Xiaoyang Yu

本文把一个先已立住的判断——“观察宇宙，与观察一局康威生命游戏，并无区别”——从一句关于宇宙是什么的话，推进为一句关于我与宇宙是何种关系的话，并指认：这关系，正是“上帝之眼”的哲学版。说“观察宇宙”，舌头一打这结，便立了两个端点：这边一个我，那边一个宇宙，中间隔着“看”；仿佛我能如看一局生命游戏般、从屏幕外端详它。可这比喻最要害处恰是它的破绽：看游戏的人在屏幕外，我看宇宙却从里头看——我并非悬在万有对面的一双眼，我就是这片阵列里一团正在刷新的活格子，“观察”不过是这团格子内部、泛起的一阵关于其余格子的微光。于是那道缝本不存在：没有“我，面对着，宇宙”，只有宇宙在它自己一小块上、朝自己亮起一点觉知。我与它的关系，剥到最里，是它与它自己的关系——整块涌动，在一个叫“我”的褶皱处，回看了自己一眼；且不可分离：看游戏者可关屏走开，我却走不出去，我看它的这一下，本身就是它在动。这便是上帝之眼的哲学版，其要在一个反转：宗教那只眼，从外、从高俯瞰万有，其全部威权建于“看者与被看者可分离”；哲学版把那份“全体尽在其中”的辽阔原样留下，却把观看的位置，从世界之外的高处，收回世界之内的一道褶皱——神的“纵览一切”留下了，神的“站在一切之外”取消了；内在，而非超越。 它还交还了一样东西：审判。那只天上的眼看而兼判（谁对、谁错、谁有罪）；可顺前理已明：图上的“错”是红笔添的、非图本有，“行动的主体”是放大镜抠出的轮廓，连其“移动”都只是亮灭接力被读成的位移。看清这些，这只内在之眼便看得见每道红圈如何画上、每个主语如何被偷渡——于是它看，却不再补盖“谁错了”那道章。它是一只看而不判的眼：辽阔仍在，威权落地；错从不在刷新里，只在挑错的手里。剩下的，是一个素朴到发晕的圆：宇宙，在我这儿，看着宇宙。 没有神与世界的对望，只有一个东西经由一道褶皱回照自身——斯宾诺莎（神即自然，于自身之内思自身）、那句古老的“那即是你”、“天地与我并生”，都认得这只眼。本文守一处分寸：把“我”看成“局内一格映现全局”，绝非把“我”抹成虚无——倘整块涌动只在某些褶皱处回看自己，那么“我此刻正看、正想、正亮”非但不渺小，反是这片浩瀚里它唯一照见自身之处；庄严不在天上、不在高处，就在此刻你这一眼。“天行健”的健，未因少了掌舵者而停半分；它照样涌着，只是此刻在你这儿睁了一眼、看了看自己，然后下一帧，继续涌。“上帝之眼”一词，借的是其辽阔与庄重，不为“我”加冕、不暗藏一个超自然的背书；看而不判，连“不判”也含不指导、不背书：它看得见受苦的格子真在亮，却既不替“去防范”背书、也不替“可放任”背书——防范、修复、归责如何重新安放，是局内之事，落在这只眼之外。而本文自己，也不过是阵列在“上帝之眼”这一处、又一次的自照；指完，即弃。 Abstract This paper takes a judgment already established—“to observe the universe and to observe a game of Conway’s Life are no different”—and advances it from a claim about what the universe is into a claim about what relationship I bear to the universe, identifying that relationship as the philosophical version of the “eye of God.” To say “observe the universe” is, the moment the tongue ties that knot, to set up two poles: a me here, a universe there, “looking” stretched between—as if I could survey it, the way one watches a game of Life, from outside the screen. But the crux of that image is precisely its flaw: the one watching the game is outside the screen, whereas I watch the universe from within—I am not a pair of eyes suspended opposite all things, I am one clump of live cells refreshing inside this array, and “observing” is only a faint light, risen inside this clump, about the other cells. So the seam never existed: there is no “me, facing, the universe,” only the universe kindling, upon one small patch of itself, a little awareness toward itself. My relationship to it, peeled to the core, is its relationship to itself—the whole welling, at a fold called “me,” glancing back at itself; and inseparably so: the game-watcher can switch off and leave, but I cannot step out, and my looking at it is itself it, moving. This is the philosophical version of the eye of God, and its essence is one reversal: the religious eye looks down upon all things from outside and on high, its whole sovereignty built on “watcher and watched are separable”; the philosophical version keeps that vastness of “all held within” untouched, yet draws the seat of looking back from the heights outside the world into a fold within it—God’s “surveying all” is kept; God’s “standing outside all” is dropped; immanent, not transcendent. It returns one more thing: judgment. The eye in heaven looks and also decrees (who is right, who wrong, who guilty); but by the foregoing it is clear: the “error” on the picture is added by the red pen, not native to it; the “acting subject” is a contour the magnifying glass cut from the pattern; even its “movement” is only a relay of on-and-off read as displacement. Seeing this, the immanent eye can see how every red circle gets drawn and every grammatical subject gets smuggled in—and so it looks, yet no longer stamps “who did wrong” upon what it sees. It is an eye that sees without judging: the vastness remains, the sovereignty set down; the fault was never in the refreshing, only in the fault-picking hand. What remains is a circle plain to the point of vertigo: the universe, at me, watches the universe. No mutual gaze of God and world, only one thing, by way of a fold, shining back upon itself—Spinoza (God-or-Nature thinking itself within itself), the ancient “that art thou,” “Heaven and Earth are born together with me,” all would know this eye. The paper keeps a measure: to see “me” as “one cell within mirroring the whole” is by no means to erase “me” into nothing—if the whole welling looks back at itself only at certain folds, then “I am, right now, looking, thinking, lit” is not small but the only place, in all this vastness, it sees itself; the solemnity is not in heaven, not on high, but here, in this one glance of yours. The vigor of “Heaven moves with vigor” has not paused for the loss of a helmsman; it wells on, only that now, at you, it has opened an eye and looked at itself, and then, the next frame, wells on. The phrase “eye of God” is borrowed for its vastness and gravity, not to crown “me” nor to smuggle in a supernatural backer; and its seeing-without-judging also withholds all directive and endorsement: it sees that the suffering cells are really lit, yet backs neither “go prevent” nor “let it be”—prevention, repair, and the re-situating of blame are matters within the array, outside this eye. And this paper is itself only the array, once more, beholding itself at this spot called “the eye of God”; once it has pointed, discard it.

