Reflections on my Dissertation Part 4: We’ve seen Clickbait Before

This one will be quite short.  Whenever I get frustrated at a piece of “news” that is high on personal attacks and short on facts, I remember that we’ve seen this stuff before.

Many right now in media have been using pre-World War II as a model for the considerable unrest occurring in the world right now — at least from a media point of view.

I think a better model is post-War of 1812 North America. Canada was caught between the Scylla of UK colonialism and Charbydis of US expansionism in terms of its media coverage — something I discuss in a previous reflection — and a large number of American loyalists became the mainstay of Canadian culture by using new technologies like pulp paper and the rotary press to develop platforms for their news. One of the main differences between then and now, however, was the near reliance of these news business on government printing in order to stay alive. But sometimes there are “special interest stories” that resemble much of what we might see on Gawker or Buzzfeed:

On friday evening last some of our official sprigs and hopeful upstarts sallied forth to make a display of their activity; which they did by making depredations upon the property of our peacable citizens. Mr. James Taylor, Merchant, had a quantity of salt barrels lying piled up in his house, where they their ingenuity by tumbling them down where some of them were broken by the fall — they proceeding from thence to pull down diverse signs, one of their member, a young gentleman by the name of Craig (who we learn was in the main well behaved) a nephew of the Archdeacon, who had clambered up for the purpose, fell from his station and injured himself so much that he died on Tuesday evening. We understand that the companions all ran away and left him except William Campbell Esq., clerk of Assize for the district. We have not learned their names but we understand that Robert Hamilton, a student of law, and one of the Archdeacon’s sons, were of the party.

What a pity our gents would not learn to behave better. Had these depredations been committed by the boys in our office there would be little doubt but they would soon have found safe lodging within the strong walls of the brick house.

We understand that young Craig stated that he had been drawn into the scrape by persuasion, and that he desired all the youth of the town to take warning by him and avoid such company.

It’s unclear to what degree poor Craig’s injuries could have been helped by modern medicine, but what strikes me about this passage is the very informal and personal style of the author. There’s a clear social-justice bent to the language and the story is very much intended to act as a warning to all those other young hoodlums who would get themselves into such a mess. By changing the language, this could easily become a Tumblr post.

The newspaper that ran this article, by the way, was the Colonial Advocate (cited in W.H Kesterton’s A History of Journalism in Canada) — the owner of the paper was the leader of the Upper Canada Rebellion, First Mayor of Toronto and First President of a declared “Republic of Canada” in 1837 five years after this article was written.  He was also the grandfather of William Lyon Mackenzie King – a future Prime Minister of Canada.

So, if the historical pattern follows, the concern that we should worry about most is not a build-up of fascism, but instead a divided public on the road to civil war.  Canada’s response was the response of a nation too small and poor to get into a big fight: a restructuring of government that followed Durham’s famous report on responsible government for Canada. Unfortunately, the differences between the more progressive and expansionist Yankees and the colonialists slave-owners to the south were not as reconcilable. One hopes that common ground can build to prevent such a catastrophe as a big war.

Reflections on my Dissertation Part 3 – What Do We Know We Know About Twitter?

As a long-time Twitter user, discussing what Twitter does or does not mean to public policy in Canada was surprisingly the hardest thing to do. The main challenge is that there is plenty about what we think we know about Twitter, but really very little we can say we know for sure. Here are some things we do know for sure.

“Ordinary” People Create News on Twitter

Content creation has long been a privilege of a certain class. This class was not always politically powerful. For instance, ee cummings was a very successful poet, but had very little influence in an institutional sense. However, they did have a sense of the institutions that matter most to political life. For instance, Joseph Howe had to be keenly aware of the rules of engagement when he criticised the politicians of his day. He also needed some serious political resources to avoid being sent to prison when he eventually overstepped.

The Joseph Howes of our day are much less attuned to the institutions that potentially affect them. First of all, the critic and the criticized may be many miles away and live under different rules, although I eventually show that with some exceptions, we can expect a Twitter policy conversation about Canada to be primarily by Canadians, and particular Canadians in regions that are most affected by the policy. International actors tend to engage in parallel conversation about Canadian policy instead.

