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.
perfect example of this occurred at the end of Book Five. In Book Five, Harry was nearly attacked on the Hogwarts Express at the end of the school year by Malfoy, Crabbe, and Goyle. Before they are able to attack him, however, several members of Dumbledore’s Army, including Hannah Abbott, Susan Bones, Anthony Goldstein, and Terry Boot, who were not friends of Harry’s prior to that year, jump to the rescue and prevent Harry’s attack (Rowling 2003). This example illustrates the differences that existed after the formation of Dumbledore’s Army: a significant increase in camaraderie and a banding together, or higher level of mobilization, could be observed.
Results The first objective of this study was to identify the presence of homophilous friendship ties by determining if young adolescents interact predominantly with friends who are similar in terms of race and gender. Organizational association, or association with house-members, was also examined. The methods used to test this procedure were fairly straight-forward. To test the hypothesis, external group ties and internal group ties were measured by running the E-I Index procedure on Ucinet. Actors were coded separately for Gender, Race, and House and were used as attribute vectors for comparing Network A and Network B.
Race was measured by coding individuals with a pure-blood status as one, and individuals without a pure-blood status as zero. E-I Index was used to measure the percentage of ties being sent outward and the percentage remaining internal. The E-I Index was run for Network A and Network B, to compare the changes in homophily after the formation of Dumbledore’s Army. For gender, the re-scaled E-I Index was found to be -.185. The negative value shows a weak tendency for internal ties. However, when the same procedure was run for Network B, the E-I index was found to be .083,
showing a higher tendency for external ties. Tables 3a and 3b show which actors have the highest and lowest internal and external ties. To further investigate gender homophily, observations were performed by simply examining the network structure of friendship ties for Network A compared to Network B. The friends in Network A consisted of a total of six females and ten males. The node Herminone Granger had a total of three ties, all of which were connected to males. The node Hannah Abbott had two ties, both of which were connected to boys. Both Milicent Bulstrode and Pansy Parkinson had four ties, three of which were to boys, and one of which was to a girl.
The nodes Lavender Brown and Parvati Patil stood out: not only do they constitute a sub-division in the network, but they are tied to only each other, showing them as isolates from the rest of the network. In the case of Network A, it does not appear that females show a tendency towards friendship with members of the same sex. However, when the boys in the network were examined, there was a considerable difference. The actors Dean Thomas and Seamus Finnigan each had three ties, all of which were directed at males. Ron Weasley had a total of five ties, four of which were directed at males. The actor with the most ties, Harry Potter, had a total of seven ties, and six of these seven ties were directed towards males.
Although there was a disproportionate ratio of males to females, it is clear that within Network A, there was a discrepancy between males' tendency to be-friend other boys, and girls' tendencies to befriend other females. These findings are therefore partially inconsistent with Shrum, Cheek, and Hunter's observations that gender homophily is strong among young adolescents: although males showed a strong tendency for gender homophily, females did not. Shrum, Cheek, and Hunter found females to be much more likely to associate with same-sex friends beginning in early adolescence, and the reverse appears to be true for females at Hogwarts.
When ties were measured for house membership, Network A was shown to have an E-I Index of -.852, representing a strong tendency towards internal ties. The results were even stronger for Network B, which had an E-I Index of -1. However, this can be attributed to Slytherin’s complete separation from the other houses (see Table 4). Slytherin was the only house after the formation of Dumbledore’s Army to not have external ties. Finally, to measure racial homophily, the E-I Index was run using the blood status attribute vector. The results from this test showed a re-scaled index of -.037, representing a very weak tendency toward internal ties for Network A.
This means that, within Network A, ties of friendship were more likely to be to people of the same race (in this case, pure-blood or non-pure-blood), but only slightly. These results changed significantly for Network B, which showed a strong tendency towards internal ties. These results show that actors were much more likely to be friends with people of the same race after the formation of Dumbledore’s Army. This is consistent with Shrum, Cheek, and Hunter's claims that racial homophily increases with age.
In order to determine if Harry was the most central and powerful actor in Network A and Network B, Freeman’s measures of Closeness Centrality and Betweenness Centrality were calculated. The idea behind closeness centrality is that actors who are closer to others have greater power. Betweenness centrality is the idea that actors who serve as intermediaries between other actors have the most power. According to Freeman’s Closeness Centrality measure, the most powerful actors in Network A were Harry Potter, Ron Weasley, and Ernie Macmillan (see Table 5). The weakest actors were Parvati Patil and Lavender Brown.
