卷:卷二《长安月下·关中六十夜》|主角:Marco·强东林(独行之夜)|地点:延安·洛川县·塬上·强家院子|典:《大学》“物有本末,事有终始,知所先后,则近道矣”、《论语·子路》“无欲速,无见小利”|主导感官:味(新摘苹果的涩、砖茶、旱烟)|过所进度:一字·「客」|金融-国学对应:Unit Economics(一笔账算到底,才知道钱漏在哪一格)↔ 物有本末,事有终始

段一·现身

那本账,是记在一个化肥袋子上的。

第七十一夜。

从第七十夜到这一夜,中间隔了三个月零九天。

马可后来才想明白这件事的意思。

卷一在成都,六十夜是连着的,一夜接一夜,中间一天没断。因为那六十夜他学的是怎么做一个人,而做人这件事,只要你在,天天都能学。

这一卷不是。

这一卷他学的是怎么做一件事,而事情是有季节的。

苹果十月才熟。

所以他在七月回了成都,八月在龙泉山,九月在锦心号,十月坐飞机回来。

第七十一夜,是十月十七。

他在飞机上想过一个问题:这算不算第七十一夜?

中间隔了三个月,杜甫一次也没出现过。他在成都的那九十几天,跟卷一之后的日子没什么两样,开会、看账、跟小周吃了两次饭、去了四趟龙泉山。

他甚至有点怀疑西安那十夜是不是真的。

可他每天早上换衣服的时候,都会摸一下上衣口袋。

那枚铜牌一直在。

上面那个「客」字,一天也没有淡过。


塬上比三个月前冷得多。

他到强家的时候是下午四点。院子里晒着苹果,不是好果,是套袋没套好、有斑的那些,晒成干。

强东林蹲在门口抽烟。看见马可,他站起来,把烟在鞋底上摁灭了。

“你真来了。”

“我说了十月来。”

他说得很轻,可强东林明显愣了一下。

后来马可才知道,这三个月里,强磊问过他爸好几次:那个外国人还来不来。

强东林每次都说:来不来的,人家那么大的老板。

他嘴上这么说,可十月一到,他每天都往院门口看两回。

强东林盯着他看了两秒,忽然咧嘴笑了:“进屋。”

进了屋,他媳妇端上来一壶茶,砖茶,很酽,有点苦。

马可喝了一口。

“老强,”他说,“咱今天什么也不干。”

“啊?”

“就算一笔账。”马可说,“去年那一笔。从头到尾,一格一格算。”

强东林愣住了:“就这?”

“就这。”

“那你跑这么远……”

“我就为这个来的。”

强东林看了他很久,忽然转身进了里屋。

出来的时候,他手里拿着一个东西。

一个化肥袋子。

洗干净的,剪开,铺平,折成四折。上面用圆珠笔写满了字。

“我的账。”强东林说,“都在这上头。”


段二·古文镜像

“物有本末,事有终始,知所先后,则近道矣。”
English Mirror: Things have roots and branches; affairs have ends and beginnings. Know which comes first and which comes after, and you are near the Way.

马可把那张化肥袋子铺在炕桌上。

字很密。有的用圆珠笔,有的用铅笔,还有几处是用红笔改过的。写得歪歪扭扭,可每一行都有数。

他看了很久。

这不是账。这是一堆数。

他翻到背面,背面也有。字更小,挤在一起。

有一行写着:“九月廿三,客商来,八毛,没卖。”

下一行:“九月廿五,第二个客商,八毛五。”

再下一行,那一行被红笔圈了起来:“卖了。

就两个字。

马可盯着那两个字看了很久。

一个人在一张化肥袋子上,用红笔圈出了自己一年里最重的那个决定。

他能看懂上面的每一个字,可他看不出一件事:这一年到底挣了多少。

“老强,”他说,“咱重新记一遍。”

“咋记?”

马可从包里掏出一个本子和那支朱笔,他没舍得用那支笔写字,换了一支普通的。

他在本子上画了一条竖线。

“左边写钱出去,右边写钱进来。”他说,“咱一样一样过。”

“从哪儿开始?”