Twitter is a “Thin” Engagement Tool

In social network theory, there is a distinction between “strong” and “weak” ties.  Strong ties are those that are long lasting and we measure them by something called “triad closure” which basically means that the friends of your friends are also your friends. “Weak” ties are more fleeting but they are important for the spreading of information (or conversely, contagious diseases) across long distances. If you have been following this series, you may notice the connection to Harold Innis.

For Twitter there is a bit of a difference, however.  I think the main difference is that in an online social network, there is always an “invisible friend” that confounds our traditional understanding of “weak” and “strong.” That invisible friend is Twitter itself.  This would include the Twitter interface, algorithms for recommending followers, “trending topics” and any other feature of the site that would connect one user to another. I call this “thin” engagement, using the language of some theorists.

While we need more research on this idea, I suggest that the “thin” elements of Twitter include a tendency to react to our experience of receiving information more than the development of a relationship. The best example I can offer of this is what happens when a celebrity’s name appears in Twitter’s trending topics. A common reaction (especially after 2016) is to fear that the celebrity has passed away, shaping what might be a response that may in turn result in a discussion. Others have argued that a tool like Twitter permits a sense of parrhesia — the ability to speak frankly about topics — so to emit a truth that otherwise could not be done in person. I argue that this parrhesia perspective is not parrhesia about society, but parrhesia about the information we experience while on these tools. For this reason, it’s not particularly helpful a source for establishing priorities for policy.

Twitter is Parasocial

Twitter may offer a thin engagement experience, but that does not mean it is not influenced by power structures. One element of its power is its parasociality. Parasociality refers to the way people experience an interaction as being reciprocal (ie. both parties feel the same about a conversation) when in fact it may not be. A non-online example is the experience of being a fan in a stadium watching a football game. When your team gets a touchdown, it feels as if you had a role in making that happen merely by cheering on your team. You get a sense of victory even though you have no business thinking you are in any way like the athletes on the field.

In my view, Twitter has this stadium-like quality that undermines the role that ordinary citizens can play in policy-making. If anything, like a Dennis Rodman, Sean Avery, John MacEnroe, Muhammed Ali or Chael Sonnen the athletes can revel in their negative press gaining notoriety as a heel in the sport, increasing their earning-power. Donald Trump appears to have used a similar approach on Twitter to his benefit as well.

These are some things I feel I have provided some (leading, if not conclusive) evidence for in my dissertation. In the next reflection, I am going to look at what I used as a method and then I will follow with some case studies.

Reflection on my Thesis Part Two: Agenda Setting

Agenda setting is confusing in the policy literature, and particularly so because technology can make us feel so close to the decision-making process. In a layperson’s sense of the term, agenda setting is about raising issues for governments to solve. Maybe someone notices that a particular crosswalk appears unsafe for pedestrians.  She calls the authorities and lets them know about it. If that doesn’t emit a response, then she goes to media. And if not media, then maybe she can drum up some concern on social media by creating a hashtag and posting some videos of dramatic close calls. That attention, we are to believe, tells politicians that they risk not winning the next election if they do not address the issue.

In policy theory, however, agenda setting is very different. Policy agenda setting is about what issues go on the agenda at an executive board meeting. Moreover, it implies something about the scope of those issues and their relative importance from the government perspective.

For example, dealing with societal problems one issue at a time creates problems of its own. Decisions on what constitutes an unsafe crosswalk and when they should be upgraded or replaced may be more relevant to government than any one crosswalk. Except this plan requires an understanding of the problem from a wide range of perspectives. Will it increase taxes? Who will we buy the new equipment from? Who will install the new systems? Will they obstruct traffic and does that matter if it saves lives? Then there will political arguments. “This city is involved in a war on cars” says a community group concerned about traffic. “People should feel safe in their communities without cars driving through them all the time.”  These groups may be determined to thwart policy no matter what, or just wanting to ensure that their voices are heard. And so on.

In short, policy agenda setting is a messy thing, potentially involving near limitless political conflict, economic & scientific evaluation of evidence, stakeholder input and pressure to both expand and contain the scope of the issue at hand. Take, for example, the role of the tragic deaths of Amanda Todd and Rehtaeh Parsons in the bringing about of legislation to prevent cyberbullying in Canada. The average Canadian will suggest that public outrage, facilitated through Twitter and Facebook were the reasons why the government had to respond to the cyberbullying issue. In reality, the federal ‘response’ was already decided a decade before these incidents during a European convention on Cybercrime in Budapest, Hungary. You see, to capture someone on the Internet sometimes or often requires international cooperation, which in turn requires agreements about the terms upon which such information can be exchanged — for example, how quickly police can and should respond to such information requests.

But wait — you may say — that is more about cyber-CRIME like international fraud schemes or child pornography rings. That goes way beyond the scope of the real issue, which is the mistreatment of young people by cyberbullies. Of course it does.  The point is that “setting the agenda” may have little or anything to do with what people get outraged by. In this case, the solution had much more to do with the giving power for police to do their jobs than the tragedy of the victims, other than to suggest that such an increase in police power would prevent cyberbullying.

On the other hand, to say the Twitter outrage played no role in agenda setting is not quite true either. In fact, social media was part of the reason the policy stalled in the early days as an anonymous privacy advocate posted court proceedings for then Public Safety Minister Vic Toews divorce proceedings. The public embarrassment demonstrated the dangers of privacy from Toews “you are with us, or with the child pornographers” perspective and the Conservative government backed down from the policy.  When nearly identical legislation passed after the Todd and Parsons tragedies, it was much less viable to argue the privacy line. Opponents could only argue that the government was passing a “Trojan Horse” law, promising to save young girls from heinous cyberbullies while undermining the privacy of ordinary citizens.

360 No Scope

In sum, I think agendas are determined largely ad hoc. There are many theories about the details of the ad hoc-ness of agenda setting, but largely people think about the issues and come up with solutions behind closed doors. Then they try to get them on the agenda in any way they can. Sometimes that means being quietly about it; sometimes reducing the scope and implementing piecemeal; sometimes finding a circumstance to demonstrate a broader policy principle; and sometimes by using an event as a window of opportunity to make a decision. I find that Twitter outrage plays a very limited role in policy development right now. In some ways, it may make one time dramatic displays of public resistance to seem rather mundane and ordinary. In much the same way that developing your own blog was seen as innovative and new in the turn of the millennium and now is boring; so goes the use of attention as an agenda setting device.  It’s not so special when anyone can make something go viral. More importantly, it’s not threatening to those in power. For those in a Democracy, we must question whether such actions increases or reduces the power of the people in the long run.

Reflections on my PhD work — Part One Harold Innis

I have been studying Twitter for the past four years, following major controversies including Idle No More, Gamergate, Trump’s election campaign, Black Lives Matter and so on. Not all are mentioned in my dissertation, but I’ve been following them nonetheless. I want to know what online people think is important and how this connects (or does not connect) to what the state thinks. Can we speak about online and offline people as if they are two different things? If so, what are the differences?

Other scholars have thought about this question as well, usually under the umbrella term “agenda setting,” although philosophers like Jurgen Habermas and political economists like Harold Innis have considered it from the perspective of political discourse and empire respectively. Innis is probably the best start, because he is Canadian and my dissertation is about agenda setting in Canada.

Harold Innis

As a country wedged culturally and politically between Europe and the United States, Canada is a country that has historically been “cultured upon” more than it has influenced the global culture. Innis argues that this reality is first the result of Canada’s economy being historically relegated to staple goods (furs, fisheries, lumber etc.) for larger powers in France, the UK and the United States. More recently, the oil and gas  and other mining industries have been the major influences on Canada’s economy. In the digital age, one might also see Canada’s service sector as an extension of the staple thesis.