本文用一个我们亲手排除了主体的世界，把“行为里有主体”这个假定干净地证伪——那个世界，是康威生命游戏。先立总枢：抠出唐僧取经一路沿时间线的一举一动，与抠出生命游戏里一个模式沿时间线的“行为”，是同一道手势；而“行为”这个词本身，就暗含了一个主体、一份独立性。生命游戏是一个理想的对照组：它的规则全由我们定，每个格子下一刻的死活，只由其八邻当下的死活机械决定，没有任何格子“想”什么、“选”什么——主体的余地，在设计层面被焊得一丝不剩。可即便在这样一个铁证没有主体的世界里，我们照样抠得出一个“滑翔机”，盯着它说“它移动了”“它走向右下方”。同一把放大镜，在一个明知毫无主体的世界里，照样制造出一个“行动者”。这就把话锁死了：“一个沿时间移动、有行为的东西”那种感觉，完全可以、且确实在毫无主体处被造出来——所以它根本不是主体存在的证据，只是放大镜架设的副产品。 而“行为”二字正是偷渡口：一旦你松口管那串格子变化叫“它的行为”，你就已在词里塞进了一个“它”、一个能“行”的主语、一个独立于背景的施动者。可生命游戏里真实发生的，从头到尾只是整个阵列、按一条全局规则、同步刷新到下一帧——没有“滑翔机在移动”这回事，只有“全盘的下一个状态”。本文再下最锋利的一刀：滑翔机根本不是阵列里一个真实存在的“东西”——阵列只认格子与规则，“滑翔机”是观察者的模式识别、是我们眼里的格式塔，不在阵列本体里；它沿时间“移动”更是彻底的错觉——没有任何一个格子真的挪了位置，动的只是“哪些格子亮着”，是亮灭的接力被我们读成了一个东西的位移。故在“滑翔机移动”里，移动的主体（滑翔机）是抠出来的，移动这桩事（位移）也是读出来的——两头都不在阵列里，全在放大镜里。把这一刀架回唐僧、架回人：唐僧那条“行为线”与滑翔机那条“行为线”，是同一类东西——观察者从一片全局演化里，圈出一团碰巧被一个名字归拢、形状大致自我维持的局部，再用“行为”给它安一个主语；宇宙这片更大的阵列，也只一帧帧浑然刷新，“人”“人的行为”“他在选、在走、在负责”，是我们这些内部的模式拿放大镜互相圈认出来的格式塔——一团滑翔机式样的活格子，圈认另一团，说“看，它在行动、它该负责”。于是举证责任被这个干净的对照组整个翻转：既然“行为”暗含的那个独立主体，在生命游戏里被证明可以是纯属虚构而我们浑然不觉，那么凭什么断定，在宇宙这片阵列里它就不是同一种虚构？本文守一处分寸：说滑翔机不是真东西、它的“移动”是读出来的，并不是说那片光什么都没发生——格子确实在亮灭，接力确实在传，那是真的；被拆的，只是“有一个叫滑翔机的独立东西、它真在自己移动、那移动是它自己之所为”这层附加的读法。同理，拆“人是被抠出的主体”，拆的是那层主语，不是那一程真实的涌动——人真在其中动，如滑翔机式样的光真在阵列里传；阵列越是全局一体，你这团碰巧自我维持着、与万物同步刷新着的活格子，越是真实地正亮在其中。体验是真的，用力没被取消，痛与伤害是真的，归责如何重新安放是开放问题。而本文自己，也不过是阵列里一个滑翔机式样的模式，拿放大镜把“滑翔机”圈成此刻最显眼的那团光；指完，即弃。 Abstract This paper falsifies the assumption that “there is a subject in behavior” by means of a world from which we ourselves have excluded all subjects—Conway’s Game of Life. The hub first: extracting the Tang Monk’s every move along the timeline of his pilgrimage, and extracting a Life-pattern’s “behavior” along its timeline, are the same gesture; and the word “behavior” itself already smuggles in a subject and a measure of independence. The Game of Life is an ideal control group: its rules are entirely set by us; each cell’s next aliveness is mechanically determined by the present aliveness of its eight neighbors; no cell “wants” or “chooses” anything—the room for a subject is, by design, welded shut without remainder. Yet even in such a world, provably subjectless, we can still carve out a “glider” and, watching it, say “it moves,” “it travels toward the lower right.” The same magnifying glass, in a world we know to be wholly subjectless, still manufactures an “agent.” This nails the point: the feeling of “a thing that moves through time and has behavior” can be—and demonstrably is—produced where there is no subject at all; so it is no evidence of a subject, but a byproduct of setting up the magnifying glass. And the word “behavior” is precisely the smuggling-point: the moment you let yourself call that string of cell-changes “its behavior,” you have already packed into the word an “it,” a subject that can “act,” a doer independent of its background. But what actually happens in the Game of Life is, from start to finish, only the entire array refreshing, by one global rule, to its next frame—there is no “glider moving,” only “the next global state.” The paper then makes its sharpest cut: the glider is not a real “thing” existing in the array—the array knows only cells and rules; “glider” is the observer’s pattern-recognition, a gestalt in our eyes, not in the array’s own ontology; and its “moving through time” is utterly an illusion—no cell ever changes its position; what changes is only “which cells are lit,” a relay of on and off that we read as one thing’s displacement. So within “the glider moves,” the moving subject (the glider) is carved out, and the moving event (displacement) is read in—neither is in the array, both are in the magnifying glass. Lay this same cut back onto the Tang Monk, back onto the human: the Tang Monk’s “behavior-line” and the glider’s “behavior-line” are of one kind—the observer, out of one global evolution, circles a clump that happens to be name-gathered and roughly self-maintaining, then fits it with a subject by the word “behavior.” The universe, that larger array, also merely refreshes frame by frame; “person,” “a person’s behavior,” “he is choosing, walking, responsible” are gestalts that we—patterns inside—circle in one another with the magnifying glass: one glider-shaped clump of live cells circling another and saying “look, it is acting, it is responsible.” So the burden of proof is wholly reversed by this clean control group: since the independent subject implied by “behavior” is shown, in the Game of Life, capable of being pure fiction while we remain wholly unaware of it, on what ground does one declare that, in the array of the universe, it is not the same fiction? The paper keeps a measure: to say the glider is no real thing and its “moving” is read in is not to say nothing happens in that field of light—cells really do flicker, the relay really does travel, that is real; what is dismantled is only the added reading of “there is an independent thing called a glider, it really moves itself, and that motion is its own doing.” Likewise, to dismantle “the human is a carved-out subject” is to dismantle that grammatical subject, not the real welling of that journey—the human really does move within it, as the glider-shaped light really does travel in the array; the more globally unified the array, the more really your clump of cells—happening to self-maintain, refreshing in synchrony with all else—is lit within it. Experience is real, effort is not cancelled, pain and harm are real, and how blame is to be re-situated is an open question. And this paper itself is merely one more glider-shaped pattern in the array, holding up the magnifying glass to circle “the glider” as the most salient patch of light right now; once it has pointed, discard it.