马可想了想。

“从去年这个时候。”他说,“从你去年摘完苹果那一天开始。”

强东林点了一根烟。

“行。”


段三·事件主体

那笔账算了六个小时。

马可后来一直记得这六个小时。因为在这六个小时里,他做的事只有一样:问,然后写。

他一句道理也没讲。


第一个小时,算的是地。

“十二亩。”强东林说,“红富士。”

“都是你自己的?”

“九亩是我的。三亩是租的,我三叔的地,他儿子在西安,不种了。”

“租金多少?”

“一亩八百。”

马可写下:三亩 × 八百 = 二千四。

“租金什么时候给?”

强东林顿了一下。

“年底。”

“哪个年底?”

“……卖完苹果给。”

马可停下笔,抬起头。

“老强,这意思是,你要是苹果没卖好,你三叔的租金就晚给?”

“那能咋办。”强东林说得很自然,“亲戚嘛。”

马可在旁边打了一个小小的记号。

这一笔,是一个隐形的贷款。

三叔在替他垫一年的资金成本,一分利息不收。

而这笔钱,在任何一张报表上都不会出现


第二个小时,算的是投入。

“化肥。”

“一年多少?”

强东林算了半天:“底肥、追肥、叶面肥……大概九千。”

“十二亩九千?”

“嗯。”

“农药呢?”

“套袋前打三遍,套袋后还得打。六千吧。”

“套袋呢?”

强东林忽然认真了:“这个大。”

“多少?”

“一亩地一万两千个袋。一个袋子连工带料,一毛二。”

马可算:十二亩 × 一万二 × 零点一二 = 一万七千二百八。

他停住了。

“老强,这一项比化肥农药加起来还多。”

“对。”强东林说,“套袋最费钱,也最费人。”

“人从哪儿来?”

“雇。”强东林说,“套袋那二十天,全村都在雇人。一天一百五,管一顿饭。”

“你雇几个?”

“六个。二十天。”

马可算:六 × 一百五 × 二十 = 一万八千。

他抬起头:“老强,那这一万七千二百八里,包不包括这一万八?”

强东林愣住了。

“……不包括。”

“那你刚才说一毛二是连工带料?”

“那个工是,说的是买袋子的钱里头有个加工费。”强东林挠头,“雇人的钱是另算的。”

马可把这两笔分开写。

套袋物料:一万七千二百八。 套袋人工:一万八千。

强东林盯着这两个数看了很久。

“这么多?”

“老强,你以前算过这个吗?”

“算过。”强东林说,“可我没这么摆着算过。”


第三个小时,算的是他自己。

马可写下一行:你自己的工。

强东林笑了:“我的工不算钱。”

“为什么?”

“我自己的地,我自己干,算什么钱。”

马可放下笔。

“老强,”他说,“咱做个算术题。”

“你说。”

“你一年在这十二亩地上,干多少天?”

强东林想了想。

“修剪一个月。疏花疏果二十天。套袋二十天,那二十天我也在干。打药十来次。摘果十五天。冬天还得刨树盘、涂白。”

他数着手指:“加起来,一百八十天有。”

“一天算多少钱?”

“……不算。”

“老强,你村里雇人一天多少?”

“一百五。”

“那你自己这一百八十天,按一百五算,是多少?”

强东林不说话了。

马可自己算了:一百八 × 一百五 = 二万七千。

他把这个数字写下来,在旁边画了一个圈。

屋里安静了很久。

强东林把烟摁灭了。

“兄弟,”他说,“你这一算,我这一年白干了?”


马可摇头。

“先别急着下结论。”他说,“咱把钱进来的那边也算完。”

第四个小时,算的是收入。

“去年打了多少斤?”

“三万八千斤。”强东林说得很准,这个数他记得清楚。

“一亩三千一百多斤?”

“对,那年下了冰雹,掉了一茬。好年景一亩四千。”

“卖了多少钱?”

