Trees, Branches, Twigs and Leaves — Mental Models for Writing

21 Dec 2024

Shreyas Prakash headshot

Shreyas Prakash

What differentiates a good essay and a great essay? — With a great essay, you can see the larger tree that connects all those branches, leaves, and twigs together into a single unit of a ‘tree’. A bad essay is where you only see the random assortment of leaves and branches. You don’t see the larger tree among the branches.

While listening to one of the podcast episodes from David Perell’s Writes of Passage where he interviews writer-operators to deconstruct their writing process, I noticed the podcast episodes to be very free-flowing, without losing the central theme of the conversation.

And this made the format engaging. In fact, the most boring podcasts I’ve listened to are the ones which start with the premise, and then the supporting arguments, and then the conclusion. That’s not what podcasts/conversations are supposed to be. A conversation has to be fluid. It should not be confined by a specific agenda/bullet points. It’s a discourse, or a ‘meeting’ in the business-sense with specific agenda / bullet points.

After listening to a couple of the episodes, I unearthed Perell’s mental model on how he approaches his interviews with his podcast guests. The tree is the mental model here.

Perell starts with the central theme (the tree), and then goes into the various sub-topics (the branches), goes deeper into the sub-topics of the sub-topics (twigs, leaves), and before it gets more deeper than that, he takes a step back, and directs the user back to the central-theme (the tree). In this way, the viewer/listener is not trapped in some version of a prime-number maze, where they’re not able to see where they are heading. They are still directed back to the tree, and can also capture the gist of what the author is saying. And this is a common pattern among other seasoned podcast hosts such as Joe Rogan, Lex Fridman and others.

And an attempt to progress is by following this principle: Go deep into the sub-topics, suspend, and take a step back and come back to the main topic. You’re the spider, and you are interweaving all the webs. Perell’s approach on podcasting has inspired me to do the same with my writing process. Essays are also supposed to be more ‘loose’ than be ‘rigid’. Even etymology wise, essays originate from the French word ‘essais’ which means an attempt.

So, I can very well add side-notes, callouts, footnotes, and make side-steps throughout my essay, but the reader has to still see the forest among the trees.

The plot should not be lost in the mess. And this is a creative tension worth pursuing.

Subscribe to get future posts via email (or grab the RSS feed). 2-3 ideas every month across design and tech

