Você está na página 1de 13

The Twitter Revolution and #HumanRights

By: Luis Rosias

Research Topic:
During the 2009-2010 Iranian election, news sites claimed something unprecedented was happening.

They claimed the Iranian people were using technological tools such as Twitter, Facebook and SMS to

organize protests which would lead to massive demonstrations around the country and the world in

order to attempt to take down the regime. Ultimately these protests were unsuccessful, but many

pointed to the power of social media in helping to fuel this movement. I became interested in following

#HumanRights in order to see how social media was or was not improving Human Rights and to get a

better sense of to what degree social media could influence social change. Human Rights are extremely

important as they are our rights as citizens of the world. If human rights are being violated elsewhere,

that diminishes our own rights. If our government is ok with allowing and participating in practices of

torture, what does that say about our own values and rights? Human Rights must be protected

universally and citizens must make their voices heard if they wish to incite change in their governments.

It is for this reason that researching the ways in which people share information and make their voices

heard matters. The hastag #HumanRights is used by members of the international community to raise

awareness of social problems and instigate change. This paper will focus on the topology of the

#HumanRights network on twitter, the types of conversations that are occurring on the network and will

discuss the effectiveness of twitter as a social platform for social change using twitter data with the

hashtag #HumanRights.
Early Hypothesis:

Like many networks, I expected certain nodes to have a lot more links than others. In my previous

experiences with social media and the web in general, Ive noticed that a few number of individuals or

organizations dominate a large amount of traffic. Google dominates the search market, while Amazon

dominates the online marketplace. The internet due to its ease of scalability and near 0 cost of an

additional person tends to be very monopolistic and networks formed on the internet tend to follow

that same trend for those reasons. In the real world, people are limited in how many friends they can

have, but on twitter sharing your opinions with 10 people is nearly identical to sharing your opinions

with 10 million. As a result I expected a few number of individuals to dominate the network.

I also expected the conversation to be fairly global and to focus on a variety of topics. I was unsure how

the fact that #HumanRights was an English hashtag would affect the distribution of communication. I

predicted that most of the communication using the hashtag would be in the United States as the

country is both technologically developed and has a large English speaking population.

Literature Review:
In his influential 1964 paper, Marshall McLuhan made the bold claim that the medium is the message.

In short this means that the medium through which information travels, largely influences the content

that message can contain [1]. This concept applies directly to our study of Twitter data as we must

consider how the medium itself influences the types of messages that we see. Zeynef Tufekci takes this

concept and applies it to caution those who seek to use Big Data to draw conclusions about the real

world. She cautions that one must pay special attention to the implicit and explicit structural biases of

the platform, the fact that our analysis is coming from a single platform, and the representativeness of

our sampling [3]. These concerns are directly related to our data analysis as Twitter has severe structural
biases that affect communications such as a character limits and how information is diffused (from

tweeter to follower). Additionally we are only using twitter data to draw out conclusions and the use of

an English language hashtag likely greatly impacts the representativeness of our data set.

The structural limitations of twitter and pitfalls in the analysis of twitter data in the media have

led some academics such as Malcolm Gladwell and Golnaz Esfandiari to disagree with popular media as

to the influence of Twitter on the so called Twitter Revolutions [6][8]. According to Esfandiari the

impact of Twitter inside Iran was nearly nil. She says journalists simply took to social media and see the

multitude of tweets with the #iranelection and conclude twitter is influential, yet through it all, no one

seemed to wonder why people trying to coordinate protests in Iran would be writing in any language

other than Farsi. She then points out that various prominent twitter accounts commenting on the

protests were based in The US, Turkey and Switzerland and not Iran itself. [8] Malcolm uses these

insights to claim that Twitter itself was not that influential in these movements and that word of mouth

was still the most powerful form of social change. He claims that social media is filled with weak links,

but that strong links are necessary to organize a movement.