Innis’s second consideration about Canada’s “cultured upon” status is his theory of communication bias too often mistakenly seen as merely a precursor to Marshall McLuhan’s ideas about “hot” and “cold” media. The difference between McLuhan and Innis is that McLuhan is much more interested in the role of media in creating a global society. Innis is much more pragmatic — he argues that an empire must be able to communicate over time and space to survive.  That is, an empire must have sufficient political influence over geographic space (very important for Canada, a country with plenty of geography) and  a set of common or at least compatible identities that “stick” for a long period of time. Different media are biased towards one or the other of these necessary empire-sustaining ingredients. For example, a book is much more durable than a newspaper and thus has a time bias. People are much more likely to have read Pride and Prejudice than a newspaper from the early 19th century. This means that Pride and Prejudice has achieved its goal of extending the culture and values of the Anglo-Saxon empire to Canadians 200 years later. For example, the value of marrying for love rather than convenience (a major theme in the book) has much more resonance for modern Canadian society than does the use of marriage for power relations more common in Renaissance Italy.

On the other hand, why is it that of the 30 highest grossing films in Japan, only 9 were originally written in Japanese? Examining 19th Century Canada, Innis noticed the role of pulp paper and the invention of the rotary press making it much easier for the United State to convince Canadians to buy American products. The power of radio, television and now digital media to promote and distribute goods across long-distances shows the power of space-biased media. Twitter and Facebook exemplify space-biased media, but with one caveat: Innis has also argued that oral communication has an important role in time-bias, exemplified by the major influence of Socratic philosophy and Judeo-Christian religions. Is it possible that social media sustains both time and space bias and defies Innis’s theory?

Time and Space?

What I discovered is that issues may have continuity online when social groups do not. What this means is that ideas can go viral on the one hand, but the deepness of the idea does not necessarily carry across time periods. When I look at social network graphs, it is as if I am seeing a large group of people yelling comments at their television, except the comments remain even though minds have changed. Innis’s work definitely applies generally to social media, but there seems to be something incomplete about the analysis. What does it mean when ideas that have sudden but fleeting relevance to society are etched in digital stone for people to view 10, 20 or 30 years later? How will historians view this information? How will we use (or not use) these ideas to identify what matters to our nations? How will this kind of data influence the way we allocate resources for better or worse?

In the next installment, I will spend more time looking at what “agenda setting” is and what it means to media, society, and the state.


I’ve been plagiarized! It’s an Outrage!

As I’ve quietly tried to ignore the GOP convention, a major controversy had developed based on the claim that Melania Trump had plagiarized her speech from Michelle Obama.

Mashable ran Trump’s speech through Turnitin and found that the speech was as much as 47% non-unique.  Some clever person did a little hunt for the first result and found that the original genius writer was no one other than yours truly.

Here’s my original phrasing:

Your word is your bond. You do what you say you are going to do and it matters to you lots when projects do not come through they way they should.

Here’s Michelle Obama’s wording:

You work hard for what you want in life. Your word is your bond and you do what you say.

And Melania’s wording:

You work hard for what you want in life. Your word is your bond and you do what you say you are going to do.

Almost exact wording makes you think there is a very serious problem, hunh?  As much of an honor as it is to be plagiarized by a current and potentially future first lady, I think it’s fair to say that this is a pretty common definition for integrity which is political jargon for “please vote for me.” The Wall Street Journal made this point which is how I eventually found out about all this as did this clever Ars Technica user.

The similarities between the two speeches as a whole do seem a little like laziness to me on the part of Melania’s speech writer. Of course, I do not approve of plagiarism and in academic papers I strongly advise against even blinking in that direction.  On the other hand, I would not use Turnitin on a wedding speech and take the results seriously either.

At least I had an excuse to write a blog post as I hadn’t in over a year! I hope all the readers of periodic nothingness are having a great summer!

Ten Implications of Social Networks That Matter

“What exactly do you do?”

I was told this would happen when I was doing my PhD.

The true answer to this question is “I write code in R, do maths, write articles, manage data and read lots of books” but this doesn’t really answer the question implicit in the subtext.

