Anssi Männistö: Trust rethought: Digital image after Boston

Anssi Männistö is a professor of visual journalism at the University of  Tampere, Finland. Recently he has been researching the use of cameraphones and tablet-computers in journalism as well as new forms of photojournalism. The topic in his doctoral dissertation (1999) was the Western myths of islam and muslims in the media during the cold war era.

 

 

 

 

Trust rethought: Digital image after Boston
– Social aspects of digital imaging may strengthen its reliability

Boston bomb attack shocked the world in April 2013. The two bombs that went off  at the marathon event killing several and injuring hundreds of people. Exploring a huge amount of images proved to be essential for pinpointing the suspects. With the help of new analyzing methods it was possible to utilise photos that were taken long before the actual moment the crime occurred. The investigation of Boston bombings highlighted the chancing nature of photos and image making in our lives.

Before the suspects of Boston attacks were tracked, various kinds of cameras and new social dimensions of photographying were essential to tracking down the Boston attack suspects. A day after the attack, investigators turned to crowdsourced photographs and videos in order to hunt down the perpetrators. (Zhang 2013a). The FBI received more than 2000 tips and sift through more than 10 terabytes of data. The data included surveillance footage collected by city cameras, local businesses, gas stations, media outlets and spectators who volunteered to provide their videos and photos. (Serrano and Bennett, 2013).

New analyzing tools

The investigation showed a glimpse of what big data and data analytics can do. Main goal of any big data projects is to pull insights from large amounts of data with the help of powerful computers. (Konkel 2013). In this case the investigation benefited from the possibility to make a continuous visual narrative from hundreds of separate photos. When hundreds and thousands of people had captured the event from different angles and moments, the investigators were able to create a three dimensional photonarrative. It enabled investigators to show where all the spectators of this large event were positioned at a certain point of time.

When the FBI published the photos of ”possible suspects” on the internet, the public very soon gave crucial tips leading to revealing the identity of the suspects. For many years in the internet there has been various kind of communities aiming at scrutinize the photos. These communities have been successful in revealing e.g. intentional manipulations. One of the most famous examples of this is the revealing of manipulations related to Lebanon war in 2006. This scandal led the news agency Reuters to introduce its very strict norms concerning news photo manipulation.

In Boston this kind of crowd sourcing produced both results and excesses. Several online communities were attempting to identify the attackers behind the Boston Marathon bombings by crowd sourcing publicly available photographs from the scene. (Zhang 2013b) This led some home detectives to publish photos in which they had tagged potential suspects  -who in the end proved to be innocent. The same happened also in the traditional media. The New York Post took one of these wrongly tagged photos and ran it on the front page with the headline: “Bag Men: Feds seek these two pictured at Boston Marathon.” (Zhang 2013c).

Besieging the suspects progressed when a victim of carjacking called police and the two thieves were identified as alleged bombers. Officers were able to track the vehicle using the drivers iPhone and Mercedes satellite navigation system. (Henry 2013). The other suspect managed to escape, but officers were able to locate him in certain neighborhood. Then, one of the most intensive manhunts in recent U.S. history begun and police roped off the area and begun to search house after house the neighborhood.

One new phenomenon was that the officers using Twitter were about to tell in real time about the advancing of the search. The public was also told not to compromise officers safety by broadcasting the tactical location of homes being searched. In the end of the operation with the help of a helicopter -mounted infrared camera the suspect was found hiding in a boat. Cameras were present also in the final episode: people living near the place where catching of the suspect happened, filmed the raid and put the material immediately to the web.

Metadata adds credibility

In the early years of 21st century the digital image-making has rapidly replaced analogue processes in the news industry. This development initially raised concerns that the fundamental nature of photography itself would alter. Because digital image was easily open to manipulation, it was widely thought this would seriously affect the truthfulness of news photos. Today the spreading of camera phones and social aspects of mobile photographing (see Brook 2012) have opened a new perspective which seem to highlight the fact that we can once again begin to trust the digital photo. The inspection of Boston Marathon attack is a good example of this.

Currently almost every event is surrounded by digital photographing. The photo itself contains more and more relevant metadata and it has become a common phenomenon to share with each other the photos immediately. Under these circumstances photos taken by several independent people can be used to verify and crosscheck the integrity of those photos. With the help of dedicated services and applications, such as Photosynth (photosynth.net), people can download their photos of a situation in the web and the service combines these photos into a three dimensional map, where one can navigate. Law enforcement offices have their own much more powerful and sophisticated tools for their use.

Camera phones have brought rich metadata as a standard feature of digital photos. Various sensors are included in high -end camera phones: GPS, gyroscope, electronic compass and accelerometer. With the help of these sensors the actual location (GPS) and direction of the lens (compass and gyroscope) can be stored into the metadata of an image file. In some cases it might be useful to know at which speed (accelerometer) the camera was moving when the shutter was released. There are several other kind of contextual information that could be stored into the metadata too. It is obvious that the image files can be manipulated in many ways, but at the same time various methods for revealing manipulations, through the so called “digital forensics”, have advanced rapidly.

In the era of camera phones an individual press photographer is usually no more the only witness of situation. Because of the new technology he or she might even stay more relaxed: as several other images are representing the situation from different angles there is no need for proofing that the situation did happen in a way the photographer shows it. Under these conditions the making of intentional manipulations seem to become less and less tempting.