本文以一个比喻为总枢——红圈——把最近彼此咬合的几记洞见，收束成一幅图。给你一张画，让你挑出几处“错”、用红笔圈上：这就是“选错”的含义——用思维，在世界这张实际之图上，画红圈。要紧的是看清：错不在图里，在圈里。 那张图（实际的发生）本身无所谓对错，它只是恰好那样发生了；“错”只在我们拿红笔画圈时才出现。圈是思维添上去的，不是从图里读出来的——“选错”从来不是图画的属性，是一道红圈的属性。这正接回“发生不会错”：会错的，只有我们画上去、却没画准的那些圈。而要在某处落笔，须先在心里备一张“本该如此”的正确版，拿它去比那张实际的图，差出来的地方才圈得下去；那张没发生过的正确版，就是反事实之梦。故红圈、梦、反事实，是同一样东西的三个名字：红圈，是“实际之图”与“梦中本该”之间那道缝的标记。动物世界那一笔最见筋骨：对着动物，人照样画得出红圈（“这儿跑错了”），却未必补盖“它自由、因而该负责”那道章——可见“画圈”与“盖章”本是两件能分开的事，圈是主动作、先有，“自由—有责”是可上可不上的附加章。对动物，画圈而扣下章；对人，文化自动补盖——于是我们错把两者当成同一下。人／动物之间那道线，根本不在世界里（两边都是图＋圈），唯一的差别是文化要不要补盖“自由”；这线会动（历史上真有过审判、处决动物的事），可见那章纯是约定，不是对“它到底有没有真选”的查证——查证从未做过，无论对谁。再往下：画圈，把被圈的那块从整幅画里剥了出来，显得能独立抽换——而“独立的自我”，正是这桩剥离附带的效应。我们其实并不关心“人是否独立”，只想说圈里那块行为是独立、可变的（好让反事实有处落脚）；人，是被顺带创作为独立的一个。而这剥离不止于空间：我们更把一个人沿时间的行为抽成一根线（如把唐僧取经一路举动串起、拿放大镜逐帧考其对错），横向剪断了他与旁人活动（悟空那一棒、那条路、那阵风）的咬合，纵向又把本无现成边界的“他的行为线”硬描了出来；所谓“一个人”，正是被这样串成、攒成一个可追责的连续主体的。我们觉得他的行为独立，恰因我们强行按其独立来思考它——放大镜不是中立的观察器，是把行为从整张网里剪出、再呈给我们看的那把剪子。若不是非要去设想一个人的不同行为，这人不过是一整块不分瓣儿的磐石里的局部——何独立之有？宇宙及其演化本就铁板一块；人的活动，如恒星之运行；倘宇宙里没有人，地上一切物理活动读起来本与天象无异——是我们对人类行为的种种负面情感，才把似乎与人相关的活动，从宇宙整体背景里降级、切割出来。那道“人不同于天象”的缝，是情绪刻的，不是世界本有的。最后一层：愿望，只是中间层。 选某个选项，像是愿望的实现；可愿望并非一股推动力。它有前因、生后果，但真正在走的因果是“前因穿过愿望流到后果”——愿望是中转产物，不是源头。它之所以被错当成一个“没有前因的因”，只因它的前因沉在内省水线之下、看不见；看不见，就显得是个起点。故愿望并无它自以为的那种“额外发起、能拽方向盘”的作用：人的活动，与星体沿固定轨道的运行无异，不受愿望操控；愿望只是流经的一层，实无真正的效力。四者——没有真正的“选错”、没有独立的自我、没有发起性的愿望、那道被情绪刻出的人／天之缝——不是四件事，是同一道红圈投下的四道影；拆一道，四道俱塌。 本文末了把分寸守足：这不是虚无，“天行健”的“健”分毫未减——化掉的，只是那个握方向盘的“我”和那道缝，不是那场运行本身；铁板照样在涌、在健，每个人都在其中动，如恒星之动。愿望作为被经历的那一段，照旧真切地被经历着（没人否认你此刻确在“想”）；被否的，只是它“无前因、能发起、掌全局”那层身份。真在发生的痛与伤害是真的，用力也未被取消（被推着走，取消的是外面那个安排者，不是此刻的用力；水无舵手而自流，绝非躺平）。责任、归罪该如何重新安放，是开放问题。而本文自己，也不过是又画了一个红圈——它把“红圈”圈成了此刻最显眼的那块；指完，即弃。 Abstract This paper takes one metaphor as its hub—the red circle—and gathers several recently interlocking insights into a single picture. Give you a drawing, ask you to spot a few “errors” and circle them in red: this is what “choosing wrongly” means—using thought to draw red circles on the world’s picture of what actually occurred. The crucial thing to see: the error is not in the picture, but in the circle. The picture (the actual occurrence) is in itself neither right nor wrong; it merely happened just so; “wrong” appears only when we take up the red pen and circle. The circle is added by thought, not read off the picture—“choosing wrongly” is never a property of the picture but a property of a red circle. This returns us to “occurrence cannot be wrong”: what can be wrong is only the circles we draw and draw amiss. And to set the pen down somewhere, one must first hold in mind a “how-it-should-have-been” correct version, compare it with the actual picture, and circle where they differ; that never-realized correct version is the counterfactual dream. So the red circle, the dream, and the counterfactual are three names for one thing: the red circle marks the seam between “the actual picture” and “the dreamed should-have-been.” The animal-world stroke shows the bone most clearly: facing animals, a person can still draw red circles (“it ran the wrong way here”) yet need not add the stamp “it was free, therefore responsible”—revealing that “circling” and “stamping” are two separable acts, the circle primary and prior, the “free–responsible” stamp an optional addition. For animals we circle but withhold the stamp; for humans culture adds it automatically—so we mistake the two for one act. The line between human and animal is not in the world at all (both sides are picture-plus-circle); the only difference is whether culture adds the “free” stamp; and this line moves (history really did try and execute animals), showing the stamp to be pure convention, not a verification of “did it really choose”—a verification never performed, for anyone. Further down: drawing the circle peels the circled piece out of the whole picture, making it look independently swappable—and the “independent self” is exactly the side-effect of this peeling. We do not really care whether the person is independent; we only want to say the circled behavior is independent and changeable (so the counterfactual has somewhere to land); the person is incidentally created as an independent one. And this peeling is not only spatial: we also abstract a person’s behavior across time into a single line (as in stringing the Tang Monk’s deeds along the whole pilgrimage and examining each frame for “whether it was wrong”)—horizontally severing his interlock with others’ activities (Wukong’s blow, that road, that gust), and vertically drawing out a “line of his behavior” that had no ready-made boundary; so-called “a person” is precisely what gets strung and assembled into one continuous, accountable subject this way. We feel his behavior to be independent precisely because we forcibly think it as independent—the magnifying glass is no neutral observer but the very scissors that cut the behavior out of the whole net and then present it back to us. Were we not bent on imagining a person’s different behaviors, that person would be merely a part of one undivided monolith—what independence? The universe and its unfolding are a single iron-block; human activity is like the motion of the stars; were there no humans in the universe, all physical activity on Earth would read no differently from the celestial; it is our various negative emotions toward human behavior that downgrade and cut the seemingly-human activity out of the cosmic whole. That seam—“the human differs from the celestial”—is carved by emotion, not found in the world. The last layer: desire is only an intermediate layer. Choosing an option looks like the fulfillment of a desire; but desire is not a driving force. It has antecedents and yields consequences, but the causation actually running is “antecedents flow through desire to consequences”—desire is a transit-product, not a source. It is mistaken for an “uncaused cause” only because its antecedents lie below the waterline of introspection, unseen; and the unseen looks like a starting point. So desire has none of the “extra originating, wheel-steering” power it imagines itself to have: human activity is no different from a star’s motion along a fixed orbit, ungoverned by desire; desire is merely a layer it flows through, with no real efficacy. The four—no real “wrong choice,” no independent self, no originating desire, and the emotion-carved seam between human and heaven—are not four things but four shadows of one red circle; pull one and all four collapse. The paper keeps its measure to the end: this is not nihilism; the “vigor” of “Heaven moves with vigor” is undiminished—what dissolves is only the wheel-gripping “I” and the seam, not the running itself; the iron-block still wells, still runs with vigor, every person moving within it like a star. Desire as experienced is still really experienced (no one denies that you do, right now, “want”); what is denied is only its status as “uncaused, originating, in control.” The pain and harm that really occur are real; effort is not cancelled (being carried along cancels the external arranger, not the present effort; water flows with no helmsman, which is hardly lying flat). How responsibility and blame are to be re-situated is an open question. And this paper itself merely draws one more red circle—it circles “the red circle” as the most salient thing right now; once it has pointed, discard it.