“客商来收,八毛五。”

马可算:三万八千 × 零点八五 = 三万二千三百。

他把这个数写在右边。

然后他把左边的数加起来:

租金 二千四 化肥 九千 农药 六千 套袋物料 一万七千二百八 套袋人工 一万八千 其他(水电、机械、运费)他问了半天,强东林说大概四千

左边合计:五万六千六百八十。

右边:三万二千三百。

马可放下笔。

他和强东林两个人,一起盯着这两个数看。

“兄弟,”强东林的声音有点抖,“这不对吧。”

“哪儿不对?”

“我要是一年赔两万四,我早不种了。”

马可点头。

“老强,你说得对。这个算法有问题。”

“哪儿有问题?”

马可把笔倒过来,指着“套袋人工 一万八千”那一行。

“这一万八,”他说,“你是怎么给的?”

强东林愣了一下。

“……换工。”


马可放下了笔。

“老强,什么叫换工?”

“就是,我给他家套,他给我家套。”强东林说,“咱这儿都这样。套袋那二十天,全村互相帮。”

“不给钱?”

“不给。”

“那你刚才为什么说一天一百五?”

“那是雇外人的价。”强东林说,“外头来的,一天一百五。村里换工,不给钱,管饭。”

马可的心跳得很快。

“老强,你去年雇了几个外人?”

“两个。”强东林说,“另外四个是换工。”

马可把那一行划掉,重新写:

套袋人工:外雇二人 × 二十天 × 一百五 = 六千。

换工四人 = 零。

他抬起头:“老强,那你欠了村里多少工?”

强东林一下子答不上来。

“我……我也给他们家干了啊。”

“干了多少天?”

“……差不多吧。”

马可盯着他:“老强,‘差不多’是多少?”

强东林抓着头,想了很久。

“我给王家干了三天,给李家干了五天,给我二哥家干了……”

他忽然停住了。

他给别人干的,是十一天。

别人给他干的,是八十天。


屋里安静得能听见外面的风。

强东林坐在炕沿上,手里的烟燃到了指头,他才反应过来,甩了甩。

“兄弟,”他慢慢地说,“这……这咋算?”

马可没有立刻回答。

他在本子上画了一个新的格子。

“老强,”他说,“咱先别管这个人情账。咱先把它记下来。”

他写:换工净欠:六十九个工日。

“这个不用今年还。”他说,“可它在。”

强东林盯着那行字,忽然说了一句话,那句话让马可后来在本子上写了整整一页:

“兄弟,我这辈子记账,从来没记过这一格。”

“可我年年都在还。”

马可放下笔,问了一个他自己也不确定该不该问的问题。

“老强,要是有一年,你还不上呢?”

强东林抬起头。

“比方说你病了,或者出门了,那一年套袋的时候你去不了别人家。”

强东林想了想,答得很平常:

“那就欠着。”

“欠多久?”

“欠到能还。”

“要是一直还不上呢?”

屋里安静了一下。

强东林把烟卷在手里转了两圈。

“那我儿子还。”他说。

马可怔在那里。

这笔账,是可以传给下一代的。

他这半辈子做的所有信贷模型,都有一个期限。三年,五年,十年。到期就结清。

而这一笔,没有期限,它跟着一户人家走,一代传一代。

它没有利息,可它也从来不会被免除


第五个小时,算的是那八毛五。

“老强,客商为什么给八毛五?”

“市场价。”

“谁定的市场价?”

强东林愣了:“……就是市场价啊。”

“你能不卖吗?”

“不能。”

“为什么?”

强东林伸出七个手指。

“摘下来七天。”他说,“七天以后就发面了。”

马可点了点头。这个数他记得,七月那次来,强东林跟他说过。

“那七天里,来了几个客商?”

“三个。”

“价格差多少?”

强东林想了想:“一个给八毛,一个给八毛五,一个给九毛,可给九毛那个只要一级果,二三级果他不要。”

马可停下笔。

“老强,那你去年一级果多少?”