However, when the same measure was performed for Network B, all members of Dumbledore’s Army had the same amount of closeness centrality, which was substantially more than the non-members. The change from Network A to Network B can be attributed to the decreased distance between
actors from Network A to Network B. In other words, all actors were at an equal distance from one another, so no one actor had a greater amount of power. Freeman’s Betweenness Centrality measure yielded very interesting results between Network A and Network B. In Network A, the most powerful actors for betweenness were Harry Potter, Ernie Macmillan, Justin Finch-Fletchley, and Ron Weasley (see Table 6). All other actors had the same amount of betweenness power. However, when these results were compared to Network B, it was found that no actors had betweenness power: all actors had a betweenness power of 0.
The reason for this is that the creation of Dumbledore’s Army removed intermediaries, neutralizing the power of Harry Potter and the other formerly powerful nodes. It should also be noted that the actors who were not a part of Dumbledore’s Army had the same amount of power between actors because they also had no intermediaries: all actors were the same distance, a distance of 1, from other actors. The results of Freeman’s Centrality measures show that Harry Potter did have the most power in terms of betweenness and closeness prior to the formation of Dumbledore’s Army, but that power was voided, or perhaps reallocated into a measure of power not devised in this study, after the formation of Dumbledore’s Army.
Discussion This study analyzed the Harry Potter series through a sociological lens, treating the characters and social network of friendship ties in Hogwarts as any other social network with basic properties and characteristics. The study examined friendship ties during Harry’s first four years at Hogwarts, and compared them to friendship ties after Dumbledore’s Army, which was formed in Harry’s fifth year. The primary objective of this study was to highlight the similarities
that exist between friendship networks in the series, and recent sociological studies on the same subject, showing that the Harry Potter series can be analyzed sociologically by identifying basic network properties and characteristics. An additional objective of this study was to test existing theories about homophily, specifically, the idea that adolescents are friends with those who are similar to them in terms of race, gender, and organization. This was accomplished by comparing changes in friendship patterns before Dumbledore’s Army, when the actors were between the ages of 11-14, and after, when the actors were 15 and older.
A final objective of this study was to determine if Harry Potter, the focal point and most recognizable character in the series, wielded the most power and influence within his social network.
The results of this study showed a decrease in gender homophily between friends over time, further supporting Shrum, Cheek, and Hunters findings that gender homophily decreases over time. However, in contrast to their research, my findings showed that males were more likely to be friends with other males prior to the formation of Dumbledore's Army, whereas females exhibited many cross-gender ties. Also consistent with Shrum’s studies on adolescent friendship, I found that racial homophily did increase over time, but this study did not perform an in-depth analysis on the nature of the relations among those actors.
Organizational homophily was not shown to have a significant change over time, but a strong tendency towards internal ties was discovered, suggesting that the placement of students into houses played a heavy role in the friendships those students formed. In the fictional setting of Hogwarts, students are required to share dormitories and common rooms, attend classes, and eat lunch with members of the same house, so it is not surprising to find that ties of friendship within houses are stronger than ties of friendships outside houses.
Conclusions An unexpected finding in this study was the discovery that among early adolescents, the males were more likely to exhibit gender homophily than the females. One of the most notable exceptions to the theory that gender homophily is strongest in early adolescence is Hermione Granger. Until the formation of Dumbledore’s Army, Hermione did not have friends of the same sex in her grade. She spent the majority of her time with Harry and Ron. Although I did not measure the intensity of friendship ties, or examine the social dynamics between close friends or “best friends,” this would be an interesting study to conduct.
The characters Dean Thomas and Seamus Finnigan, Lavender Brown and Parvati Patil, Harry, Ron, and Hermione, and Malfoy, Crabbe, and Goyle, are understood to be best friends within the series, and an analysis of the similarities between those dyads, or triads, could further support sociological studies pertaining to the nature of adolescent friendship.
Another useful study that could be conducted would be to examine the break-down of gender in the friendship networks at Hogwarts. There is a disproportionate number of males to females mentioned within the friendship networks studied, and a comparison study could be drawn up between the Harry Potter series and other works of juvenile fantasy. What social roles do the female characters occupy in this series in comparison to other stories? What about males? Although I did prove my hypothesis that Harry Potter held the most power and connection in his friendship network as a young adolescent, it was interesting to discover that the level of power dispersed as Harry lost his position as an intermediary.