Weak ties are powerful because they allow us to diffuse information to a greater network; the

majority of interactions in your personal network are with people who you have formed strong ties. Due

to the triadic closure property which states that if A and B have a strong tie with C then A and B with

develop a strong tie between each other. Due to this property, your network and the network of your

close connections overlap which makes it unlikely for you to receive new information from a strong

connection. This is where weak ties are powerful because they are where new information likely enters

your personal network as your weak ties have a vastly different network from you. [5]

While Gladwell has recognized the power of weak ties [7], he believes that social change

requires strong ties because there is a great deal of risk involved for those starting a movement and
weak links generally have very low involvement or accountability. He believes a hierarchy and structure

are needed to keep people accountable and make difficult strategic decisions and that a networks lack

of hierarchy prevents it from making decisions that will be followed by the group [6]. However some of

his claim that no hierarchy exists seem contrary to insights by Barabasi into the nature of networks.

Barabasi claims that the degree of nodes in a network follows a power-law distribution;

that is, in networks the nodes with many connections are more likely to develop more connections while

those with fewer connections are less likely to develop new connections. This results in certain nodes

dominating the number of links in the network while the majority of nodes have only a few connections.

In a way, this forms a type of hierarchy where the nodes with the most links have the greatest amount

of influence [4]. However this does not mean that these two viewpoints are not reconcilable.

It seems very possible for organizations to form using the hierarchical structure that Gladwell

mentions to organize events and the be spark for the movement, but use social media and the power of

weak ties to spread word of the event and fuel the movement. However it seems this method is reliant

on the prevalence of information communication technologies and an information rich society. A society

which does not have access to these technologies might be forced to rely on traditional methods of

mouth to mouth communication. This is not to say that all a country needs is access to communication

technology in order to increase their level of human development however.

Research by Stephanie and William Birdsall showed that there was some statistical correlation

between human development and digital access, however there were multiple countries stronger in

human development than digital access while others had high digital access and low development. They

confirmed a statistical correlation within regional clusters that suggested that cultural and economic

factors could play a significant role as well [2]. As a result one must not only examine the technology,

but the people behind that technology and invest in human capital as well.
Overall I would posit that Twitter as a platform is a tool with various limitations that cannot

replace traditional methods of social change. However, if used effectively under the right conditions, it

can enhance and propel a movement forward. Keeping in mind the limitations of twitter and the ways in

which we have examined it is effective and not effective, let us now look at the results of our twitter

data collection and attempt to draw some conclusions from the data.

Methods of Data Collection:


5000 Tweets were collected per day for a month using a Hawksey script (https://tags.hawksey.info/).

This script was used because it allowed us to get data from each day for a specific hashtag. This allowed

us to get tweets that had #HumanRights. The data was not collected at consistent times which may have

led to certain geolocation favoring certain time zones. Tweets were collected all at one time on each

day, and the hashtag #HumanRights had enough tweets to immediately get all 5000 tweets. For data

involving geolocation such as maps, only data with geotags were used. Network analysis was done by

connecting people through the @ sign (which signifies messaging a specific individual) and through

retweets. Certain days were skipped due to procedural difficulties so data used to study trends in word

usage uses as many sequential days as possible.

Additionally Data was used from the CIRI Human Rights Data Project (humanrightsdata.com) to

construct a visualizations that compares physical integrity in 1981 and 2011. This was chosen to help

illustrate the progression of human rights and highlight the concerns in the modern age.

Analysis and Discussion:


The above visualization shows the physical integrity distribution for all countries in 1981 and 2011. In

this metric compiled by the CIRI Human Rights Data Project, a 1 represents the lowest respect a

government can show for individual human rights while an 8 represents a country with the highest

respect for human rights. The above visualizations demonstrate how the distribution of countries has

changed over time. While there were more countries with the lowest score of 1 in 1981, the number of

countries within the range of 1-3 has increased from 21% to 24%. Additionally, the number of countries

that have the highest respect for human rights, a score of 8, has dropped from 27% to 16% while the

aggregate of the top 3 scores has remained relatively the same (dropping from 50% to 49%). The United

States during this time has dropped for an 8 to a 6. These facts reveal that human rights are not by any

means improving universally as some might think. Some countries are improving while others are

becoming worse.
The above is a word cloud generated using Voyant (voyant-tools.org). It takes the most frequent words

in our dataset and assigns them a size relative to the number of times the word has appeared in tweets.