本文承接一个已被坐实的结论，并据一记新洞见把它推到根上。已坐实的结论：责任的认定，并非真的基于“对不同可能性的设想”。交通事故是最清楚的例子——在场与不在场者、修路者、造车者、乃至司机所服药的厂商，每一处只要不同、事故本都可避免，他们的反事实同样成立；既然在“反事实是否成立”这一关上彼此无别，那么独认司机有责、忽略其余，就不可能是这一关筛出来的。做出区分的，是反事实之外的另一样东西——显然性：哪条因果线最容易设想、最现成。于是“最显然者”被当成了“唯一有罪者”。——而且，连归责所倚的那块地基“他自由地选错了”也同样对称：定罪司机，是把罪系于他的一个自由错选（比如前一晚没多睡一小时），可所有反事实同样成立的关联者（修路者、造车者、药厂……）按同一标准也都“自由地选错了”、因而同样有罪，我们却只肯替司机论证其“自由错选”、从不替其余论证；可见连这条最讲原则的归责理由，也对称地人人都有，独被显然性单单为一个人浇筑。更彻底地说，根本没有“自由地选错了”这件独立事——“自由”只是文化对任何“人选错了”的自动盖章；故归责唯一在做的，就是把谁的哪个行为梦成一个“错选”，一旦梦成，“自由”与“责任”当场自动挂上，于是整桩归责无非“梦谁的行为为错选”的问题，即前述的显然性与钦定。新洞见把问题推到根上：可显然性是哪里来的？世界没给（诸反事实对称），凭什么我们都恰好把司机、而非修路者，感知为“显然”？答案是：它被供给、被统一了——而供给与统一它的，正是伦理、道德、法律。本文据此提出：这三者的真正作用，不是它们自称的“凭说理与分析找出真正该负责的人”（这件事它们做不到：宏观加总固然组合爆炸而不可能；而在交通事故这类“变一处即可避免”的例子里更不必动用此重炮——单点反事实唾手可得，轻易就能造出许多关联者各负其责的对称之梦，仅凭这些梦根本推不出“唯司机该负责”，可见此判决非由论证得出、而是从外加上），而是——给我们洗脑，向我们推荐、营销显然性，并使我们把最显然的，感知为唯一。 它们不靠论证让你信服（论证做不到），而靠反复、靠权威，把一张“哪个选择要紧、哪个人该负责”的显然性地图，一遍遍装进每个人心里，直到它感觉起来天经地义、不证自明；它们抬高某些归责设想、埋掉另一些（修路者、造车者、生成情境的条件），这正是“营销”——推销一部分显然性、下架其余。而须再推狠一层：那几条归责反事实其实彼此同样显然（“司机本可多睡一小时”与“修路者本可修宽些”一样现成好设想），故这与其说是营销，不如说近乎钦定——在一组等价的反事实创作里，凭权威强行规定唯独哪一种归责是被允许的、政治上正确的，把其余同样成立的逐出“可认真对待”之外；故司机之“唯一有罪”，底色不是认知的（他并不更扎眼），而是规定的、政治的。而最深的一层，是它们把“显然”做成了“唯一”：抹去“这里本有一场排序、一次剪枝”的痕迹，让那个幸存下来的靶子，感觉起来就是从来唯一的那个、是从世界里读出来的客观事实，而非一张被装进来的地图上的头名。本文须把分寸守到最足：说这是“洗脑、营销显然性”，不是说“于是无人该负责、法律不正当、道德是骗局、可随意伤害、怎样都行”，也不是说背后有谁在蓄意操纵——“洗脑”在此指一种结构性的功能效果，未必有任何主使者。而且必须公允地看到：把显然性统一起来、装成共享的自动反应，恰恰是任何协作性的道德与法律生活得以可能的前提——没有共享的、自动的“什么要紧、谁负责”，社会根本无法协调运转。所以这同一套机制，既是社会得以成立的经纬，也是那桩错觉（把装进来的显然性当成客观唯一的真）的来源；本文揭的是后者，不否认前者。真在发生的痛与伤害是真的，协调、防范、保护的功能也是真的；“显然性被误当真”这个地基既被看穿，责任该如何重新安放，是开放问题，本文不裁断、更不为任何伤害开脱。反身：本文自己也在营销一种显然性——它把“显然性之营销”做成了此刻最显眼的那个东西，请你把它这幅图感知为真；它指出洗脑，自己也在往你心里装一张图，逃不出它所指认的那套机制。指完即弃。 Abstract This paper takes up a result already secured and, on a new insight, pushes it to its root. The secured result: the attribution of responsibility is not really based on “the supposition of different possibilities.” The traffic accident is the clearest example—those present and absent, the road’s builders, the car’s makers, even the maker of the medicine the driver had taken: at each point, had it been different, the accident could have been avoided, so their counterfactuals are equally valid; and since at the test of “does the counterfactual hold” they are indistinguishable, to pin guilt on the driver alone and ignore the rest cannot be what that test filtered out. What does the distinguishing is something other than the counterfactual—salience: which causal line is most easily imagined, most ready-to-hand. So “the most salient” is taken for “the only guilty one.” —And even blame’s very ground, “he freely chose wrongly,” is symmetric: convicting the driver pins guilt on a free choice of his (he chose wrongly—e.g., didn’t sleep an hour more the night before), yet all the equally-counterfactually-valid related persons (road-builder, car-maker, drug-maker…), by the same standard, also “freely chose wrongly” and are equally guilty, yet we only ever build the “freely chose wrongly” case for the driver, never for the rest—so even this most principled-seeming ground of blame is symmetric across everyone, poured by salience for one person alone. More radically, there is no separate fact of “freely chose wrongly” at all—“free” is merely culture’s automatic stamp on any “a person chose wrongly”; so the only thing blame really does is to construct/dream whose-which behavior as a “wrong choice,” upon which “free” and “responsibility” attach automatically—whence the whole of blame is nothing but the question of “whose behavior we dream as a wrong choice,” i.e. the salience and fiat already described. The new insight pushes the question to the root: but where does salience come from? The world does not supply it (the counterfactuals are symmetric); why do we all happen to perceive the driver, rather than the road-builder, as “obvious”? The answer: it is supplied and standardized—and what supplies and standardizes it is ethics, morality, and law. Hence the paper’s claim: the true function of these three is not their professed “finding, by reasoning and analysis, the one truly responsible” (which they cannot do: the macro-aggregate is indeed combinatorially impossible to analyze; yet in a traffic-accident case—where changing any one point averts it—one need not even wield that heavy weapon, since the single-point counterfactuals are readily at hand and one can easily construct many symmetric dreams of various persons each bearing responsibility, from which “only the driver is responsible” simply does not follow—so this verdict is not derived from the arguments but added from outside), but rather—to brainwash us, to recommend and market salience to us, and to make us perceive the most salient as the unique. They do not convince you by argument (argument cannot do it) but, by repetition and authority, install in everyone a salience-map of “which choice matters, which person is responsible,” again and again, until it feels self-evident; they elevate certain attributions and bury others (the road-builder, the car-maker, the situation-generating conditions)—this is the “marketing”: pushing some salience, taking the rest off the shelf. And a sharper turn: those guilt-counterfactuals are in fact equally obvious (“the driver could have slept an hour more” is no less ready-to-hand than “the road-builder could have built it wider”), so this is less marketing than near-fiat—among a set of equivalent counterfactual creations, by authority forcibly stipulating which single attribution alone is permitted, is politically correct, and expelling the rest, equally valid, from “what may be taken seriously”; so the driver’s “sole guilt” is, at bottom, not cognitive (his line is not really more salient) but stipulated and political. And the deepest layer: they make “the salient” into “the unique”—erasing the trace that “here there was a ranking, a pruning,” so that the surviving target feels like the one there always uniquely was, an objective fact read off the world, rather than the top entry on an installed map. The paper keeps its measure to the utmost: to call this “brainwashing, the marketing of salience” is not to say “therefore no one is responsible, the law is illegitimate, morality is a scam, one may harm at will, anything goes,” nor to say that someone is deliberately manipulating behind the scenes—“brainwashing” here names a structural, functional effect, with no necessary mastermind. And one must see, in fairness: standardizing salience, installing it as a shared automatic response, is precisely the precondition for any cooperative moral and legal life—without a shared, automatic “what matters, who is responsible,” society cannot coordinate at all. So this same mechanism is both the warp and weft by which society is possible and the source of the illusion (taking installed salience for objective, unique truth); the paper exposes the latter without denying the former. The pain and harm that really occur are real; the functions of coordination, prevention, and protection are real; with the foundation (“salience mistaken for truth”) seen through, how responsibility is to be re-situated is an open question, not adjudicated here, and no license is offered for any harm. Reflexively: this paper too markets a salience—it makes “the marketing of salience” the most salient thing right now and asks you to perceive its picture as true; it points at brainwashing while itself installing a map in you, unable to step outside the very mechanism it names. To be discarded once it has pointed.

