But how do we know? — “CSI” and beliefs about visual evidence. • A different CSI Effect — what are the modes of thought used in relation to visual evidence, and how might they interact with attributes of the CSI program? • A theoretical musing, totally free of empirical evidence, and filled with many and diverse speculations. The photographs, and the question • (Here begins what may appear to be a diversion away from the topic, but stay with me: this is about mind-sets in understanding evidence.) • Roger Fenton was an early (mid-19thC) British photographer, known best for his travel pictures from exotic locations (e.g. China) and his war photographs from the Crimea. • Attached (embedded) to British Army in 1855 during Siege of Sevastopol. Main task to take photographs that showed war in a good light. • On April 23, 1855 he ventured up into the “Valley of the Shadow of Death” (not the location of the Charge of the Light Brigade — named for the frequency of Russian cannonballs ending up in it). Location for the photographs • But how do we know? — “CSI” and beliefs about visual evidence. • Fenton and his assistant , Marcus Sparling arrived at the Valley at around 3pm, and made two exposures from the same tripod position: The Photographs: “on”… • …and “off.” • But how do we know? — “CSI” and beliefs about visual evidence. • I’ll refer to them as ON and OFF (cannonballs ON road, cannonballs OFF road) as so as not to bias the question about them that is the subject of the investigation described here. • Both have been published, but “ON” more frequently than “OFF;” thereby hangs this tale. • Errol Morris, eminent documentary film maker, became interested in the question of why/how the two versions existed, and in the confidence expressed by some writers on history and photography, about the two photographs. • (Full story is told in Morris’s three-part series in the NYT, starting at: http://opinionator.blogs.nytimes.com/2007/09/25/which-camefirst-the-chicken-or-the-egg-part-one/ More about the photographs • In Susan Sontag’s last book, Regarding the Pain of Others,: “Not surprisingly many of the canonical images of early war photography turn out to … have had their subjects tampered with … in the first version of the celebrated photo he was to call ‘The Valley of the Shadow of Death’ … the cannonballs are thick on the ground to the left of the road, but before taking the second picture – the one that is always reproduced – he oversaw the scattering of the cannonballs on the road itself.” [italics added] • Morris’s reaction to this: How did Sontag claim to know which photo was made first, and that he “oversaw” placing the balls on the road? • Set out to answer the questions: which photograph was made first — and how do we know? (A self-contained microcosm) • Between them, these two photographs are: • The crime (or the misdemeanour); they cannot both be ‘found,’ unmanipulated photographs; • the scene of the crime (or a representation of it), and • the evidence (almost all of it). Who’s on first? • Morris questioned Sontag’s source, a book about the visual history of the Crimean War by Ulrich Keller, who had written: • “The first variant obviously represents the road to the trenches in the state in which the photographer found it … In a second version … Some round-shot is now … distributed all over the road surface – as if the balls had just been hurled there, exposing the photographer to a hail of fire. Not content with the peaceful state of things recorded in the first picture, Fenton obviously rearranged the evidence … to create a sense of drama and danger that had originally been absent...” [emphasis added] • …and who says, in conversation with Morris: • “Well, I can see a motivation for him to take the balls out of the ditch and put them in the middle of the road. That makes sense to me. It’s something that I think is plausible for someone to do. The other way around, I don’t know why anyone would do that. I don’t think it’s likely. “ • MORRIS:” Is it the absence of a psychological explanation that makes “the other way around” unlikely or implausible?” • KELLER: “Yes.” Top-down and bottom-up • In Keller, we see (?) one mode of reasoning that is also used in relation to crime: construction of a psychological theory of motivation, which is ‘backfilled’ with suitable material evidence, consistent with the theory. This is the “Top-down” approach. • A model frequently used in crime TV, where we see character motivation before we see a crime. • Keller/Sontag model of Fenton, as someone who exaggerated risks, may be unfair. Cannonballs were fired into the valley while they were there (Fenton’s letters), and his assistant Sparling feared for his life. They had to move the camera back from an initial position because of the incoming cannonballs. Morris’s search continued • (A long story, and you should read the whole, but): •Morris consults other experts, with variable results: • Gordon Baldwin (Getty): the balls were harvested by soldiers, to fire back; • Malcolm Daniel from MOMA — agrees with this. • An earlier letter from Fenton (April 4), following a reconnaissance of the same area, indicates that the road had been ‘covered’ with balls at that earlier time — and thus that the “on” photograph was an attempt to represent that earlier condition. • Note that these are largely character defences of Fenton, based on presumed intentions of the photographs. •The quest for an authoritative witness is futile — Morris looks for answer within the photograph. The search within • Morris attempts to determine the direction in which Fenton’s camera faced, so as to read any changes in shadows/sun direction. Inconclusive — some experts are certain camera faced