New ethical dilemmas

In Boston the authorities and the media were interested in different kind of photos. Traditionally for the media the most valuable photos are ”a moment after the explosion” –type of scenes. At this time the authorities were more interested in collecting photos that were taken hours before the attack, and also from locations further away from the spot. (Vänskä 2013). Photos and videos taken by different spectators from different vantage points and angles help investigators crosscheck people they might find suspicious. (Ackerman 2013.)

Lauren Grabbe, a freelance contributor to Wired-magazine, was one spectator who photographed near the Boston Marathon finish line about 90 minutes before the bombs went off. A panoramic photo, Grabbe took with her camera phone, interested the FBI. The photo showed a lot of people, what they are wearing and where they are positioned. The FBI analyzed this along with other photos using biometric tools like facial recognition. With this data, investigators can assemble an understanding of what normal behavior at the Marathon looks – and perhaps what it doesn’t look like. Abnormal behavior might be e.g. running in the opposite direction as almost everyone else. This kind of approach helps to track down the ones that just don’t fit the pattern. (Ackerman 2013.)

Using of photos taken by spectators to support criminal investigation raises several ethical questions. In Boston the photos and videos were asked to send to the FBI. Are there situations where authorities may try to require the photos for themselves? If an ordinary citizen who as a passer-by records on the event, which afterwards proved to be a crime scene, is he or she responsible of the integrity of photographic data he might provide for the authorities? In that situation would the intentional – yet harmless – manipulation of a photo be considered as falsification the evidence?

Let’s think about the Boston Marathon finish line right before and after the attack. Before the bombs went off someone had recorded with his camera phone a friend arriving at the finish line. These photos show the happy runner standing in the foreground, while nearby are two grim-looking guys walking rapidly carrying backpacks. Luckily the camera phone features a function which enables the photographer to erase particular people or just the moving objects from the scene. For the runner this ”cleaning of the scene” might be ok – he is after all the main subject and maybe it is just nice that some disturbing details are erased. Then, when bombs explode, the same photographer begins to record the scene again. If used as an evidence how to treat his photos, when some of the photos include all the data recorded while the shutter was released and the other photos don’t?

Photos reveal the ”minus-time”

Crimes and accidents have their zero-hour, the moment when it took place. I will call the time before zero-hour, the minus-time while the time after the event is plus-time. For revealing the identity of the Boston bombers it was crucial to scrutinize massive amount of data that was recorded at the minus-time.

It is an age-old tradition that police asks the audience for eyewitness stories considering crimes. Would the police at some time open a line for photographs that depicts the minus-time of old, unsolved crimes? Big data -analyze might in some cases bring new tips for solving crimes. Police might ask photos taken all over the town and taken, lets say five hours or days, prior to the zero-hour of a crime. In this sense old photos and negatives might yield a surprising amount of evidences. It ought to be remembered that e.g. facial recognition soft wares can be used for analyzing images scanned from photos originally taken on film. One interesting – yet fictive – example of reconstructing an event with the help of decade old photos is given by Stig Larsson in his book Men who hate women.

When president Kennedy was assassinated in Dallas 50 years ago, there were hundreds of spectators, but only very few cameras around the crime scene. It is obvious that all the images and every bits of data has been explored by the best investigators during many decades. Still, many think there might be some mysteries around the assassination. A layman may not know whether or not the big data –method might reveal some new aspects of that crime. Or if it has all ready been used? Maybe it is not unthinkable that at some point people are asked to send all the photos taken everywhere in Dallas during the minus-time – week or month – prior to the assassination and those photos would then be scanned and put to the big data –analyze. By the way, do your relatives happen to have photos taken in Dallas in November 1963?

Sources:
Ackerman, Spencer (2013), ”How this photo of Boston marathon gives the FBI a bounty of data”, wired.com 18.4.2013.
Brook, Pete (2012), ” Photographs Are No Longer Things, They’re Experiences”, 15.11.2012, http://www.wired.com/rawfile/2012/11/stephen-mayes-vii-photography/
Henry, Jim (2013),”M-B tech led cops to Bostons suspets’ car”, www.autonews.com 29.4.2013.
Konkel, Frank (2013), ”Boston probe’s big data use hints at the future”, fcw.com, 26.4.2013.
Serrano, Richard A. and Bennett, Brian (2013), ”Possible pressure cooker bomb believed hidden in black backpack”, 16.4.2013, http://articles.latimes.com/2013/apr/16/nation/la-na-nn-boston-bombings-pressure-cooker-0130416
Vänskä, Olli (2013) ”Paljastaako Bostonin pommittajan näin yksinkertainen tekijä?”, 18.4.2013, talouselama.fi.
Zhang, Michael (2013a), ”Boston Marathon Bombing Investigators Using Crowdsourced Photographs”, 17.4.2013, http://petapixel.com/2013/04/17/boston-marathon-bombing-investigators-using-crowdsourced-photographs/#QByMit5jvwfRiQ8w.99
Zhang, Michael (2013b), ”Reddit and 4chan Working to ID Boston Bomber Using Available Photos”, 18.4.2013, http://petapixel.com/2013/04/18/reddit-and-4chan-working-to-id-boston-bomber-using-available-photos/#68vmUQyrtrc6CoO6.99
Zhang, Michael (2013c), ”NY Post Uses Photo of Innocent Teen as Boston Bombing Cover Photo”, 19.4.2013, http://petapixel.com/2013/04/19/ny-post-uses-photo-of-innocent-teen-as-boston-bombing-cover-photo/#Kh8fXcKT2lpvkvzM.99