The Scienciness of danah boyd

2007 June 26
by Cornelius

Edit (4/21/2010): This post and another one published a bit later stand here as testament to the smartassery and smugness of a young PhD student (me). While some minor points I made back then might not be entirely moot, the overall tone strikes me as arrogant from today’s perspective and I don’t see myself using the adjective “sciency” lightly ever again when talking about a colleague, least of all danah.

Anyhow, I’m not deleting or changing this either. Sins of youth and all. The tagline to this is “The Intaweb — reminding people of dumb stuff they thought and did since 1995.”

Hmm, funny how these things go. Originally, I just wanted to write a short entry about an essay by danah boyd, but then things turned out differently, in the sense that the entry became anything but short (a blog essay, if you will) and that it ended up concerning more than just one piece of writing and its reception.

Let’s see.

boyd’s recent post on class divisions on MySpace and Facebook caught my attention because it raises a number of interesting questions (the full essay is available here) and I was eager to blog a detailed comment.

But on the Web time works against you and others were a lot more prompt with their responses than me. Many others. So many, in fact, that boyd was taken off guard by the mass of reactions, which were provided by both the mainstream press (1, 2, 3, 4 – many more if you search) and of course by countless bloggers (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, …. the list is endless). So instead of adding my nitpicking to what has already been written – much of which is strongly decontextualized and hyperbolic – I want to look at the dynamics of the situation instead.

  1. A PhD candidate writes up her impressions regarding the socioeconomic status of users on two social networking sites and puts the resulting text on her web site.

  2. The new and old media pick up the piece and present it as the result of scientific inquiry (see below).

  3. People hotly debate whether a) the claims made in the piece are accurate and b) what the (negative) implications are.

  4. The essayist qualifies her writing, calling it “problematic” [1] and noting that “as an academic… I feel guilty” [2].

Looking at this course of events, I realized that I was actually more interested in the reception and interpretation of boyd’s essay and in what it means for the relationship between science and society in the digital era than I was in the content of the essay per se.


Because it is a perfect example of how genre conventions can be creatively subverted to gain prestige and authority (though I am not implying that this is boyd’s goal). Furthermore, the essay’s reception demonstrates how the lack of reliable formal criteria to distinguish between emergent digital genres can cause uncertainty, and that the news media reinforces this uncertainty through shoddy and inaccurate reporting.

The result of this process is a sort of Scienciness – a set of opinions and impressions which are supported by the experience, intuition and (most importantly) prestige of an academic, but not by empirical data. This last fact is not problematic per se, since there are many interesting questions that simply cannot be answered in a satisfactory way by relying purely on quantitative data. In countless areas of investigation the ethnographic methodology as employed by boyd is ideally suited to the question – just not here, where the question is a simple socio-demographic tidbit: are the user communities of Facebook and MySpace economically stratified or not?

In the following paragraphs, I want to look at how boyd characterizes and classifies her own writing and how it is in turn interpreted and (badly!) recontextualized by some of the news sites that cite her. In that context it is worthwhile to point out that the popular reception of the essay is partly due to its usefulness in confirming what we suspected all along.

Is anything a scientist says scientific per definition? Is it our job to find proof for what people suspected all along (for example, about the unwashed masses and the digital Las Vegas they have created on MySpace, to use boyd’s example)? In other words, is what we (as in “we scientists”) do about finding the truth, or about finding answers; the difference being that the truth can be rather fuzzy and inconclusive and answers are, by contrast, meant to satisfy the inquirer?

Let’s start by looking at some of the points made in boyd’s essay and blog entry:

There is indeed a change taking place, but it’s not a shift so much as a fragmentation. Until recently, American teenagers were flocking to MySpace. The picture is now being blurred. Some teens are flocking to MySpace. And some teens are flocking to Facebook. Which go where gets kinda sticky, because it seems to primarily have to do with socio-economic class.

(from the blog)

The assertion made is simple enough: the migration of people from one site to the other is not age-related, but influenced by socioeconomic factors. boyd notes that she lacks both a concise definition of class and the empirical data needed to back up her claim that there is a migration from one platform to the other and that this is indeed socioeconomically motivated:

(I want to take a moment to make a meta point here. I have been traipsing through the country talking to teens and I’ve been seeing this transition for the past 6-9 months but I’m having a hard time putting into words. Americans aren’t so good at talking about class. It’s sticky, it’s uncomfortable, and to top it off, we don’t have the language for marking class in a meaningful way. So this piece is intentionally descriptive, but in being so, it’s also hugely problematic. I wish I could just put numbers in front of it all and be done with it, but instead, I’m going to face the stickiness and see if a point can be made through it all. Hopefully it works. If not, sorry.)

