卷：卷二《长安月下·关中六十夜》｜主角：Marco·强东林·强东林媳妇·强磊·李叔·八十一岁的老太太｜古人：贾思勰（说《齐民要术》）｜地点：洛川·强家十二亩果园｜典：《齐民要术·序》“顺天时，量地利，则用力少而成功多”；《养羊篇》“余尝畜羊二百口……一岁之中，饿死过半”｜主导感官：触（一整天捏果柄，第三天手指起茧）｜过所进度：四字·「客」「侣」「本」「渠」｜金融-国学对应：The Constraint Is Labour, Not Knowledge（人人都知道要疏果；卡住的从来不是知识，是十五天里凑不出五十六个工）↔ 顺天时，量地利
段一·现身
四月二十一日，早上六点二十。
塬上。
花落完了大概十天。
每一个花序底下，坐了三到五个小果。
每个小果，比黄豆大一点。
强东林递给马可一把剪子。
很小，比指甲刀大不了多少。
刃口是弯的。
“这个咋用？”
“不用。”
“那给我干啥？”
强东林说：“给你壮胆。”
“那用啥？”
强东林伸出右手，用拇指和食指做了一个动作。
很轻，往外一推。
“手掐。”
“为啥不用剪子？”
“剪子快，可你剪不准。”
他往树上一指。
“一个花序五个果，你要留哪个？”
马可看了半天。
五个果，长得几乎一模一样。
“中间那个。”
强东林摇了摇头。
“中间那个叫中心果，多数时候留它。”
“那什么时候不留？”
“它要是歪了，或者屁股上有锈，就不留。”
“那留哪个？”
“留最壮的那个侧果。”
“怎么算壮？”
强东林把一根枝子拉下来。
“你看果柄。”
马可凑过去。
五个小果，柄的粗细确实不一样。
最粗那个，比最细那个粗了大概一半。
“柄粗的，往后长得大。”
“为啥？”
“柄是管子。管子粗，走的东西多。”
马可伸手，掐了第一个果。
那一下用力过了。
果掉了，可连着的那根短枝也断了。
强东林在旁边看着，没有说话。
过了一会儿，他说了一句：
“那个是果台副梢。”
“是啥？”
“明年结果的地方。”
马可站在那儿，手停在半空。
“老强，我掐掉了明年的？”
“一个而已。”
强东林走过来，把那根断枝拿在手里看了看。
“一棵树上有几十个。”
他把断枝丢了。
“你要是一天掐断十个，一年一百棵树，就是一千个。”
“那明年少多少果？”
强东林摇了摇头。
“不是明年。”
“那是啥时候？”
“后年才看得出来。”
马可后来在本子上记了一行：
「今天犯的错，后年才罚。」
「这就是为什么果园里的坏习惯改不掉。」
第二天他掐得很慢。
慢到强东林都看不下去。
“兄弟，你别看着一个果想三十秒。”
“我怕掐错。”
“你怕掐错，一天就三棵。”
强东林把一根枝子拉过来。
“掐错一个，损失一个。”
“不掐，损失一树。”
段二·古文镜像
“顺天时，量地利，则用力少而成功多；任情返道，劳而无获。”
English Mirror: Follow the season, measure the ground, and effort is small while the yield is large. Go against it on impulse, and you labour for nothing.
