Saju and bazi: what Korean and Chinese four pillars share, and where they part

Saju and bazi are the same eight characters, read by two traditions that grew apart. The calendar underneath is shared down to the minute. The lineage, the clock corrections, the vocabulary and the questions people arrive with are not.

Is saju the same thing as Chinese bazi?

At the level that matters most, yes. Saju (사주), the Korean reading of the four pillars of birth, and Chinese bazi (八字) are built from the same sexagenary calendar, the same ten stems and twelve branches, the same twenty four solar terms, and the same four pillars of year, month, day and hour. Give one birth moment to a Korean reader and to a Chinese reader and, conventions aside, the eight characters come out identical.

What differs is everything after that: emphasis, vocabulary, how much weight the auxiliary lists carry, and what the client came in wanting. If saju is new, start here.

Korea, sajuChina, baziJapan, shichu suimei
What the name meansFour pillarsEight charactersInferring fate from four pillars
Clock correctionAbout 30 minutes, near uniformZero to over two hours by citySmall, running both ways
Day turns atCommonly 11:30 at night11 at night or midnight, disputedCommonly 11 at night
Ten gods calledSipseong, sipsin, yukchinShi shenTsuhensei
Symbolic starsWidely usedUsed, and argued overUsed, with the twelve stages
Marriage readingGunghap, once part of the wedding procedureHehun, a private matterAisho, informal

Where did saju, bazi and shichu suimei come from?

One root, three arrival dates. The method took its present shape in Song dynasty China and is called the Ziping method after Xu Ziping, credited with adding the hour pillar. The manuals that followed are still the shared library, so Korean and Chinese readers quote the same books today.

Korea has known the method since at least the Goryeo period, then did something with it that China did not: Joseon made it a government subject. Myeonggwahak (명과학) was taught in the royal astronomy office and examined under the state examination for yin yang specialists, with Xu Ziping among the set texts, and it covered fortune, divination and the choosing of auspicious dates together. Korean practice still keeps date selection close to chart reading, because for centuries they were one civil service subject rather than two trades.

Japan received it later and through translation. In the early nineteenth century Sakurada Komon, a Confucian scholar of Sendai, produced the Suimeisho, the oldest documented Japanese rendering of the Yuanhai Ziping. Sources disagree on the year, some giving 1801 and some the Bunsei era two decades later. Japanese schools then developed alone for two centuries and disagree among themselves as much as with China.

Why do the three traditions correct clock time differently?

Geography, not doctrine, and this is where comparison pages go vague. All three read from the sun, so clock time has to be turned back into solar time before the hour pillar is drawn, and the size of that correction depends on where a country sits relative to its meridian.

Korea sets its clocks by 135 degrees east while the peninsula sits nearer 127.5, so the correction is about thirty minutes and much the same wherever you were born: Gangneung on the east coast and Mokpo in the south west differ by roughly ten minutes of solar time. China runs one clock across some sixty degrees of longitude, so the correction is almost nothing in the east and over two hours in the far west, and when it is noon in Beijing the sun over Urumqi is at about ten in the morning. Japan's standard meridian passes through the country, near Akashi, so Tokyo runs about nineteen minutes ahead of the clock and Fukuoka about eighteen behind.

That is a difference of geography mistaken for one of belief. A Chinese practitioner who insists on your city is being sensible, because there the hour pillar can move two places; a Korean reader who asks more casually is too, because here it moves minutes. Not everyone applies the correction at all, and one objection holds that correcting strictly by longitude ought to move the solar term moments too. I apply it and say so when it has changed anything.

Both countries also had daylight saving time. The page on why readings differ works through all five splits.

When does the day change, and why is the Korean answer half an hour later?

All three argue about the same two hours, the ja hour from eleven at night to one in the morning, and a whole pillar rides on the answer. Chinese practice is openly split between turning the date at eleven and holding it to midnight, a debate framed as early ja against late ja, with no settled position. Japanese practice commonly starts the birth day at eleven. Korean almanacs put the turn at half past eleven.

The Korean half past eleven is not a third position. It is the eleven o'clock position with the thirty minute longitude correction already folded in, so 11:30 on a Korean clock is eleven o'clock by the sun. Korea and Japan are saying one thing in two clocks, and the real disagreement is the two way one inside China.

This is the one place a Korean and a Chinese chart for the same person genuinely differ. Born in those two hours, your day pillar can land on either of two days, so two of the eight characters change and one is the day master, the stem the reading is anchored to. Ask any reader which convention they use.

Do Korean and Chinese readers mean the same thing by the same words?

Mostly yes, under different names. The ten relationships between the day master and the rest of the chart are shi shen in Chinese and tsuhensei in Japanese, while Korean uses sipseong, sipsin and yukchin depending on whether the reader treats them as functions or family roles.

