click here to skip the menu and go to the page content

rebecca's pocket

about / archive / syndicate

.: April 2006 --> More on Flickr-powered collaborative photojournalism

More on Flickr-powered collaborative photojournalism

» Emily Turrettini notices the difference between the Flickr slideshow of the CPE protests I linked here last week, and other CPE photo compilations on the Web, which depict a much more peaceful event. She wonders whether the slideshow photos depicting vandalism were taken during the French riots in November, and then deliberately mis-tagged "CPE", but I don't think that's the case. The slideshow I linked was organized by "interestingness", which is likely to skew to the sensational. The "most recent" slideshow presents a much less dramatic series of images. Note that neither is a measure of "importance" or "fairness", values that will likely always require human editorial judgement.

Of course it would be easy to deliberately mis-tag photos as they were uploaded. I posted here about the inherent limitations of tagging back in January 2005. At the time, others joined me in commenting on the potential to game the system. But Flickr is still a relatively unknown phenomenon, and I would be surprised if, at this point, anyone is trying to game it for political gain. That will likely change once Flickr becomes more widely known. Perhaps the automatic inclusion of GPS and time/date information when photos are uploaded would provide enough information to allow viewers to make more accurate assessments.

Emily is correct when she says "Caution and good judgment must prevail, not only toward the traditional media, but with regard to collaborative citizen journalism as well" — but that applies equally to the other sites she links. Organizations that participate in an event will document just one version of the event: their own. This version might be carefully constructed to present a particular narrative of the event and of the organization's role (think of the narratives routinely presented by political parties). At its least contrived, organizational records will consist of of "our favorite moments" — the parts participants themselves most want to look back on. This is right and natural and how we all organize our personal memory-markers. It is one reason I've argued that narratives must be written by a third party in order to be classified as "journalism" instead of memoir.

People photograph that which they think is "interesting". Photographers then apply another filter of "worthness" before they upload their photos to the server — or show them to friends. Every photograph has a point of view. Every series of photos creates a narrative. It will never be complete, or unbiased. Even so, barring a large-scale misinformation campaign, Flickr photos of any event should, in aggregate, represent a relatively impartial account of what could be be captured on film.

That lack of editorial control provides some protection against any person or organization seeking to control the narrative of any given event. Everyone's photos are published, regardless of their political standing or intent. Given enough participants, Flickr's inclusive nature will work against anyone deliberately skewing coverage of an event.

We need to be on guard against fraud. We need to create technological systems that will support transparency and reduce distortion. But in the end, one of our best weapons against deliberate manipulation and misinformation may be the simple, non-technical principle of inclusion. In fact, the framers of our Constitution were onto this 200 years ago. It's a little thing they liked to call "a Free Press".

 [ 04.17.06 ]


I beg to differ with your interpreation of my comments. The CPE slideshow on Flickr depicts (with few exceptions) images of the violent riots in the suburbs held in November, and not the student demonstration in Feb/March. I did not suggest there were two versions of the same event (CPE - one peaceful, one violent). These are two very separate and very different events that have shaken up France. They have been assimilated in this slideshow under CPE - which is misleading.

Emily, if the photographs that depict rioting were publicly taggable - as photos sometimes are on Flickr - it would be easy for a vandal to go through and mis-identify them as photos of the CPE protests. But as far as I can see, none of the photos in question is publicly taggable.

Going through the CPE photoset arranged by "interestingness" as a page of thumbnail photos, it's easy to see who uploaded each of them. In the first 10 pages of results, I identified 7 photographers whose photographs depicted vandalism. Each of those photos resides in a photoset, dated, and labelled or annotated by the photographer as depictions of the recent CPE protests. In several cases, the photographer was careful to note that the vandalism of the day was performed by outside provocateurs, not by the main group of protesters.

- Hugo*'s photoset: Demonstrations against the CPE.

- Gonzale's photoset "Paris in riots".

- Stuartisett's photoset: CPE Protests: France.

- Philippe Leroyer's photoset: Demonstration & Riots

- Bachellier's photoset: Manifestation anti CPE 18 mars in Paris.

- Spencer Olinek's photoset:28/3 CPE Protests.

- PE Weck's photoset: CPE.

If you don't accept my explanation that the more sensational photos have percolated to the top of the "interestingness" pile, then you are accusing Hughes Leglise-Bataille, Charlotte Gonzalez, Stuart Isett, Philippe Leroyer, Alain Bachellier, Spencer Olinek, and Pierre-Emmanuel Weck of deliberately falsifying information about their photographs when publishing them on the Web.

That's a very serious charge. If you have proof if that, I think it's very important that you produce it.

I'm not implying anyone has falsified any information. I'm simply pointing out that in the flickr slide show, under the tag "CPE" are included photos of TWO very different events - one NOT related to the CPE (but related to the november riots) and one related to the CPE (student demonstrations).

In my title "More on false tagging in Flickr, "false" is meant as "arising from error; "a false assumption"; "a mistaken view of the situation" - not "false" as "deliberate deception".

