marshall: 854–55; Fitzmyer: 1478–79; Brown), not only emphasizes, for example, the enmity between and reconciliation of Pontius Pilate and Herod Antipas or the declarations of Jesus’s innocence by both the Roman governor and the Galilean ruler (Brown: 777; Bond; Jensen). This study uniquely highlights how the presence of Herod Antipas in Luke’s trial narrative reshapes Jesus’s social network in such a way that Herod serves as a bridge that draws together many of the civil and religious authorities involved in Jesus’s arrest and trial, which had been grouped in a separate cluster in Matthew and Mark, into a larger cluster with Jesus, other authorities, and many stigmatized people. Why does Luke present these additional interconnections among the civil and religious authorities which are absent in the other synoptic Gospels? A second example involves Andrew and Philip, who are not only mentioned more frequently in John (Freedman et al. 1992a: 242; Freedman et al. 1992b: 311) but who are also accentuated in John’s relational structure due to their unique role of bridging between Jesus and other actors. Why does John emphasize and present Andrew and Philip, in particular, in these bridging roles?
This study has limitations similar to those of my previous studies (McClure 2016, 2018). First, the Gospels are “ancient biographies” and are not “objective or balanced” (Powell: 7). They are also not comprehensive in providing information about every aspect of or period in Jesus’s life (ibid.). They do not include every interaction surrounding Jesus’s life and ministry; thus, the social networks analyzed in this study are not complete and do not contain all of Jesus’s relationships (McClure 2016: 15). However, these Gospels are “some of the earliest sources we possess” about Jesus (White: 98). The canonical Gospels present a particular orthodox Christian perspective about Jesus (Evans & Tov; Jenson), and there are debates about the extent to which they are historically accurate (Ehrman; Bauckham 2017). The social networks analyzed in this study reflect how the Gospels depict Jesus and the social relationships in which he was embedded (McClure 2016, 2018). Additional weaknesses have resulted from my decision to maintain consistency with my previous studies (McClure 2016, 2018). The data do not reflect more critical insights about, for example, the identity of the Beloved Disciple (Powell: 117; Charlesworth) or narratives that were not present in some early manuscripts (Coogan et al.: 1824–25, 1895–96). These decisions should not be interpreted as a rejection of critical perspectives, which I value. The analyses, like those in my previous articles (McClure 2016, 2018), also focus on the whole social network. However, I encourage other scholars to use social network analysis in order to examine, for example, relational dynamics in smaller vignettes, the development of ties over time, and other promising directions that I have not pursued in this study.
In conclusion, this study has sought to understand how the four Gospels depict the relational dynamics surrounding Jesus’s life and ministry using social network analysis. Investigating network overlap, the prominence of different actors, and the relational structures of the Gospels, the results indicate high levels of similarity between Matthew and Mark, some differences between Luke and the other synoptic Gospels, and substantial differences between John and the synoptic Gospels. While these statements should not be surprising to biblical scholars, this study supplements their textual, historical, and theological observations by exploring relational and structural patterns not previously examined. Social network analyses provide rich and intriguing illustrations of the Gospels’ relational dynamics and structures that hopefully will spark further discussion among biblical scholars.