The Facebook Project

Interpretation and Analysis

In a Facebook group full of 200,000 thousand people to answer the question of just who resides there with qualitative methods would likely take a lifetime.  Certainly there are some key players and consistent posters but explain who’s there overall is definitively different then who there is most active or influential.  In stride with my belief in multi-method research I actually took to answering this question with a very simple statistical analysis.  I used a random number generator to summon up a listing of 38 different page numbers to visit in the group member index and noted exactly 383 different profiles.  With each profile I noted its gender (members self-identify on their own profiles) and network affiliation.  By examining 383 profiles I could say at a 95% confidence level that my statistical observations of the group were within 5% of the actual numbers.  More important than this accuracy, however, is what the observations revealed substantively.  In all the group actually has many more women than men, females outnumber males approximately 2 to 1.  Assuming that most members believer in the title and rule/suggestion listings on the site (which is as yet unproven)¹ this implies that women are more likely to believe in the founding principles of the group.  The next point of interest requires a couple of figures to drive home the relevance:

The group, as it would happen, is dominated by people in high school and contains many people who have profiles affiliated with places outside of the US.  In fact the largest single portion of the population is those with no network or who have a geographical affiliation.  This actually disrupts the common notion that Facebook is filled with mostly college students—tens of thousands of these members are younger and from outside of the country.  There seems to be some amount of turbulence in membership amongst high schoolers but we can also observe a definitive trend of fewer members as you get higher in age.  This might be a function of membership on Facebook in general – the service has only been around since 2004 and with the majority of first joiners coming in high school or college (they had to, your email ID was required) naturally overall on Facebook there may be fewer older people in general.  Regardless we can observe two really big things about what types of people join this Facebook group: they are more likely to be female, and they are more likely to be young.  They are also more likely to be associated with a network other than a college network.  This backdrop of the membership provided an interesting point of contrast for me when it came to discourse analysis.

In any face to face world ethnography researchers carefully observe their environment and become a sort of unit of measurement.  They learn over time to become comfortable and interact with the various people going about their business in the site of research.  I did much of this online but was able to do so initially in an invisible fashion, observing without others knowing it.  I was, in a sense, their invisible audience, until I revealed myself.  I was able to do this because of the interface at hand.

Just like in advertising when sometimes the presentation of a product even alters how customers think it tastes (Gladwell 2007), the interface, layout, demeanor, and interaction mediated by the systems of Facebook play a crucial role in fabricating the space for participant actors.  The arrangement of the group and the availability of information there helps to determine the importance of certain elements.  Based on my collected observations I identified three key themes at play in the design of the group: sharing and connection, activity and involvement, and the mediating factors of interface; sometimes known as the dialectic of content and form.

In truth there are too many examples to count, so I will present a few here to give a feeling for what I mean.  The interface pushes users to share and connect with one another in a variety of ways.  In the case of this specific group it literally contained the instructive text reading “This is an open group. Anyone can join and invite others to join.”  Through this the interface outright says this is an open and welcoming environment.  Membership is not determined by investigation or approval, but instead is an open door to all that choose to step through.  If there were a listing of related groups beneath this item, as there are in many other groups, users would be able to see ways others in the group are connected.  They are encouraged to connect to others by sharing the group.  A literal ‘share’ button exists beneath the join/invite area that can be used to loop in others who aren’t on Facebook.  The groups categorization is even listed as a “Common Interest” among friends, displaying a shared perspective or connection.  Whether or not the participants believe in these values the interface encourages and applies them.

The same could be said for the portrayal of activity and involvement.  Those who find the group in a search queue or check up on it in their groups page are confronted with a report on its latest activity—in my case it was the gain of several members and new discussion posts.  The interface gives statistics like this all over the place to relay non-specific interactions that are afoot in the group.  When I visit the page I can see the activity that has taken place recently in wall posts and read about the sheer numbers of people who are doing things (discussing, posting, joining) with the group.  Photos, videos, and posted items are all presented to me to reflect activity that could (or has) taken place in the group.  The search options for the forum are regulated by post popularity and how recently it was contributed (to).  All of these elements give a lively picture of the group.