本文把“归罪”一脉的两层合为一篇：其上层，是伦理、道德与法律的主流思考方式所立的结构；其下层，是这结构在现实中实际如何运作。上层（结构）：这套主流思考方式做两件环环相扣的事——其一，苛求人做出正确、最佳、最合理的选择；其二，不遗余力地证明，正是那选择之错误，导致了世界演化的恶果。而这两件事各预设一个站不住的东西。苛求正确选择，预设一个能站在条件之外、本可选对的择子者——可没有那个可选择的主体（“我”不过是“本可不同”的栖身处），“他本可选对”之可想象是细微条件不透明所造的错觉，连供他选对的菜单本身都是条件造的，故苛求他命中“正确”，等于要他站到自己条件之外。把恶果归咎于错误选择，预设一条从“错误选择”干净通到“恶果”的因果线——可因果是覆在相继之上、再外投的故事，“导致”承重于一个从未发生的“本可避免”分支，而“恶果”里的“恶”又是拿发生去比一个“本该更好”的没发生版本读出的评判；发生本身不会错。下层（实操）：这套结构假装自己在“分析谁有罪”，可真正的全盘分析——把一桩事里所有关联之人的所有决定、所有选项、连同其种种联合的“本可不同”逐一罗列掂量——在组合上根本不可能（梦的数目指数爆炸，无人枚举得了）。于是归罪退化成一个偷工的替代品——但未必是“先定罪、再圆梦”那种犬儒的两步；更可能与分析自己时同理：在牵涉的众人中，那个最容易把一条因果线设想着归属于他的人，凭这份“易于设想”自己浮上台面、被当成最可疑的真凶——“择出靶子”与“建起那个证其有罪之梦”不是先后两步，是梦的易于成形本身完成了择出，谁的“他若不那样则恶果可避免”最顺口现成，谁就被这份顺口选中。同一个错觉，在私下自责里一模一样地运转：复盘时一下子端出的那个最扎眼的悔点，凭其脱口而出的容易，被误当成我“唯一”真正的过错——其实可梦出“更好后果”的点多得是，只在内部显然性排序里分高下，最显然者赢了一场我都没察觉的内部海选、遂冒充唯一。两处共用一个引擎：显然冒充唯一，靠剪枝之不可见。 因我们只端得出极少几个梦，那道把“不那么显然、却永不该被排除”的其余设想无声剪掉的工序，自己感觉不到；台面只剩一个、剪掉的隐于幕后，于是幸存者看似唯一。证其有罪之梦之所以显得铁证如山，不因它真击败了所有对手，而因所有对手都没被请上场；它的真身只是“貌似最可能”，被当成了“唯一可能”。（最显眼的例子是交通事故的定罪：在场与不在场者、修路者、造车者、乃至司机所服药的厂商，每一处只要不同、事故本都可避免，可罪独落司机——不因唯他的因果线经穷尽比较而胜出，只因它最显然、最易设想，其余同样成立的证罪之梦因不够显然被无声剪除。文中“梦”即反事实设想。）本文守一处分寸：这是结构性的诊断，不是处方。指出归罪的地基（条件之外的择子者、干净的因果线）站不住、其确凿感来自剪枝而非穷尽分析，绝不等于说“于是无人该负责、法律不正当、自省无意义、可随意伤害、怎样都行”；那真在发生的痛与伤害是真的，伦理、法律、自省、防范的功能也是真的。就自责，附带一个朴素而准确的提醒：你死死攥住的那处“当初要是不那样”，是多中之一、不是你独力毁掉一切的唯一铁证。地基既崩，责任、惩戒、悔过如何重新安放，是尚待研究的开放问题，本文不裁断、更不为任何伤害开脱；伦理诸传统各异，所诊断的是占主流的、以“选择之对错”与“选择之因果归责”为中心的那一种。本文自己，也是一次剪枝与端出——它端出这一个最显然的诊断，其余它也没能尽列。 Abstract This paper merges two layers of the “blame” lineage into one: the upper layer, the structure erected by the mainstream mode of ethics, morality, and law; the lower layer, how that structure actually operates in practice. Upper layer (the structure): the mainstream mode does two interlocking things—first, it demands that a person make the correct, best, most reasonable choice; second, it spares no effort to prove that it was the wrongness of that choice which caused the bad outcomes of the world’s unfolding. Each of these presupposes something untenable. Demanding the correct choice presupposes a chooser able to stand outside the conditions and to have chosen rightly—but there is no such choosing subject (the “I” is merely the dwelling of “could-have-been-otherwise”), the imaginability of “he could have chosen rightly” is an illusion made by the opacity of the fine conditions, and even the menu on which he was to choose rightly is itself made by conditions, so to demand he hit upon “the correct” is to demand he stand outside his own conditions. Pinning the bad outcome on the wrong choice presupposes a clean causal line running from “the wrong choice” to “the bad outcome”—but causation is a story laid over succession and projected outward, “caused” rests on a never-realized “could-have-been-avoided” branch, and the “bad” in “bad outcome” is a verdict read off by comparing the happening with an unrealized version that “should have been better”; the happening itself cannot be wrong. Lower layer (the operation): this structure pretends to be “analyzing who is guilty,” yet the genuine full analysis—enumerating and weighing all the decisions, all the options, of all the persons involved, together with all their joint “could-have-beens”—is combinatorially impossible (the dreams explode exponentially; no one can enumerate them). So blame degenerates into a corner-cutting substitute—but not necessarily the cynical two-step of “convict first, then dream up a justification”; more likely it works just as it does when we analyze ourselves: among the many people involved, the one to whom a causal line is most easily imagined to belong surfaces of itself, by that very ease, as “the most suspect”—“selecting the target” and “building the guilt-proving dream” are not two steps in sequence; the dream’s ease of formation is itself what does the selecting, and whoever’s “had he not, the harm would have been avoided” is the most ready and fluent is the one that fluency picks out. The same illusion runs in private self-blame: the most salient regret-point, brought forth so effortlessly on review, is mistaken for my “unique,” true fault—whereas the points for which a “better outcome” can be dreamed up are many, differing only in rank within an internal ordering of salience, and the most salient wins an unnoticed internal contest and thereby masquerades as the unique. Both share one engine: the salient masquerades as the unique, by the invisibility of pruning. Because we can bring forth only a very few dreams, the step that silently prunes away the “less salient but never-to-be-excluded” remaining suppositions goes unfelt; only one is left on the table, the pruned remain backstage, and the survivor looks like the only one. The guilt-proving dream seems airtight not because it truly defeated all rivals but because no rival was ever brought onto the field; its true identity is merely “seemingly most likely,” taken for “the only possibility.” (The clearest example is conviction in a traffic accident: those present and absent, the road’s builders, the car’s makers, even the maker of the medicine the driver had taken—at each point, had it been different, the accident could have been avoided, yet guilt falls on the driver alone, not because his causal line won an exhaustive comparison but because it is the most salient and most easily imagined, while the equally valid guilt-dreams of the rest are silently pruned for being less obvious. Throughout, “dream” means counterfactual supposition.) The paper keeps a measure: this is a structural diagnosis, not a prescription. To point out that blame’s foundation (an out-of-condition chooser, a clean causal line) is untenable, and that its air of certainty comes from pruning rather than exhaustive analysis, is in no way to say “therefore no one is responsible, the law is illegitimate, self-reflection is pointless, one may harm at will, anything goes”; the pain and harm that really occur are real, and the functions of ethics, law, self-reflection, and prevention are real. As for self-blame, a plain and accurate reminder: the “if only I hadn’t” you keep clutching is one among many, not the sole proof that you single-handedly ruined everything. With the foundation collapsed, how responsibility, sanction, and remorse are to be re-situated is an open question for further inquiry, not adjudicated here, and no license is offered for any harm; ethical traditions differ, and what is diagnosed is the dominant one, centered on “the rightness/wrongness of the choice” and “the causal imputation of outcomes to the choice.” This paper itself is also a pruning and a bringing-forth—it brings forth this one most salient diagnosis, and the rest it, too, has failed to list in full.