Read more

  1. Notes on Londontravel
  2. 18 key ideas from reading Henrik Karlsson this monthwriting
  3. Stop arguing, and start drawing circles togethermathematics
  4. The whole world is just the snake eating its own tailmental-models
  5. Life lessons and hot takes from my 30slifestyle
  6. Building a skill for coherent science illustrationsscience
  7. My agentic engineering workflow (step by step)agentic-coding
  8. Hammock driven developmentagentic-coding
  9. Peculiar ways number three fits into our funny little brainsmental-models
  10. AI sandwich as a defacto principle for anything agentic engineering relatedagentic-coding
  11. Authority in the guise of evidencecritical-rationalism
  12. Map is not the territoryphilosophy
  13. Self hypnosis as a manifestation ritualmeditation
  14. Hegelian dialectic for structured reasoning with AI agentsphilosophy
  15. How I prepare for tough negotiations nowadaysnegotiation
  16. When should we steelthread somethingproduct-development
  17. Breadboarding, shaping, slicing, and steelthreading solutions with AI agentsproduct
  18. Healthy conflict in teams have a tipping pointteam-building
  19. How I deslopify AI writingwriting
  20. How I started building softwares with AI agents being non technicalagentic-coding
  21. Read raw transcriptswriting
  22. Legible and illegible tasks in organisationsproduct
  23. L2 Fat marker sketchesdesign
  24. Writing as moats for humanswriting
  25. Beauty of second degree probesdecision-making
  26. Boundary objects as the new prototypesprototyping
  27. One way door decisionsproduct
  28. Finished softwares should existproduct
  29. How I periodically rank my rough draftsobsidian
  30. Flipping questions on its headinterviewing
  31. Vibe writing maximswriting
  32. How I blog with Obsidian, Cloudflare, AstroJS, Githubwriting
  33. How I build greenfield apps with AI-assisted codingagentic-coding
  34. We have been scammed by the Gaussian distribution clubmathematics
  35. Classify incentive problems into stag hunts, and prisoners dilemmasgame-theory
  36. I was wrong about optimal stoppingmathematics
  37. Thinking like a shipmental-models
  38. Hyperpersonalised N=1 learningeducation
  39. New mediums for humans to complement superintelligenceagentic-coding
  40. Maxims for AI assisted codingagentic-coding
  41. Virtual bookshelvesaesthetics
  42. It's computational everythingtrends
  43. Public gardens, secret routesdigital-garden
  44. Git way of learning to codeagentic-coding
  45. Style Transfer in AI writingagentic-coding
  46. Understanding codebases without using codeagentic-coding
  47. Vibe coding with Cursoragentic-coding
  48. Virtuoso Guide for Personal Memory Systemsmemory
  49. Writing in Future Pastwriting
  50. Publish Originally, Syndicate Elsewhereblogging
  51. Poetic License of Designdesign
  52. Idea in the shower, testing before breakfastsoftware
  53. Technology and regulation have a dance of ice and firetechnology
  54. How I ship "stuff"software
  55. Writing is thinkingwriting
  56. Song of Shapes, Words and Pathscreativity
  57. How do we absorb ideas better?knowledge
  58. Read writers who operatewriting
  59. Brew your ideas lazilyideas
  60. Trees, Branches, Twigs and Leaves — Mental Models for Writingwriting
  61. Compound Interest of Private Noteswriting
  62. Conceptual Compression for LLMsagentic-coding
  63. Meta-analysis for contradictory research findingsdigital-health
  64. Proof of workproduct
  65. Gauging previous work of new joinees to the teamleadership
  66. Task management for product managersproduct
  67. Beauty of Zettelswriting
  68. Stitching React and Rails togetheragentic-coding
  69. Exploring "smart connections" for note takingwriting
  70. Deploying Home Cooked Apps with Railssoftware
  71. Repetitive Copypromptingwriting
  72. Questions to ask every decadejournalling
  73. Balancing work, time and focusproductivity
  74. Hyperlinks are like cashew nutswriting
  75. Brand treatments, Design Systems, Vibesdesign
  76. How to spot human writing on the internetwriting
  77. Can a thought be an algorithm?product
  78. Opportunity Harvestingcareers
  79. Everything is a prioritisation problemproduct
  80. How I do product roastsproduct
  81. The Modern Startup Stacksoftware
  82. In-person vision transmissionproduct
  83. How might we help children invent for social good?social-design
  84. The meeting before the meetingmeetings
  85. Design that's so bad it's actually gooddesign
  86. Lessons learnt interview prepping for product rolesinterviewing
  87. Obsessing over personal websitessoftware
  88. English is the hot new programming languagesoftware
  89. Better way to think about conflictsrisk-management
  90. The role of taste in building productsdesign
  91. Dear enterprises, we're tired of your subscriptionssoftware
  92. Products need not be user centereddesign
  93. World's most ancient public health problemsoftware
  94. Pluginisation of Modern Softwaredesign
  95. Let's make every work 'strategic'consulting
  96. Making Nielsen's heuristics more digestibledesign
  97. Startups are a fertile ground for risk takingentrepreneurship
  98. Insights are not just a salad of factsdesign
  99. Minimum Lovable Productproduct
  100. Methods are lifejackets not straight jacketsmethodology
  101. How to arrive at on-brand colours?design
  102. Minto principle for writing memoswriting
  103. Importance of Whytask-management
  104. Quality Ideas Trump Executionsoftware
  105. Why I prefer indie softwareslifestyle
  106. Use code only if no code failscode
  107. Self Marketing
  108. Personal Observation Techniquesdesign
  109. Design is a confusing worddesign
  110. A Primer to Service Design Blueprintsdesign
  111. Rapid Journey Prototypingdesign
  112. Visualise detailed file structures on CLIcli
  113. Do's and Don'ts of User Researchdesign
  114. Design Manifestodesign
  115. Complex project management for productproducts
  116. How might we enable patients and caregivers to overcome preventable health conditions?digital-health
  117. Pedagogy of the Uncharted — What for, and Where to?education
  118. Future of Equity with Ludovick Petersinterviewing
  119. Future of Ageing with Mehdi Yacoubiinterviewing
  120. Future of Tacit knowledge with Celeste Volpiinterviewing
  121. Future of Mental Health with Kavya Raointerviewing
  122. Future of unschooling with Che Vanniinterviewing
  123. Future of Rural Innovation with Thabiso Blak Mashabainterviewing
  124. Future of work with Laetitia Vitaudinterviewing
  125. How might we prevent acquired infections in hospitals?digital-health
  126. The why to endure any howentrepreneurship
  127. Design education amidst social tribulationsdesign
  128. How might we assist deafblind runners to navigate?social-design