5 ms·

> A huge fraction of us believe in fundamental truths that transcend reportable fact, of the power of first principles reasoning and of a web of credibility ext

> A huge fraction of us believe in fundamental truths that transcend reportable fact, of the power of first principles reasoning and of a web of credibility extending from sources to journalists to our favored analysts. I find this world view terrifying, but I understand why other people have it. I would argue that there's some (good) reasons to have this kind of belief though, given what we've seen. The very basis of the CIA's clandestine operations is to abuse people's usual reliance on those mechanisms by subverting some part of that process. That's "plausible deniability" in a nutshell -- that there exists at least one place you can't actually connect that logical chain you suggested. Once you "lift" every step of the chain to include the idea of "this step is likely being manipulated by experts to misinform me", the minimal explanatory model is often not the same as the minimal explanatory model would be without that. I would argue that this is the discrepancy that you're pointing out when people deal with the CIA/NSA/etc: the minimal model people think of must account for the massive amount of lying, including to their nominal overseers like Congress; the minimal assumption for lying is that they're up to something, not that they're just lying, because lying (even for them) does have a slight cost, and historical data shows that when they were lying in the past... they were usually experimenting on humans, building up blackmail profiles of civic leaders, illegally performing their duties, etc. Ergo, you get people with models that have HUGE priors that those agencies are up to something targeting various communities, and so stories which doesn't have the usual amount of evidence are still enough evidence to suggest that the CIA/NSA/etc is up to something targeting that group. I'm not arguing that most people are reaching the correct priors -- just that the overreactive priors are "less wrong" than the null priors your complaint suggests they should use. I would argue that the insistence on "background" priors, in the face of decades of consistent and premeditated behaviors on the part of the CIA/NSA/etc, is the exactly opposite of journalistic integrity: it's specifically refusing to report what you know (and is good for the public to know, too), by getting rid of your knowledge of what the real priors are and not reporting on that. I find a lot of the same "dishonesty" in reporting on issues of culture or race, again, often presented behind a veneer of integrity, when really, it's capitulation to social power -- either to not offend the government or cultural norms.

tptacek 10y ago
I understand what you're saying and I understand why people believe this to be the case. But help me understand something: How are you not arguing that journalists should report what they feel, rather than what they know? I'm very ready to believe you aren't saying that, but I can't reconcile what you wrote with that question.
- SomeStupidPoint 10y ago
  > How are you not arguing that journalists should report what they think, rather than what they know? I changed your question slightly, but I hope I preserved the intent of it. If not, please let me know! I guess my point is that "knowledge" I have is just a special class of "thoughts" (or "beliefs") I have, where my surety in them is above some particular threshold. It's certainly not the case that most of my knowledge is absolute and couldn't (theoretically) be wrong. (At least, as I think you're meaning it, we could reasonably say I "know" my bed is in my bedroom, even though, strictly speaking, I'm in my living room and it might be elsewhere.) I would argue that the journalists are then reporting knowledge: that they have good (established, documented) reason to think that the priors are some particular way (which would be knowledge), and given those priors with some facts, the CIA/NSA/etc is up to some class of actions, of which the told story is an example member, which is again knowledge, because we can have a stronger belief that the class of actions is correct than the particulars. An example might be: you're not sure that I cut the victim with a knife, but the slash marks all over his body indicate some kind of cutting instrument. Is it not useful to talk about me slashing him? Similarly, I would argue that I expect journalists to report knowledge in the usual sense, but allow them to present knowledge about things like priors or that something is in a class of behaviors, rather than just knowledge that a specific behavior occurred. (Which is what I get out of your text, but I don't think is actually what you're trying to claim.) It's certainly common in mathematics to introduce talking about a class of objects with a few definite examples, and I feel that the reporters telling "fictionalized but consistent with facts" stories to exemplify the class of things we know the CIA is up to (or capable of) is meaningful and useful journalism. That kind of writing helps us picture and understand what the class of (likely) actions really looks like, particularly if we read multiple sources which picked different examples from the class. Of course, I think we both agree that journalists could be clearer about when they're doing that kind of thing, and the field of journalism could obviously do better on epistemology in general. I guess my argument could be summarized as: the knowledge being shared by those kinds of reports isn't about the specific actions of the CIA being discussed in the story, but rather, the CIA priors versus background priors (in context and historical pieces) and the knowledge that the CIA is up to something of the flavor in the article. I also have a lot of sympathy for your position, and get where you're coming from, but let me ask you this: how do you think journalists should professionally convey information about CIA behavioral priors (eg, that they're systemic liars and career criminals) and sufficient evidence (in light of those priors) to suggest the CIA is doing something against a community when they don't have sufficient evidence for any particular claim, eg, that it was actually CIA officers selling crack versus enabling selling crack by claiming drug dealer phone numbers were agency ones to police versus just not doing anything to stop it because it served their ends? Surely you agree that journalists should be able to present the claim that the CIA willfully failed in their duty to protect black communities from foreign drug interests, and even probably actively worked against them, without having to get bogged down in the specifics of which actions you think they perpetrated when if there's sufficient circumstantial evidence that something unusual occurred. Otherwise, how would you ever write a story about the actions of experts in plausible deniability?

