Morsi and global communication November 23 to December 2

On the 22nd of November, 2012 president Morsi of Egypt declared himself above the law. At least he declared that his actions were beyond the reach of the judiciary. Since there was no legislature the rejection of intervention by the judiciary made him de facto dictator. That shocked Egyptians and people around the world, and it produced an outporing of protest 'on the ground' and communication 'in the cloud.'

The primary focus of this report is on the global character of Twitter communication about Mr. Morsi's actions. But I will start with the volume of that communication stream.

Number Twitter messages per day mentioning Morsi

This is a sample of messages drawn from the Twitter streaming api. It is not certain what the sample is relative to the total stream, but generally it is thought to be a one percent sample. So it is the changing quantity over time that is important rather than the number of messages captured.

Morsi acted on November 22, and the collection of tweets began on November 23. The pattern is a spike on the 23rd and 24th and then a gradual decline over time. The decline was interrupted on December 1 which was the day after the Constituent Assembly approved the final draft of the new constitution. They had worked through day and night to get the document completed so there could be a vote on the constitution by the people on December 15. The speed with which the document was produced and rejection of the document by people not of Morsi's party prompted an invigorated protest and increased messages.

The search

There are two points to be made about the search and the volume of messages found.

First, the search term used from November 23 through December 1 was Morsi. It turns out there is more than one way to write Mr. Morsi's name in western script. There is also Mursi and Morsy. When I discovered the multiple renditions of the name I began parallel searches to estimate the difference made when collecting one versus the three renditions. Between November 28 and December 6 I did parallel searches. The result was 225,515 Twitter messages containing Morsi and 309,357 messages containing either Morsi or Mursi or Morsy. That suggests the numbers in the figure above should be about 1.37 times larger.

Second, on December 2 I launched a new search with a more complete set of search terms. Any Twitter message containing Morsi or Egypt or 25Jan or Tahrir or MuslimBrotherhood was captured. Part of the great increase from December 1 to December 2 can be accounted for by the change in the search term.

The point is that search terms can make a significant difference in what is captured from the stream of Twitter messages. There is no 'correct' search term. It is, therefore, important to be as explicit about the search as possible.

Global communication

In "What? A new dictator in Egypt?" I looked at the first two days -- November 23 and 24 -- and outlined the procedure for tracking the geography of the messages. (Boynton, 11/25/2012) Twitter assigns a version of time zones to accounts when they are set up, and these 140+ geographic locations can be used to track the location of users and the locus of messages.

Between November 23 and December 1 the search captured 217,722 messages by 78,217 user accounts. That was an average of 2.8 messages apiece for the period. Seventy-two percent of the user accounts could be identified by the Twitter time zones. The location of the user accounts posting the messages is presented in the table.

They are sorted in the table on the basis of the increase from the first two days to the last seven days. The three regions that are most heavily Muslim are Jakarta, the combination of nations in the Middle East, and Kaula Lumpur and they show the greatest increase in users posting messages to Twitter mentioning Morsi. The United States, which has the largest number of users posting to Twitter also has the smallest increase. The size of the increases reflects the gradual decline between November 25 and November 30.

How global is the communication? The US has the largest number of users, but it is only 37% of the total. Since we know that half of all messages originate in the US this is a smaller number than expected. There are more users posting from the heavily Muslim regions than from any region other than the US. But in terms of geography they are widely dispersed. Greenland stands out as a locus of users interested in the conflict in Egypt. Europe, South America, and Australia also have users posting about Morsi.

What is missing is the far east, Russia and Africa. Taipei is 30 users. Osaka is 16. Tokyo is 153, which is very few given the thriving use of Twitter in Japan. Seoul is only 21. Moscow is only 56 users, and St. Petersburg is only 15. A time zone for West Central Africa had 164 users posting to Twitter about Morsi, and only 65 are posting from Harare.

The numbers are not huge. There is nothing like the 31 million messages congratulating President Obama on his re-election on November 6. But the concern is widespread. Much of the globe is covered. But there are areas of the globe that Twitter has not yet reached.

Global Communication: Recapitulation

These are explorations in global communication via Twitter. I want to assemble earlier results here for comparison.

Ten days in November -- Barack Obama is re-elected

Between November 1, 2012 and November 10, 2012 I collected Twitter messages mentioning President Obama. The search terms were Obama, barackobama, and obama2012. Barackobama is the user name of the Twitter account the president has had at least since he was first elected. Obama2012 was the Twitter account specifically for the 2012 campaign. The collection accessed the Twitter streaming api which is, in principle, a one percent sample.

The table displays the number of user accounts found in the Twitter messages mentioning the president each day. These are not messages. They are the user accounts sending the messages. The distributions in the table show that a very substantial percentage of the people mentioning the president were from outside the United States. On November 7 and 8 more than 50% of the people posting messages congratulating the president were from locations outside the U.S.

Mr. Netanyahu -- the aftermath

Israel launched an attack on Hamas on November 16, 2012 and a cease fire agreement was reached on November 21. (Federman and Laub, 11/21/2012) I collected mentions of Prime Minister Netanyahu for the next two days. I wanted to look at global the communication about the war and cease fire. The search term was Netanyahu, and the collection ran the two days -- November 22 and 23. The collection totaled 11,805 Twitter messages mentioning Netanyahu. A total of 6,630 user accounts posted messges to Twitter.

A first look is at user accounts from the region.

The total from the region was only 346 out of the 6,630 users who posted a message to Twitter. And there are only a few from each Twitter time zone. Only 116 users in Jerusalem posted a message mentioning Netanyahu, and that was more than double most of the rest of the time zones in the region.

The U.S. and the E.U. dominated the Twitter communication.

The total for these regions is 5909, which is 89% of the total user accounts. The U.S. is the largest with 44% of the users in the six regions. Greenland is again a location of a number of active user accounts. The far east is again very low compared with other regions. And the nations of the former Soviet Union are hardly present.

© G. R. Boynton, 12/13/2012

References

Boynton, G. R. (11/25/2012) What? A new dictator in Egypt?

Federman, Josef and Karin Laub (11/21/2012) Cease-Fire Begins Between Israel and Hamas, AP