“What do you do?” actually means “what do you accomplish” or, perhaps in even more market-ey jargon “what difference do you make in the world?” In answer to that question, it is probably better that I should say: “I describe political groups in order to help them identify potential blind spots, function better internally and connect their ideas to resources.” Also, I tend to focus on the habits of social groups in online social movements. And generally, I use social network analysis to do that.

There are tutorials and classes about the main elements of social network analysis if people actually care about things like degree centrality, clustering and path length. It’s all quite fun stuff, especially because I enjoy things like Linear Algebra. However, the details of social network analysis are different from their implication, which are often more interesting. So, I thought to offer ten core implications that make Social Network Analysis Matter.

1. Your Thoughts and Ideas Exist in a Social Context

Sorry for the jargon, but this basically means that your ideas are at least partly a function, not only of who you know, but also of who also knows the people you know.

2. It Matters Whether A Social Context Runs Both Ways

You get ideas from others, and sometimes you get to share them back directly. Other times, you share the ideas and it goes nowhere. Even other times you share the idea and it goes through a round of “telephone” before it comes back to you looking like something completely different than what you said.

3. Social Costs are Important

The more attention your ideas get, the more likely it is that there will be some social costs that come with that attention. Some people are more free to share their opinion in public forums than others. Therefore, we should not be surprised to see only a few voices left over after a big online drama.

4. Most Major Social Problems Online are Actually Local New York (or big city) problems.

Manuel Castels famously theorized that flows of information across networks are a source of power. More specifically, he argued that virtual space is guided by a “space of flows” that focuses attention on public problems. This space — namely the space of ideas that have mainstream appeal — is usually dominated by those who are the best connected overall. That’s why catcalling ends up being a salient women’s issue that captures more attention than the atrocities of boko haram. People in New York are overall better connected than those in Nigeria and have much more power to spread their messages whatever they may be.

5. Both Friends and Acquaintances Matter to Us

Friends give us both social support and hold us accountable when we don’t behave as we should. Acquaintances help us access communities that we otherwise wouldn’t have access to. Friends will give us their shoulders when we lose our jobs, but acquaintances will most likely help us find a new job. So says Mark Granovetter.

6. The Org Chart Rarely Tells the Whole Story

There are people with cushy executive titles and then there are people that we trust to get the work accomplished. Sometimes you need to get the latter person onside before you can get approval from the former.

7. Networks are Usually Dynamic

Networks in real life are often like an amoeba. There is a nucleus that stays more or less stable, but the overall the shape can change dramatically from one moment to the next. Perhaps there may be a way to tell how the shape of the amoeba changes based on the actions of people in the nucleus. Interesting hypothesis – I might test that out some day.

8. As Time Goes By, The Default Network Shape Will Be A Star

It may not be the best situation for our communities as a whole, but most networks will feature a few “representatives” that with a whole bunch of followers. In social network graphs this looks like a “star” with one person in the center and everyone else following that central person.

Screen Shot 2014-08-19 at 10.09.13 AM(Notice how everyone seems to follow that one orange ball in the centre?)

9. “Brokers” fill the holes in our relationships and Make Us Stronger as a Group

Inside any social group, there are those few people who will show up to a biker bar in the three-piece suit, or attend a Symphony concert in a mohawk. Watch these people because they connect groups in new and interesting ways and encourage innovation. So says Ron Burt (more or less).

10. Social Network Analysis can find Rabble-Rousers

It’s a double-edged sword, but using social network data, you can find the person who matters most in a network. Consider the hypothetical example of British Colonialists looking for this revolutionary named Paul Revere. In short, there are a lot of ethical questions to consider when using social network analysis. Like any tool, it’s bound to benefit those with resources to use it the most against those who are capable of using it the least. Edward Snowdon seems to have a very important point here.

That’s 10 things. Hopefully it helps provide a bit of context for people who know nothing about social network analysis. At the least I hope it shows the potential power the method has to describe the invisible ways that our friends friends’ friends have a real impact on our daily lives. And I did it by showing only one social network graph!

Anyone else have any ideas on some of the big picture things that social network analysis can tell us?