第一天，马可疏了三棵树。
三棵。
强东林那天疏了十一棵。
他媳妇疏了九棵。
中午吃饭的时候，马可问了一句。
“老强，这十二亩，一共多少棵？”
强东林算了一下。
“五百零四。”
“一亩多少？”
“四十二。”
“都是乔化？”
“九亩是老树，乔化。三亩是租三叔的，矮砧。”
马可掏出本子。
他开始算。
五百零四棵。
强东林一天十一棵。
他媳妇一天九棵。
两个人一天二十棵。
五百零四除以二十，二十五点二天。
他抬起头。
“老强，疏果有几天时间？”
强东林把碗放下。
“落花以后，十五到二十五天。”
“多少天？”
“满打满算，十五天。”
马可看着本子上那个数。
二十五点二天的活，十五天干完。
“老强，这个干不完。”
强东林没有说话。
他媳妇在旁边接了一句：
“年年都干不完。”
下午两点，马可跟着强东林媳妇干了两个小时。
他一直在看她的手。
她的动作跟强东林不一样。
强东林是站定了看一圈，然后一口气掐五六下。
她是走一步掐一下，眼睛几乎不停。
“婶子，您这样不怕掐错？”
她没有抬头。
“错就错了。”
“那您咋比他还快？”
她笑了一下。
“我不是快。”
“那是啥？”
“我是不回头。”
马可停住了。
“啥叫不回头？”
她这才直起腰。
“他掐完一棵，要回头再看一遍。”
“您不看？”
“不看。”
她把手在裤子上擦了擦。
“看了也白看。掐都掐了。”
那天晚上，马可在本子上算了一笔。
强东林一天十一棵，他媳妇九棵。
看着是他快。
可强东林那十一棵里，有大半是他回头改过的。
他媳妇那九棵，一遍过。
马可后来问了老周一句。
老周说了一句很实在的话：
“小马，疏果这活儿，回头看的那一遍，值不值，取决于第一遍错多少。”
“错多少算多？”
“一成以内，别回头。”
“一成以上呢？”
“一成以上，你就不该自己疏，你该请人。”
段三·事件主体
贾思勰是第三天下午来的。
马可那天的手指已经起了茧。
掐了两天，右手拇指和食指的侧面，磨掉了一层皮。
贾思勰从地垄那头走过来。
五十来岁，穿一件很旧的青布袍，下摆掖在腰带里。
裤腿上全是土。
他手里拿着一根木尺。
“先生是，”
“贾思勰。”
他把木尺往地上一插。
“你在算工。”
“先生怎么知道？”
“因为你手上那个本子，写的是数，不是果。”
马可把本子递过去。
贾思勰接过来，看了一眼。
“五百零四棵。二十五天的活，十五天。”
“是。”
“那你怎么办？”
“雇人。”
“雇几个？”
马可算过了。
“要在十五天里干完，一天得三十四棵。”
“他们两口子一天二十。”
“还差十四。”
“一个熟工一天九到十棵，得再来两个人。”
“多少钱？”
“塬上四月的工价，一天一百五。”
“两个人十五天。”
马可报了一个数。
“四千五。”
贾思勰把本子还给他。
他没有评价这个数。
他反问了一句：
“疏了以后，多打多少斤？”
马可停住了。
“先生，疏了以后，是少打。”
“少多少？”
马可翻了翻本子。
这个数他昨天问过老周，是打电话问的。
“按老周说的，疏到位，总斤数掉一成半到两成。”
“老强去年打了三万一千斤。”
“疏到位，今年大概两万五到两万六。”
“掉五千斤。”
“是。”
“一斤多少钱？”
“去年塬上收购价，八毛七。”
贾思勰在心里算了一下。
他算得很快。
“四千三百五。”
马可点了点头。
“先生，所以疏果这件事，账是这样的。”
他把本子摊开，一行一行念：
「雇工：四千五百。」
「少打五千斤，按八毛七：四千三百五十。」
「合计：少收八千八百五十。」
贾思勰听完，问了一句。
“这两笔，哪一笔更难受？”
马可想了一下。
“先生，四千五那一笔更难受。”
“为什么？现钱少，掉的斤数多。”
“因为那四千五要先掏出去。”
马可说，“四月掏，十月才见回头钱。”
他抬起头。
“先生，这个塬上，四月手里有四千五现钱的人，不到一半。”
贾思勰点了点头。
“这个我懂。”
他把木尺竖起来。
“我那本书里有一句：‘凡人家营田，须量己力。’”
“什么意思？”
“别照着最好的法子做。”
他停了一下。
“照着你手上有的人和钱做。”
“先生，这不是让人将就吗？”
贾思勰摇了摇头。
“不是将就。”
“那是什么？”
“是别让一个人，为了做对一件事，把明年的本钱赔进去。”
他把木尺插回土里。
“马先生，我见过太多人，第一年照着书做得很好，第二年就没了。”
“为什么没了？”
“因为第一年他借了钱。”
马可把这句话记了下来。
他在旁边加了一行自己的话：
「所有的技术推广，都假设农户有一笔闲钱。」
