The Sociology of Harry Potter: An Analysis of Adolescent Friendship Networks in the Harry Potter Series
The Sociology of Harry Potter: An Analysis of Adolescent Friendship Networks in the Harry Potter Series
The Sociology of Harry Potter: An Analysis of Adolescent Friendship Networks in the Harry Potter Series Katie Christie The College of Charleston Department of Sociology
Tables and Figures Table 1. Actor Degree Before Dumbledore’s Army…p. 22 Table 2: Reachability Before Dumbledore’s Army…p. 23 Table 3a. E.I. Index- Homophily by Gender Before Dumbledore’s Army… p. 24 Table 3b. E.I. Index- Homophily by Gender After Dumbledore’s Army… p.25 Table 4. E.I. Index- Homophily House Before Dumbledore’s Army… p. 26 Table 5. Closeness Centrlity Before Dumbledore’s Army… p.
27 Table 6a. Betweenness Centrality Before Dumbledore’s Army…p. 28 Table 6b. Betweenness Centrality After Dumbledore’s Army… p.29 Figure 1a. Friendship Ties Before Dumbledore’s Army…p.30 Figure 1b. Friendship Ties After Dumbledore’s Army… p.31
The Sociology of Harry Potter: An Analysis of Adolescent Friendship Networks in the Harry Potter Series Abstract Recent studies in sociology of literature have focused on the ways in which readers construct meaning in their lives based on the stories that they read. This study adopts the idea that literature can be analyzed sociologically, applying social theories to characters and plots, by examining stories as social models. Using the internationally acclaimed Harry Potter series as an example, this study will focus on two networks of adolescent friendship ties within the fictional school of Hogwarts, analyzing basic social network properties, such as density, reachability, and centrality.
Particular emphasis will be placed on the presence of homophily in friendship ties at Hogwarts. The character Harry Potter will also be singled out and analyzed in terms of power and connectedness. The main objective of this paper is to provide a model for analyzing contemporary fantasy novels geared towards adolescents, utilizing sociological network methods and procedures in order to put a spin on literary analysis.
Introduction In 1999, Harry Potter and the Sorcerer’s Stone was released in the United States. Since then, the seven-part series has gone on to sell millions of copies worldwide, earning author J.K. Rowling international fame. Millions of children around the world have become devoted fans to the series, and the term,“Potter-mania” has been used to refer to the way in which the series has swept through the world of children’s literature and burst forth into mainstream society. The character Harry Potter has become a cultural icon, and the record-breaking series has left its mark on contemporary culture.
The Harry Potter series is important not only because of its commercial and financial success in the publishing, marketing, and entertainment industries, but also because it has created characters with which contemporary children heavily identify. The Harry Potter series is more than a collection of popular children’s books; it can be viewed as a mirror of contemporary society. As such, the characters and events in the story take on a new meaning. The characters have special friendships and relationships with one another, and the majority of the events take place in a school setting. If Rowling’s fictional world is viewed through a sociological lens, the wizarding school, Hogwarts, where Harry and his friends attend, can be thought of as its own network environment.
The actors in this social network are the students at Hogwarts, and the relations that link the students are ties of friendship. To give the world of Harry Potter true sociological significance, it must be viewed as possessing the characteristics of any other social network. To achieve this, I review the literature that examines adolescent friendship networks, many within the context of schools.
To properly examine Hogwarts as a social network, the Harry Potter series must be analyzable and frameable in a sociological context. Wouter de Nooy (2001: 364) proposed
studying stories as social facts or models and applying social theories to characters and plots in the same way that those theories would be applied to social institutions. De Nooy argued that texts and stories offer models for interpersonal relations, and fictional stories serve as models in everyday life, shaping people's self-perceptions and the way in which they share experiences (de Nooy 2001: 365).
Similarly, Griswold argued that recent advances in the sociology of literature have focused on readers' construction of meaning, and research has indicated that there’s a “psychology of reading” involved whereby readers transform the words being read into mental associations that relate the story to real life experiences (Griswold 1993: 458). Therefore, stories and novels, such as Harry Potter, can be viewed as fields of research in need of further exploration.