“大概一半。”

马可开始在纸上画。

如果卖给九毛那家:一级果一万九千斤 × 零点九 = 一万七千一百。 剩下一万九千斤二三级果,只能等,可只剩四天了。

他抬起头。

“老强,你去年是不是算过这个?”

“算过。”强东林说,“我算了两天。”

“最后为什么选了八毛五那家?”

强东林把烟又点上了。

“因为他全要。”


马可在本子上写下了这一夜最重要的一行字。

他不是在选价格。他是在选,确定性。

那多出来的五分钱一斤,是不确定性的价格

一万九千斤二三级果,如果四天内卖不出去,就是零。

为了避免那个零,他放弃了每斤五分钱。

三万八千斤 × 五分 = 一千九百块。

他花一千九百块,买了一份保险。

而这份保险,没有任何人给他定过价,也没有任何人告诉他这是保险。

他忽然想起自己在纽约做过的一件事。

他们组卖过一种结构化产品,本质就是这个:客户放弃一部分上行,换取下行的保护。他们管它叫“领口策略”。

那种产品的说明书有四十七页。定价模型用的是布莱克-舒尔斯,波动率取的是三十日隐含。他们收百分之一点二的费。

每一个买那个产品的客户,都清楚地知道自己在买保险。

而这个塬上的人,每一年、每一户、每一次开秤,都在做同一笔交易。

没有说明书。没有定价。没有人告诉他们那叫什么。

他们只知道一句话:那个客商全要。

“老强,”马可说,“你去年那个决定,是对的。”

强东林抬起头:“真的?”

“真的。”马可说,“换了我,我也这么选。”

“可你要是……”他停了一下,“可你要是有一个库呢?”

强东林的手停在半空。


第六个小时,算的是那个库。

“老强,你前年存过库,对吧?”

“存过。”强东林的脸沉了下来,“赔了。”

“讲一遍。”

“前年一块七收的,我存了一万五千斤。库费一万。到二月,市场跌到一块四。我少卖了九千。”

“加上库费一万,”马可说,“一共赔一万九。”

“对。”

马可在纸上写:前年存库净亏 一万九千。

然后他写下一行:如果前年不存库:一万五千斤 × 一块七 = 二万五千五。

存了库:一万五千斤 × 一块四 = 二万一千,减库费一万 = 一万一千。

他把这两个数并排放着。

强东林看着,脸色越来越难看。

“兄弟,”他说,“你别算了。”

“老强,我得算。”

“为啥?”

马可放下笔,看着他。

“因为你这三年,一直觉得自己那年做错了一个决定。”

强东林不说话。

“可你没做错。”马可说,“你只是,赌输了一次。”

“这俩有啥不一样?”

马可想了很久,才说:

“做错了,是你不该那么干。赌输了,是你该那么干,可这一次没赢。”

他指着那两个数:“老强,前年二月要是涨到两块呢?”

强东林愣住了。

“那我就……”

“那你就挣两万。”马可说,“同一个决定。同一个你。”

“不一样的是那年二月的价,而那个价,你定不了,我也定不了,客商也定不了。”

强东林盯着那两个数,忽然问:

“兄弟,那你说,我该不该存?”

马可摇头。

“我不知道。”

“你不知道?”

“我真不知道。”马可说,“我要是说该存,那我就是在替你赌。赢了算你的,输了也算你的。”

他把笔放下:“老强,我今天不给你出主意。”

“那你能给我什么?”

马可想了很久。

“我能给你一样东西。”他说,“我能告诉你,你那一次赌的是什么。

“那年你赌的是二月的价会比十月高,高到能盖住一万块库费。”

“这是一个明明白白的赌。它有可能赢,也有可能输。”

他指着那张化肥袋子:“可你这三年,一直以为自己那年是做错了事。”

“错事和输赌,人的反应是不一样的。”

“做错了事,人会改。”

“输了赌,”马可停了一下,“输了赌的人,会不敢再赌。”


强东林在炕沿上坐了很久。

天已经黑了。他媳妇进来开了灯,又出去了。

“兄弟,”他忽然说,“你今天来,是不是要跟我谈合作?”

