jan d via matrix 
Collection of thoughts:
> we in software don't really use metaphors. Or: we use them in a maximally impoverished way.
Oh, interesting point. I sometimes mention to non-programmer people that computing language is very metaphor rich. But I also agree, it is very impoverished, in the sense that concepts are supposed to be non-leaky, and the interesting thing about a lot of metaphors is that they come with connections and associations (which might play a large role in coming up with good abstractions, but once that is stabilized, they probably cause problems. (Garfinkel/Sacks in "On Formal Structures of Practical Actions" mention "Mockups" as things that partially explain something, maybe worth a look)
digraph zettelkasten { layout=twopi; rankdir=LR; node [shape=rect, style=filled, fontname="Helvetica"]; edge [fontname="Helvetica"]; // Existing Nodes (1-18) 1 [label="1: brianmarick\nFirst thoughts about fedwiki narratives"]; 2 [label="2: brianmarick\nMetaphor wiki goals (reference + through-lines)"]; 3 [label="3: brianmarick\nFedwiki notes insufficient for narrative insights"]; 4 [label="4: brianmarick\nSoftware metaphors are 'impoverished' (Aha!)" fillcolor="#ffcccc"]; 5 [label="5: brianmarick\nRethinking scripts vs. wiki rework burden"]; 6 [label="6: brianmarick\n'Kill your babies' analogy (wiki as storage)"]; 7 [label="7: Ward\nChallenge: non-metaphorical computing terms"]; 8 [label="8: brianmarick\nExisting text as 'spike' (write-to-throw-away)"]; 9 [label="9: Ward\nConflicting metaphors (conveyor vs. relay)"]; 10 [label="10: Ward\nNew Relic reorganization example"]; 11 [label="11: brianmarick\nComputational metaphors lack 'oomph'" fillcolor="#ffcccc"]; 12 [label="12: Ward\nMetaphors as unspoken frameworks (Aha!)"]; 13 [label="13: Ward\nCanyon aside (unresolved analogy)" fillcolor="#ffcccc"]; 14 [label="14: brianmarick\nConfusion about canyon reference" fillcolor="#ffcccc"]; 15 [label="15: Brian\nPraise for FedWiki's sprawl + courage"]; 16 [label="16: Brian\nNew wiki for narrative + link to old"]; 17 [label="17: Brian\nParagraphs as index cards (vs. essays)"]; 18 [label="18: Brian\n'Writing to think' vs. 'writing for audience'"]; // New Matrix Conversation Nodes 19 [label="19: jan d\nMetaphor analysis:\n- Software: 'non-leaky'\n- Natural: rich associations\n(ref Garfinkel/Sacks)", fillcolor="#ccffcc"]; 20 [label="20: jan d\nNarrative order:\nConstraint → Insight", fillcolor="#ccffcc"]; 21 [label="21: brianmarick\nWhy metaphors fail:\na) Overly concrete (Lakoff)\nb) Avoid inconsistency\nc) Lack context-setting", fillcolor="#ffcccc"]; // Theoretical References 22 [label="22: [STUB]\nGarfinkel/Sacks\n'Mockups'", fillcolor="#ccccff"]; 23 [label="23: [STUB]\nLakoff's\nConceptual Metaphors", fillcolor="#ccccff"]; // Primary Path 1 -> 2 -> 3 -> 4 -> 5 -> 6 -> 7 -> 8 -> 9 -> 10 -> 11 -> 12 -> 13 -> 14 -> 15 -> 16 -> 17 -> 18; // Matrix Conversation Links 3 -> 20 [xlabel="Matrix:\n'constraint-driven insights'", style=dashed, color="#666666"]; 4 -> 19 [xlabel="Matrix:\n'impoverished vs associative'", style=dashed, color="#666666"]; 19 -> 22 [label="cites", style=dotted, color="#0000ff"]; 21 -> 23 [label="references", style=dotted, color="#0000ff"]; 21 -> 4 [label="extends", color="red"]; 21 -> 9 [label="contradicts", color="red"]; 20 -> 3 [label="explains", style=dotted]; // Alternative Paths 2 -> 4 [style=dotted]; 4 -> 7 [style=dotted]; 7 -> 9 [style=dotted]; 9 -> 11 [style=dotted]; 1 -> 16 [style=dotted]; 6 -> 15 [style=dotted]; // Inconsistencies 4 [fillcolor="#ffcccc"]; 9 [fillcolor="#ffcccc"]; 11 [fillcolor="#ffcccc"]; 13 [fillcolor="#ffcccc"]; 14 [fillcolor="#ffcccc"]; 21 [fillcolor="#ffcccc"]; }