Inkwell: Authors and Artists

    
Inkwell welcomes Joseph Reagle, Assistant Professor of Communication
Studies at Northeastern. He’s been a resident fellow at the Berkman
Center for Internet and Society at Harvard (in 1998 and 2010), and
he taught and received his Ph.D. at NYU’s Department of Media,
Culture, and Communication. Dr. Reagle is joining us to discuss his
latest book, _Reading the Comments: Likers, Haters, and Manipulators
at the Bottom of the Web_ (MIT Press, 2015). He's also author of
_Good Faith Collaboration: The Culture of Wikipedia_ (MIT Press,
2010).

As a Research Engineer at MIT’s Lab for Computer Science he served
as an author and working group chair within the IETF and W3C on
topics including digital security, privacy, and Internet policy. He
also helped develop and maintain W3C’s privacy and intellectual
rights policies (i.e., copyright/trademark licenses and patent
analysis). Dr. Reagle has a Computer Science degree from UMBC, a
Masters from MIT’s Technology and Policy Program, and Ph.D. from
NYU’s Department of Media, Culture, and Communication. Dr. Reagle
has also worked on short consulting projects for Open Market
(electronic commerce protocols), McCann-Erickson (Internet and
interactive media), and go-Digital. He has been profiled,
interviewed, and quoted in national media including Technology
Review, The Economist, The New York Times and American and New
Zealand Public Radio. His current interests include life hacking,
geek feminism, and online culture.

Our conversation with Dr. Reagle will take place over the next two
weeks, and is led by Jon Lebkowsky, who has been active in digital
culture and media for over 25 years, and is currently focused on
strategic digital consulting and development as member and CEO of
the Polycot Associates web development cooperative. He is also known
as an activist, sometimes journalist, and an authority on the future
of the Internet, digital culture, media, and society.
  

    
We're having this discussion on the WELL, which is mentioned in an
early section of the book, where you mention Fred Turner's argument
that "much of the Internet's culture is rooted in the West Coast
movement from 'counterculture to cyberculture.'" Can you trace the
origin of the "comment culture" of today's Internet to online
communities like this one?
  

    

Yes, this coincidence is quite "meta"! Indeed, the whole project has
been rather meta. For example, my treatment of the Amazon Vine
Reviewer program has been widely discussed by Amazon Vine Reviewers,
since we participated in that program.

With respect to The WELL, it's not so much about tracing a linear
origin, but seeing rhizomic tendrils. As I write in the book,
"likers" are those who share recommendations rooted in love and
experience.

When people first employed digital networks, I didn't think they
anticipated that the technology would be used in this way. But the
motive to talk about stuff that you liked was apparent even in the
earliest days of digital communication. Internet pioneer Vint Cerf
noted that "when e-mail showed up in 1971 on the ARPANET, we
discovered instantly that e-mails were a social network phenomenon."
How so? With the quick appearance of two email lists dedicated to
"book reports and restaurant reviews" (i.e., the SciFi-Lovers and
Yum-Yum lists).

With respect to The WELL, I think we see these impulses to share in
the careers of Stewart Brand (the Whole Earth Catalog), Kevin Kelly
(Cool Tools), and Mark Frauenfelder (Boing Boing). Each of these
guys have worked together, which creates a span of commenting about
things loved from 1960s print, to 1980 zines, to early BBS systems,
blogs, and online recommendations today.
  

    
If you look at early blogs that focused on reviews and
recommendations, like Boing Boing, you can see the influence of the
Whole Earth Catalog and Coevolution Quarterly/Whole Earth Review in
the structure of those posts: a brief review and a pointer to the
book or product, in this case by linking. And the WELL Gopher was an
early example of curation of nontechnical content on the pre-web
Internet. So I think the influence of the "Whole Earth" project on
the evolution of the Internet was substantial. Katie Hafner also dug
into that background for her book about the WELL.

Did research for your earlier book about Wikipedia, _Good Faith
Collaboration_, have any bearing on your research for the current
book? Said another way, how did you decide to write about online
comments and commenters?
  

    
In a way, the Wikipedia book did lead me to the commenting book.
However, it wasn't research about online community as much as my
surprise at part of *Good Faith Collaboration's* reception. The
Wikipedia book was well and widely received. However, there was one
group of folks who hassled me; these were former (embittered, often
banned) users of Wikipedia who could not tolerate positive reports
of the Wikipedia community. Hence, some of them left bad book
reviews without having read it, and, more oddly, photoshopped my
likeness into pornographic comics. (The comic was slightly clever in
that they used images of penises found on WikiCommons; the
presumption was that if I liked the Wikimedia projects, I must like
these penises as well.)