Finally, the interface also mediates the ways the group can be used.  There are restrictions, such as the left alignment style of the site (a western setup) or the constrained navigation options, which are emphasized in their priority by listing and location.  By placing the terms of use link at the bottom of the page Facebook does not direct much attention to the legal underpinnings of the site, but by putting pictures of members in a highly visible spot they guide them to engage with others.  This is not an insidious design necessarily, and in many ways might be functional.  It could, however, constrain the ways users can engage with the site.  For instance, many appear to use the wall like someone could use a live-chat service like AOL instant messenger, by posting back and forth within the span of a few seconds to respond to the previous poster.  Facebook did not design it to be used this way as each time you go to post the form is refreshed; furthermore you can delete your own wall posts, unlike statements made in live-chat.  Wall and discussion posts also require that the poster’s picture be placed next to them, forcing a verification of identity (even if imprecise).  The form fields don’t filter out swear words or certain languages, which might imply that they rest upon the value of free speech, and yet at the same time you can only enter in a certain number of characters before your post is turned away.  Overall, users have only a limited level of control over the interface – they can dictate how they use many parts of it—such as what content they put up or the way the engage each item—but they cannot alter the underlying boundaries of the code.  If the color blue for some reason deeply offended them the best they would be able to do is report the page to Facebook, who would probably brush them off as out of their mind.  While taking issue with the color of the page might be rightfully dismissed, more important elements, like say whether or not a user can list their ethnicity or not state their email address, hold more significant implications.

This debate of course, is to some extent not very new.  Old print-media came with interface limits in the size and design of paper or colors available.  A reader might have been able to write back to the author demanding a change, or supplying a contribution.  The digital architecture of the medium, as discussed earlier, alters this system in regards to time, distance, bodies, and identity.

If this Facebook group were a room, it’d be pretty strange.  Over 200,000 some people from all over the place would claim to have a membership to this room, but only a minute portion (at least 145 of them) would visit the room with any frequency.  Anybody entering would be immediately confronted with a giant list of instructions stapled to the wall.  Some might not read them, basing their membership off the title on the door and others might hang out without reading the list.  Regardless outsiders could and would visit and hold the group accountable to the list that describes the purpose of the group.  In the case of this Facebook group the list presents a very strong indication of what members (should) believe in.  The entire description is contingent on a 40 point series (going from 0 to 39, point 0 was likely added later).  A later update to the group indicated that they were taken to be suggestions, but this contradicts the way most of them are written, as will be seen.  There were several common themes inherent in the list:

1. Each item was given with limited rational, sometimes in a contradictory manner; they were interpretive because they were vague, but inflexible because of the use of strong commanding words.
2. Many items make assumptions about the audience reading them (presumably the group members).
3. They were often related to compassion, care-taking, and gratitude.
4. To a lesser extent they alluded to tradition, religion, and employed condescending language.
5. And of interest to this study, they relied on many female and male gender roles and stereotypes.

The purpose of this ethnography was not to do a mere content analysis of the group’s description, so for the extent of this paper I’ll only give analysis of a few excerpts from the list.  Readers can look at the Appendix and find many more instances if they’re interested.
Perhaps one of my favorites in the grouping was point number four, which goes as follows:

“4. Play one of the songs that would make any woman weep like the little girl she once was (but in a good way). A brief list includes, but certainly isn't limited, to:
"You & Me" by Lifehouse
Anything by Frank Sinatra
Any rendition of "Everything I Do, I Do it for You"
"Collide" by Howie Day
"Out Of My League" by Steven Speaks
And MOST IMPORTANTLY "Question" by the Old 97's (if you propose to a girl with this song, she is putty in your hands).”
("Putty in your hands" is not meant to promote "using women" in any way. This group does not encourage guys to be polite in order to get her into bed.)

First, take note of the language used in the statement.  Readers are instructed to play.  Not ‘you might play’ or ‘you could play’ but a definitive command to play songs that make women weep and become little again.  The rationale given for why a guy should play these songs seems to be so that they can cause her to weep and become vulnerable but it is not understood or explained why a man would want a women in this state of being.  Readers are left to interpret when they should play these songs but also know that they must play one of them.  There seem to be a few contradictions here, too.  The list includes, but is not limited to, and yet there is a strong command to play.  If it were a set of suggestions likely such strongly commanding language would not be used, there seems to be a conflict of emphasis.  The author also inserts several defensive statements within parenthesis.  The “but in a good way” seems to imply a shared understanding of what it means to be a crying little girl (and how that’s good).  It’s also interesting to see how the author switches back and forth in using woman and girl.  In my own experience the referral is tied to age, as most would expect, but the use of the term girl may extend into adulthood.  Rarely would anyone refer to a mother as a girl, but often female youths in their 20’s seem to be referred to as girls instead of women.  Suddenly age and maturity enter into the question – a woman is reduced to putty child, weakness incarnate, by a man.  She is stripped of all of her own agency and becomes comparable to a moldable form.  Remember those countless instance in pop culture (and worse in past paradigms of academe) of women brushed off as irrational beings driven and masterminded by their emotions?²  The defensive comment at the end only brings in another stereotype about men—it assumes the role of moral regulation, advocating that men not make women vulnerable to have sex with them.  The fact that a woman is referred to as an object to be molded, or set into a vulnerable state constructs an image of femininity as weak and emotional.  Additionally, if you read over the list of songs you’ll notice they’re really intended for a certain audience.  Women 60 years of age or who are deaf or who live in India and speak English (I observed at least 6 profiles belonging to Indian women) probably aren’t going to be reduced to tears from You and Me by Lifehouse.  The list is associated with what I would term as a mainstream white, middle-class American youth audience and really constructs a very constrained view of what romantic songs can be. 