(from the essay)

(A “meta point” that could be made here is that her piece is not so much descriptive as it is impressionistic. In order for something to be descriptive it has to accurately describe something.)

Now, although having empirical data on the users of Facebook and MySpace and their demographic and socioeconomic status at hand would be convenient, lacking such information by no means precludes any kind of analysis. boyd has conducted countless interviews and analyzed a large number of profiles in the course of her studies. The unique advantage of an ethnographic methodology is that it allows the researcher to capture contextual information that is simply omitted when just looking at numbers on a spreadsheet.

But the problem is that the kind of question boyd asks doesn’t require or even allow an ethnographic methodology. It requires a large-scale, representative quantitative analysis. I fully agree with boyd when she notes that, to pick a prominent example, income is not a single sufficient indicator of “class”. But this does not impact the central issue at all: when making judgements on two internet communities with a size exceeding 125 million users having conducted a series of interviews seems grossly insufficient.

It is interesting that boyd reacted one day after initially publishing the essay by appending the following to her foreword (without marking the change in any way):

For the academics reading this, I want to highlight that this is not an academic article. It is not trying to be. It is based on my observations in the field, but I’m not trying to situate or theorize what is going on. I’ve chosen terms meant to convey impressions, but I know that they are not precise uses of these terms. Hopefully, one day, I can get the words together to actually write an academic article about this topic, but I felt as though this is too important of an issue to sit on while I find the words. So I wrote it knowing that it would piss many off. The academic side of me feels extremely guilty about this; the activist side of me finds it too critical to go unacknowledged.

(from the essay)

Several aspects of this qualifying paragraph are interesting. Firstly, she specifically addresses the academic community when stating that her piece is not an academic article, which almost creates the impression that this caveat does not apply to those outside of academia who interpreted it as such a text (e.g. the news media). Secondly she presents the defect of the article as a linguistic problem – she isn’t able to find the correct terminology for the phenomena she believes to have noticed. Thirdly, her motivation for writing and publishing the text despite of its self-attested weaknesses is that of the activist who finds the issues “too critical to go unacknowledged”.

But what exactly is it that shouldn’t go unacknowledged? That the use of social networking sites may be socially stratified? Not only would that be hardly surprising (if it could be proven), but it would also hardly serve as a grounds for any kind of activism (should we ban Facebook for elitism, or MySpace for appealing to the proles?). Or, does boyd mean it is an important issue that American society is shaped by its socioeconomic makeup, or that this makeup should be changed, or that this is a new, pressing issue that needs to be urgently addressed?

All of these things seems ridiculously implausible.

And what about the second problem – that she is lacking the words to describe the problem. The claim frames scientific inquiry as the application of obscure jargon to an issue, a jargon that wouldn’t change the quality of the facts presented in any way. A discussion about the implications of how the user bases of social networks are socially stratified may be highly political, but the data itself certainly isn’t. In other words, the issue isn’t what to call things, the issue is whether or not they are there in the first place. Accurate quantitative data precludes a careful description – there is no language fix to that problem.

This “message to the academic readership” echoes similar remarks made in boyd’s blog:

I think some folks misinterpreted this piece as an academic article. No doubt this is based on my observations from the field, but this is by no means an academic article. I did add some methodological footnotes in the piece so that folks would at least know where the data was coming from. But I didn’t situate or theorize or contextualize this at all. It’s more like publicizing field observations. There’s much work to be done before this can be anything resembling an academic article. The “citation” note at the top of my pieces also confuses this. That was meant for when people picked it up and stole it whole from my page or when people got to it indirectly. I put that as a standard for my blog essays a while back because of this issue. I guess I see my blog as a space to work out half-formed ideas. I just didn’t expect 90K people to read it. Blog essays to me are thoughts in progress, blog entries that are too long to be blog entries. But I can see where there’s confusion.

(from the blog)

The last highlighted remark is what genuinely puzzles me. When is a blog entry too long to be a blog entry? The software imposes no natural limit on how much you can publish in one entry. In fact, I have individual posts in my corpus that exceed 5,000 running words of text. Whether that is typical is another question, but there is certainly no technical reason to limit the length of a blog post any more than the length of any other piece published on-line.