The stems trip people up, because each language hides something different. Gap (甲), yang wood, is Jia in Chinese; Byeong (丙), yang fire, is Bing; Gye (癸), yin water, is Gui. The trap is that each collapses a different pair. Korean romanisation writes both 辛, the yin metal stem, and 申, the monkey branch, as Sin. Pinyin stripped of its tone marks, which is how it usually reaches English, writes both 戊, the yang earth stem, and 午, the horse branch, as wu. An English chart is ambiguous at a different point depending on the language it came through, so ask for the character, not the spelling, and the chart reading page lays the eight out in order.

How much weight do sinsal get in each tradition?

This is the clearest difference of emphasis and the one most often overstated. Sinsal (신살), the symbolic stars, are shensha in Chinese: named patterns such as the peach blossom or the travelling horse, each with a meaning of its own. The classical Ziping stream in China pushed them to the margin, holding that the eight characters and the relations between them carry the reading and that the star lists accumulated later. Korean practice kept them in far wider circulation, to the point where several sinsal names are everyday Korean words.

It is a tendency, not a border. Plenty of Chinese practitioners use shensha freely, plenty of Korean readers barely touch them, and sources disagree on how wide the gap is. My own use is conditional: a symbolic star tells me where to look, then the eight characters and the cycles have to agree before it goes in the report. The sinsal page takes the common ones one at a time.

Why do Koreans check gunghap before a wedding?

Because in Korea it was part of the procedure rather than a private curiosity. China has its own marriage reading, hehun, but the Korean version sat inside the formal negotiation between two families: the groom's side sent over a saju danja, a paper carrying his four pillars, and the bride's family took it to a reader before answering with a wedding date.

The custom has faded, and now survives partly as a graceful way of declining a match without anyone giving a reason. Gunghap (궁합) splits into a surface reading from the branches and a deeper one from the five elements, and it shows where one chart supplies what the other is thin in, where the two double up on an element and are likely to compete, and which years press on both at once. That is information two people can use, and it is not a verdict. The gunghap page goes further into it.

Is a reading the same experience in Seoul and in Taipei?

Often not, though these are habits rather than doctrine. Chinese language practice leans towards the consultation: an appointment, a specific question about a move or a business, an answer inside the hour. Korea has that too, at the cheolhakgwan and in the saju cafes, and it also has a seasonal habit with little equivalent elsewhere: many Koreans have a reading at the new year whether or not anything is wrong.

One warning these comparisons never give: that new year reading is often not saju at all. Tojeong Bigyeol, a nineteenth century Korean book of annual fortunes, is a separate practice with its own method, so check which it was before comparing notes with a Korean friend.

Written reports are a recognised Korean format, and the one I work in. I build the eight characters into a single scene of natural images, and write it out at length. That is my own choice, not a national standard; the sample report shows its shape.

So which tradition should I go to, and what can this page not tell me?

I will not rank them, and I would treat anyone who does with caution. The three read the same eight characters and have never been tested against one another in any way that would settle the question. Choose on practical grounds: whether you can follow the reader's vocabulary, whether they will name the conventions they work in, and whether the questions they usually answer are yours. Weighing this against a Western birth chart is a different comparison again.

What this page cannot do is say anything about you. It is general explanation of how two traditions handle a shared method. Which convention moves your own day pillar depends on your exact date, time and birthplace, and what your chart is short of depends on all eight characters and the cycles passing over them, which the five elements page describes. A reading gives tendencies and timing in one chart, a smaller claim than these pages usually make and the only one I will stand behind.

Common questions

Is saju just the Korean word for bazi?
It is the Korean name for the same four pillar method, taken from a different part of it: saju means four pillars, while bazi means eight characters. The chart, the calendar and the solar terms are identical. What differs between the traditions is lineage, vocabulary, how much weight the symbolic stars carry, and the conventions around the day boundary and clock correction.
Will a Korean reader and a Chinese reader give me the same eight characters?
Usually yes, and where they differ there is normally one reason. A birth between eleven at night and one in the morning can land on either of two days depending on which day boundary convention is used, and a birth near an hour boundary can move if one reader applies a longitude correction and the other does not. Ask each of them which conventions they work in and the difference explains itself.
Do I need to know my birthplace for a Korean reading the way I would for a Chinese one?
It helps, but it matters far less. Korea is narrow from east to west, so the solar time difference between the east coast and the south west is about ten minutes, which changes nothing unless your birth time sits right on an hour boundary. China spans a single time zone across sixty degrees of longitude, where the same correction can run past two hours, so a Chinese practitioner has to ask.
Is Korean saju better than Chinese bazi, or the other way round?
Neither claim can be supported, and since the two traditions read the same eight characters there is not much for such a difference to sit in. The useful question is which reader explains what they are doing, tells you which conventions their chart is built on, and answers the kind of question you came with. Saju and bazi both describe tendencies and timing rather than settled outcomes.

Read next

What does your chart look like as a landscape

Reports are written in Korean today.
An English edition is in preparation.

One email when English readings open. Nothing else.

Waitlist