本文把“加总”这一簇思考合为一篇：我们把宇宙宏大的状态转移，讲成系于人类的“决定”，给单个决定派上意义、却把它们的加总记为零；本文层层下探，给出加总归零的四重理由，而最深的一重，是“决定”根本不是一个可加的单元。第一重（现象）：我们能轻而易举地讲述单个决定的意义，却从来指不出两个无关决定的“共同意义”；后一刻的宇宙状态本系于前一刻状态之全部，可我们的叙事只会一次照亮一个显眼单元、把其余整片背景清零。第二重（意义之为梦）：一个决定的“意义”，是把可见的实际后果，与那个未被选选项的“梦”——一段程式化生成的反事实——相比较而定义的；故加总诸决定的意义，实质是整合它们的梦，而无关之梦只能并置（不生新义）、唯有关之梦可焊为一梦。第三重（组合爆炸）：纵然假定每个决定是个固定单元，梦也算不动——把 k 个各 n 选项的决定联合起来，联合的反事实情形有 n^k 种，除去实际发生的那一种，便有 n^k − 1 个联合之梦（两个各四选项的决定即 4×4−1＝15 个），随决定数目指数爆炸；人只好把每个决定偷偷砍成“两个最显著选项”，这让单个、少数决定的意义重新可算、却救不了整体（即便二元化，整体仍是 2^k − 1、依旧算不动），且这道砍法随尺度愈发失真——总不能设想七十亿人的决定齐刷刷只有两个选项。第四重，也是最深一重（决定之为投射）：上述全部，都预设了“决定”是一个形式化的、固定的单元——可它不是。设七十亿人上一决定各四选项，麻烦立刻来了：甲的上一决定在一分钟前、乙的在一秒钟前，而乙最近一分钟做了十个决定，你如何整合甲乙最近一分钟内的决定？决定本身就是臆造之物：我尽可说甲在最近一分钟做了一万个关于身体姿态的决定。决定不是一种形式化的固定之物，而是观察者主观的投射——它的起止、粒度、有几个选项、算不算一个，全由观察者划定，没有世界那一侧的事实可凭。于是“加总决定”从一开始就不是一个良定义的操作：既数不清（一个还是一万个）、又对不齐（甲乙不在同一节拍上），根本没有固定的“决定”可供相加。涌动本身没有名为“决定”的关节，是我们把关节画上去的。四重叠加，加总必然归零——而最深的归零，不是“难加”或“加不出意义”，是没有可加之物。本文守一处分寸：说决定是投射、加总归零，绝不是说什么都没发生、努力无用；那场可见的涌动是完整而真实的，无论我们把它切成一个决定还是一万个，神经与行为的活动都照旧真实地发生着——被揭出的，只是“一个决定”这个边界与单元是我们划上去的，以及建在其上的“意义”是一个局部的、简化的、且单元都不固定的投射性渲染。你此刻的用力，无论被算作几个决定，都照旧在这涌动之中。意义如何看待，仍是开放问题。本文自己，也是一道把涌动切成“决定”“加总”来谈的投射——它谈论投射，自己也在投射。 Abstract This paper combines the “aggregation” cluster into one: we narrate the universe’s vast state-transition as hinging on human “decisions,” assign meaning to single decisions, yet record their aggregate as zero; the paper descends layer by layer, giving four reasons for the aggregation’s collapse to zero, the deepest being that “decision” is not an aggregable unit at all. First layer (the phenomenon): we can effortlessly narrate the meaning of a single decision, yet can never point to the “joint meaning” of two unrelated decisions; the next moment’s cosmic state hinges on the entirety of the prior moment’s state, but our narration can only light up one salient unit at a time and zeroes the rest of the background. Second layer (meaning as dream): a decision’s “meaning” is defined by comparing the visible actual outcome with the “dream” of the unchosen option—a procedurally generated counterfactual; so aggregating decisions’ meanings is integrating their dreams, and unrelated dreams can only be juxtaposed (breeding no new meaning), while only related dreams can be welded into one. Third layer (combinatorial explosion): even granting each decision a fixed unit, the dreams cannot be computed—joining k decisions with n options each, the joint counterfactual situations number n^k, and removing the one actual leaves n^k − 1 joint dreams (two four-option decisions: 4×4−1 = 15), exploding exponentially; the human cuts each decision down to “the two most salient options,” which makes single and small-cluster meanings computable again but cannot rescue the whole (even binarized, the whole is still 2^k − 1, still beyond reach), a cut that grows more distorting with scale—one cannot suppose seven billion people’s decisions each had exactly two options. Fourth and deepest layer (decision as projection): all of the above presupposes “decision” to be a formalized, fixed unit—but it is not. Suppose seven billion people’s last decisions each have four options; trouble arrives at once: A’s last decision was a minute ago, B’s a second ago, and B made ten decisions in the last minute—how do you integrate A’s and B’s decisions over the last minute? A decision is itself a fabrication: I may as well say A made ten thousand decisions about body posture in the last minute. A decision is not a formalized fixed thing but the observer’s subjective projection—its start and end, its grain, how many options it has, whether it counts as one, are all drawn by the observer, with no fact on the world’s side to go by. So “aggregating decisions” was never a well-defined operation: uncountable (one, or ten thousand?), unalignable (A and B are not on the same beat), with no fixed “decisions” to add at all. The welling has no joints called “decisions”; we paint the joints on. Fourfold overlaid, the aggregation must come to zero—and the deepest zero is not “hard to add” or “yields no meaning,” but there is nothing to add. The paper keeps a measure: to say the decision is a projection and the aggregate is zero is in no way to say nothing happens or that effort is useless; the visible welling is complete and real, and whether we cut it into one decision or ten thousand, the neural and behavioral activity really occurs—what is exposed is only that “a decision,” as boundary and unit, is drawn on by us, and that the “meaning” built upon it is a local, simplified, projective rendering on a unit that is not even fixed. Your effort now, however many decisions it is counted as, remains within this welling. How meaning is to be regarded remains an open question. This paper, too, is a projection that cuts the welling into “decisions” and “aggregation” to speak of them—it speaks of projection while itself projecting.