Future studies could place more emphasis on the power of intermediaries in social networks within literature, perhaps using Bonacich’s power measure as a basis for determining actors’ power.
The biggest weaknesses in this study related to the methodological procedures. The procedure used to establish ties of friendship was an entirely subjective measure: an arbitrary measure was drawn up to link actors as friends, and no distinction was made between different levels of friendship. In other words, an actor could be tied to his or her best friend and someone classified as a friend who is not as intimately linked, i.e. does not share secrets with this person, and no distinction would be made. Because only friends in Harry’s grade were included in the sample, any friendship ties with older or younger students were ignored.
It is possible that patterns of friendship would be different if all ties of friendship were examined. Additionally, there were a number of procedures that would have been helpful to use in examining ties of friendship, such as transitivity, block densities, and point connectivity, but these were not included. Due to the limited amount of information available about the actors in Network A and Network B, details from the first four books were often lumped together into a “before” category, despite the fact that each book experienced a passage of time and featured emotional changes in characters.
The behavioral, emotional, and psychological aspects of the sociological parameters in this study were not discussed, but this topic would be fertile ground for further study.
References Berndt, Thomas J. 1982. "The Features and Effects of Friendship in Early Adolescence." Child Development 53:1447-60. Berndt, Thomas J. and Keunho Keefe. 1995. "Friends' Influence on Adolescents' Adjustment to School." Child Development 66: 1312-29. Cairns, Robert B., Man-Chi Leung, Lisa Buchanan, and Beverly D. Cairns. 1995. "Friendships and Social Networks in Childhood and Adolescence: Fluidity, Reliability, and Interrelations." Child Development 66: 1330-45.
Crosnoe, Robert. 2000. "Friendships in Childhood and Adolescence: The Life Course and New Directions." Social Psychology Quarterly 63: 377-91.
De Nooy, Wouter. 2001. “Stories and Social Structure: A structural Perspective on Literature in Society.” The Psychology and Sociology of Literature: 359-75. George, Thomas P. and Donald P. Hartmann. 1996. "Friendship Networks of Unpopular, Average, and Popular Children." Child Development 67: 2301-16. Griswold,Wendy. 1993. “Recent Moves in the Sociology of Literature.” Annual Review of Sociology 19: 455-67.
Hartup, Willard W., Doran C. French, Brett Laursen, Mary Kathleen Johnston, and John R. Ogawa. 1993. "Conflict and Friendship Relations in Middle Childhood: Behavior in a Closed-Field Situation." Child Development 64: 445-54. McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001. "Birds of a Feather: Homophily in Social Networks." Annual Review of Sociology 27: 415-44. Rowling, J.K. 1997. Harry Potter and the Sorcerer’s Stone. U.S.A.: Scholastic Books. Rowling, J.K. 2003. Harry Potter and the Order of the Phoenix. U.S.A.: Scholastic Books. Shrum, Wesley and Neil H. Cheek, Jr. 1987. "Social Structure During the School Years: Onset of the Degrouping Process." American Sociological Review 52: 218-23.
Shrum, Wesley, Neil H. Cheek, Jr., and Saundra MacD. Hunter. 1988. "Friendship in School: Gender and Racial Homophily." Sociology of Education 61: 227-39. http.//en.wikipedia.org/wiki/Harry_Potter, accessed November 15, 2007. http.//en.wikipedia.org/wiki/Dumbledore’s_Army_Members, accessed November 15, 2007. http://en.wikipedia.org/wiki/List_of_ characters_in_Harry_Potter_books, first accessed November 25, 2007.