tptacek 10y ago

I understand what you're saying and I understand why people believe this to be the case. But help me understand something: How are you not arguing that journalists should report what they feel, rather than what they know? I'm very ready to believe you aren't saying that, but I can't reconcile what you wrote with that question.

SomeStupidPoint 10y ago
> How are you not arguing that journalists should report what they think, rather than what they know? I changed your question slightly, but I hope I preserved the intent of it. If not, please let me know! I guess my point is that "knowledge" I have is just a special class of "thoughts" (or "beliefs") I have, where my surety in them is above some particular threshold. It's certainly not the case that most of my knowledge is absolute and couldn't (theoretically) be wrong. (At least, as I think you're meaning it, we could reasonably say I "know" my bed is in my bedroom, even though, strictly speaking, I'm in my living room and it might be elsewhere.) I would argue that the journalists are then reporting knowledge: that they have good (established, documented) reason to think that the priors are some particular way (which would be knowledge), and given those priors with some facts, the CIA/NSA/etc is up to some class of actions, of which the told story is an example member, which is again knowledge, because we can have a stronger belief that the class of actions is correct than the particulars. An example might be: you're not sure that I cut the victim with a knife, but the slash marks all over his body indicate some kind of cutting instrument. Is it not useful to talk about me slashing him? Similarly, I would argue that I expect journalists to report knowledge in the usual sense, but allow them to present knowledge about things like priors or that something is in a class of behaviors, rather than just knowledge that a specific behavior occurred. (Which is what I get out of your text, but I don't think is actually what you're trying to claim.) It's certainly common in mathematics to introduce talking about a class of objects with a few definite examples, and I feel that the reporters telling "fictionalized but consistent with facts" stories to exemplify the class of things we know the CIA is up to (or capable of) is meaningful and useful journalism. That kind of writing helps us picture and understand what the class of (likely) actions really looks like, particularly if we read multiple sources which picked different examples from the class. Of course, I think we both agree that journalists could be clearer about when they're doing that kind of thing, and the field of journalism could obviously do better on epistemology in general. I guess my argument could be summarized as: the knowledge being shared by those kinds of reports isn't about the specific actions of the CIA being discussed in the story, but rather, the CIA priors versus background priors (in context and historical pieces) and the knowledge that the CIA is up to something of the flavor in the article. I also have a lot of sympathy for your position, and get where you're coming from, but let me ask you this: how do you think journalists should professionally convey information about CIA behavioral priors (eg, that they're systemic liars and career criminals) and sufficient evidence (in light of those priors) to suggest the CIA is doing something against a community when they don't have sufficient evidence for any particular claim, eg, that it was actually CIA officers selling crack versus enabling selling crack by claiming drug dealer phone numbers were agency ones to police versus just not doing anything to stop it because it served their ends? Surely you agree that journalists should be able to present the claim that the CIA willfully failed in their duty to protect black communities from foreign drug interests, and even probably actively worked against them, without having to get bogged down in the specifics of which actions you think they perpetrated when if there's sufficient circumstantial evidence that something unusual occurred. Otherwise, how would you ever write a story about the actions of experts in plausible deniability?