I can laugh about it now, but I did find it distressing at the time.
This was one instance in which I realized how we live in an
environment of ubiquitous comment. There is almost nothing,
literally, that cannot be commented upon, rated and ranked. This is
why I was so amused by the (satiric) Jotly app in which users could
rate trash cans, slides, ice cubes. "Take for example this glass of
water, which ice cube is the best? According to Jotly: that one,
it's melting the slowest. Helpful!" Even though it was a parody,
there were similar apps out there, and eventually Jotly became a
real app. All of this is to say that comments and reviews had
existed since humans have been communicating, but the ubiquity,
speed, and ease with which we can now make them does signal a
significant social change.
  

    
How do you think about the signal to noise ratio, when there's a
firehose of comments? Thinking of Reddit, for instance, where a post
may have hundreds or event thousands of responses. There are many
comments, but who's reading them?
  

    
Once a community grows beyond scale (i.e., some variation of
Dunbar's number) you will inevitably have the the signal-to-noise
ratio problem. Typically sites deal with this via some form of karma
or up/down voting. Of course, that brings its own difficulties.

As I write in *Reading the Comments*, I came to appreciate this over
a decade ago on the social news site Slashdot. Like many sites,
Slashdot permits active users to rate comments, which have a
cumulative score from worst (-1) to best (5). Readers can then
filter out comments below a threshold. While the system keeps the
worst comments from being visible, I noticed that some of the most
informed comments were also hidden. This *rush and slash effect*
privileged comments written within the period most people were
likely to rate them. Since those likely to rate comments were the
most active users, they typically did this within hours of a story's
posting. After that, the raters had moved on. In fact, I found that
the average age of a comment with a rating of "4 or higher" (where I
set my reading filter) was just over an hour. I typically wouldn't
see *any* comments older than eight hours. Early comments often get
more attention than they deserve.

Also, once you deploy a voting system, cabals of one sort or another
arise. Elsewhere, I write about this problem at the site photo.net
[1]. They grappled with how to prevent people from "mate-rating"
friends and "revenge-rating" enemies in evaluations of  photos. From
this community, I discerned a handful of characteristics of
evaluation in the digital age. The third of which was that "fixes"
to manipulation have their own, often unintended, consequences and
are also susceptible to manipulation. In this case, one they moved
from anonymous ratings to non-anonymous ratings (to increase
accountability), the ratings of new photos were much more inflated.
Consequently, when you searched for the top photos, you'd see no
works before the move away from anonymity---it's as if all those
photos now sucked!

[1]: http://reagle.org/joseph/2013/photo/photo-net.html
  

    
Administrivia: 

If you want to share this discussion, the short link is
http://bit.ly/inkwell-reagle.

If you're not a member of the WELL and you have a comment or
question, one option is to join the WELL, and join the conversation
directly: http://www.well.com/join.html  However you can also
participate by sending your comment or question to inkwell at
well.com. Let us know if you're okay with us posting your full or
first name.
  

    
Following up on Slashdot, at some point they included meta
moderation (i.e. rating moderators' decisions), which was supposed
to mitigate the effect you describe (informed comments voted down).
Did you study the effectiveness of meta moderation?
  

    
Could someone please define "Dunbar's number" for the non-tech crowd?
  

    

<https://en.wikipedia.org/wiki/Dunbar%27s_number>

Dunbar's number is a suggested cognitive limit to the number of people with
whom one can maintain stable social relationships.
  

    
Hi, Joseph.  I will start by just thanking you for providing lots of
data describing the reasons for my knee-jerk distrust of much
internet commentary over the past 30 years.
  

    
This is how I describe Dunbar's number in the book:

> The origins of YouTube and Facebook demonstrate that people like
to talk about one another: we gossip. While gossip might seem like a
trivial thing, evolutionary psychologist Robin Dunbar argues it is
central to understanding our humanity. Even if you don't know who
Dunbar is, if you participate in online communities you might've
heard of his eponymous number: 150. People invoke *Dunbar's number*
when a community, such as an email list in which everyone used to
know most everyone else, gets too big.... However, Dunbar didn't set
out to coin an aphorism about online community; he sought to answer
the question of why primates, especially humans, are so smart. Why
is it that our brains are about nine times larger, relative to our
body size, than other animals? Some suggested it was related to our
environment, our use of color vision to find fruit, the distance
traveled while foraging, or the complexity of the omnivore's diet.
When Dunbar looked at all of these variables among primates, he
found no such pattern. But the size of primates' neo-cortex did
correlate with the size of their groups and the time spent grooming
one another. 
  

    
@jonl, I think meta-moderation has provided a useful layer. However,
it too might contribute to the rush-and-slash effect (no new
comments are voted up) and you'd still want to be careful about the
formations of cabals.
  

    
@krome, I hope I don't come off as cynical or overly pessimistic.
There are informing, helpful, insightful, and funny comments out
there. It's just not the default. :-)
  

    
administrivia, Jon that bit.ly link is no good.
  

    
Joseph, thanks for being with us. This should be great. Could you
talk a bit about how the move from anonymous comments to known
commenters has or hasn't changed the world of commenting? 
  