But instead of publishing it in her blog, boyd evokes a new genre of writing: the blog essay. Googling for that compound term results in 26,500 hits, a tiny fraction of the number of results found when searching for either “blog” or “essay”. It can thus be assumed that boyd is either an early adopter of the term, or possibly its inventor. The blog essay is published on boyd’s website, which also contains a full biography and information on her to-date research. By contrast, her blog apophenia is located at The meta-description she provides in the side-bar there reads: “I use this blog to express random thoughts about whatever i’m thinking about.“ By contrast, the essay is stored at – the only text currently in that folder. The folder is linked to in several places and referred to as “my research papers” and “formalized reflections” [1, 2].

Here’s the layout of the essay’s header, as published at

Viewing American class divisions through Facebook and MySpace

danah boyd
June 24, 2007

Citation: boyd, danah. 2007. “Viewing American class divisions through Facebook and MySpace .” Apophenia Blog Essay. June 24 .

(If you have comments, please add them to the related entry on my blog. Thank you.)

[text body]

The citation, as noted by boyd [1], serves to reinforce any assumption on part of the reader that this is a scholarly article, or at least the draft of one.

Let’s summarize: the text is on a separate HTML page, located in a folder that contains research papers. It is deliberately published in another place than the blog (that is used to publish thoughts and opinions) and instructions are given for how it should be cited.

And yet it is not an academic article.

It appears that those citing the piece are under an entirely different impression (and I’m not saying that this isn’t largely their own fault). The BBC mentions a study, calling the essay a “preliminary draft”, to Mashable it’s a “case study”, BoingBoing also uses the term “draft”, blogger John Scalzi uses the more specific label “draft of an academic paper” and finally, calls it “a new six-month interview-based study”.

Looking again at the novel genre label of blog essay doesn’t help to clarify the issue, as the semantics of the term are confusing (to me at least). Looking up definitions of essay nets explanations such as these:

an analytic or interpretive literary composition

Essay, a short work that treats of a topic from an author’s personal point of view, often taking into account subjective experiences and personal reflections upon them.

A short literary composition on a single subject expressing a personal view.

I think most people would agree that these definitions are equally applicable to a blog, with the difference that the essay is a pre-digital genre and one that is usually associated with traditions of learned or academic writing. boyd makes a choice when calling her text a blog essay: she assigns it a certain degree of authority by delineating it from the pieces published in her blog. Together with her status as an academic – at least that is the role that virtually all news sources assign her – the implication for most readers is clearly that what she is saying is the result of scientific research. In other words, it is an assessment based on empirical facts that tells us something about the world we live in. It is verifiable, objective and the result of thorough inquiry.

I know I’m applying the most drastic view possible here, but you can probably see the problem. boyd has – whether deliberately or not – positioned her blog essay in an ontological nirvana between scientific research and impressionistic argument. She has creatively remixed different forms of publishing in a way made possible only by the push-button power of the Net. By placing her blog next to her archive of research papers, she has created a hierarchy – here a few loosely connected thoughts, there a collection of structured ideas. The blog essay is the hybrid form that reaps the benefit of both text types but is under none of the constraints. To those who find it convincing it can be research, to those who are skeptical it is supposed to be just an opinion piece.

It’s like science, just without all that rigor.

Maybe that is what is causing boyd’s uneasiness about the impact of the piece, expressed in her closing note:

I also need to get my head around the fact that sharing something problematic has sparked more of a conversation and reflection than being precise. In some senses, this bothers me. At the same time, inciting people to think is exactly what I want. So I am feeling very bewildered. Is the way to make change to present something problematic so that people have to engage by disagreeing? Hmm..

(from her blog)

It’s the truth vs. answers conundrum again. Scientific investigation leads us to more questions, not to any definite, final answers that are easy to understand and confirm the stereotyped expectation we have, based on our own anecdotal experience. It isn’t science’s job to spark “conversation and reflection”, as harsh as that may sound. I don’t mean that it’s not great when that happens – I mean that that shouldn’t be the primary goal. Because if it is our goal, we are putting social gratification before investigating the truth, since in the end “conversation and reflection” largely accomplishes social ends. It’s not a popularity contest, it’s a truth contest. But of course I’m probably looking at this from an entirely wrong angle.