「而这个塬上，四月手里有闲钱的，不到一半。」
他抬起头。
“先生，那怎么办？”
贾思勰把那根木尺拔出来，掂了掂。
“你手上有几户？”
“七户。”
“七户手上有几个人？”
马可算了一下。
“十五个。”
贾思勰把木尺往地上一戳。
“那你算的就不该是钱。”
“那该算什么？”
他说了两个字。
“算人。”
“那多的是什么？”贾思勰问。
“一级果的比例。”
“从多少到多少？”
“老周说，从三成六到五成。”
“多多少斤一级？”
马可算了一下。
“两万五千斤里，一级从九千斤变成一万二千五。”
“多三千五百斤。”
“一级果贵多少？”
马可没有立刻答。
他在那片果园里站了几秒钟。
然后他说了一个字。
“零。”
贾思勰把插在地上的那根木尺拔起来。
“零？”
“先生，塬上收购不分等。”
“一级、二级、统货，一个价。”
贾思勰站在那儿，很久没有说话。
他把那根木尺横过来，在手里掂了掂。
“马先生，那你告诉我，为什么还要疏？”
马可摇了摇头。
“先生，我这三天，一直在想这件事。”
“你想出什么了？”
“我想出来的是：所有教人疏果的人，都没有算过这笔账。”
他把本子合上。
“先生，我上个月在西安，听一个研究员讲了七分钟。”
“他讲什么？”
“他讲疏果的技术要点。叶果比，间距，留中心果。”
“讲得对吗？”
“每一句都对。”
马可停了一下。
“他没有讲的是：这件事要五十六个工，而农户手里只有三十个。”
贾思勰点了点头。
他在那条地垄上蹲下来。
他伸手，从一棵树上摘了一个小果，放在手心里看。
“马先生，我写过一本书。”
“《齐民要术》。”
“你读过？”
“读过序。”
贾思勰笑了一下。
“读序的多。”
他把那个小果丢在地上。
“我问你一件事：那本书里，有一段写我自己养羊。”
“先生养过羊？”
“养过。二百口。”
“后来呢？”
贾思勰答得很平常。
“一年之内，饿死过半。”
马可怔住了。
“为什么？”
“因为我没备草。”
“先生不知道要备草？”
“知道。”
贾思勰站起来。
“书上都写着。人人都知道。”
他拍了拍手上的土。
“可我那年，就是没备够。”
“为什么？”
贾思勰看着这片果园。
“因为割草要人。”
“我那年手上的人，都去种地了。”
他停了一下。
“种地是眼前的，割草是冬天的。”
马可站在那儿，一句话也没说。
“先生，那您为什么把这一段写进书里？”
贾思勰转过身。
“因为要是不写，后人看我那篇《养羊》，会以为养羊很容易。”
“只是这个原因？”
贾思勰想了一下。
“还有一个。”
“什么？”
“我那篇里，写了怎么备草，写了一口羊一冬天要多少。”
他抬起头。
“要是我不写我自己饿死过一半，后人会说：这人纸上谈兵。”
他把木尺插回地里。
“马先生，一本教人做事的书，最要紧的不是它教得对不对。”
“那是什么？”
“是写书那个人，有没有在这件事上栽过。”
马可想了一下，问了一个他一直想问的问题。
“先生，您那本书，写了多少年？”
“十来年。”
“写完以后，有人用吗？”
贾思勰把手插进袖子里。
“我活着的时候，没有。”
他说得很平常。
“一个都没有？”
“我那时候是高阳太守。”
“下面的人不听？”
“他们听。”
贾思勰摇了摇头。
“听完就完了。”
“为什么？”
贾思勰蹲下去，捡起刚才那个小果。
“马先生，我这本书里，最要紧的东西，不能靠听。”
“那靠什么？”
“靠一年一年做。”
他把那个果捏了一下。
“我写的‘二月上旬耕’，一个农人听一遍，记不住。”
“他做过三年，就不用记了。”
“那书有什么用？”
贾思勰抬起头。
他说了一句让马可记了很久的话。
“书是给那个没有三年的人用的。”
“谁没有三年？”
贾思勰把那个小果放在马可手心里。
“接手的那个人。”
他站起来，拍了拍膝盖。
“马先生，一个村里最会种果的人，不需要我这本书。”
“要这本书的，是他儿子。”
他停了一下。
“或者，是他死了以后，接这十二亩地的那个人。”
马可拿着那个小果，站在地垄上。
他忽然想起强东林那个八十一岁的娘。
这四天，那位老人每天都在地头坐着。
她坐在一个小马扎上，什么也不干。
她就看。
第二天下午，马可去给她递过一次水。
“大娘，您看啥？”
老太太没有看他。
她指了指前面那几棵树。
“那几棵，是我栽的。”
“哪一年？”
“七九年。”
马可算了一下。
四十七年。
“那时候塬上种苹果的多不多？”
老太太摇了摇头。
“没人种。”
“那您为啥种？”
她想了很久。
“公社来了个人，发苗子。”
“发多少？”
“一家二十棵。”