In order to put the character, Harry Potter, his friends, and Hogwarts School of Witchcraft and Wizardry into a sociological context, I first examine literature that relates to adolescent friendship networks. Unfortunately, much of the research that exists on friendship networks is disconnected and suffering from inconsistencies in methodological procedures. Crosnoe (2000) reviewed contemporary research on the topic of adolescent friendship, and suggested using the concept of life-course theory in order to fuse existing data. Crosnoe focused on four broad themes found in contemporary literature.
Among his themes are friendships as developmental contexts, the nature and structure of friendship, and the social context of friendship (Crosnoe 2000: 377). In organizing my literature, I chose to use most of his themes, and expanded upon certain elements within each category. I also created sub-categories within the category of ‘Friendships as developmental contexts’ and ‘The social context of friendship.’ It should be noted that the themes identified by Crosnoe are by no means exclusive, as I found
multiple studies in which the research bridged these categories. However, for organizational purposes, I found Crosnoe’s themes to be extremely useful in putting existing data into themes. Friendships as Developmental Contexts This category deals with literature that focuses on individual analysis; in other words, the literature discussed uses a micro-level approach in analysis. There is a wide range of literature dealing with friendships and how individuals learn cooperation, forge new roles, identities, and shape their views based on these friendships (Crosnoe 2000:378). Berndt (1982) cited three explanations for the significance of adolescent friendships.
He found that biological, social, and cognitive forces are at work in shaping adolescent friendships (Berndt 1982: 1447-48). The social environment of adolescents is of particular interest and relevance. In early adolescence, the adolescent reaches a unique position because he is no longer a child, but also doesn’t yet have all the responsibilities or freedom that come along with adulthood (Berndt 1982: 1447). The majority of social interactions in this stage exist between close friends and peers, and friendships are egalitarian (Berndt 1982: 1447-48). In the fictional setting of Hogwarts, social interaction occurs at an even higher rate: the children are removed from their homes and are forced into constant interaction with friends and peers.
Due to the unique setting at Hogwarts, an environment that serves as both school and home for the students, socialization and friendship play an even larger role in the daily lives of the students. It is therefore important to consider the ways in which friendships are formed, and the role friendships play in the setting of Hogwarts. Berndt (1982, 1995) discussed the role of intimacy within friendships, and how adolescents build self-esteem by sharing feelings and information with friends. In a later study, Berndt (1995: 1327) examined the influence of friends
on development as it related to school adjustment and found that adjustment is affected by the characteristics and quality of friendships.
The Nature and Structure of Adolescent Friendship Research in this category can be divided into two categories: the nature of adolescent friendships, which can be further divided into the elements or characteristics of friendship and the quality of friendships, and the structure of friendships, which has slightly more complicated sub-divisions (Crosnoe 2000: 377). Studies have been done in which the focus is on particular features of adolescent friendship, such as George and Hartmann’s 1996 study of popularity. Characteristics or elements of adolescent friendships may include age, gender, reciprocity, identity, similarity, and stability.
George and Hartmann (1996) conducted a study where students were given the opportunity to rate their classmates on how much they were liked by their peers. The purpose of this research was to examine three popularity groups: children were defined as popular, average, or unpopular, in order to determine friendship network characteristics and friendship prevalence (George and Hartmann 1996: 2304). George and Hartman found that unpopular children were friendless in comparison to other children. However, they also concluded that prior research done on popularity overestimated friendlessness in unpopular children (George and Hartmann 1996: 2311).
Another interesting discovery was that approximately 75% of children who were popular had friends who were approximately the same age (George and Hartmann 1996: 2312). Unpopular children, on the other hand, were more likely to be friends with children who were different ages (George and Hartmann 1996: 2312).
There is also a significant amount of research done on the structure of adolescent friendship. I chose to highlight social networks and friendship groups, as they are most relevant to my topic. Social network analysis includes information on cliques and isolates. Shrum (1987) argued that examining liaisons in the social networks of adolescents, rather than focusing solely on cliques as the hub of peer relations, is important in social network analysis. A liaison can be defined as an individual belonging to a social system, who serves the purpose of linking outside members to the system (Shrum 1987: 218).
