Conversation

Good analogy. I’m always exhausted by crackpots dumping evidence on me I’m not competent to evaluate and which looks like it will take me 6 months to even behind doing my “own research” with.
Quote Tweet
What the crank is doing is ultimately a lot more condescending—the equivalent of giving a child a fake cell phone so they can “make calls” just like mom & dad. They’re pretending not to ultimately rely on trust, and so they get trust.
Show this thread
3
64
Actually reading research even in a field you have competence in is harder than reading code written by someone else. My advisor’s standard for a phd was: collect and survey 100 papers, shallow read 20, deep read 5. Which is actually an ambitious target for the 3-5 years it takes
1
34
You’re not “doing your own research.” It’s some sort of academic tsundoku. At *best* you’re scanning for resonance with something you already believe.
2
31
Research is also harder than building in industry. Just because a lot of phd academic researchers who do it badly are dumber than people in tech doing simpler things doesn’t make the *task* easier. The smartest people don’t necessarily get selected to work on hardest problems.
1
22
Most “do your own research” discourse is at the level of: A: Hey, look at this infinite chocolate bar trick B (smugly): Ah, the Banach-Tarski theorem! Neither of them understands the theorem or how it relates to the chocolate bar trick (it doesn’t really)
1
17
This is increasingly becoming a real problem. I suspect it’s due to people having too few reading modes. I guess when even establishing a single kind of literacy is a challenge, expecting 2 or 3 is unrealistic.
1
15
I wonder if there’s a relation to Americans being mostly monolingual. Many Americans seem to genuinely believe that English spoken really slowly, with careful enunciation and exaggerated gesticulation, somehow equals entirely different languages.
Replying to
Research reading isn’t just like reading tweets really slowly. It’s a distinct literacy. And unlike many literacies, like picking up on the tropes of a new pop fiction genre, where creators kinda work with your natural literacy acquisition, “research” is an unnatural literacy.
1
12
It is NOT a natural way to read or process anything. Breaking down claims, identifying necessary/sufficient conditions, following formal logic, examining how words are defined and used, tracing critical path… it’s all very unnatural.
1
17
And typically since every paper, especially in STEM, rests on others, you’ll typically only “understand” the 20% that’s new in the paper. The other 80% you’ll take on faith unless you’ve studied it before. Just as programmers take on faith that the chip underneath works.
1
14
Going through the “stack” a result depends on and identifying gaps or circular support structures etc is approximately as hard as forensic analysis, a whole deeper level of scrutiny beyond basic reading or even peer review.
1
12
Not saying you must have a phd to have the literacy, or even have the exact same kind of literacy academia develops. But if autodidact, you need to have done a comparable amount of work acquiring an equivalent outsider literacy.
2
13