Home | Hegemony | Archives | Blogroll | Resume | Links | RSS Feed | subscribe by email    


to Reason


blog roll

    Washington Consensus: Crumblin..., 2002-11-20 02:00:00 | Main | Prospects for Liberty: No War ..., 2002-11-20 02:00:00

    Some notes on Roger Bournival's, aka the Angry Cyclist, criticism of Herold's tallying of civillian Afghanistan casualties (why I'm bothering to note anything is a good question, given that this has apparently already been beaten to death, but I still haven't seen the corpse of the bogus horse): Herold's decision "to do the study because I suspected that the modern weaponry was not what it was advertised to be" does not mean that Herold "starts his research with an apparently predetermined conclusion" as B argues, it means he started it with a predetermined suspicion - hardly a specious cause for study. At the same time I wouldn't call collecting press clippings "research", but it's something like it - in Herold's case he can hardly be accused of cargo cult science as there's no gravy train to hop onto for small liberal arts college professors determined to make the US government look bad. B is mostly correct to my knowledge that "there is no effort on anyone's part to determine the actual casualties". At least in this he and Herold are in agreement, "given the lack of official interest".

    It's quite possible that Herold's biases are affecting his judgement, that's a fair question, but it's equally possible that B is guilty of letting his biases affect judgement: "17.5 percent of his data points contained multiples of ten, extremely strong evidence of rounding, and, in my estimation, rounding up". What's the basis for that estimation? It also might be helpful to include the number of casualties represented by that 17.5 precent to given some indication of it's impact on the end tally. Nevermind that Herold's tally is based entirely on press reports - given the lack of media access to the war zone and military intelligence it hardly seems outrageous to suggest that such a tally, assuming it were credible (an assumption that remains dubious, I concur), would be something of an undercount.

    Furthermore in acknowledging that there are "reports from potentially biased sources such as the Afghan Islamic Press, Pakistan Observer and the Guardian of London" one has to level the same criticism at US sources, given in particular the substance of some internal memos commanding the US press corps to downplay civillian suffering and the otherwise radical inflation of the value of patriotism and jingoistic pride that has been occasionally acknowledged and was a) understandable and b) obvious. There's propaganda, and then there's facts. The "constant revision of [the] database" presumably speaks of some degree of scrutiny of orginal reports' accuracy, quite likely not enough, but there's no clear basis on which to judge how such scrutiny would affect the end tally, given the lack of it. I'm unaware if Herold has taken all double counting into account (he's stated that double counting due to different spellings of place names has been), but assuming a huge influence on the results that would render them "hardly reliable" "based even on this limited sample" doesn't really speak any better of the critic's background in statistics than does Herold's unrelated academic field, nor do their efforts to follow up reports via the web and dropping the issue when it doesn't provide the necessary archived information really say much about the quality of their own "research".

    Regardless Herold doesn't include estimates for casualties due to the interupted flow of aid because of the bombing campaign, and not many others have taken up the challenge (the Guardian has), as again there is little official interest. It's not unreasonable to argue that not taking this into account renders some critics upwards of irrelevant in terms of their self-vaunted dedication to "truth" and calls into question declarations of the minimal effect on civillian life and the corresponding success of the campaign, given that no notable kills were made on the organizers of 9/11 and that CIA/FBI reports indicated that said campaign if anything increased the US security risk. Just like shit you can flush them out, but there's always more to flush if you keep shitting.

    Finally, all the defamations back and forth between contributors to this discussion accusing eachother of racism, low valuations of Afghani lives and/or freedom, fabrication, of being "dumb fucks" and insinuating that one or another individual should be fired and/or is guilty of gassing Kurds is terribly entertaining, but given the prominence of such comments in what passes for the argumentation I'm prone to judge it more as an exemplary indication of how little intelligence has actually been applied to this subject. Bravo everybody.

:: posted by buermann @ 2002-11-20 02:00:00 CST | link

    go ahead, express that vague notion

    your turing test:

journals, notes,
other curmudgeonry

- A Timeline -

Oil for Nothing:
US Holds On Humanitarian Supplies
Iraq: 1997-2001

the good book
and other cultural

The Autobiography
Mother Jones

Contact Info: