Digging Up the Narrative – Forensic Practices between Objectivity and Interpretation
Fading Away by Henry Peach Robinson was shocking to its audience when first shown in 1858. ‘How,’ people asked, indignantly, ‘could the photographer dare to intrude upon the intimate moment of a family mourning its dying daughter?’ (cf. Newhall 2012, p. 74, 76) But what they were looking at wasn’t a lack of respect for the dead; it was a product of craftsmanship and a sense of artful composition. Robinson had merged five staged photographs together in order to create the effect of the perfect moment.
That reaction may seem antiquated or naïve, but it does have its equivalent in our current ‘digital society’: In early 2015, photographer Giovanni Troilo had to return his World Press Photo Award because he was found guilty of having provided his portraits of the industrial Belgian town Charleroi with captions that told a story deviating from actual events: but some of the series’ photographs were taken in other cities and depicted people whose claimed identities (like a guy in a car having random sex that turned out to be Troilo’s cousin) didn’t coincide with their ones in real life (cf. World Press Photo 2015b).
Similarly, there has been a recent media hype emerging around the ‘getting-real’ of teen Instagram-star Essena O’Neill, who publicly revealed that the authentic look of her posts were the result of a lot of effort and generous sponsorships of fashion labels (cf. O’Neill 2015). Over 150 years after Fading Away, we still believe in the unaltered, somehow natural picture.
These examples sum up photography’s complicated relationship with reality: it is traditionally seen as a medium which depicts ‘the world as it is’, while at the same time undermining that very promise by holding the potential of being altered, staged or to unfold diverse meanings in different context. This tension between revealing and concealing is a constitutive characteristic of the medium, and an important aspect of photography’s appeal.
And not only in regard to the digital, for the manipulated photograph belongs to a lengthy tradition: In the mid-and-late 19th century, retouching was an important argument for the defenders of the artfulness of the machine-made picture; In the Soviet Union, Trotsky, fallen out of grace, magically disappeared from the scene; Spiritual (ghost-)photography tried to capture what the human eye (and thus the camera) couldn’t see (because it was never there in the first place); Journalists chasing iconic pictures haven’t hesitated to re-enact dull looking events in a dramatic manner or give them an Instagram-y feel in post-production; As well, in the realm of jurisdiction, forgers are busy faking facts.
So, why do we still trust imagery which has been doubted, deconstructed and proven to be dependent on context, human gestures and interpretation, as well as the defaults of the camera? And why do we still feel insulted and betrayed when the fakery is revealed? The answer lies in photography’s specific truth claim which is a product of social discourses as well as the indexical quality of the camera“ (Gunning 2004, p. 48). The idea of photographic indexicality can be traced back to Henry Fox Talbot’s description of the then new technique as the “pencil of nature”, launching the idea of the medium’s mechanical imagining mode, which became closely tied to the notion of objectiveness and authenticity. This was a circumstance was further cemented by scientific and artistic practices, as well as the use of photographs as proof in the courtroom.
As Lorraine Daston and Peter Galison have shown (cf. Daston & Galison 2010), sciences as well as arts entrusted themselves to the photograph. The artists praised its detailed-ness and how it made visible what the human eye couldn’t perceive, such as the way a horse moved at full gallop or the flight path of a bullet. The medium’s ability to visualize even the fastest, smallest or slowest movements was appreciated by scientists as well, but even more importantly, photography for them became the ray of hope in terms of rendering their research-evidence trustworthy. The French medical researcher Alfred Donné was one of the first scientists to use photographs along with drawings of microscopic displays like blood, milk or sperm in a publication (cf. Daston & Galison 2010, p. 58). Using the photographic medium, he tried to upgrade the truth-promise of another mechanical imaging technique (the one of the microscope) which didn’t seem convincing enough in its own right.
This doubly verification-gesture speaks volumes about the problems of visual representations in the middle of the 19th century, and also in general, since the trust Donné tried to gain wasn’t the one that was undermined by willingly manipulated evidence, but by visual evidence (like drawings) that was prone to misinterpretation. Scientists who couldn’t rely on their handmade imaging techniques placed their bets on a medium, which came with the promise of ‘objectiveness’, hoping that photographs would clear up doubts of their colleagues as well as their own. The mere possibility to depict something without risking an interpretative contamination changed scientific practices of reasoning and, at the same time, strengthened photography’s truth-claim.