north, some south. • Consults experts who superimpose one photo over the other to see where the balls on the road have been taken to, or where they have been taken from. Position doesn’t indicate timing. • The sky in both photographs is washed out because of the emulsion’s over-sensitivity to blue; though one photograph is mostly clear sky, the other cloudy-bright, they look the same. • Shadows, esp. long ones, are almost entirely absent. Shadows of the balls are inconclusive; because one photo is cloudy-bright sky, it shows diffuse highlights on the balls where the other has specular highlights, foiling a direct comparison. The search without • Morris does not give up easily… • He goes to the Crimea, and finds the Valley of The Shadow Of Death (which had not been done before): Search and Create • Morris borrows a cannonball from the Panorama Museum, and takes it to the site and photographs it, with camera pointing in the right direction, at the right time of day: Shadowland • …and calls in new experts to assist with reading shadows: • R. BOUWMEESTER & ASSOCIATES Sun & Shadow Position Specialists with Modeling Applications in Accident and Crime Scene Reconstruction, Urban Development, Site Planning and Building Design • And also: Fovea Pro – a forensic photography program for processing and measuring images. • Neither is helpful (contrast differences, uncertainty about whether laterafternoon light would make a hillside brighter or darker, etc etc) • (During this process there are two reversals of opinion by the experts about whether ON or OFF comes first.) • The issue is finally resolved by information within the photographs, but not related to any of these issues… Some rocks and a hard place • Detail from OFF: • Some rocks and a hard place • Detail from ON: • Some rocks and a hard place • Superimposition: • All of which means…? • Firstly, if a verdict (on or off) was the dependent variable here, then both the ‘over-arching character-motivation theory’ approach, and the bottom-up nitty-gritty evidence approach have led to the same conclusion. ‘Verdict’ not a helpful measure. • So, we need to be very interested in how and why decisions are arrived at, not just what decisions are arrived at. • But what else? Morris (et al’s) search seems to tease out some of the cognitive dispositions that might be present in a juror’s mind when considering the meaning of visual evidence. Maybe it’s possible to tease out more. • It seems potentially — heuristically — useful to compare these cognitive dispositions in Morris (et al’s) search with the attributes of CSI that distinguish it even within TV crime shows. CSI • (from: Panse, Silke: ‘The Bullets Confirm the Story Told by the Potato’ Materials without Motives in CSI: Crime Scene Investigation. In Michael Allen (Ed) Reading CSI: Crime TV Under the Microscope. London, I.B. Tauris & Co, 2007, 153-166.) • First crime series to reject a plot driven by psychological motivation, in favour of accumulation of empirical evidence (153) • Much of plot is driven by images rather than words (153) • Evidence is not transparent — must be revealed through technology (fingerprint machine, DNA, UV light) • Close focus on science of materials, rather than character construction; • Objects, not subjects, tell stories (Grissom, talking to person but referring to corpse: “you don’t have to talk to us. He’ll talk to us;” and “The bullets confirm the story told by the potato.” • Dealing with objects is portrayed as rational, talking to people as irrational. • Cases are more like jigsaw puzzles or riddles than motive-driven character narratives. Cognitive styles in the tale of two photos • We’ve seen that some of the experts in photography use inferential arguments about personality/character, rather than attributes of the photograph, as the default means of arguing for a certain meaning. In fact most people mentioned in the Fenton photograph story do this, at some stage. (CSI: never used). • We can see contrasting views of meaning: meaning ‘within’ the evidence; meaning ‘discursively constructed from’ the evidence; meaning conferred from outside [some of] the evidence. (CSI concentrates on the first of these). •The truth/meaning of visual evidence may indeed be located ‘within’ the image itself, but some contextualising is almost always essential. (CSI: a machine will extract/reveal meaning for us). • Grateful to Neal Feigenson and Christina Spiesel for pointing out how the notions of ‘top-downers’ and ‘bottom-uppers’ interact here. It may appear that CSI inclines viewers toward bottom-up approaches, — but I suspect CSI enthusiasts’ enjoyment of the show comes from the tension between their topdown tendencies and the show’s bottom-up revelations. Ideas for more research • Two other things play into this (and confirm need for further research): • Jane Goodman-Delahunty has shown that keen CSI viewers are simultaneously: • Less knowledgeable about forensic science techniques; • More confident about their knowledge, and • Less educable (about forensic science). • This appears to confirm the likelihood that rather than understanding in detail, they form an overall hypothesis (top-down), and wait to see it confirmed or denied by the revelations of CSI technology. • The notion of a generalized CSI effect is a misunderstanding of media effects generally, caused (alas) by media effects research. • Advertising (for example) works — but not because everybody is slightly influenced by it. Some people are influenced by it because their interests coincide with it in some fashion, but the eventual result is a complex— and unpredictable(?) — mix of attributes of the show and cognitive attributes of the viewer. • So it is, probably, with CSI.