马可摇头。

“不谈。”

“那你算这一天,图啥?”

马可看着那张化肥袋子,看着自己本子上密密麻麻的六页字。

他忽然想起七月那一夜,司马迁在未央宫的土台上跟他说的话。

善者因之。

“老强,”他说,“我今天什么也不图。”

“我就想知道,你这一年,钱到底是怎么进的,又是怎么出的。”

“为啥想知道?”

马可老老实实说:“因为我不知道。”

“我在成都做了七个村,那是桃子。桃子跟苹果不一样,四川跟陕西不一样,龙泉山跟这个塬也不一样。”

“我要是不知道你这笔账,我给你出的任何主意,都是,瞎说。”

强东林盯着他看了很久。

然后他做了一件马可完全没料到的事。

他站起来,走到墙边,把墙上挂的一个日历翻了过来。

日历的背面,也写满了字。

“兄弟,”他说,“这是我前年的账。”

他又翻出一张:“这是大前年的。”

三张纸,三年。

“我从来没给人看过。”强东林说,“我媳妇都不知道我记这个。”

马可看着那三张纸,一张化肥袋子,一张日历背面,一张水泥袋。

他忽然眼眶发热。

这是一个人,用三年,用三张废纸,问了一个他自己回答不了的问题。

“老强,”他忍不住问,“你为什么记这个?”

强东林蹲下来,把那三张纸摊在地上,一张一张摆平。

“我爹不记。”他说,“我爹种了一辈子地,从来不记账。他说记那个干啥,能吃能穿就行了。”

“那你为什么记?”

强东林用手把纸上的褶子抹平。

“我不服。”他说。

马可怔住了。

“我不服什么呢?”强东林自己也说不太清,“我就是觉得,我这一年干下来,累成那样,到年底手里剩不下几个钱。我想知道是哪儿出了岔子。”

“记了三年,找到了吗?”

强东林摇头。

“没找到。”他说,“我数是记下来了,可我看不出个道道。”

他抬起头看着马可:“今天你这一算,兄弟,我看出来了。”

“你看出什么了?”

强东林指着本子上那个被圈起来的数字:二万七千。

“我最大的一笔本钱,”他说,“是我自己。”


马可在本子上写:Six hours in a courtyard on the Luochuan plateau. Nothing sold, nothing proposed, no terms mentioned. I asked and I wrote down the answers. That is the entire content of night seventy-one.

The ledger came out of the inner room on a washed-out fertiliser sack, ballpoint over pencil with corrections in red. It is not a ledger. It is a pile of numbers that cannot answer the only question it was made to ask.

What the six hours surfaced, in order:

The uncle's rent, 2,400, paid after the fruit sells. That is an interest-free one-year working capital loan from a relative, and it appears on no statement anywhere. Every smallholder economy I have ever modelled has this line missing.

Bagging: 17,280 in materials, more than fertiliser and chemicals combined. I nearly booked 18,000 of labour against it before he corrected me — two hired hands at 150/day, four on 换工, reciprocal village labour, unpaid. So the cash line is 6,000, not 18,000.

And then the thing that stopped the room. He owes eighty days of received labour against eleven days given. Net sixty-nine man-days outstanding. His words, and I want them recorded exactly: I have never once written down that line, and I have been repaying it every year of my life.

His own labour: 180 days, unpriced, because "it's my own land." At the village hire rate that is 27,000 — larger than any input line on the sheet. When I wrote it down he asked whether he had worked a year for nothing.

He had not. But nobody had ever shown him the number.

On the 0.85. Three buyers in the seven-day window: 0.80, 0.85, and 0.90 for grade-one only. He took 0.85 because that buyer took everything. He was not choosing a price. He was buying certainty — 1,900 yuan, or five fen a jin, to avoid the possibility that 19,000 jin of grade two and three went to zero with four days left. That is an insurance premium. Nobody priced it for him and nobody told him he was buying it. His decision was correct and he has never been told that either.