“别人种了吗？”
“大半没种。”
“为啥？”
老太太笑了一下。
她的牙掉了大半。
“那时候地里得种粮。”
“那您为啥种？”
她说了一句很短的话。
“我那二十棵，种在地边上。”
马可怔住了。
“地边上不占地？”
“占一点。”
她用手比了一下。
“一条边，二十棵。”
她把水碗还给马可。
“娃，那时候我不敢占好地。”
“要是占了呢？”
“要是那年粮不够，我男人要骂死我。”
马可蹲在她旁边，问了最后一句。
“大娘，那二十棵，几年结的果？”
老太太伸出四根手指。
“四年。”
“四年以后呢？”
她把手放下。
“四年以后，全村都往地边上栽。”
她停了一下。
“再过三年，就不栽地边了。”
“栽哪儿？”
老太太用下巴指了指那一整片塬。
“栽这儿。”
马可站起来，看着这一整片塬。
四十七年。
从一条地边上的二十棵，到今天满塬的苹果。
他这四个月，一直在找一个能让人改行为的杠杆。
他试过讲道理，试过写规矩，试过算账。
而四十七年前，那个杠杆是什么？
不是公社发的苗子。
苗子大半没人种。
是一个不敢占好地的老太太，在地边上栽了二十棵。
四年以后，那二十棵结了果。
全村人看见了。
马可在本子上写了一行：
「四十七年前，说服全村的不是一句话，是二十棵长在地边上的树。」
「那二十棵为什么能说服人？」
「因为它们种在最不占地的地方，所以失败了也不要紧。」
他停了一下，又加了一行：
「我那九百八十四斤，就是二十棵地边上的树。」
段四·钩子结尾
第四天早上，马可做了一件事。
他把七户人家都叫到了强东林的园子里。
加上强磊、李叔，一共十四个人。
他在地上摆了一块黑板。
是从村小借的。
他在黑板上写了三行数：
「强家：504棵。两口子一天20棵。要25天。只有15天。」
「七户合计：约3100棵。」
「七户合计劳力：15个。一天约140棵。要22天。」
“各位，这个是七户加起来的账。”
“还是干不完。”
七户人家蹲在地上，没有人说话。
李叔忽然开口了。
“年轻人，你这个算法有个毛病。”
“李叔您说。”
“你把七户加起来，可七户的花期不一样。”
马可怔住了。
“咋不一样？”
李叔指了指塬上。
“背阴那几家，花落得晚三到五天。”
“向阳这几家，早三天。”
“李叔，这个差多少天？”
李叔想了一下。
“头一家跟末一家，差七八天。”
马可蹲下来，重新算。
要是七户的疏果窗口不是同一个十五天，而是错开七八天，
那整个村的窗口，就变成了二十二到二十三天。
而七户十五个劳力，正好需要二十二天。
他站起来，在黑板上写了一行：
「不是十五天，是二十二天。因为花期错开七八天。」
七户人家还是蹲着。
有人开始明白了，有人还没有。
强东林先开的口。
“兄弟，你的意思是，换工？”
李叔在旁边说了两个字。
“变工。”
马可看着他。
“李叔，什么是变工？”
李叔在地上蹲下来，用一根树枝画了几个圈。
“八十年代，我们村有变工队。”
“怎么变？”
“你家收麦，全队去。我家收麦，全队来。”
“记工分？”
“记工，不记钱。”
他把树枝插在地上。
“一天算一个工。年底算总账，谁欠谁，用工还，不用钱。”
“后来为什么没了？”
李叔想了很久。
“因为后来各家的地不一样了。”
“怎么不一样？”
“有人种苹果，有人种玉米。”
他把树枝拔出来。
“苹果地跟玉米地，忙的不是同一个时候。”
“忙的不是同一个时候，就变不成工。”
他抬起头，看着这七户人家。
“可是这七户，”
他停了一下。
“这七户，全是苹果。”
那天上午十点，七户人家在强东林的园子里，定了一件事。
从四月二十五到五月十六，二十二天。
十五个劳力，按花期先后，从向阳那三家开始，往背阴那四家走。
一天一个园子，做不完的第二天接着。
强磊拿了一个新本子。
他在第一页上写了四栏：
日期、哪一家、来了几个人、疏了多少棵。
李叔在旁边看了一眼，说了一句：
“娃，加一栏。”
“加啥？”
“加‘谁没来’。”
强磊愣了一下。
“李叔，这个不好写吧。”
李叔摇了摇头。
“娃，八十年代那个变工队，最后是怎么散的？”
“咋散的？”
“就是因为没人记‘谁没来’。”
他站起来，拍了拍裤子上的土。
“头两年都来。第三年有一家总不来。”
“没人说他？”
“没人说。大家都是一个村的。”
他往园子外面走了两步，回过头。
“第四年，就没有变工队了。”
马可蹲在那儿，把这件事从头到尾想了一遍。
变工队这个东西，八十年代有过。
它不是被谁取消的。
它是自己散的。
散的原因有两个：一个是各家种的不一样了，二是没人记谁没来。