Clique members are defined as participants in a close-knit group of interacting peers (Shrum 1987: 218). Shrum argued that distinguishing liaisons from cliques is especially important for contemporary theories of the school as a context for developing peer relations (Shrum 1987: 219). Social networks can also be viewed as changing, adaptive processes (Cairns, et al 1995: 1330). Cairns, et al found (1995: 1343) that fluidity in the strength of friendships and group members is an important feature of social relationships.
Social Context Research with an emphasis on social context examines various aspects of adolescent friendship as they are embedded in a larger social context. Examples of this include homophily, race, gender, and physical location/geography. McPherson, et al (2001) defined homophily as the tendency for people to have more contact with people who are similar to them than to people who are dissimilar. In the United States, the biggest divide in social networks is between race and ethnicity (McPherson, et al 2000: 420). The idea of homophily is important because it says that people are more likely to confide and share with individuals who are similar to them.
When the principle of homophily is applied to adolescents, the literature suggests that teenagers tend to
associate with other teenagers who are similar in terms of behavior or achievement (McPherson, et al 2000: 428). Sources of homophily include space, family ties, and organizational ties (this could include schools) (McPherson, et al 2000: 429-434). Shrum, Cheek, and Hunter (1988) examined racial and gender homophily in children in association with grade level. Their findings showed that racial homophily increased as grade level increased, although the reverse was true for gender homophily (Shrum, Cheek, and Hunter 1988: 237). For gender homophily during elementary school, boys were more likely to be friends with other boys than girls were with other girls, but this reversed in middle school.
In other words, middle school girls were found to have a significant increase in same-gender preference, leading to the conclusion that there is more exclusion and intimacy in girl friendships (Shrum, Cheek, and Hunter 1988: 236). One of the most interesting and significant findings of this study was that children develop an early awareness of racial identity, and as they age, there is an increasing tendency to develop homophilous friendships (Shrum, Cheek, and Hunter 1988: 236). The topic of racial homophily, in particular, is an interesting one to examine in the environment of Hogwarts, because racial divides are not based on skin color, but on blood type.
Racial homophily based on blood type, along with gender homophily, will be discussed at length later in this study. Summary of Literature Based on the substantial amount of literature available on adolescent friendships, several themes appear to be consistent. Building on Berndt's idea that the majority of social interactions in early adolescence are between close friends and peers, and these friendships tend to be
egalitarian in nature, this study will examine whether or not this is true within the setting of Hogwarts, and how or if this changes over time. In other words, this study will determine the extent to which homophily plays a role in the friendship ties among students in Hogwarts by examining the interactions occurring between friends who are similar in terms of race, gender and organization. For the purpose of this study, race will consist of purebloods, or individuals belonging to a genealogy that consists of no "muggles," or humans, and non-purebloods, which constitutes individuals who have mixed parentage.
By combining Berndts' ideas with Shrum, Cheek, and Hunter's findings that young adolescents are more likely to spend time with members of the same sex and same race, it is expected that this study will confirm the theory that young adolescents interact primarily with friends who are similar to them, especially in terms of race and gender. A second objective of this study is to determine how much power and influence the character, and actor, Harry Potter has in his network of friends. The actor Harry Potter is also expected to have the most connections in the network. The final and most important objective of this study is to test social network procedures on the friendship networks present in Hogwarts, in order to set the stage for further sociological analysis of literature.
Because Harry Potter is a prominent series in contemporary society, it is important to recognize the power that the books have in shaping children’s ideas about friendship, love, and even notions of good and evil. Using friendship ties as the relations linking the students (the actors) of Hogwarts (the network) to one another, particular emphasis will be placed on homophily, connectedness, network size, and distance between actors. Although there is a great deal of literature available on the subject of adolescent friendship, many studies highlight inconsistencies in similar studies. Also, weaknesses appear in the methodological procedures used.
There is a great deal of research available on cliques and group level analysis, but literature
consisting of liaisons or other linking actors could be useful. Literature consisting of sociological analyses of children’s books seems to be rare, but the importance of Harry Potter in youth culture cannot be denied. The aim of this research is to bridge the world of fantasy that children find so compelling and highlight the similarities that exist between actual adolescent friendship networks and the social networks in the fictional setting of Hogwarts. Methods The data collection process of this study involved unique procedures. Two networks were examined, and the actors in the networks were students at Hogwarts School of Witchcraft and Wizardry.