In a way, Robert Scoble summarizes the appeal of boyd’s blog essay chimera aptly:

Here’s why I was interested in her paper last night where some of her other, more “researched” writings just leaves me cold.

1. It clearly defined a conflict. And a big one at that between two classes of people.
[drama makes for great entertainment]

2. It fit my already pre-defined stereotypes. My brother, Ben, for instance, is on MySpace [...] Of course these kids totally fit into danah’s post last night.
[stories are told to reinforce our view of the world]

3. Most of danah’s posts are written for an academic audience. Put a little simpler: they are information dense and hard to get through. The one last night had a breezy, conversational feel to it. It was more approachable than her usual writings. I think that in our RSS “J, J, J, J” fast track world we just give up on posts that are too academic and not interesting to us as humans.
[please don't make me think]

Oh danah, do you forget that we live in a world that pays 1,000,000 times more attention to Paris Hilton in jail than we do to whatever our President is doing? And you wonder why your article yesterday got so much attention? You hit the same nerve that Paris Hilton does.

Perhaps it is just me, but I am not entirely sure boyd enjoys being framed as the Paris Hilton of internet sociology. At least, I can see that not working out so well with her feminist activism. Mass appeal is just one form of currency and in the academic world it’s not the one with the highest value.

In the end this is not about one piece of writing, its author or how it’s being received though. It is about how we communicate what we do as researchers in a networked world. When everyone suddenly has a voice, authority is crucial – and absolutely invaluable. Scientists have authority not because of academic titles or the prestige of universities, but because they play by certain rules and present their work in specific formalized ways. These ways are changing, invariably, along with the way we communicate digitally. The academic community (and especially an innovator like mrs. boyd) should actively engage this change. Academic publishing won’t take place on paper in the future and the genre labels we’ll use will be different. But we will need clear labels nonetheless; we will need peer review and we will need other measures of authority that are as objective as possible, because otherwise there is the danger that authority and popularity become synonymous and that our job is merely to validate what the majority finds plausible. We should let the priests and the politicians do that.

I know that in the context of the Web as a democratizing force this probably all sounds medieval, backwards and elitist. But my concern is trust, plain and simple. We are trusted for basing our claims on facts and we should do all we can to prevent incompetent journalists from misrepresenting what we do. I know that that’s easier said than done, but to me at least it’s something to strive for.

2007 June 27

But what exactly is it that shouldn’t go unacknowledged? That the use of social networking sites may be socially stratified? Not only would that be hardly surprising (if it could be proven), but it would also hardly serve as a grounds for any kind of activism (should we ban Facebook for elitism, or MySpace for appealing to the proles?) … All of these things seems ridiculously implausible.

Here you’re either being “truthy” or you’re stunningly uncreative. I work as a Web developer and something of on online organizer for a non-profit that’s very concerned with reaching out to a diverse audience. As a designer, I’ve always looked down on MySpace for reasons that should be obvious. As soon as Danah’s paper connected the dots for me, however, I realized that we should be making an effort not to discount the MySpace audience. Boyd’s analysis, adademically rigorous or no, rings true enough to make a difference in my professional behavior and provide me with some ammunition when it comes time to justify it.

(Note on capitalization: Sorry, Danah, personal names start with capital letters.)

2007 June 27

Good point, I admit I was being far too general there. Information on the make-up of user groups on MySpace and other networks are certainly interesting and relevant for a number of reasons. And I can’t stress enough that I too find the idea plausible that MySpace=lower class and Facebook=upper class (simplifying a great deal) and that this is also reflected in the look, feel and “culture” of the sites.

“Boyd’s analysis, adademically rigorous or no, rings true enough to make a difference in my professional behavior and provide me with some ammunition when it comes time to justify it.”

The ammunition part is what irks me. It’s (better) ammunition if it is based on (more) substantial research. That’s not your problem though, but ours.

2007 June 27

As an amused observer of the danah controversy (now the ‘danah effect’), and someone in an information-related field, I have been reading/browsing/scanning commentary on various blogs and stumbled into yours. You make a lot of really good points, especially about trust which I think increases in proportion to your academic and blog notoriety. Without saying so, I think you suggest that we should, as bloggers, adhere to a set of codes or guidelines for ethical use of the blogosphere – are you? Fresh off her spearking stint at Harvard, danah may have been emboldened to do something that would catch attention for, as the Scobleizer wrote yesterday, we live in a Paris Hilton news-saturated world. It can’t hurt an academic’s career to stir up controversy, and in using the media in this way I’m sure danah completely knew what she was doing.