第一个原因，今天这七户没有。
七户全是苹果。
第二个原因，今天还在。
他站起来，走到黑板前面。
他在最下面写了一行：
「变工的账，跟库里那本账，是一本账。」
七户人家有人问：“年轻人，啥意思？”
马可指了指那块黑板。
“各位，去年冬天那本台账，为什么老周肯信？”
没有人答。
“因为里面写了坏的。”
他说，“断电四个小时，写了。想过少报，也写了。”
他把粉笔放下。
“这本变工的账，一样。”
“谁来了，写。谁没来，也写。”
他看了一圈这十四个人。
“要是只写来了的，这本账三年就散。”
“要是两样都写，”
他停了一下。
“两样都写，它就能当钱使。”
李叔在旁边点了点头。
“娃说得对。”
他补了一句：
“当年我们那个队，最后是拿工分换粮的。”
“工分能换粮，是因为有人记。”
他把那根树枝丢了。
“没人记，工分就是一句话。”
那天中午，马可一个人在地头坐着。
他手上贴了两块创可贴。
四天，他一共疏了二十九棵。
强东林四天疏了四十四棵。
他掏出手机，看了一眼。
崔处长的回复是四月十号来的，已经十一天了。
那条消息很短：
「马先生：第九条已加，原文照录。第一条改为‘不低于八万’。第五条改为‘由厅方委托第三方，名单报备，不公示’。第二条未通过。」
马可当时看到「第二条未通过」这五个字，在那间六楼的房子里坐了一夜。
第二条是：五座试点库的收储协议，第三条按原文，不改为三成。
他这四个月，兜了一个大圈子。
从二月九日那面墙上的四种情形，到二月十日的一成零九，到二月十二日的一千一百四十七公里，到二月十八日那间会议室。
结果那一条，还是没进去。
他把手机收起来，站起来，往那片园子走。
走到一半，他停住了。
因为他忽然想到一件事。
他刚才想的是「兜了一个大圈子」。
可这四天，他手上疏掉的那些小果，
他回过头，看了一眼身后那三十棵树。
每一棵树上，他都掐掉了大概三分之二的小果。
掐掉的那些，落在地上，已经开始发黄。
一棵树上原来有八百个果。
他留下了不到三百个。
那五百个，不是失败。
马可站在那条地垄上，忽然明白了一件他这四个月一直没想通的事。
他掏出本子，写了一行：
「疏果和写条件，是同一件事。」
「七条报上去，过了三条，改了两条，没过一条。」
「我一直算的是没过的那一条。」
「可疏果的人从来不数掐掉的果。」
写完，他又加了一行。
这一行他写得很慢：
「掐掉的那些，不是被否掉的。是让剩下的能长大。」
他又想到一层。
七条里，过的那三条是：第一条的岗，第三条的台账归属，第四条的公示。
改的两条是：第一条的钱数，第五条的第三方。
没过的是第二条。
他把这七条重新排了一下顺序。
过的那三条，全是关于「谁在干」。
没过的那一条，是关于「谁赔钱」。
马可在本子上写：
「他们肯给人立岗，肯让记账人拥有台账，肯把名字公示。」
「他们不肯的，只有一样：不肯把损失接过去。」
「这不是他们坏。」
「这是因为‘接损失’这件事，在他们那套账里没有一格。」
「而我这四个月，一直在替这一格找一个能签字的人。」
「找了四个月，只找到一个。」
「就是我自己。」
他把本子合上，走回那棵树下。
下午他疏了十一棵。
这是他四天里第一次，达到强东林第一天的数。
夜里九点，马可回到县城那间六楼。
手机响了。
是老周。
“小马。”
“周老师。”
“你在洛川？”
“在。疏果。”
电话那头顿了一下。
“你自己疏？”
“自己疏。四天二十九棵。”
老周在电话里笑了很久。
笑完，他说了一句正经的。
“小马，我给你打这个电话，是有个事。”
“您说。”
“你二月份问我，一级果为什么在塬上卖不上价。”
“是。”
“我这两个月，替你问了一圈。”
马可站直了。
“周老师，您问到什么了？”
电话那头安静了两秒。
“小马，不是没人肯为一级付钱。”
“那是什么？”
老周说了一句让马可当晚没睡着的话。
“是塬上从来没有人，能把‘一级’这两个字，说成同一个意思。”
欲知后事如何，请看下回，《一级两个字定不下来》
附录一·Dongpo's Note
Jia Sixie came up the row on the third afternoon in a worn blue robe with the hem tucked into his belt, trouser legs filthy, carrying a wooden measuring rod which he stuck into the ground.