Simultaneous to the developments within sciences, lawyers discovered photography for their modes of creating evidence. Before the 1850s, the witness was basically the only evidence lawyers could produce, since forensics were still in their infancy and drawings and other visual imaging techniques were not considered trustworthy. Meanwhile, photography was being promoted from very early on as more convincing than even the most eloquent words of the lawyers. This was a trend which eventually lead to the establishment of a whole new type of visual proof, the “demonstrative evidence” (Mnookin 2013, p. 8), which changed the practices of the courtroom tremendously, as Jennifer Mnookin states: “Now, an attorney or witness could not only locate evidence, but could create it himself. […] These forms of visual evidence were especially persuasive because jurors and judges could see the evidence for themselves.” (Mnookin 2013, p. 5) But in the realm of jurisdiction, the photographic image was never considered more or less reliable than an eyewitness, since it always carried the risk of being manipulated.
And this is a risk that seems to have become even higher as so-called new media have rendered the photograph into an unstable and chronically unfinished object, easily alterable, since the idea of continuous light-inscription onto a material carrier has been replaced by the malleable, immaterial pixel. But, as there have been more and easier ways to ‘manipulate’ photographs, so has there been an increase in the ways to detect them, which today puts digital forensics in the position of re-establishing ‘reality’ as a referential point by tracing every step of the process of alteration or by revealing buggy defaults, turning the dubitative image into one that is doubt-free once its metadata has been analyzed. Of course, this emphasis on data as the new savior ties in with recent acclamations of a data-driven society, which have their roots in the old belief of data being unbiased, as well as in more recent developments which started in the 1940’s in cybernetic circles.
As Orit Halpern argues in Beautiful Data (2014), our data-serfdom stems from cybernetic theorists such as Norbert Wieners, who “dreamed of a future where there is no ‘unknown’ left to discover, only an accumulation of records that must be recombined, analyzed and processed.” (Halpern 2014, p. 12), implying that this dream could be achieved if machines did their job right, dropping responsibility as well as interpretational sovereignty on computers and statistical approaches.
While Wieners ultimate goal, to compare and eventually equate computational processes and the human brain, was unsuccessful, his and other cybernetician’s theories where involved in triggering a “shift in scientific practice and standards of truth from a focus on discovery, archiving, and documentation of new facts about nature to an emphasis on pattern-seeking, analysis, and recombining data.” (Halpern 2014, p. 15) This alleged power of data in terms of knowledge production is reflected not only in the proclamation of the ‘data-driven’ society and the creation of terminology like big data, but also in the establishment of scientific fields such as the digital humanities, informationvisualization, and digital forensics.
Digital Forensics combine the objectivity-aspirations defined in the scientific realm in the 19th century with the one brought about by techniques of data-analysis, which tend to emphasize the impartiality of the computational and quantitative tools the experts are using. This is especially interesting when it comes to photographic pictures which -- as I have shown -- come with their own truth-claims, and whose (mis-)interpretation can have serious consequences such as when they are used in courtrooms. So in order to reveal or dismiss a possible fakery, a picture, like the one that shows Lee Harvey Oswald with the rifle he is said to have killed President Kennedy with, can be examined by using quantitative methods which, in this case, showed that the shadow beneath Oswald’s nose, is not defying physical law, as had been implied.1 A reassuringly clean result, which Hany Farid, accredits to mathematics, because it liberates us from relying on our flawed human perception: “We don’t have to just guess.” (Farid 2012)
But is this the whole story? Isn’t every data-analysis also a mode of abstraction which relies on isolation, re-organization and re-contextualization of data as well as on the subjective decisions of the expert? In the following I will pursue these question by taking a closer look at the investigation of the Abu Ghraib torture scandal. Brent Pack, Army Special Agent with the military investigation unit and specialist for digital forensics, was in charge of the analysis of photographs taken in the prison throughout 2003 by members of the U.S. army, which eventually led to the conviction of those involved.
Based on the metadata of the photographs, Pack created a timeline, a data-visualisation which was used as demonstrative evidence in the actual court hearings. Forensic experts entrusted with tasks like Pack’s have to follow a strict protocol and record all their steps in order to render their work as transparent as possible. Pack was also diligently interviewed by Errol Morris for his documentary Standard Operating Procedure (2008) in which he’s painting a picture of the occurred events at Abu Ghraib, the motifs of the involved soldiers, and the role of the pictures they had taken for the investigation as well as the changing of the public view and the course of politics regarding the war in Iraq. The story behind this crucial line of argument is therefore not only well-known, but also very well-documented.
When the affair entered the public conscience in 2004, military investigators tried to make sense out of the photographic material they had secured. But even though the “pictures spoke a thousand words”, they, according to Pack, “didn’t know what the story was.” (2008) In order to ‘listen’ to them more carefully, Pack was given ten CDs with incriminating photographs. His task was to go through them and reorganize them according to the events that had occurred. He soon became aware that the problems he was facing were due to flawed defaults, like “file time stamps, [that] were all off: anywhere from a year to plus a couple of hours. And every time they got copied to a CD, or from one computer to another, the times would change based on that computer’s time setting.” (Pack 2008) In order to find out who was involved in what incident, and what happened when and lasted for how long, he turned to “the one time setting that would stay constant”: the metadata.