Table 1. Actor Degree- Before Dumbledore’s Army Actor Degree Harry Potter 7 Ron Weasley 5 Vincent Crabbe 4 Draco Malfoy 4 Hannah Abbott 2 Lavender Brown 1 Parvati Patil 1
Table 2: Reachability- Before Dumbledore’s Army L S H N P H D R H J E M V G D P Brown 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 Finnigan 0 0 1 1 0 1 1 1 1 1 1 0 0 0 0 0 Granger 0 1 0 1 0 1 1 1 1 1 1 0 0 0 0 0 Longbottom 0 1 1 0 0 1 1 1 1 1 1 0 0 0 0 0 Patil 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Potter 0 1 1 1 0 0 1 1 1 1 1 0 0 0 0 0 Thomas 0 1 1 1 0 1 0 1 1 1 1 0 0 0 0 0 Weasley 0 1 1 1 0 1 1 0 1 1 1 0 0 0 0 0 Abbott 0 1 1 1 0 1 1 1 0 1 1 0 0 0 0 0 Finch-Fletc 0 1 1 1 0 1 1 1 1 0 1 0 0 0 0 0 Macmillan 0 1 1 1 0 1 1 1 1 1 0 0 0 0 0 0 Bulstrode 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
Crabbe 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 1 Goyle 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 Malfoy 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 Parkinson 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 Table 3a: E.I. Index- Homophily by Gender- Before Dumbledore’s Army Actor Internal External Lavender Brown 1 0 Seamus Finnigan 3 0 Hermione Granger 0 3 Neville Longbottom 2 1 Parvati Patil 1 0 Harry Potter 6 1 Dean Thomas 3 0 Millicent Bulstrode 1 3
Table 3 b. E-I Index- Homophily by Gender- After Dumbledore’s Army Actor Internal External Lavender Brown 5 10 Seamus Finnigan 9 6 Hermione Granger 5 10 Neville Longbottom 9 6 Parvati Patil 5 10 Harry Potter 9 6 Dean Thomas 9 6 Ron Weasley 9 6 Hannah Abbott 5 10 Susan Bones 5 10 Justin Finch-Fletchley 9 6
Ernie Macmillan 9 6 Terry Boot 9 6 Michael Corner 9 6 Anthony Goldstein 9 6 Padma Patil 5 10 Millicent Bulstrode 1 4 Vincent Crabbe 3 2 Gregory Goyle 3 2 Table 4- E-I Index- Homophily by House- After Dumbledore’s Army Actor Internal External Lavender Brown 7 8 Seamus Finnigan 7 8 Hermione Granger 7 8 Neville Longbottom 7 8 Parvati Patil 7 8 Harry Potter 7 8 Dean Thomas 7 8 Ron Weasley 7 8 Hannah Abbott 3 12 Susan Bones 3 12 Justin Finch-Fletchley 3 12
Ernie Macmillan 3 12 Terry Boot 3 12 Michael Corner 3 12 Anthony Goldstein 3 12 Padma Patil 3 12 Millicent Bulstrode 5 0 Vincent Crabbe 5 0 Gregory Goyle 5 0 Table 5.
Closeness Centrality- Before Dumbledore’s Army Actor Closeness Harry Potter 12.397 Ron Weasley 12.097 Ernie Macmillan 12 Justin Finch-Fletchley 12 Dean Thomas 11.905 Hermione Granger 11.905 Neville Longbottom 11.905 Seamus Finnigan 11.905 Hannah Abbott 11.450 Vincent Crabbe 8.333 Draco Malfoy 8.333
Milicent Bulstrode 8.333 Gregory Goyle 8.333 Pansy Parkinson 8.333 Lavender Brown 6.667 Parvati Patil 6.667 Table 6a. Betweenness Centality- Before Dumbledore’s Army Actor Betweenness Harry Potter 16.190 Ron Weasley 2.857 Ernie Macmillan 2.857 Justin Finch-Fletchley 1.905 Dean Thomas 0 Hermione Granger 0 Neville Longbottom 0 Seamus Finnigan 0 Hannah Abbott 0
Vincent Crabbe 0 Draco Malfoy 0 Milicent Bulstrode 0 Gregory Goyle 0 Pansy Parkinson 0 Lavender Brown 0 Parvati Patil 0 Table 6b. Betweenness Centrality- After Dumbledore’s Army Actor Internal External
Lavender Brown 0 0 Seamus Finnigan 0 0 Hermione Granger 0 0 Neville Longbottom 0 0 Parvati Patil 0 0 Harry Potter 0 0 Dean Thomas 0 0 Ron Weasley 0 0 Hannah Abbott 0 0 Susan Bones 0 0 Justin Finch-Fletchley 0 0 Ernie Macmillan 0 0 Terry Boot 0 0 Michael Corner 0 0 Anthony Goldstein 0 0 Padma Patil 0 0 Millicent Bulstrode 0 0 Vincent Crabbe 0 0 Gregory Goyle 0 0 Draco Malfoy 0 0
Figure 1a. Friendship Ties Before Dumbledore's Army
Figure 1b. Friendship Ties After Dumbledore's Army