Because the two networks examined did not consist of people or organizations, but characters in literature, the samples were chosen based on the characters' relevance to the story. In other words, characters were picked as actors according to the amount of information that was available in the series about the character, and whether or not they were featured frequently enough to be considered key actors. To begin collecting data, a pool of potential actors was devised. A list of actors for the entire network of Hogwarts could not be determined, because no complete list of characters is provided within the book.
Hogwarts consists of seven grades, grades one through seven, but the number of students within each grade is not specified. The data collection process, therefore, involved identifying a pool of actors to be examined, based on their relevance in the series. For the first set of data, or Network A, an actor was classified as relevant and included if he or she was mentioned throughout the series, beginning with the first book. A preliminary list of actors was drawn up by examining each character who was named in the first book during the sorting ceremony at Hogwarts. The sorting ceremony
alphabetically listed the names of many of the characters in Harry's grade. From this pool of actors, relevance was determined based on which characters went on to be featured regularly in consecutive books. Any character who did not appear in each story was considered to lack sufficient information for further analysis and was excluded from the list of actors. In order to determine which characters were mentioned frequently enough to be included, the online source Wikipedia was used. Although Wikipedia is not an academic source, it provided basic information on each of the characters, including which books featured particular characters, which houses they belonged to, etc.
Wikipedia featured an alphabetical list of all the characters mentioned in the Harry Potter series, and by searching the list, information could be found in sufficient detail about each actor.
The next step in the data collection process was to identify the relational ties between actors in the Hogwarts network. With few exceptions, the actors who are identified in the series are mentioned in relation to Harry Potter, and the vast majority of characters with regular mentions are in the same grade with him. For this reason, all actors chosen were in the same grade as Harry Potter. The character Harry Potter starts Hogwarts in the first book of the series, Harry Potter and the Sorcerer's Stone, at age eleven. In the six books that follow, Harry Potter ages one year, and advances one grade.
All students in the series belong to one of four houses, which is determined in the Sorting Ceremony in the first year at Hogwarts. The four houses are Gryffindor, Hufflepuff, Ravenclaw, and Slytherin. Rowling identifies twenty-three characters during the Sorting Ceremony for Harry's grade in Book One, but only sixteen characters play an active role in the series in the first four books (Rowling 1997). These sixteen characters are the actors comprising the first data set, and the relations linking these actors are ties of friendship.
An actor was classified as a friend, and a tie was formed when actor A spent time outside classes, engaged in conversation, shared leisure time, or otherwise interacted in an intimate setting, with actor B. Being members of the same house or sharing classes together was not considered sufficient grounds for a friendship tie. It should be noted that this measure of friendship ties allowed for friendships to exist inside as well as outside houses. However, due to the structure of Hogwarts, there are considerably more opportunities for members of the same house, or inter-house members, to interact than members outside of houses, or extra-house members.
Members of the same houses eat lunch together, share a common room, and attend classes together.
Two networks were examined in this study. During the preliminary stages of research, it was observed that friendship patterns change throughout the series, which is to be expected as the characters age. However, differences in friendship ties became especially pronounced after the formation of Dumbledore’s Army in Book Five. Dumbledore’s Army is a club that was formed in Harry Potter and the Order of the Phoenix by Harry Potter, Ron Weaslely, and Hermione Granger for students wishing to practice using defensive spells. The first network, Network A, therefore, consisted of actors who played a prominent role throughout the series, and were friends before the formation of Dumbledore's Army.
Network A also represented early adolescence, as the actors in this network were between the ages of eleven and fifteen in the first four books. Network B referred to the ties of friendship formed after Dumbledore’s Army. In Book Five, Rowling chose to introduce six new characters in Harry’s grade. These characters go on to be featured regularly in the series, so they were included in Network B. The two friendship networks consisted of symmetric, undirected ties and a binary scale of measurement. To measure friendship ties between nodes, the presence or absence of friendship ties was determined
based on the question, "Is node A friends with node B?" A score of 1 was given for a tie of friendship, and a score of 0 was given where a tie was absent. All ties of friendship were symmetric and undirected. In other words, if actor A considered B to be a friend, B also considered A to be a friend. The network size for the first data set, friends before Dumbledore's Army, or Network A, was found to be sixteen by performing a simple count of nodes. The network diameter was found by running the Ucinet procedure for identifying geodesic distances. Geodesic distances refer to the shortest possible path from one actor to another, usually resulting in the most efficient connections between actors.