According to Pack, “Metadata is a big two dollar-word for information about information”, which is embedded “inside the file” (Ibid.) such as the ones that “tells you about when that file was created, what software created it, the exposure settings, and the date and time the camera thought it was when it took the picture.” (Ibid.) With the help of this information, Pack was able to see that three cameras had been involved -- belonging to specialists Charles A. Graner, Ivan Frederick and Sabrina Harman -- which had taken pictures of the same incidents but from different angles. However, his unbiased and constant metadata wasn’t enough for the purpose of putting them into the right order. It needed to be compared with the actual content of the pictures: “I found a total of eight separate time-synch incidences, and I was able to say, this camera thought it was this time, this camera thought it was that time.”(Ibid.) And when he started to organize the pictures according to the scenes they depicted, “all the pictures just seemed to line up.” (Ibid.) He confirmed his findings by using one of the oldest monitoring tools there is, the guard log, “that recorded incidences that occurred in the jail. It actually confirmed that the timeline was accurate.”(Ibid.)
After completing these first steps which not only revealed a timeline, but also “how much actual efforts did these people put in what they were doing to the prisoners”, and who “else was there at the time it occurred”(Ibid.). Pack started to go through the pictures again, trying to decide which ones showed so-called Standard Operation Procedures and which ones committed crimes: “You had to look at exactly what the pictures depict. It was important to separate those that were criminal acts and those things that were not criminal acts.”(Ibid.)
Pack, relying only on what his material ‘depicts’, based his analysis on what he had learned during his time as a member of the U.S. Army, emphasizing his expertise by stating “I’ve been in the army for twenty years. I’ve been to Desert Storm One. I’ve spent four month on Guantanamo Bay.”(Ibid.) While at the same time, denying others the right to doubt his abilities: “people who haven’t been where I have been, I can’t expect them to see the pictures in the same way.”(Ibid.)
Pack is also well aware that in his position he has to put on an indifferent face, upholding his ethics as an unbiased expert: “All you can do is present what you know to be factual, you can’t bring emotions or politics into the court.”(Ibid.) But like the metadata, that doesn’t work without the interpretation of the pictures’ visible content. Pack mixes practices he claims to be objective and at the same time based on personal experiences, paradoxically using this conflation to verify his findings. His final statement in Errol Morris’ documentary involuntarily reveals the conflict: “A photograph depicts what it is. You can put any kind of meaning to it, but you’re seeing what happened at that snapshot in time. You can read emotion on their face and things in their eyes, but it’s nothing that can be added in the fact. All you can do is report what’s in the picture.”(Ibid.) The reason why his line of argumentation still persuades doesn’t stem so much from the notion of objective metadata, but from making his working process transparent and comprehensible, using yet another gesture of authentification: data visualisation as part of the narrative of a documentary directed by Errol Morris. What is being revealed here is not the only possible true story of actual events, but the practices of rendering evidence trustworthy, by fueling old claims with new techniques.
References:Daston, LJ. & Galison, P 2010 Objectivity. Zone Books, New York.
Farid, H., 2011, What's a Picture Worth? Available from: https://www.youtube.com/watch?v=zVOYInKgy8k. [11 November 2015].
Gunning, T. 2004, 'What’s the Point of an Index? or, Faking Photographs' in Digital Still/Moving: between Cinema and Photography eds. K. Beckman & J. Ma, Duke University Press, Durham, pp. 23- 40.
Halpern, O. 2014, Beautiful Data: A History of Reason since 1945. Duke University Press, Durham.
Mnookin, J.L. 2013, 'The Image of Truth: Photographic Evidence and the Power of Analogy' in Yale Journal of Law & the Humanities, vol. 10, no. 1, pp. 1-75. Available from: Digital Commons. [10 November 2015]
Morris, E. 2008, Standard Operating Procedure (video file), Available from: <http://putlocker.is/watch-standard-operating-procedure-online-free-putlocker.html. [11 November 2015]. Newhall, B 2012, The History of Photography. Museum of Modern Art, New York.
O’Neill, E. 2015, Why I really am quitting social media (video file), Available from: <https://www.youtube.com/watch?v=EE5K7gx2VZw ""- Original Video.> [11 November 2015]. World Press Photo 2015a, Entry Rules, Available from : <http://www.worldpressphoto.org/contests/entry-rules. [11 November 2015].
World Press Photo 2015b, World Press Photo Withdraws Award For Giovanni Troilo’s Charleroi Story, Available from: <http://www.worldpressphoto.org/news/2015-03-04/world-pressphoto- withdraws-award-giovanni-troilo’s-charleroi-story>. [11 November 2015].