This week smintheus has been going above and beyond his usual stellar work - and that was all by Tuesday night. As I read smintheus' most recent work, I thought: wouldn't it be fun to have a sit-down a la Talmudic scholars and just have a real go at the Bush Admin documents from every angle - use of language, use of stats, use of graphics. Since that is not possible in reality, I'm going to try to do something close.
First, here are smintheus' recent pieces, so you can see what motivated this post. They are in chronological order.
Words of wisdom from Fred Kagan
Pentagon tries to dump the disastrous news
An update on the Pentagon report on Iraq
If you want to read the Daily Kos versions, complete with lots of comments, including both congrats and thanks as well as substantive additions, go to smintheus' DK page where you can also check out the rest of his oeuvre.
Reading If-icial Documents
All of smintheus posts this week point to flaws in if-icial documents and punditry, as well as if-iciousness in general. You are a smart group and will notice that I have replaced official or officious with if-icial and if-iciousness. I use this as a way to label a heap o' rhetorical, statistical, and reality sins we are seeing in the administration's documents and statements and those of its boosters.
Put aside the vacuousness and mendacity of their arguments and "ideas" for a moment, and look only at the language - the phrasing of ideas.
One thing I noticed is the degree to which they use if and the way they use it so they are not lying but are still able to suggest a reality that does not exist here. This is what I refer to as "if-icity". Here is classic if-icity: A statement begins "if x happens". The statement then takes a twist and finally ends as though x had been established, that if had happened, or that it definitely would come about. Tom Friedman is a master of if-icity.
Here is another example - often accompanied by an arrogant tone, which can become pity for those too soft-headed, sentimental, or not in "the know" to understand these high issues. The arrogant tone effectively says that if-iciousness analysis / statement is the only proper form of statement / analysis. It begs all sorts of relevant questions. For example, from smintheus' Fred Kagan piece, we have this quote:
The ground forces must accept longer tours for several years. National Guard units will have to accept increased deployments during this period.
This is the sort of if-icousness reasoning that only the elite can engage in. Its use of "must" and "will have to accept" eschews the need to examine what it suggests are mere details. But those details are anything but "mere."
Here are a few details that matter.
Why should we impose and continue to assume we should impose these burdens only on the good citizens who joined the National Guard and are actually serving. [unlike certain other, well, former national guard members]
And why do we hear nothing about the need to spread this responsibility to all of us?
Isn't this staying in the fantasy realm where it is absurd to demand jenna/barbara-ity service?
Isn't this assuming that the mode of what is "good enough for proles" is what must be?
In other words, the if-icial tone and story rule out asking: Must they stay? Must they serve? And if so: Why?
For an explanation, all we get is if-iciousness.
Another analytical tack
Yes, statistics. I have discussed the use of statistics in general before, and here is the link, if you are interested. Here are a few suggestiongs of things to look for.
Are they changing the unit of measurement?
Are they fudging what a unit of measurement means?
Here are some examples.
1.Footnotes.
In the Iraq study smintheus dissects, on p.31 the table is entitled: "Current Number of Trained* Iraqi Security Forces" However the * footnote completely undercuts any significance that figure has. It says: "* These numbers are not the same as those present for duty." So we seem to have some issue with security forces who have been trained (which is where this reports wants our attention to be focused) versus the numbers who actually show up, which is pretty darned important.
Except in if-icial reports.
2. Units of measurement.
In that same report they seem to make changes in the unit of measurement as it suits them - e.g., number of units v. number of personnel within each unit or total numbers of police units.
3. Choice of baseline.
Problems with baseline figures are common in political reports. If-icial reports will choose a baseline so they can try to show the greatest possible evidence of improvement or whatever it is they want to show, rather than one that makes sense.
4. Labeling bad news as really good news.
smintheus pointed out examples of this. Tom Friedman likes to do this too. It fits with the if-icious view of hard-headedness that eschews sentiment. And, as Friedman is fond of saying, we'll all see that in about 6 months.
5. Leaving out bad news.
smintheus pointed out that they said nothing in the Iraq report about the situation of women, which as we all know has become awful from a situation a few years back that for most was relatively decent. Iraqi women now fear to leave their houses at all, fear to leave them - if they go out - unless they are clad in black robes.
smintheus:
And let's not forget that the things it does not bother to report, to survey, and to investigate are perhaps more important than what it does address. The things it skates over may be a lot more telling than some of the trivial things it stops to focus on.
For example, I remember precious little in it about how the lives of Iraqi women have deteriorated terribly in the last year. How do you measure the fear of appearing outdoors? This report doesn't seem to have even tried. That's a salient fact.
Imagine, sisters, if this were done to us here. These women exist and live under an oppressive system we would find intolerable here - and one that was not so long ago more similar to life here now.
6. Grouping questions together that were actually asked separately and/or even dropping categories, which would then inflate the % remaining.
Again, smintheus gives examples of both these techniques in the Iraq report.
I've seen both these tricks used plus mischaracterizing what people said and carefully picking the respondents, which is referred to as selection bias.
7. In the case of polling Iraqis, it must be quite a trick to get people to answer given that they must have some concern that the answers might go straight to the folks who carry you away in the middle of the night never to be seen again.
8. Ask people's opinions and then using those opinions as standing for reality.
And, that, my friends, is the if-icial story.
Comments
Add Comment