For me, the value of the 4636 system was never been just about the actionability of the data — I saw the reports were profoundly flawed as soon as they began to arrive. There were amazing sparks of actionability — glimpses of something we can all fumble toward somehow — but it was always clear that these were no replacement for well-tested logistics plan.
I know that you and others working in the crowdsourcing ecosystem in Haiti must have recognised this. Yet it didn’t seem like the public statements at the time reflected this; it doesn’t seem as if any of the interviews or conference presentations or media articles reflected this; and most importantly of all, it didn’t seem as if your message to the people of Haiti reflected this. It’s interesting that the rhetoric now – after the funding is in place, after the conference invitations are in, after the media attention has been secured – is much more circumspect than it was previously.
And what exactly was the value of the 4636 system from your perspective? This is what I’ve been really interested to hear.
But, while I was quite aware of how utterly limited the data were, this never made me question the basic value of an open communications system during this crisis.
I guess this is where my old-school sensibilities are a handicap, because my view is that if you’ve set up an information system and the data it’s producing is not that useful, there’s a problem with that system. If we agree for a moment that there’s nothing wrong with your technology, that means there’s a problem with the inputs – because as we all know, the basic rule is garbage in, garbage out – yet the inputs are exactly what you’re claiming are special about your system.
When I am in a crisis, I want strong systems in place that let me text anything I want.
The problem here is that your needs are not necessarily a useful guide to the needs of other people. I’ll say that again, because it’s a fallacy that affects nearly everybody starting out in relief or development: your needs are not necessarily a useful guide to the needs of other people. It’s nothing personal: my needs aren’t a useful guide to the needs of other people either. Hell, my needs probably aren’t a useful guide to your needs, or vice versa.
However I agree that open communications (along several tracks) are valuable, and that accessible communications systems can be as important as accessible food distribution networks. Different criteria apply in a disaster, though; the food distribution networks run by Wal-Mart aren’t as effective in South Sudan. This doesn’t mean that we can’t learn something from Wal-Mart – just that we need to be careful about what we learn.
The same holds for technology; context is vital and we can’t just transplant things that work really well in one place to another. And besides, there’s already a strong system in place that lets you text anything you want – it’s called the mobile phone network.
There may be a limited amount of resources on a particular day in a particular crisis — but in the long term I believe that there are millions of people who could be engaged in productive and powerful work through diverse and open platforms.
I also believe that there are millions of people who could be engaged, it’s just that I don’t believe that millions will be engaged. There’s already ways in which those people could be involved if they wanted to, and the vast majority of them simply don’t want to. I long ago ceased to be naïve enough to believe that everybody shared my value set, so I’m not condemning anybody here; I just fail to see why anybody thinks that technology will magically make people more altruistic.1
The only vaguely convincing argument I’ve seen says that the internet – and in this case crowdsourcing – has lowered the barrier to entry to get involved in humanitarian activities. Unfortunately that doesn’t mean that more people will get involved, because the barrier to entry was also lowered on a huge range of other, more interesting activities that people will choose instead of the difficult and depressing work of humanitarian activities. A lower entry barrier is also a lower exit barrier, as Ushahidi found out somewhere between Haiti and Pakistan.
The bitter truth is that you may have to face the possibility that the crowdsourced response to Haiti was possible mainly because Haiti is within flying distance of the US, has a long history with the US (including a large diaspora community), and was a particularly media-friendly disaster. I’m not saying that this was definitely the reason why that particular crowdsourced response ever got off the ground, but I’m interested to hear what this discussion looks like from the other side, rather than just get fed another inane marketing video.
Particularly I see a great promise in networks which help people in crisis see their own neighborhood, to help each other, rather than always relying on the aid worker. So, information which seems predictable and “not novel enough” to an aid worker might be quite powerful to someone who lives down the street.
I agree completely. We are already seeing this (something I’m going to talk about soon in a follow-up article) but this isn’t the model of crowdsourcing that was developed in Haiti. I remain unconvinced that the Haiti model is a useful model for anything else – in fact I remain unconvinced that the Haiti model was a particularly useful model in Haiti. All I’ve been asking for is some evidence that it warrants the massive amount of hyperbole that surrounded it – and the opportunity costs of implementing it.
The other point to note here is that the fact that different types of information have different value to different people is trivially true. That being the case, then who gets the value in the type of information that Ushahidi et al gathered in Haiti – the aid worker or the person who lives down the street? Because it seems to me that the information gathered in Haiti might well be useful to somebody who lives down the street, but that’s not where the project was driving (because it was subject to exactly the same constraints as everybody else in reaching the street).
In this sense, I have a hard time understanding how you could question whether it has “any substantive value.” To me the value is self-evident, if complex, and problematic for traditional patterns of response.
The value may be self-evident to you, and others working on crowdsourcing projects like 4636 or Ushahidi, but they’re not self-evident to others. You can’t accept people to accept your claims just because you really, really believe; you need to present a convincing narrative. The narrative around Ushahidi is really convincing to technologists because it tells them that technologists can really really make a difference.
That narrative is a lot less convincing to people who actually do the work. All I wanted was for proponents of crowdsourcing to step up to the plate with something more than the usual vague gushing about the revolutionary power of the crowd, with some more awareness of the embedded technocratic ideology that comes with the package, and to be backed up by slightly more evidence than second-hand anecdotes from a US marine. I freely admit that I could be wrong – that was why I wrote the article – but I need to be shown to be wrong.
All tools are suitable for some tasks, and not for others. It strikes me as strange that Ushahidi is proposed as a mapping solution for citizen journalism; for election monitoring; for human rights monitoring; for emergency response; for clearing snow from American cities; and for pretty much any activity you care to mention. It strikes me as singularly unlikely that it’s a solution for all of them.
Generally speaking, the rule of tools is this: the wider the range of tasks that a tool can perform, the less effective it is for any given task within that range. There is no shame in anybody anywhere standing up and saying, “it turns out that our tool isn’t suitable for this particular taskâ€
- I have to refer back to Shirky again. The opening chapter of Cognitive Surplus contrasts Ushahidi and Lolcats as two ways in which that surplus can be engaged; he judges one to be more constructive and one less constructive, but both are deemed more constructive than watching the television. This is bullshit. You learn more from watching Unreported World than you do from watching X Factor, but a lot more people watch the latter. Ushahidi maps to Uncovered Stories, Lolcats to X Factor, and you can do the math about how many people are likely to contribute to any given Ushahidi instance. [↩]
Related posts:
As usual, excellent points and again, it’s a shame that Ushahidi is not taking the time nor bother to answer you, which would seem to point to the fact that they have no answers. “The narrative around Ushahidi is really convincing to technologists because it tells them that technologists can really really make a difference.” is especially poignant.
Pingback: Disaster Response 2.0 report is out « From the Edges In
>It strikes me as strange that Ushahidi is proposed as a mapping solution for citizen journalism; for election monitoring… and for pretty much any activity you care to mention.
This attitude is what led me to try using Ushahidi for an environmental research map. The civil engineering department at my university collected data from state and federal isources by traditional means, and tried to combine it on the map with public feedback. In the end, the best information was from our official sources and realtime data, which were easier to present in a Google Map.
I work on plugins that close the gap between crisis maps and scientific GIS, similar to my link that appeared on the Ushahidi blog: http://mapmeld.com/bookmarklet-home.html Ultimately the concept of a crowdsourced map could help existing activists, such as those in Kettleman City, CA, to explain what they’re observing and to put their message on equal footing with EPA maps. I suspect the future of crisis maps is closer to this “give activists tools for data and publicity” rather than “enlist people to use apps”.
Nick – thanks for sharing your experience. It’s a good point that these types of maps might actually be more useful as a counterweight to “official” sources of information that activists can use in campaigning. Such an approach doesn’t require that the data has any operational use, or even that it’s completely accurate or comprehensive.
It’s also worth noting that a lot of data on a lot of Ushahidi maps are from official sources anyway, manually entered by volunteers. This kind of situational awareness is something that I think is a valuable contribution (although obviously it doesn’t have to be on an Ushahidi installation).