Protocol · study · 18:30 to 19:30, then wind down
English and Chinese, outdoors, walking, audio only, no screen. The hour does two jobs at once, and that is why it is placed here rather than in the morning.
Receptive language work is auditory and motor rather than executive, so it does not spike arousal the way problem solving does. It belongs where your brain is already used up.
And with a five minute commute you have no buffer at all between desk brain and sleep brain. Arousal does not fall from a standing start in an hour: that is the most likely reason you will lie awake, and it will feel like the bedtime is wrong when the transition is what is missing. Doing this on a walk manufactures the buffer. You get the language hour and the decompression in the same 25 minutes instead of trading one for the other.
Producing language out loud is mildly arousing: vocal effort, breathing changes. Fine at 18:45. Not fine at 19:25.
The pitch roleplay. Performing under pressure is genuinely sympathetic-activating, which is the entire point of the exercise. It cannot live at 19:30.
It goes in a weekend morning block, where you can go hard without paying for it at bedtime.
Sunset in Rome is close to 20:50 in late June, so at 19:00 you are walking in full daylight.
Bright evening light is a phase delay signal. Without sunglasses this walk quietly undoes what the morning lamp did, every night, for four months.
Thirty minute weekday fragments do not build a language. A full hour, alternating, does more with the same total. English 4h, Chinese 3h.
Three evenings gets you to 3h. The fourth is a weekend morning block, and it is also where the pitch roleplay and the monthly recording live, because both are too activating for the evening.
Wednesday used to be a cooking evening, which ate this slot every week. Enit cooking means all six evenings are clean.
Hiring a cook bought back roughly 50 hours of language a year. That is more than it looks.
Because you already have the language. Shadowing maps sound you can already parse onto your own motor output, and that mapping is exactly what is missing at 10 out of 100. This is not true of Chinese, which is the section after this one.
| Mode | Share | What it is | The one rule |
|---|---|---|---|
| Shadowing | 60% | Play a native-speaker interview and repeat a beat behind, out loud, matching rhythm and stress rather than meaning. | Pick people you want to sound like. The cadence you copy is the one you get. |
| Reading aloud | 25% | Read a paragraph out loud, then reread it fixing one specific thing: a vowel, a linking, a stress pattern. | One target per session. Fixing everything at once fixes nothing. |
| Free output | 15% | Explain something you already know, unscripted, as if mid-conversation. | Do not stop to correct yourself. Fluency and accuracy are trained separately. |
Record a spoken pitch on the same topic every month, unscripted, same length. Weekend morning, not the evening.
Controlling the topic is what makes the improvement visible: it removes the only variable that could explain the difference away. Twelve recordings a year and you cannot argue with them. Fix the topic once, now, and never change it, or the benchmark is worthless.
This is the single most important line on the page, and it contradicts what you planned to do.
Shadowing works by mapping sound you already partly parse onto your own motor output. With no tonal system in place, native-speed Mandarin is undifferentiated noise. You would be copying something you cannot yet hear, which trains nothing and feels like progress, which is worse than doing nothing.
English shadowing works from day one because you already have the language. Chinese does not get the same treatment for another three months.
The twenty two-syllable combinations: four tones against four tones, plus the four with a neutral syllable. Drilled until they are automatic.
Why this is the whole game. Get tones wrong early and you spend years being not-quite-understood. It is the one beginner mistake that is genuinely expensive to fix later, because by then it is a motor habit and not a knowledge gap. Undoing a motor habit costs an order of magnitude more than building the right one.
Method: listen to a minimal pair, say it, record it, compare. Short daily reps beat long weekly ones. Alongside this, graded comprehensible input: material built for beginners, not native material played slowly.
Once the tone system is in place, everything you could not use before becomes usable. This is the gate, and it is worth waiting for rather than jumping. The three months are not a delay, they are what makes the next two years work.
It was the explicit "might as well", which is exactly what makes it the cheapest thing to lose when the hours do not close. And given Italian, it stays nearly free to pick up later, so postponing it costs almost nothing.
Revisit once Chinese clears the tone-pair phase, not before.
Reading, meditation, or nothing. Meditation and boredom need no slot of their own. They live here and on the walks.
The 12:30 walk is the boredom quota, already paid. Do not podcast it.
Same design rule as the body ones: anything that needs a judgment call to score will not survive three months. Baseline these on 5 and 6 September, before the schedule has done anything to you.
| Test | Protocol | Score | Why it works |
|---|---|---|---|
| The Recall | Immediately after listening to something, write a sealed checklist of the 10 points that mattered. The next day, summarise from memory and score against the sealed list. | /10 | The seal is the mechanism. You are graded against what you thought mattered before you knew you would be tested. Without it you unconsciously grade yourself against whatever you happened to remember. |
| Boredom Sit | Sit doing nothing. No phone, no music, no plan. Stop the clock at the first urge you act on. | minutes | It turns "be bored daily" into an integer, and it is the closest thing you have to a direct read on whether attention is recovering or just being rearranged. Most people do not clear four minutes cold. |
| English pitch | Monthly recorded pitch, same topic every time, unscripted. | the tape | Twelve a year, and the comparison is the score. |