On the cold store two years ago: bought at 1.70, stored 15,000 jin, 10,000 in fees, market at 1.40 by February, down 19,000 all-in. He has carried that as a mistake for three years. It was not a mistake. It was a bet that lost. He had never heard anyone distinguish the two, and when I said the same decision makes 20,000 if February prints 2.00, he could not answer.

Then he turned over the calendar on the wall. The reverse side is the year before. Then a cement bag: the year before that. Three years of accounts on three pieces of rubbish, and he has never shown them to anyone, including his wife.

That is a man who has been asking a question for three years with no way to phrase it.

I did not mention Jinxin Hao. I did not mention terms. Sima Qian's instruction in July was: go in October, settle that ledger, and do nothing else. Followed exactly. — M.


段四·钩子结尾

九点多,马可要走了。

强东林送他到院门口。

“兄弟,”他忽然说,“你等一下。”

他跑回屋,出来的时候手里拿着那张化肥袋子。

“给你。”

马可怔住了:“老强,这是你的账。”

“我记得住。”强东林说,“你拿去。”

他把那张纸塞进马可手里,忽然又说了一句:

“兄弟,你今天跟我说的那句,‘赌输了’跟‘做错了’不一样,那句话,我想了一下午。”

“怎么?”

强东林蹲下来,又点了一根烟。

“前年赔了那一万九以后,我媳妇没说我,我儿子没说我。”他说,“可我自己,三年没敢再存库。”

“今年呢?”

“今年也不敢。”

他抽了一口烟,抬起头:

“可我今天忽然想明白一件事。”

“什么?”

我这三年,不是不敢存库。

“是我不敢再做一次自己拿主意的事。”

马可站在院门口,忽然说不出话。

“前年那一下,把我打怕了。”强东林说,“从那以后,客商说八毛五,我就八毛五。他说八毛,我也八毛。”

“反正我自己拿主意,就输。”

他把烟摁在地上。

“兄弟,我今年五十四。我还能种十来年。”

“我不想再这么种十年。”

他说完这句,两个人在院门口站了很久,谁也没再开口。

塬上的风把院里晒的果干吹得沙沙响。

最后是强东林先动的。他站起来,拍了拍裤子上的土。

“兄弟,你啥时候再来?”

马可想了想。

他本来想说“明年春天”。可他忽然改了口。

“老强,我这次不走。”

强东林愣住了:“不走?”

“我在县城租了个房子。”马可说,“租了三个月。”

“你租房子干啥?”

“我想在这儿住到年底。”

强东林盯着他看了很久,忽然说:

“兄弟,你住到年底,也帮不了我什么。今年的果都卖完了。”

“我知道。”

“那你住着干啥?”

马可站在黑暗的院门口,很实在地答:

“我想看看你们这儿的冬天。”


马可从塬上下来的时候,是晚上十点。

他站在路口等车,四周全是黑的。远处塬下有一小片灯,是县城。

风很大,天上星星特别多。

他掏出手机,打开那个叫“错题簿”的文件。

打了第四条:

第四错:我七月来的时候,以为我该说服他。 谁发现的:我自己,今天。 为什么没有早点发现:因为我一直以为我手上有一个答案。今天我才知道,我手上连他那笔账都没有。一个连账都没看过的人,凭什么有答案。

打完,他又新建了一条:

第五错:我这十年做的所有模型,从来没有一格叫“换工”。 谁发现的:强东林。 为什么没有早点发现:因为那一格里没有钱流动。而我们那套账,只认钱流动的东西。可他一年里最大的一笔支出,是六十九个欠着的工。


他抬起头,忽然发现路口站着一个人。

杜甫。

老诗人穿着那件旧青布袍,站在风里,手里拎着一个布包。

“先生。”

“老弟。”

马可忽然有点急着想说什么。他这一天憋了太多东西。

“先生,我今天,”

“我知道。”杜甫说,“我一直在。”

马可怔住了。

“您在?”

