Writing Is Thinking: A Cognitive Approach to Learning

One special thing about myself is I really like to code. I would spend countless hours to solve a problem. When I am coding, time goes really fast!. It is mainly because I really enjoy it.

On the other hand, I hate writing. It is not my first language. When I was a kid, I had really-bad grades for English. I didn’t know grammar, and I struggle a lot! When time goes by, I improved my English. I knew some amount of grammar, and I could write something, but it was never perfect!

So, now I am in my third year in my PhD, and I realized that I have to write a lot. Slowly, but I learned why it matters to learn to write. It helps us to organize our thoughts, understand concepts, and many more. I learned grammar past couple of months, quite well, but my thoughts are not flowing, and I could not write and forget grammar. At this point I want to improve my writing so bad, I want to become that independent researcher, who writes most parts of a paper within a day, like a champ! As my advisor suggested, I am going to write consistently, no matter what! I know for a fact that consistency is everything! It will make me a better writer!

---

What to write?

I wake up at 4.00 am, every weekday. I decided to use this time to write! Question is, what am I going to write, how to choose something to write?

To make things easy, I decide my topic or what I am going to write a day before! Follow below topics:

• Category 1: Systematic Review Writing ((for academic writing, directly building my paper))
  • Summarize a paper – write a Gopen-style summary (context → method → finding → gap).
  • Compare two papers – Explain how their data sources, NLP methods, or objectives differ.
  • Write a synthesis paragraph – Combine findings across multiple papers into one coherent argument.
  • Draft a paper section – Add 150–200 words to your background, methods, or discussion.
  • Reflect on a limitation or bias – Write about what the literature is missing and why it matters.
• Category 2: Research Development Writing
  • Explain a concept simply – Write a short post that answers “What is ___?” (e.g., “What is Persistent Homology?” or “What Are Transformer Models?”).
  • Summarize a paper conversationally – Describe a new research paper as if explaining it to a peer over coffee.
  • Bridge theory to application – Write “How this idea connects to my work” (e.g., “How Topology Can Improve Opioid Risk Modeling”).
• Category 3: Style & Structure Training
  • Rewrite an old paragraph – focus on topic–stress positions and clarity.
  • Sentence surgery – pick 5 sentences from a paper and rewrite them in active voice.
  • Gopen exercise – take one idea and write it in three different sentence orders.
  • Transition practice – write 5 sentence pairs that connect ideas smoothly.

---

Core writing cycle

Phase	Time	Purpose
Night (Study phase)	1 hour	Learn something: read one paper, tutorial, or concept (e.g., persistent homology, attention in LLMs, uncertainty quantification in NLP). Capture 5–8 key ideas in your notes.
Morning (Writing phase)	4–6 AM	Write a short, structured piece (400–700 words) turning that idea into a clear post. Focus on explaining, not summarizing.
Later (Polish & Publish)	10-15 minutes	Edit tone for general readers; add a simple graphic or hook; post on LinkedIn or blog.

Why writing phase matters in learning? Writing helps us to recall what we have learned. People (including me) re-read documents/theories just to understand(memorize). But it is scientifically proven that re-reading is the worst learning technique, poor retention. Instead of re-reading, try to recall what you have learned!

• Take few minutes off summarize what you have learned.
• Sketch down a mind map(important concepts that you have learned).
• Draw it in a white board; Explain it loudly!
• Most importantly, write about it!

This is where writing comes in handy. While reading one blog post(I will try to find it), I saw this statement about writing: “You really don’t know it, until you can explain it!”. It hit home for me, I start to write about everything I learn!

---

First draft system

First thing first, don’t worry about grammar nor perfectionism. Just read about the theories or anything that you want to learn and sketch down and summarize. In the next stage, don’t read again, just keep a blank mind and write about whatever you learned.

1. Capture.

   • The Goal: Capture, don’t compose.
   • Maybe create a specific folder in mac, and document on notability.
   • Summarize on notability.

2. Write the first draft.

   • Goal: Get your brain’s raw material onto the page before your inner editor wakes up.
   • Aim for Completion, Not Perfection
   • Write to discover what you think. Edit to discover how to say it clearly.
   • Your first draft is a thinking draft — not a finished product.
   • What to focus on?
     • Don’t worry about grammar, spelling, or perfect word choice yet.
     • Focus on ideas — what you want to say and in what order.
     • If a sentence feels rough, keep going. You’ll fix it later.
     • Use placeholders like [add citation] or [find example later].
   • Use Prompts to Stay Focused
     • If you get stuck, ask yourself:
       • What do I really want to say here?
       • What question am I answering?
       • Why does this matter?

---

Post-writing editing system

1. Step 1: Read loud (You will understand, if something feels off)
2. Step 2: Take a paragraph; Read and check the flow and structure.
   • Read the topic sentence and final sentence of a paragraph. If the topic sentence and final sentence of each paragraph make sense when read together, your structure works.
3. Step 3: Read sentence by sentence and make sure these two things:
   • (first) Make sure “old before new”, reader familiar old information first, then the new content.
   • (second) Most of the time use active voice.
   • Reminder: If you have to use passive to keep the “old before new”, go for it. Also, the stress position of a sentence should be in the end.
4. Step 4: Now read the paragraph again(the meaning).
   • Does it convey what you want to say?
   • If not add or remove sentences.
   • Go to step 1.
5. Step 5: Check the grammar
   • Correct active passive voice.
   • Correct time.
   • Article use.
   • Avoid verb nouns.
   • Use tools: Grammarly

Additional checks:

• Goal: ensure your writing says exactly what you mean — nothing vague, nothing extra.
• Is your main idea crystal clear in the first few sentences?
• Are all technical terms defined at least once in plain words?
• Are your claims supported with evidence or reasoning?
• Could a non-expert follow your argument?
• Delete filler words (very, really, in order to, due to the fact that).