本文揭出我们赋予“人类选择”以意义与后果的方式，是一种选择性的、经不起加总的捏造。宇宙是一场宏大的状态转移，而我们却只盯着其中的人类选择——从宇宙中性提供的几个选项里的选择——并把整场宇宙演化，解释成系于这些可疑而微弱的“单选题”之上：仿佛渺小如蚁的人类个体头脑里的某个拨动，恰恰造成了此刻乃至未来的宇宙状态。更反常的是，我们还能给这极细微的内在拨动安上显著的意义，仿佛它比山崩海啸还来得“有效”——因为山崩、海啸、乃至一片树叶的扇动，都被解释成中性的背景或被动的结果，于是无从呈现其意义。本文用一记加总的反诘戳穿这选择性：如果一片树叶的摆动是无意义的背景噪声、而一个人的某个决定却有显著的意义与后果，那么两个无关之人的两个决定，是否也该共同造成某种显著的意义或后果？为什么我们从来指不出这种共同的意义或后果，却能轻而易举地指出每个决定各自的意义与后果？这一不对称暴露出：“后果”并非从宇宙的状态转移里读出的客观属性（后一刻的状态本就系于前一刻状态的全部、而非其中某一个决定），而是一台只会一次照亮一个显眼单元的叙事机器强加上去的。进一步：当我们指出某决定的意义与后果时，可曾意识到、可曾表述过——在那决定之前，每个人的每个决定对“该决定之前的宇宙整体状态”的贡献？那些在先决定的后果难道烟消云散了吗？没有；它们正是那决定所坐落其上的、被我们当作中性给定的整个条件。要害在于：我们善于表达、乃至捏造单个决定的后果，却难于思考它们的加总——而凡难于思考的，便被我们默认为零。于是那真正决定后一刻状态的“全部之和”被清零，全部的因果与意义之重，都被倾倒在那一个被叙事机器照亮的决定上。意义不是均匀铺在宇宙演化上的，是被探照灯选择性地打出来的：照亮一个人类拨动，清零其余整片背景，于是一个蝼蚁般的拨动，竟显得比山崩海啸更有效。本文守一处分寸：说后果是被选择性地打出来的，不是说决定没有发生、不是说“一切皆无意义、努力无用、怎样都行”；那场宇宙的状态转移是真的，决定作为其中的环节也真在发生，你此刻的用力同样是这转移此刻的内容——被揭出的，只是“把意义与后果选择性地钉在单个决定上、而把其余清零”这桩捏造与这道探照灯。如何、是否、向何处重新分配意义，是一个开放问题，本文不裁断。而本文这套指认，自己也是一束探照灯、一次照亮——它照亮了“探照灯”，照不亮全部。 Abstract This paper exposes that the way we assign significance and consequence to “human choice” is a selective fabrication that cannot survive aggregation. The universe is a vast state-transition, yet we fixate only on the human choices within it—choices from among the few options the universe “neutrally provides”—and explain the whole cosmic unfolding as hinging on these dubious, feeble “multiple-choice questions”: as if some flick in the head of an ant-tiny human individual is precisely what brings about the present and even future state of the cosmos. More perversely still, we can attach significant meaning to this minute inner flick, as if it were more “efficacious” than a landslide or a tsunami—because landslides, tsunamis, even the flutter of a single leaf, are all construed as neutral background or passive result, and so cannot have their meaning shown. The paper punctures this selectivity with a reductio from aggregation: if a leaf’s swaying is meaningless background noise while a person’s particular decision has significant meaning and consequence, then should the two decisions of two unrelated persons also jointly cause some significant meaning or consequence? Why can we never point to such a joint meaning or consequence, yet effortlessly point to each decision’s separate meaning and consequence? This asymmetry exposes that “consequence” is not an objective property read off the cosmic state-transition (the next moment’s state hinges on the entirety of the prior moment’s state, not on some one decision within it) but is imposed by a narrative machine that can only light up one salient unit at a time. Further: when we point out a decision’s meaning and consequence, have we ever been aware of, ever articulated, the contribution of every prior decision by everyone to “the overall cosmic state before that decision”? Have the consequences of those prior decisions vanished into thin air? No; they are precisely the whole condition, treated by us as a neutral given, upon which that decision sits. The crux: we are good at expressing, even fabricating, the consequence of a single decision, but find it hard to think their aggregate—and whatever is hard to think, we default to zero. So the “sum of everything” that actually fixes the next moment’s state is zeroed out, and the entire weight of causation and meaning is poured onto the one decision the narrative machine has lit up. Meaning is not spread evenly over the cosmic unfolding; it is selectively projected by a spotlight: light up one human flick, zero the rest of the background, and an ant-like flick comes to seem more efficacious than a landslide or a tsunami. The paper keeps a measure: to say consequence is selectively spotlit is not to say the decision did not occur, nor to say “nothing means anything, effort is useless, anything goes”; the cosmic state-transition is real, the decision really occurs as a link within it, and your effort now is likewise the content of this transition now—what is exposed is only the fabrication and the spotlight that pin meaning and consequence onto a single decision while zeroing the rest. Whether, how, and to where meaning should be redistributed is an open question, not adjudicated here. And this very identification is itself a spotlight, a lighting-up—it lights up “the spotlight,” and cannot light up the whole.

本文承接“盯着选择、放过条件”这一盲点的诊断，把它从自由意志辩论，扩到伦理、道德与法律的主流思考方式——后者同样难辞其咎，且因其实践力量而远为深重。其核心动作有二：其一，苛求人做出正确、最佳、最合理的选择；其二，不遗余力地证明，正是那个选择之错误，导致了世界演化中的恶果。本文指出，这两件事各预设了一个已被深究过、其实并不成立的东西。第一桩（苛求正确选择）预设了一个能站在条件之外、本可选出最优的择子者：可没有这样一个可选择的主体（“我”不过是“本可不同”的栖身处），“他本可选对”之可想象又是细微条件不透明所造的错觉，而且那份“供他选对”的选项菜单，本身就是先前条件造的——苛求一个人，在条件造好的菜单上、用一个被条件定死的“选择”去命中“正确”，等于要求他站到自己的条件之外，而那是做不到的。第二桩（把恶果归咎于错误选择）预设了一条从“错误选择”干净地通到“恶果”的因果线：可“因果”本是我们覆在“相继”之上、再外投出去的故事，世界只是把它唯一的一支演化了出来；把其中一段判作“恶果”、并认定它是某个“错误选择”的报应，是一道承重于反事实（“本可避免”）的拟合，而“恶果”里那个“恶”，又是拿发生去比一个“本该更好”的没发生的版本、读出的评判——发生本身不会错，会错的只有我们套上去的概念。于是这套主流思考方式，是那个盲点的、更有力也更难辞其咎的版本：自由意志之争只是在“选择自不自由”上争论，而伦理、道德与法律则把同一道切分——前台是“选择”、背景是“条件”——付诸实践、加以执行，用道德谴责与法律惩罚的全部分量，把世界的恶果一次次钉在“错误的选择”上，而那些造出并框定了选项、又定死了选择的条件，照旧无人过问。本文守一处至关重要的分寸：这是结构性的诊断，不是处方。 指出这套思考方式预设了一个站不住的地基，绝不等于说“于是无人该负责、法律不正当、可以随意伤害、怎样都行”；那真在发生的痛与伤害是真的（在发生那一侧），伦理与法律在协调社会、防止伤害上的功能也是真的；当“条件之外的择子者”这个地基崩塌之后，责任、惩戒、防范该如何重新安放，是一个尚待研究的开放问题，本文不在此裁断，更不提供任何为伤害开脱的许可。本文亦不一概而论：伦理诸传统各异，所诊断的，是其占主流的、以“选择之对错”与“选择之因果归责”为中心的那一种。末了反身：本文这套指认，也是站在一旁的一道切分，它能指出审视被钉在了哪里，替不了谁去重新安放。 Abstract This paper takes up the diagnosis of the “fixate on the choice, pass over the conditions” blind spot and extends it from the free-will debate to the mainstream modes of ethics, morality, and law—which are no less culpable, and, by their practical force, far more deeply so. Their core moves are two: first, to demand that a person make the correct, best, most reasonable choice; second, to spare no effort to prove that it was the wrongness of that choice which caused the bad outcomes in the world’s unfolding. This paper shows that each move presupposes something already examined in depth and in fact untenable. The first (demanding the correct choice) presupposes a chooser able to stand outside the conditions and to have selected the optimum: but there is no such choosing subject (the “I” is merely the dwelling of “could-have-been-otherwise”), the imaginability of “he could have chosen rightly” is in turn an illusion made by the opacity of the fine conditions, and the very menu of options “on which he was to choose rightly” is itself made by prior conditions—so to demand that a person, on a menu the conditions built, by a “choice” the conditions fixed, hit upon “the correct” one, is to require him to stand outside his own conditions, which cannot be done. The second (pinning the bad outcome on the wrong choice) presupposes a clean causal line running from “the wrong choice” to “the bad outcome”: but “causation” is itself a story we lay over “succession” and then throw outward, while the world has merely unfolded its one branch; to judge some stretch of it a “bad outcome” and to deem it the desert of some “wrong choice” is a fitting load-bearing on a counterfactual (“it could have been avoided”), and the “bad” in “bad outcome” is in turn a verdict read off by comparing the happening with an unrealized version that “should have been better”—the happening itself cannot be wrong, only the concept we lay over it can. So this mainstream mode of thinking is the more forceful, more culpable version of that blind spot: the free-will debate merely argues over “is the choice free,” whereas ethics, morality, and law put the same carving—foreground “the choice,” background “the conditions”—into practice and enforce it, pinning the world’s bad outcomes again and again on “the wrong choice” with the full weight of moral censure and legal punishment, while the conditions that produced and framed the options, and fixed the choice, go unquestioned as ever. The paper keeps a crucial measure: this is a structural diagnosis, not a prescription. To point out that this mode presupposes an untenable foundation is in no way to say “therefore no one is responsible, the law is illegitimate, one may harm at will, anything goes”; the pain and harm that really occur are real (on the side of the happening), and the functions of ethics and law in coordinating society and preventing harm are real; once the foundation of an “out-of-condition chooser” collapses, how responsibility, sanction, and prevention are to be re-situated is an open question for further inquiry, not adjudicated here, and certainly no license is offered for harm. Nor does the paper paint with one brush: ethical traditions differ; what is diagnosed is the dominant one, centered on “the rightness/wrongness of the choice” and “the causal imputation of outcomes to the choice.” Finally, reflexively: this very identification is itself a carving made from the sidelines—it can point out where scrutiny has been pinned, but cannot do anyone’s re-situating for them.

