🌲 zettels

synthesis Z: Scholarly synthesis is argumentation

Z1a: different building blocks synthesis

Z1b: emotion and levels of processing

Z1c: levels of processing in brain

Z: Doctoral students struggle to effectively synthesize literature

Z: Synthesis is a creative act

Synthesis is sensemaking

Z: Scholarly synthesis is argumentation

Scholarly synthesis is problem framing

Synthesis is learning

A causal model is a kind of synthesis

Interdisciplinary synthesis must integrate diverse evidence

systematic reviews are a narrow subset of synthesis

systematic reviews are infrequently updated, despite often being outdated very quickly

Writing review papers is impactful but discouraged

Claims are a usefully granular unit of knowledge reuse in design

Atomic statements are a fundamental unit of scientific discourse

Decontextualized knowledge is easier to store and refind

Z: Compression is necessary for synthesis

[[CLM - Contextualizability is necessary for synthesis]]

People must recontextualize knowledge before they can reuse it

[[CLM - Specifying context for future reuse is costly]]

flexible compression can mitigate costs of context specification for knowledge reuse

Analytic provenance is necessary for effective intelligence analysis

[[CLM - Compression and contextualizability are in tension]]

Z: Composability is necessary for synthesis

Z: Some degree of formalism is necessary for compositionality

Z: Complex compositionality can arise from surprisingly simple units and connection dynamics

Z: Gracefully integrating formalism into interactive systems is hard

Requiring formalism in interactive systems can be harmful

[[CLM - incremental formalization can mitigate risks of formalism in interactive systems]]

People are reluctant to formalize

Tacit knowledge cannot be formalized

Interactive systems that require formalism frequently struggle with user adoption

Useful structures and formalisms vary significantly across people, situations, and tasks

Fundamental ontologies of disciplines change as the disciplines evolve

Ontologies create an unhelpful illusion of consensus and certainty

Z: Infrastructure is invisible until it fails

Z: Infrastructure doesn't serve everyone

Z: Prefer to accrete new infrastructures on top of existing ones

Z: Infrastructure is sociotechnical, not just technical

Z: Infrastructure work is frequently underappreciated

[[CLM - Scientists read strategically, not linearly]]

People very frequently search for articles they have found before

Citation is a noisy signal

Most systematic reviews are done by temporary volunteers

Many systematic reviews are done in an ad-hoc mix of tools

Researchers today spend less time reading each paper than in the past

Researchers today read more papers than in the past

Researchers today read more than twice as many different journals as they did in the past

Annotations

Scientists repurpose qualitative data analysis software to do synthesis

Z: Sensemaking requires iterative loops of (re)interpreting data in light of evolving schemas

There is a small and finite set of sensemaking representations

People start making sense of large collections by forming temporary clusters

Data extraction is the main cost of sensemaking

Ghost theories hinder scientific progress

Bullshit is very adaptive in environments with noisy and costly evaluation

Leeway is a critical enabler of interdisciplinary work

Intellectual lineage constrains the questions scientists ask

Information is often fragmented across media

People like folders

People often spend most of their 'thinking' time getting into a position to think

Piles of documents often have well-defined semantics

The process of capturing information may be more valuable than the information that is captured

Z: Visual thinking interfaces are tricky to get right