“在院子里。”杜甫说,“你们算账的时候,我坐在门槛上。”

他望着黑暗中的塬:“老弟,你今天做对了一件事。”

“哪一件?”

你一句道理也没讲。

马可低下头。

“先生,我不是不想讲。”他老实说,“我讲不出来。我今天才知道,我以前那些道理,全是在不知道他这笔账的情况下讲的。”

杜甫点了点头。

“那就对了。”

他把手里那个布包递过来。

“这是什么?”

“干粮。”杜甫说,“明夜路远。”

马可接过来,那布包沉甸甸的,是几个还热的馍。

“先生,明夜去哪儿?”

杜甫往西边看了看。

“你今夜算清了一笔账。”他说,“可你算完之后,能干什么?”

马可怔了怔。

“……我一个人干不了什么。”

“对。”杜甫说,“所以从明夜起,是第二关。”

“第二关是人。”

他往那条黑路上走去,声音在风里传过来:

明夜我带你去见一个人。他一辈子只有三十六个人,可他把整个西域,定了三十一年。


欲知后事如何,请看下回,《三十六人定西域》

附录一·Dongpo's Note

Foreign friend — Du Fu, and I sat on the threshold of that courtyard for six hours without either of you knowing.

You did one thing correctly tonight and it is not the arithmetic. You did not explain anything.

In July you came to this province with thirty-seven pages and a method, and you have spent ten nights being taken apart for it. Tonight you arrived with a notebook and asked a man how his money comes in and how it goes out, and then you wrote down what he said. That is all. The Great Learning says: things have roots and branches, affairs have beginnings and ends, and knowing the order is most of the Way. You have finally got the order right — the account comes before the advice, and it took you three months and a great deal of humiliation to learn a sentence a child can recite.

What you turned up in that room is worth more than the whole of your research. The uncle's rent, deferred without interest, which is a loan nobody records. The sixty-nine days of borrowed labour he has been repaying his entire adult life and has never once written down. His own hundred and eighty days, priced at nothing because the land is his. And the five fen a jin he paid to the buyer who would take everything — an insurance premium, correctly bought, which no one has ever told him was insurance or that he was right to buy.

But the thing to hold is what he said at the gate. He has not stored fruit for three years, and it is not the cold store he is afraid of. He is afraid of being the one who decides. One bad February and a man stops deciding, and after that the buyer says eight-five and he says eight-five.

That is what you are actually up against here, and no ledger shows it. He is fifty-four. He has ten years of planting left in him and he has just told a stranger he does not want to spend them that way.

Now the harder part. You can settle one man's accounts in six hours. There are four hundred households on that plateau alone. You cannot sit in four hundred courtyards, and the second gate begins tomorrow for that reason. It is people. Not persuading them — having them.


附录二·Marco's Memo

Memo to file. Qiang household, Luochuan plateau, 16:00–22:00. Night seventy-one, 17 October.

Subject: unit economics of twelve mu of Red Fuji, reconstructed from a fertiliser sack.

Method: asked and recorded. No product mentioned, no terms discussed, no attempt to close anything. Per instruction.

Findings, in order of how badly I had them wrong:

1. Invisible financing. Three mu rented from an uncle at 800/mu, paid after the crop sells. A twelve-month interest-free working capital facility from a relative, extended annually, recorded nowhere. Correct treatment: this is the cheapest capital in the system and it is entirely relationship-collateralised. Any product I bring competes against a zero-rate incumbent lender who cannot foreclose.

2. 换工 — reciprocal village labour. I had booked 18,000 of bagging labour. Cash cost is 6,000; the other four workers are paid in kind. Then the number that stopped the room: he has given eleven days and received eighty. Net sixty-nine man-days outstanding. His comment, verbatim: I have never written that line down, and I repay it every year. This is a liability with no maturity, no interest, no documentation, and near-absolute enforcement. It is the largest obligation on his balance sheet and it does not exist in any agricultural model I have built or read.