Qiang Donglin gave me a small curved pruner on the first morning and told me not to use it. It is to give you courage. You thin with thumb and forefinger, one light push outward, because shears are fast and you cannot be accurate with them.
Five fruitlets to a cluster, all of them apparently identical. Keep the king fruit unless it is crooked or russeted at the calyx; otherwise keep the strongest lateral, and the way to tell is the stem. The stem is a pipe. A thick pipe carries more. Across five fruitlets the thickest stem is about half again the thinnest, and once he pointed at it I could not stop seeing it.
Day one: I thinned three trees. He did eleven and his wife nine.
The arithmetic at lunch, which is the whole chapter. 504 trees. Twenty a day between the two of them. 25.2 days of work. The window from petal fall is fifteen days. I said it cannot be finished. He said nothing. His wife said: it is never finished.
Hired labour would close it — fourteen trees a day short, two skilled hands at nine or ten each, 150 a day, fifteen days: 4,500. And proper thinning drops total tonnage by 15–20%: on last year's 31,000 jin that is about 5,000 jin, at 0.87 another 4,350. Total cost of doing it right: 8,850.
Against which the gain is grade one from 36% to 50% — on 25,000 jin, roughly 3,500 jin moved up. Jia Sixie asked what grade one is worth extra. I stood in that orchard for several seconds and then said the true number, which is zero, because the plateau buys flat.
So every extension officer in the province is teaching a technique that, at current prices, costs the household about nine thousand yuan to perform correctly. The researcher in Xi'an spoke for seven minutes on leaf-to-fruit ratio, spacing and king fruit, and every sentence was correct. What he did not say is that it takes fifty-six labour-days and the household has thirty.
Then Jia Sixie asked whether I had read his book. The preface, I said. He laughed: most people read the preface.
He kept two hundred sheep. Within one year, over half starved. He knew about laying in fodder; everybody knows; it is written in the books. He simply did not cut enough that year, because cutting fodder takes hands and his hands were all in the fields. Planting is now; fodder is winter.
He wrote it into the book. Two reasons. If he had not, a reader of his chapter on sheep would think sheep are easy. And: he had written out how to store fodder and how much a sheep eats over a winter, and without the record of his own losses, posterity would say the man was theorising.
His formulation: the important thing about a manual is not whether the instruction is correct. It is whether the person who wrote it has failed at the thing.
Day four I put a blackboard borrowed from the village school on the ground and gave the seven households the combined figure: ~3,100 trees, fifteen able hands, ~140 trees a day, twenty-two days needed against a fifteen-day window. Still impossible.
And Li said my arithmetic had a fault in it. The seven households do not share a window. The shaded plots drop petals three to five days late, the sunny ones three days early, and between first and last there are seven or eight days. So the village window is not fifteen days. It is twenty-two or twenty-three. And fifteen hands need twenty-two days.
Qiang Donglin said it first: rotate the labour. Li gave it its old name: biàngōng. In the eighties the village had a labour-exchange team; you count days, not money, and at year end whoever is short repays in days. It died because the households stopped being busy at the same time — apple ground and maize ground peak differently, and where the peaks do not coincide, labour cannot be exchanged. These seven are all apples.
Fixed that morning: 25 April to 16 May, twenty-two days, fifteen hands, starting with the three sunny plots and working round to the four shaded ones, one orchard a day.
Qiang Lei opened a new book with four columns: date, whose orchard, how many came, how many trees. Li told him to add a fifth. Who did not come. Qiang Lei said that would be awkward to write. Li asked how the eighties team finally broke up: nobody recorded who did not come. First two years everyone came. Third year one household stopped, and nobody said anything, because it is one village. In the fourth year there was no team.
Cui's reply came on the tenth and I have been carrying it eleven days. Clause nine added verbatim. Condition one reduced from 120,000 to not less than 80,000. Condition five weakened: third party commissioned by the department, panel filed but not published. And condition two — clause three restored for the pilots — not approved.
I sat up a whole night over those five characters. And then on the fourth afternoon, walking back up the row past thirty trees I had stripped of two-thirds of their fruitlets, I understood something I have been slow about all winter. A thinner never counts the fruit he removes. Seven conditions went up; three passed, two were amended, one was refused. I have spent eleven days counting the one.
The ones you pinch off are not defeats. They are what lets the rest get big.
I thinned eleven trees that afternoon, which is Qiang Donglin's first-day number.
Zhou rang at nine. He laughed for a long time about four days and twenty-nine trees, and then said he had spent two months asking around on my February question. And the answer is not that nobody will pay for grade one. It is that on this plateau nobody has ever been able to make the words "grade one" mean the same thing twice. — M.