This word cloud helps reveal the intent of users through the words stop, false, and lying. Users are

trying to expose human rights abuses and raise awareness in an attempt to stop these injustices. The

conversation over this time period is also dominated by Middle Eastern countries and locations such as

Iran, Syria, Saudi Arabia, and camp liberty. The word cloud also reveals a secondary conversation in

women and girls as womens rights are also coupled into human rights. The word cloud also reveals

some key figures within the community such as the Dalai Lama, @amnesty(Amnesty International), and

@freemindsar. The Dalai Lama, just by looking at a quick sample of tweets, is a very controversial figure

with some praising and some vilifying him.

This visualization made using Voyant shows the frequency of words over a 5 day span. This shows a

surprising trend as activity seems to have spiked on one of the days and many different countries are

correlated. While it is unclear why all these different countries are correlated, it is interesting to note

that #Iran is constant throughout this time revealing consistent issues in that country. This visualization

shows that while certain events or news might prompt quick spikes that are quickly forgotten, certain

issues are remembered and discussed consistently over time, signifying that twitter can hold on to a

cause for an extended period of time.


The above visualization shows the locations of tweets throughout the world and the bubble size

represents the frequency of those tweets in that area. While it seems that there are tweets coming for

many areas of the world, it seems that tweets are concentrated primarily in Europe. We see here

potentially the influence of choosing an English language hashtag and the technological divide. It is

possible that many others are tweeting about human rights issues in their native languages and are not

being captured by this data. Additionally, there may be many people without access to twitter who are

suffering human rights abuses who cannot make their voices heard due to a lack of access. However the

visualization is also hopeful because while the density is centered in Europe, there seems to be

widespread use of this hashtag nearly everywhere in the world albeit with lower frequency.
The above is a social network analysis of users in the network made using Gephi. Connections denote

retweets and uses of the @symbol. The relative size of the node denotes its centrality in the network. As

we can see above the network is dominated by a few entities, primarily ncri_women_comm and

amnesty. This is very interesting as these groups likely have a large influence on the conversations

happening within the network. It is important to note not only the size of amnesty, but its widespread

reach. It doesnt just have many connections in its local cluster, but rather many connections touching a

wide variety of clusters. This means that information from amnesty spreads rapidly throughout the

network as nodes retweet information within their clusters. women_comm, despite being a large node

itself doesnt have as wide a reach as amnesty. It is also important to note the relative smallness of

nodes compared to amnesty and women_comm. These nodes are central to the network and have the

ability to disseminate information quickly throughout the network.

Visualization Summary:
Together, the visualizations help reveal who is communicating, why they are communicating, and how

they are communicating. The people communicating using this hashtag are from all over the world but

are centered near Europe. They are trying to spread the word about human rights abuses all over the
world in order to raise awareness and help instigate change. It seems however that this platform is not

being used to stage direct change outside of the internet. The absence of the words protest or event

and the wide distribution of users shows that this hashtag is being used primarily for informational

purposes. Additionally 70% of tweets had links in them which supports that people are primarily trying

to provide each other with more information. Due to the wide spread of users, there seems to be little

interpersonal communication, and people are primarily tuning in to the major sources of information

,@amnesty and @ncri_women_comm, to be informed about world events. The visualizations dont

necessarily do a good job of showing us the smaller networks and voices that are getting drowned out

by the major players. In the visualization showing where people are tweeting from, we saw a large

number of people tweeting from the Caribbean yet issues from the Caribbean are not being widely

talked about. It seems likely that the direction of the overall discussion is being influenced by the major

players. The findings are important because they help identify how communication is happening

throughout the network and the capabilities of the network. The network can likely be very effective at

handling international issues where followers can correspond with UN representatives to advocate for

change but less effective in creating change on the ground. The centrality of amnesty and

women_comm means that they can raise awareness of issues and try to accrue resources from a

widespread population that is now aware of these issues. Additionally, these organizations can drive the

dialogue in such a way so that major issues can be tackled and not forgotten quickly. This network does

not seem effective at setting up protests or physical action as the regions they are focusing on are not

speaking English and are largely unrepresented in the network potentially due to language and

technological barriers.