2007 June 27

“Without saying so, I think you suggest that we should, as bloggers, adhere to a set of codes or guidelines for ethical use of the blogosphere – are you?”

Yes, though I’m not suggesting that I have the magic formula to resolve certain ambiguities, e.g. is there *always* a clear difference between a blog entry, an essay, a paper, a study etc (I think most of the time there is, but sometimes things inevitably get fuzzy.

New technology makes new forms of expression possible and that’s absolutely a good thing (I love my blog). But by labeling something in a deliberately unclear way we can make it look like something it isn’t, or we can make it easy for others (i.e. old and new media) to misrepresent our work and create a false image of its authority.

“It can’t hurt an academic’s career to stir up controversy, and in using the media in this way I’m sure danah completely knew what she was doing.”

Absolutely, and I honestly have no issue with looking out for media attention. I just think we need to check the validity of our claims constantly, because the BBC & co certainly won’t do it.

2007 June 28
CaptainBooshi permalink

You raise a good point about the need for more discussion about the changing methods of communication. I don’t believe that danah boyd was doing this all purposefully, but she was doing everything you said, and this is something we need to talk about, especially during the ongoing advent of open-access digital journals. Do you have any ideas how we can bring this into a broader focus besides just discussing it more?

2007 June 28

I think conferences, workshops etc about scholarly publishing and communication are a start – for example I hope there will be room to discuss this at the PKP Conference ( There’s a tendency to look at these things from a purely technical perspective (e.g. how do we implement an institutional repository? how do we assure permanent access to research results?) but I don’t think anyone has thought about ‘best practices’ or ‘rules of conduct’ for academic Web publishing yet.

2007 June 28

[...] now, after an exciting trip into the world of science blogging, we return to our regular scheduled [...]

2007 July 1

[...] concerning socioeconomic status and social networking sites with a second post (read the first one here). As with my previous piece, I’m especially interested in the reception and in how boyd [...]

2007 July 13

Hmm. Not sure I agree. Firstly, I’d quibble (and yes, we’re talking opposing ontologies here) that danah has to do substantive qualitative work to back up her conclusions. One could equally rule out quantitative analysis on the grounds that the criteria for analysis would only be valid in the beholder. Now I’ll grant that that’s a position which is hard to sustain in terms of medical research about the success rates of a vaccine or treatment method but when it comes to something as fuzzy and subjective as people, I don’t see that a fuzzy, subjective position which acknowledges and reflects on it’s limitations is an invalid one. It may well be the most appropriate position to take since in all things social, we’re dealing with theorising about people’s perceptions of themselves. To be more explicit, I’d say that danah was theory building. Maybe quantitative research is one way of testing that theory but her initial lack of it isn’t necessarily a problem.

My second quibble is as to the authority that scientists actually have – history suggests that facts tend to be most mutable in the hands of those who most vigorously attest to their concreteness and validity – fine when you’re talking about the tensile strength of a steel cable (if your sums are right), not so fine when it comes to issues of society and class.

On the other hand – I do agree that danah’s mix and match of academic and blog discourse is on shaky ground. Whether she likes it or not, she’s an academic, she presents herself as such and she can’t put off the privilege and responsibilities of academic speech at will any more than a judge can visit a strip joint. So to that extent, I’d say that the BBC’s reaction was not unfair (I’m not contradicting myself, I’m developing an argument, ok?) This is turning into an essay so I’ll stop! But the meta-issues have certainly become more interesting than the ostensible ones.

Note on capitalization: Sorry, LS. Personal names start with whatever the owner wants them too. But that’s just me.

2007 July 14

Your point about the problems with measuring class using quantitative data is well made – I agree that there is no definite answer to that question and I find danah’s observation quite persuasive.

As noted at some point in my blog entry, her method is perfectly suited for answering all sorts of questions. I’m just not sure if it is well-suited for answering that question in particular. But you’re absolutely right that that point is open to debate. The other aspect – that presenting her thoughts in a certain way lead the news sites to label it as a study (without consulting danah first) – is what makes this such an interesting case.

The meta-issues are indeed interesting and lead to all sort of tricky questions in conjunction with how academics communicate in the age of the social web.

Comments are closed for this entry.