附录二·Marco's Memo
Memo to file. Qiang orchard, Luochuan, 21–24 April. Night one hundred and five. Resident fifteen days.
1. Technique, recorded because I will forget the specifics. Thin by hand, not shears — accuracy over speed. One fruit per cluster. Keep the king fruit unless crooked or calyx-russeted; otherwise the strongest lateral, identified by stem calibre, which varies about 1.5× across a cluster and predicts final size. The stem is a conduit.
2. The arithmetic, which is the finding.
— Qiang household: 504 trees over 12 mu (9 mu standard rootstock, 3 mu dwarf, rented). — Throughput: him 11 trees/day, wife 9. 20/day. — Requirement: 25.2 labour-days. — Window from petal fall: 15 days. — Shortfall: ~5 days of work with no days to put it in. His wife: it is never finished.
3. Cost of doing it correctly, at current prices. — Hire 2 skilled hands × 15 days × ¥150 = 4,500 — Tonnage loss 15–20% on 31,000 jin ≈ 5,000 jin × 0.87 = 4,350 — Total: −8,850
Gain: grade one 36% → 50%, i.e. ~3,500 jin moved up on a 25,000 jin crop. Value of that gain on the plateau: 0, because buying is flat across grades (established ch. 99, seventeen calls).
So correct thinning is, under the prevailing pricing regime, a reliably loss-making investment of about nine thousand yuan per household — and it is the single most-promoted practice in the extension programme.
Restating the general form, because this is the third time I have hit it this year: the recommended practice is correct; the constraint is not knowledge; and the party issuing the recommendation bears none of the cost of following it. Under-thinning is not backwardness. It is a household with thirty labour-days doing a fifty-six-labour-day job.
4. Jia Sixie's method, which I want to copy exactly. He recorded that he kept two hundred sheep and starved more than half of them in one year, in the same chapter in which he explains how to lay in winter fodder. He knew the rule. He failed on labour allocation: fodder-cutting competes with planting, and planting is now.
His two reasons for writing it down: without it the chapter makes sheep look easy; and without it, a reader is entitled to suspect the author has never done the thing. His rule: the value of a manual is not the correctness of the instruction but whether the author has failed at the task.
Note against myself: my six iron clauses contain no record of my own failure. The four provenance pages do, but they are about money, not about operations. Next season's ledger format should carry a standing section for what the recorder got wrong. Qiang Lei has already done this voluntarily twice (the outage, the thought of under-reporting). I should make it a field rather than a virtue.
5. The fix, and it did not come from me.
My combined figure: 3,100 trees, 15 hands, 140/day, 22 days needed, 15-day window. Li corrected the model: the seven households do not share a window. Shaded plots drop petals 3–5 days late, sunny plots ~3 days early; spread first-to-last is 7–8 days. Village window is therefore 22–23 days, which exactly matches 22 days of requirement.
The constraint I had treated as physical was an artefact of aggregating seven staggered windows into one.
Mechanism: biàngōng, labour exchange, revived. Days counted, not money; settlement in days at season end. 25 April – 16 May, one orchard per day, sunny plots first.
Why it died in the eighties, per Li, and this is the part to protect: cropping diversified, peak labour demands stopped coinciding, and where peaks do not coincide labour cannot be exchanged. These seven are all apples, so the mechanism is available again — and it is available only as long as they stay all apples, which is a fragility to note.
6. Ledger design, and Li's amendment. Qiang Lei's four columns: date / orchard / attendance count / trees done. Li: add a fifth, who did not come. Qiang Lei objected that it is awkward.
Li's account of the eighties team's collapse: no attendance record; two good years; in year three one household stopped turning up; nobody said anything because it is one village; in year four there was no team.
General form, and it is the same finding as clause three and the same as the seventeen calls: an obligation nobody records is an obligation that decays silently, and the decay is invisible until the institution is already gone. Column added.
7. Appointment terms, received 10 April, acted on today. — Clause nine: added verbatim. (So they were buying the appointment, not the ledger. Decision rule from ch.104 satisfied.) — Condition 1: 120,000 → ≥80,000. Acceptable; it is still a budgeted post. — Condition 5: third party commissioned by the department, panel filed not published. Materially weaker. Note it. — Condition 2 (clause three restored for the five pilots): not approved.
Eleven days spent on the refusal. Correction applied in the orchard: a thinner does not count what he removes. Three passed, two amended, one refused. The removed fruit is not a defeat; it is the condition of the rest getting big.