Clearly my assessment is pretty harsh and comes from a more literal interpretation.  Even if I were to take it loosely it feels as if the author is suggesting a guy (not man or boy) should lure a woman into a state of vulnerability and insecurity with the use of a romantic song.  Even based on such a loose interpretation I get the impression that the vision is based on an idea of women being weak or vulnerable and men being those who initiate or control relationships.

The list has a number of other good examples, here characterized are two related to casting men into stereotypical roles:

“29. Just because you're a guy doesn't mean you are completely incapable of calling when you say you will, it just means you are highly incapable of it. There are few acceptable answers to, "Why didn't you call?", & being male is not one of them.”
“34. At least do everything in your power to keep cursing to a minimum while around her. If you can, cut it out period while around her, or cut it out of your vocabulary. Women don't want to hear it, guys don't care about it, adults don't want to hear it, it doesn't impress employers, and you sure won't want your children or someone else's to hear it!”

The statement “Just because you’re a guy” seems to imply that it’s a matter of ontology that males are incapable of a fundamentally social action—calling based on a prior communication.  To allege that because someone is a male (socialized) they are highly incapable of something is a stereotype and assumption.  The author doesn’t state why he believes this and instead lays a sort of condescending smack-down moral regulation on men.  Here he characterizes men as insensitive or irresponsible, which I feel is based on gender stereotypes.  The next statement hinges on another theme found throughout the list: men as the ones who hold the power (take initiative).  The statement makes all kinds of assumptions about the audience here – what women want, what guys care about, what adults or employers wish to hear, and the way the audience would want their children raised.  Mostly I just don’t think it’s fair to assume that only men swear and women don’t, and to suggest that women would be bothered by cursing.  To me it indicates another instance of implying men are insensitive or irresponsible, and an elaboration of women as weak—men must keep harsh swear words from their meek ears.

And to finish this all off, I’ll give just one last set of examples:

“27. When she feels at her worst, tell her she looks her best.”
“24. Offer her your jacket/sweatshirt. (Note: you may not see that particular item of clothing for a while, if ever again)”
“6. Find out what her favorite flower is and buy them for her randomly (regardless of the situation you might be in). A simple yet profound truth: a single rose says more than dozens of anything else. (I encourage the women to not allow a guy to "prove himself worthy" through gifts and flowers and such. Trust is a precious thing and it should take a good chunk of time before he gains it back in your heart.)”

Feeling does not match appearance.  Our appearance might reflect our feelings and vice versa, but not always.  This assumes that appearance is so important to the woman that she will feel better when she looks better – especially in times of great emotional stress. To be honest I don’t know that telling a woman she’s pretty after her mom has died is going to do anything but feel like an insult.  Again, context matters.  Moreover the statement implicitly reinforces women being valued by their appearance.  It doesn’t tell them to tell her she’s smart or caring or insightful or funny or any other darned compliment, just that she looks good.  The second one is also related strongly to context but I extracted it not to bash its lack of specifics but instead note the way it implies it is permissible for women to steal items of clothing without asking.  And then finally point six seems to have some flimsy logic.  A single rose may not be worth more than a dozen diamonds, but more importantly it presents women as gate keepers who determine a man’s worth (or worthiness of trust).  It illustrates a picture of men as wrong-doers who must be assertive to regain entry or access to a woman.