    
And another question. Is it just the digital world that shows such
muck in commenting - do people behave differently online than in
flatland? Or is it more broadly human at all levels of interaction?
  

    
The bitly link is kind of correct, but somehow the system included
the period at the end of the sentence, which makes it incorrect.

Here it is again:
http://bit.ly/inkwell-reagle
  

    
@tcn Anonymity is probably one of the most discussed and studied
aspects of online communication. There's decades of writing and
research on it. In fact, in the book I talk about Gyges (thousands
of years ago) and Gollum (decades ago) and their rings of
invisibility. In any case, I'd summarize as follows.

- Anonymity serves an important role and realname policies can be
problematic.
- But that doesn't mean every discussion platform has to support
anonymity.
- When folks have moved to systems like the Facebook comment plugin
[1] I think they generally see the quality of commenting improve.
(Although I had to see the Web dependent on a proprietary platform
for netizens to discuss.)
- Even identifiable people can be asses online.

I try to synthesize all of the research on aberrant online behavior
by way of two categories.

1. *good people acting badly*: absent a rich media, social cues are
filtered out leading to an attenuated social presence such that we
don't appreciate our effects on others and feel less accountable.
2. *bad people acting up*: some of what we see online is the
disproportionate effect of a difficult minority.
  

    
That's very interesting.  Perhaps there are some additional
complexties. 

I have observed what looks like bad anonymous actions by people whom
(I expect) have always been good out of a fear of being caught or a
desire to have an excellent reputation.  Taking away any social cost
means you don't have to bite your tongue in certain interactions. 
Even if you are "good."
 
  

    
There's an interesting dimension of exploration in this part of the
book that brings in the "obedience to authority" experiment, which I
thought was conducted by Stanley Milgram but in the book is
attributed to Philip Zimbardo. In this experiment, people thought
they were administering electric shocks to others. "Participants who
wore large lab coats and hoods were more willing to shock others
than participants who wore name tags." This was seen as
"deindividuation: a loss of a sense of self and social norms." There
was also an experiment by Zimbardo where some students were given
the role of prisoner, and others, guard. They so got into their
roles that the experiment was called off.

I can see how this relates to that idea of "good people acting
badly." This is not so much about anonymity as it is about the
adoption of a role, but the two might be connected: a person who
feels anonymous might also feel free to slip into a less civil
"self" and act out a bit, unable to see potential consequences (as
in others' damaged feelings). 

I wonder about the "lack of social cues" referenced above and in the
book. Are some people better than others at expressing and reading
social cues in an environment that's primarily text? "Reading
between the lines," so to speak?
  

    
What I don't understand is how people get bopper off FB. I mean I
regularly go off on the cops, but no one has ever threatened to kick
me off for a month. I've got a few friends who have been blocked for
far less.
  

    
Caught two things off my twitter stream today pertinent to this
discussion:

http://arstechnica.com/tech-policy/2015/09/jeweler-tries-to-sue-anonymous-woma
n-who-left-1-star-yelp-review/

Apparently anonymity can't keep you from being sued for poor
comments. Wondering about the repercussions of that. Also, from your
book I gather 1st amendment rights have not been clearly settled as
to whether or not comments fall under free speech or how they are
delineated.

http://www.niemanlab.org/2015/09/what-happened-after-7-news-sites-got-rid-of-r
eader-comments/

This seems new, moving to social media sites as a forum for
comments. Any thoughts on if this really the direction media will be
adopting?
  

    
I've been blocked for the past month and have no idea why. Doesn't
bother me as i can read my FB but every time I want to LIKE a
country decor site or some such, it comes up with this pink bar
telling me I can't. Have sent messages but not motivated enough to
call or anything.
  

    
@gail on the question of anonymity and accountability, I think you'd
be interested in a study by psychology researcher Tatsuya Nogami. He
attempted to:

> tease apart the differences between identity, such as your name,
and anonymity, an inability to associate you with your behavior. As
part of a clever take-home assignment, he asked over a hundred
university students to flip a coin twice. Some students were asked
to identify themselves on their coin-flipping report and some
received a reward (a coupon book) for getting two tails, which
should happen 25% of the time. Nogami considered all subjects to be
anonymous since he could not know if any individual cheated.
Instead, he could only infer cheating based on the aggregate
statistics. Of the non-identified subjects who could get a reward,
46% reported flipping two tails. Some of them obviously cheated.
However, he was surprised to find that those who had identified
themselves (but still could not be associated with cheating) did not
choose to cheat: only 21% reported getting two tails, statistically
indistinguishable from the expected 25%. Those in the latter group
were just as "anonymous" and unaccountable as those of the first
group because no individual could be blamed: any single person
getting two tails isn't that unlikely. Nogami concluded by
suggesting that asking people to identify themselves perhaps
prompted them to be more self-ware and cognizant of their ethical
standards even when they couldn't be linked to unethical behavior.

You can read it here:
http://opensiuc.lib.siu.edu/cgi/viewcontent.cgi?article=1018&context=tpr