8. Zhou Jianguo, 21:00. Two months of enquiry on the February question. His answer, which reframes everything in chapter 99:
Not that no one will pay for grade one. That nobody on this plateau has ever made the words "grade one" mean the same thing twice.
If that is right, then the seventeen-call table diagnosed the wrong constraint. The buyers who "do not grade" may not be indifferent to quality; they may be unable to transact on a grade claim they cannot verify. That is a standards problem, not a preference problem, and standards problems are cheaper to solve than preferences.
To pursue tomorrow.
附录三·English Highlight
"Follow the season, measure the ground, and effort is small while the yield is large."
Jia Sixie's agricultural manual is fifteen hundred years old and it contains, in the chapter on sheep, an account of the author keeping two hundred sheep and starving more than half of them within a year.
He knew the rule. Laying in winter fodder is in every book. He did not cut enough, because cutting fodder takes hands and his hands were all in the fields — planting is now, and fodder is winter. He wrote it down for two reasons: so that a reader of the chapter would not think sheep are easy, and because without it a reader would be entitled to suspect he had never kept sheep at all.
The value of a manual is not whether the instruction is correct. It is whether the author has failed at the task.
The instruction in question that April was fruit thinning, which every extension service in the province promotes and which every specialist explains correctly: one fruit per cluster, keep the king fruit, judge the laterals by stem calibre because the stem is a pipe and a thick pipe carries more.
The household in question had 504 trees, two pairs of hands doing twenty trees a day, twenty-five days of work, and a fifteen-day window from petal fall. Hiring the gap costs 4,500. Thinning properly also drops total tonnage by fifteen to twenty percent, which at the flat farm-gate price costs another 4,350. And the gain — grade one rising from thirty-six percent to fifty — is worth exactly nothing, because the plateau buys every grade at one number.
Correct practice, correctly taught, was a reliably loss-making investment of about nine thousand yuan. Under-thinning was not ignorance. It was thirty labour-days against a fifty-six labour-day job.
The fix came from a seventy-year-old who pointed out an error in the aggregation. The seven households do not share a window: shaded plots drop petals three to five days late, sunny plots three days early, and the spread across the village is seven or eight days. The village window is twenty-two days, and fifteen hands rotating between orchards need twenty-two days.
They revived a practice from the eighties in which labour is settled in days rather than money. And the same old man insisted on one column in the new attendance book — who did not come — because the last such team collapsed in its fourth year, having never recorded it.
附录四·Dongpo's Cheat Sheet
| 今日之事 | 出处 | 现代对应 | 一句话 |
|---|---|---|---|
| 剪子是给你壮胆的 | 强东林 | 手掐，不用剪 | 剪子快，可你剪不准 |
| 看果柄 | 五个果选一个 | 粗细差约一半 | 柄是管子，管子粗走的东西多 |
| 三棵对十一棵 | 第一天 | 他媳妇九棵 | 两口子一天二十棵 |
| 二十五天的活，十五天 | 504棵 | 落花后十五到二十五天 | 他媳妇：年年都干不完 |
| 雇工四千五 | 两人十五天 | 一天一百五 | 补上一天差的十四棵 |
| 少打五千斤 | 疏到位掉一成半到两成 | 八毛七 | 四千三百五 |
| 一级果贵多少 | 贾思勰问 | 马可答一个字 | 零 |
| 疏果正确做法＝稳赔八千八 | 全省都在推 | 研究员七分钟每句都对 | 他没讲的是：五十六个工，农户手里三十个 |
| 饿死过半 | 贾思勰养羊二百口 | 他知道要备草 | 割草要人，种地是眼前的，割草是冬天的 |
| 为什么写进书里 | 两个理由 | 不写就显得养羊容易 | 不写，后人会说这人纸上谈兵 |
| 教人做事的书 | 贾思勰的标准 | 不看教得对不对 | 看写书的人有没有在这件事上栽过 |
| 七户不是同一个窗口 | 李叔纠正 | 背阴晚三到五天，向阳早三天 | 十五天变二十二天 |
| 变工 | 八十年代 | 记工不记钱，年底用工还 | 忙的不是同一个时候，就变不成工 |
| 加一栏“谁没来” | 李叔 | 强磊说不好写 | 第三年有一家不来没人说，第四年就没有变工队了 |
| 第二条未通过 | 四月十日崔处长 | 九条加了，一条降到八万，五条改为不公示 | 他为这五个字坐了一夜 |
| 疏果的人不数掐掉的果 | 第四天下午 | 七条过三改二否一 | 掐掉的不是被否掉的，是让剩下的能长大 |
| 一级两个字 | 老周夜里来电 | 不是没人肯付钱 | 是塬上没人能把这两个字说成同一个意思 |