What is your evidence that the aggregated CPE slideshow contains photos from two different events? Can you point me to specific pictures that can be identified as having come from the November riots?

I'm not sure what you are suggesting when you refer to tagging that "arises from error" or "a false assumption". When they assembled their photosets, the photographers identified when and where the photos were taken. Either they are telling the truth, or they are not telling the truth.

Clicking on each photo brings up a description by the photographer of where events occured, and indeed, the violent ones that I checked are described as violent outbursts on behalf of the CPE.

The car burning photos, so emblematic of the November riots, is one instance is described as "Place de la Nation: due to outside provocators, the anti CPE demonstrators turned violent and a car has just been set on fire."

So I have obviously been wrong in saying the CPE photos were falsely tagged and that it includes pictures from the November riots.

The overall impression of this slideshow though is one of extreme violence over the CPE, when, over a two month period, the students where generally demonstrating peacefully.

Our discussion has prompted me to wonder whether, sometime in the future, Flickr will want to institute a "Trust" system of some sort. I think it would be cumbersome and probably unnecessary right now, but I can foresee a time when it would become an important mechanism for viewers to gauge the credibility of photographers/taggers and for photographers to leverage their reputations as trusted sources.

I'm also wondering whether a set of "Best Practices" for photo-reporters will emerge (if one hasn't already). For example, allowing anyone to tag your party photos is probably an efficient way to add metadata to your photoset. But in a photoset of a news event, photographer-tagged photos provide for more accountability.


I'm one of the Flickr photographer you mentioned, covering the CPE demonstrations. I think some comments are both right and wrong. First, the issue of the "fake" tags is clearly wrong: people use tags to track their own pictures, and I can't see why they would tag "CPE" an image of a burning car taken months ago. But nothing ressembles more to a burning car than another burning car, so the shot in itself, isolated from the rest of the report, isn't very interesting. Note that the automatic inclusion of date/time already exist on Flickr (as part of the EXIF data), although you can easily hide or alter it (and that will always be the case I'm afraid).

On the predominance of violent pictures, I think the way you searched the pictures (by the CPE tag and then by "interestingness", which is a mysterious Flickr algorithm) biased the results. Many people covering the event did take many peaceful photos. However (and it's a big however), the violent photos attracted more attention from the viewers and therefore, became more "interesting" according to Flickr. So it's not necessarily the photographers who are showing a taste for violence, but the public ! Not much of a surprise unfortunately... On the other hand, demonstrations take place almost every week in Paris and are usually very peaceful, so no wonder than when it gets violent, and therefore "extra-ordinary", people find it more "interesting" than the "ordinary" bunch of people marching with banners !

To further refine this, I should also say that the repetitive violence of the CPE demonstrations was something that characterized them: if you showed the events without the violence, you would not only hide the truth, but also ignore what was a specific feature of these demonstrations. The CPE demonstrations WERE violent (and having been there, I can tell you that the pictures don't even reflect the scale of the violence, like in the Invalides). They were also peaceful, but not only, it was just one side of it, whether we like it or not.

I think it's also wrong to say that it can bias the perception that people might have of the events: most of the photographers on Flickr have repeatedly insisted, in their description and comments, that the "casseurs" (trouble makers) were NOT the protesters, and some went to great length trying to explain what was really happening.

On the other hand, I (and others) have reported on events that were either not or wrongly reported by the main media. For instance, a "protester" blocking the Périphérique on April 6 had his leg crushed by a car in front of the police and was hospitalized. This was not reported in the news, and accordingly to a cameraman I met at the scene who later tried to offer footage to TV channels, the material was rejected because it was deemed "unfit". The main media have their own "filters" of the information, so the bias in treating the news also exist on their side... Another example was the repeated mistake, in many media (TV and print), of presenting a man beaten up by 2 "casseurs" on April 4 as a photographer. This was wrong, the guy was drunk and provoking everybody (police and protesters alike), but nobody corrected it...

Personnally, this was my very first experience at photojournalism, and I was surprised by the massive presence of photographers on the scene. My guess is that more than half of them were amateurs. Why not showing their photos to the public ? Some are as good if not better than the ones shot by the pros. Another key element is that the news, because of space limitations, only show a very small part of the events (for a daily, you would get 2 or 3 pictures at most). It's impossible to accurately illustrate such an event with so few pictures, and the choice of these (the editing) represents therefore a huge "bias". In contrast, the dozens if not hundreds of shots posted on the web gives a much better and broader coverage. This is a great advantage. I even know some professional photojournalists who covered the demonstrations for foreign newspapers but still posted their photos on Flickr (one told me it was somehow frustrating to report a full event only to have a couple of photos shown by the paper ; he felt that at least, with Flickr, he had an opportunity to show more -and he also enjoyed the feedback very much :-)

Hope this help and congrats on the articles !

Hughes Leglise-Bataille



» primary link / supplemental information / internal link

my book

» the weblog handbook
amazon editors' best of 2002, digital culture

recent posts