In total these forty points make several assumptions about the audience.  They seem to suggest romantic relations between a man and woman, of those of a heteronormative character.  They seem to be geared for youth, and directed and men, and many rest on shared understandings.  At root many have compassion, care-taking, or gratitude as underlying values (it’s nice to give someone your jacket if they’re cold or play a song that reminds them of a joyful childhood) but articulate these values in such a way that they reinforce gender roles and stereotypes that in turn reflect inequalities in society.  These include characterizing women as weak or vulnerable, women being gate keepers (in control of reception), and the importance of appearance in women.  They also include characterizing men as immoral or in need of sexual regulation, men as insensitive or irresponsible, and men as in control of initiation.  Overall the group paints this notion of a gentleman that is overwhelmingly restrictive: a white male of middle to upper class who is able bodied.

Talk in the group seems to take on a few themes.  Due to time constraints I couldn’t really engage in a lot of discussion or observation of discourse.  I only observed the titles of the various discussion posts (remember there are thousands) and amongst them I identified a few themes:

1. Those having to do with relationships and romantics, often started by those seeking advice, complaining about the opposite gender, or indicating desired qualities in one gender.
2. Some dealt with sexuality, such as sexualized traits, like breasts or penis size, preferred sexual actions, the notion of homosexuality as a lifestyle, or virginity.
3. Some significant and hot topic posts discussed religious and political issues, such as the existence of God, the truth behind evolution, abortion, or the upcoming election.
4. And perhaps the largest count belonged to random items, such as post strings where one poster ‘rated’ the person above or below them, advertisements and music preferences.
5. A few posts featured resistance (discursive opinions) to the list, either addressing single items or the list on the whole.

By the time I got through observing the qualities of the environment my project was already winding down.  I only had one week of formal fieldnotes pertaining to discussion analysis.  To begin talk I decided to ask the general group what their political views were and read and respond to the results.  I did it in part to try to unearth controversial subjects, and also to see if I could tease out the issue of gay marriage, which I feel is very related to gender roles and stereotypes.  I got a few bites (gay marriage did come up, but feminism never did) and I responded with a lot of questions and pretty soon the conversation ballooned to 53 posts by 21 people in the span of 3 days.  Looking back on it I wish I had asked about feminism instead.
The responses to my series started out with a little variance – almost nobody stated their political in a general fashion but instead broke them down by issue.  Others rejected politics and claimed it to be stupid.  Most of the people who responded to me were other college students, which was interesting, especially given that the group is filled with mainly high schoolers and those no in college networks.  What was surprising was how they responded to me.  Many of them, when I asked about their opinions about gender roles and the interpretation of the list, rejected the notion of an inflexible set of regulations.  I promised them that I would not put quotes into publication and so won’t feature them here but several of the most articulate and extensive responses suggested that respondents thought the list communicated a set of foundational ideals, such as communication, mutual respect, affection, and more.  Many also felt that items were reversible on the basis of these founding ideals and that traditional gender roles are irrelevant or outdated.  Several assaulted the notion of strict guidelines and literal interpretations.  Somehow these individuals could exist in a group like this without a sociological consciousness of the sexist norms it perpetuates with its contextual framing.  I don’t know if they lurked just to pick fights with people or if the group was the site of their developing perspective, they just thoroughly surprised me.  Their resistance led me to reconsider some of my own presumptions (members’ perceptions of gay marriage as it relates to gender roles).  Out of it all I mostly feel like I ought not dabble in an environment so upsetting that moral judgments might not only mediate scientific ones, but determine them.
I’d be more happy to relate their deviant perspectives if there weren’t one tragic flaw: none of them explained why such general ideas (compassion, care-taking, gratitude, etc…) needed to be articulated or translated into such inflexible list statements with limited or no rationale.  I related all of this in my final revealing (hopefully not snobby or condescending!) post to the group:

“I actually agree that the underlined ideas of the list (communication, showing respect, affection and whatever blend of the three) are great guidance - but then why not make a list of just those instead of the relatively strict set of pointers here (even the group name states quite harshly that there are prescribed behaviors that men should adhere to). And if it becomes an issue of could vs. should (a list of things you could do vs. should do) then why limited it so? A list of 'could' would span all kinds of types of people and contexts. Simply by making a limited, static list you [advocate] for certain behaviors, even if they're all ideas of things you 'could do.' If anything it ought to be dynamic, emergent, and determined by the collective.”

I did reveal at the end of my observations my position as a researcher and assured participants that I wouldn’t refer to any of them specifically to preserve confidentiality.  I felt a bit obligated to explain what was going on from my end, even if it made no difference to the actors in the grand scheme of things.