The most compelling visualizations are the social network analysis and the word cloud as they reveal the

majority of the dialogue in the network and the key players within the network that are influencing the

dialogue.
These findings tie into our discussion earlier of the effectiveness of twitter as a tool for social change. As

we can see amnesty is an organization that has the rigid structure that Gladwell said is needed for social

change, but they are using twitter as a tool to get more followers to their cause, spread information, and

obtain additional resources. The network analysis supports the insights of Barabasi and showed that a

network develops a form of hierarchy on its own if left unchecked where the rich nodes get richer over

time. The concerns of Esfandiari were confirmed within our study as well. The majority of the people

participating in this network are largely not in the countries being discussed. As a result this grants some

validity to the claim that Twitter did not have a huge impact on the Iranian Protests of 2010 and that as

Tufekci warned, one must look at the context of the data and the limitations of the structure used in

order to draw conclusions. Many reporters saw that a large number of people were tweeting about Iran

as many are tweeting now but failed to understand the context from which these people were tweeting

from.

Some of the hypotheses proposed were confirmed while others were disproved. The network

was very global, but was focused around Europe and not the United States. Like predicted the network

followed a power law degree distribution, but this is not really surprising as Barabasis findings are fairly

universal in any internet related network. The English language nature of the tweets did affect the

results but not in the way I expected. I didnt expect there to be so few people using the hashtag in the

places the network was discussing. I expected the geotags to be more representative of the discussion,

but this was not at all the case as the discussion is guided by a few crucial nodes. I also expected there to

be more activism occurring within the hashtag, but the hashtag ended up being mostly informational

and less about action.


Conclusion:
Like Gladwell states, to create social change you need strong connections, organization and

determination, but he is too dismissive about the impact that Twitter as a tool can have. While Iran may

have had technological restrictions such as the government stopping cell communications and slowing

the internet, other places do not have such stringent restrictions and the population has widespread

access to information communication technologies. In these places the tool that is Twitter can be more

aptly used to help raise awareness of issues and events. Just because a tool is not useful in a specific

situation does not mean that tool is useless. We saw from the #HumanRights that even without having

people in the physical location of issues, organizations such as amnesty can go to social media and raise

awareness and get the resources they need to make real change in those areas. Organizations and social

networks are not exclusive, and used together can be more powerful than either alone. Although it can

be negative to have many issues raised by common person be drowned out by these larger

organizations, these organizations can help provide focus and help people tackle issues rather than

being overwhelmed by a sea of woes where nothing is likely to be solved. The fact that #Iran was able to

stay high in tweet frequency shows organization emerging from a network and the ability to focus on

issues for the long term. Overall I was pleasantly surprised that the hashtag was working to generate

change and that social media and organizations could work together to protect #HumanRights.
Works Cited:
1. McLuhan, Marshall. 1964. The Medium Is the Message. In Understanding Media: The
Extensions of Man, 111. New York: McGraw-Hill.
2. Birdsall, Stephanie, and William Birdsall. 2005. Geography Matters: Mapping Human
Development and Digital Access. First Monday 10 (3).
http://firstmonday.org/ojs/index.php/fm/%20article/view/1281/1201.
3. Tufekci, Zeynep. 2013. Big Data: Pitfalls, Methods and Concepts for an Emergent
Field. Social Science Research Netrwork. SSRN Scholarly Paper ID 2229952. Rochester,
NY: Social Science Research Network. http://papers.ssrn.com/abstract=2229952.
4. Barabsi, Albert-Lszl and Rka Albert, "Emergence of scaling in random
networks", Science, 286:509-512, October 15, 1999
5. Granovetter, M. (1983). "The Strength of Weak Ties: A Network Theory
Revisited". Sociological Theory 1: 201233. doi:10.2307/202051

6. Gladwell, Malcolm. Small Change New Yorker, October 4, 2010. Accessed May 12,
2015, http://www.newyorker.com/magazine/2010/10/04/small-change-3
7. Gladwell, Malcolm. 2000. The tipping point: how little things can make a big difference.
Boston: Little, Brown.

8. Golnaz Esfandiari. The Twitter Devolution Foreign Policy, June 8, 2010 Accessed May
12, 2015, http://foreignpolicy.com/2010/06/08/the-twitter-devolution/

Você também pode gostar