3. Own labour: 180 days at the local hire rate is 27,000, exceeding every input line. Unpriced because the land is his. When priced, the year looks like a loss. When not priced, the year looks fine. Both statements are true and the difference is the whole question of whether this is a business or a way of living.

4. The 0.85 decision, reclassified. Three buyers in the seven-day window: 0.80 (all), 0.85 (all), 0.90 (grade one only). He took the all-taker. Not a price choice — a purchase of certainty at 1,900 yuan against the tail risk of 19,000 jin going to zero with four days of shelf life left. He bought insurance, correctly, unpriced and unnamed. He has never been told he was right.

5. The cold store, two years ago. Bought 1.70, stored 15,000 jin, 10,000 in fees, sold at 1.40. Down 19,000. He has carried it as an error for three years. It was a bet that lost. I told him the distinction and he had never heard it. Same decision, 2.00 in February, up 20,000.

6. The finding that matters more than all five. At the gate: he has not stored since, and it is not the store he is avoiding. He is avoiding being the person who decides. Since that February the buyer names a price and he takes it. He is fifty-four, has roughly ten seasons left, and told me he does not want to spend them like that.

That is not a financing problem. Whatever we build has to give a man back the standing to make a call — which is, I now notice, what my ten characters have said all along without my understanding them.

He gave me the fertiliser sack. He then produced two more years off the back of a wall calendar and a cement bag. Three years of accounts kept on rubbish, shown to nobody, including his wife.

Error book entries four and five logged at the roadside.


附录三·English Highlight

"Things have roots and branches; affairs have ends and beginnings. Know which comes first and which comes after, and you are near the Way."
The line from the Great Learning is usually read as counsel about priorities. On the Luochuan plateau it turns out to be a procedural instruction: the account comes before the advice.
Marco arrived in Shaanxi in July with thirty-seven pages of provincial research containing no individual's name, and was dismantled for it across ten nights. He returned in October with a notebook, sat in one courtyard, and spent six hours asking a fruit grower how money entered and left his year — proposing nothing.
What surfaced does not appear in agricultural models. An uncle's rent, payable only after the crop sells: an interest-free annual working-capital loan, collateralised entirely by kinship. Eighty days of village labour received against eleven given — sixty-nine man-days outstanding, the largest liability the household carries, undocumented, uncharged, and repaid faithfully for a lifetime. A hundred and eighty days of the grower's own work, priced at nothing on the reasoning that the land is his.
And a decision the grower had judged himself for over three years. Offered 0.90 for grade-one fruit alone or 0.85 for everything, he took 0.85 — spending 1,900 yuan to eliminate the chance that 19,000 jin of lower-grade fruit reached zero inside a seven-day shelf life. He bought insurance. Nobody had priced it, named it, or told him he was right.
The sharpest exchange came at the gate. He has not used a cold store since a bad February two years ago, and he is not avoiding the store. He is avoiding being the one who decides. Since then the buyer names a number and he accepts it.
Which reframes the entire problem. What is missing on that plateau is not capital. It is the standing to make a call and survive being wrong once.

附录四·Dongpo's Cheat Sheet

今夜之事 出处 现代对应 一句话
化肥袋子记账 强家三年账 数据不等于账 每一个字都看得懂,可它答不了自己那个问题
三叔的租金年底给 洛川租地惯例 隐形零息贷款 最便宜的那笔钱,任何报表上都没有
换工净欠六十九个工 关中互助习俗 表外负债 他年年在还一笔他从没记过的账
自己一百八十天不算钱 农户成本习惯 机会成本 算了像亏本,不算像还行,两句都真
八毛五 vs 九毛 七天窗口三个客商 确定性溢价 他买的不是价格,是保险,而没人告诉他这叫保险
存库赔一万九 前年冷库 赌输≠做错 同一个决定,二月两块就是挣两万
三年不敢再存 院门口那句话 决策权的丧失 他怕的不是库,是再当一次拿主意的人
一句道理没讲 杜甫评语 先账后策 连账都没看过的人,凭什么有答案