The diameter of the network was the furthest distance between actors, and in the case of friendship ties before Dumbledore's Army, the number was two. The second network, or Network B, was comprised of all of the actors present in the first network, with the addition of six actors who were not featured until Book 5. The network size of the friends after the formation of Dumbledore's Army was twenty-two. The diameter of this network was one. The change in the diameter between the first network and the second can be attributed to the removal of steps between actors when more friendships were formed due to the actors' involvement in Dumbledore's Army.
These results will be further explored and discussed in the Analysis and Results section of this study.
Analysis and Results In order to analyze the two friendship networks at Hogwarts, basic network properties of connection and distance were first examined. A total of four measures were used relating to
connection and distance: network size, actor degree, density, and reachability, where network size and actor degree were considered to be basic demographics, and density and reachability were measures of connection. To determine the size of Network A and Network B, a simple count of nodes was performed. The total number of nodes for Network A was sixteen.
This means that a total of sixteen ties were possible, since the network is symmetric. In other words, by using the equation, k* (k-1) / 2, where k is number of nodes, and plugging in the count of nodes, 16, the total number of ties was found. It is important to establish a count of ties in order to determine the networks' density. The total number of nodes for network B was twenty-two, and the total number of ties, therefore, was twenty-two. The difference in sizes between networks A and B is important because as the number of nodes increases, so does the number of potential relationships.
This generally means that the level of complexity increases as the size of a network increases. Figures 1a and 1b show the differences in network structure before the formation of Dumbledore's Army and after.
Actor degree can be defined as a count of ties linking a node to other nodes, where the highest possible value is (k-1). Actor degree is useful in determining how connected an individual is in relation to the overall social structure. The actors with the highest degree for Network A (see Table 1a) were Harry Potter, and Ron Weasley, with Crabbe, Malfoy, and Goyle coming in closely behind. However, if we compare Network A to Network B, we see that the actor degree changes significantly: actors 1-17, i.e., the actors in Dumbledore’s Army, had a degree of fifteen, and actors 18-22, the actors not in Dumbledore’s Army, had a degree of five.
These results are not surprising, because with the formation of Dumbledore's Army, friendships that did not exist initially within the houses of Gryffindor, Hufflepuff, and Ravenclaw, sprang
up, so that every actor in those three houses was connected with every other actor. It therefore follows that these actors would become more connected with one another. Density is the proportion of ties present to the total number of ties. Density is an important trait to examine, because generally the denser a network is, the greater the mobilization of resources and spread of information. The density of Network A was found to be .2250, which means that 22.5% of all possible ties were present. As would be expected, the density increased for Network B, which had a density of .5844. These results show more than a 30% increase from Network A to Network B.
The final measure conducted for connection was reachability. Reachability can be defined as the set of connections or links that exist between actors, even if those actors are not adjacent or directly connected. In a symmetric network, if actors are not reachable, the network can become divided and sub-populations can form. A division was encountered in this study: a major rift existed between the actors belonging to the Slytherin house, and all other actors. This is true for both Network A and Network B (refer to Figures 1a and 1b). For Network A, the actors Brown and Patil are not reachable to any other actors, and gaps exist between actors within the house of Gryffindor.
However when Network B is examined, (see Table 2) we see that all actors are equally reachable, and a natural block formation appears, showing the divisions between Dumbledore's Army members and non- Dumbledore’s Army members. The results from the Reachability measure lead us to the conclusion that the friendship ties that sprung up due to the formation of Dumbledore's Army led to a cohesion in the friendship network, further resulting in direct paths being formed between all actors in that group. Therefore, due to the high level of connection between members of Dumbledore's Army, we would expect the flow of information and mobilization of resources to be high.