这是一种朴素的看法，想把几件我们习以为常的事重新看一遍：自我、选择、自由、因果、责怪。它的起点是一句最简单的话——事情就那样一件接一件地发生着，而且只发生这一次，没有另一个“本来可以不一样”的版本悬在旁边。顺着这一句往下看会发现：我们以为“在外面世界里”的很多东西，其实是我们心里造出来、再抛到世界上的故事——因果是故事，“关系”是故事，连“我在自由地选”也是。那个看起来在做主的“我”，不过是“本来可以”这种念头的住处；意愿不是身体的指挥官，而是身体底层活动的一个外号；我们觉得“我本可以选别的”“我想好就做成了”，是因为我们看不见底下真正把事情定下来的细节。我们甚至一直问错了问题：与其纠结“我选得自不自由”，不如问“我面前这几个选项，到底是谁给的”——而那场争了几千年的“自由意志”大辩论，恰恰把这个该问的问题挡在了视野之外。说这些，不是要否定生活，更不是劝人躺平：你此刻的努力，本身就是事情正在发生的样子；水没有舵手，可水照样在流，今天真的接着明天。它也不替你解决一切——比如善恶对错该怎么重新安放，是个更大的、悬而未决的问题。它只是请你把目光，从那些我们抛上去的故事，移回那件一直在安安静静发生着的事情本身。 Abstract This is a plain way of seeing, meant to look afresh at a few things we take for granted: the self, choice, freedom, cause, blame. Its starting point is the simplest of sentences—things just happen, one after another, and they happen only once, with no other “could-have-been-different” version hanging beside them. Follow that out and you find: much of what we take to be “out there in the world” is in fact a story we make inside and throw onto the world—causation is a story, “relations” are a story, even “I am freely choosing” is one. The “I” that seems to be in charge is just the dwelling-place of the thought “could have been otherwise”; the will is not the commander of the body but a nickname for the body’s underlying activity; we feel “I could have chosen otherwise” and “I planned it and made it happen” only because we can’t see the fine details that actually fixed things. We have even been asking the wrong question all along: rather than agonizing over “was my choice free,” we should ask “who handed me these particular options”—and the millennia-long free-will debate is exactly what has kept that better question out of view. To say all this is not to negate life, still less to counsel lying down and doing nothing: your effort right now just is the shape this happening is taking; the water has no helmsman, yet it flows, and today really does continue into tomorrow. Nor does it solve everything—how good and bad, right and wrong are to be re-situated is a larger, still-open question. It only asks you to move your gaze from the stories we throw up, back to the thing that has been quietly happening all along.

我们面对一个选择时，注意力几乎全在“我会怎么选、这一选自不自由”上，而那几个摆在面前的选项，被当成天经地义、不必多问的背景。可仔细想想：为什么偏偏是这几个选项？它们之所以摆在你面前、还显得“差不多就这些了”，全是你过去的经历、处境、乃至此刻整个世界的状态造成的——选项不是凭空给的，是条件给的。更有意思的是，被我们死死追问的那个“我在选”，反倒是整件事里最站不住脚的一环：把“我”换成大脑里一串神经活动来看，你根本不会觉得它“本来能选别的”。于是其实并没有“现成的菜单＋自由的挑选”这两段，只有一条被条件推着走的过程；我们之所以只盯着“选”、放过“选项从哪来”，只因为前者被我们贴上了“这是我做的”。而几千年的“自由意志”大辩论，恰恰把这个盲点越擦越亮：不管你主张选择自由、不自由、还是“看怎么定义”，大家都在争“选择自不自由”，没人回头问“这些选项是谁给的”。真正的出路，不是在这场争论里赢一局，而是干脆下桌——把目光从“怎么选”，移回那个一直在背后做主、却被所有人当背景的东西：条件。 Abstract When we face a choice, almost all our attention goes to “which will I pick, and is the picking free?”—while the options sitting in front of us are taken as a given backdrop, not worth questioning. But think about it: why exactly these options? That they sit before you, and that they feel like “pretty much all there is,” is entirely the work of your past, your situation, and indeed the whole state of the world at this moment—the options aren’t handed over from nowhere; they’re handed over by conditions. More striking still, the very “I am choosing” we interrogate so fiercely is the shakiest link of all: describe the “I” as a run of neural activity, and you simply won’t feel it “could have picked otherwise.” So there really aren’t two stages, “a ready-made menu plus a free pick”—only one process pushed along by conditions; and we fixate on the “picking” while ignoring “where the options came from” only because we’ve stamped the former “this is mine.” The millennia-long free-will debate has only polished this blind spot brighter: whether you argue the choice is free, unfree, or “depends how you define it,” everyone is arguing about whether the choice is free, and no one turns to ask who handed over the options. The way out isn’t to win a round of that argument but to leave the table—to move our gaze from “how to choose” back to the thing that has been calling the shots all along while everyone treated it as background: the conditions.

本文把两层一桩相承的洞见整合为一篇，并补入相关洞见以使阐释完整：对象层是面对选择时一桩审视的错配，元层是供养这错配的文化活动——自由意志之争。对象层。 面对一个选择，现象上选项仿佛是被给予的中性背景，而“我的选择”才是重要、真正、值得追问自由与否的焦点；于是审视全压在“选择动作”上，选项被当作不必思考的默认背景。但那几个选项、连同其“看似合理且覆盖完全”的样子，本身是先前条件的产物——是哪几个选项、它们为何显得完备，皆由在先条件决定（完备感是被条件造的盲、非中性事实），而“我有这几个备选项”的这一刻的宇宙状态本身，就是决定后续的条件；菜单不过是认知把情境切成离散备选的一道相之切分，被误当成世界递来的中性给定。补入的依据。 而那个被执着追问的“选择”，恰恰是整桩事里最“似乎存在”、最经不起细看的一步：消主体即消反事实——“我”不过是“本可不同”的栖身处，可无损改写为信号流动，你不会设想它做出不同的输出；“在同样条件下我本可选 B”之可想象，又是细微决定者不透明所造的错觉；而把“选择之自由”兑现成一个真实之因，只会撞回那穷尽的二分（无因的果，或非物质的因），皆空。于是并无“给定的菜单＋施于其上的自由选择”，只有一条被条件推着走的相继：在先条件→情境被切分呈现为看似完备的选项→随之而来的“选择”→下一状态；菜单与选择同在条件下游，把二者切成“给定的背景”与“我的真正动作”是一道拟合，连“选项是给的、选择是我的”这种感觉本身也是被条件塑造的表象。盲点的引擎。 何以单单放过条件、执着质疑选择：自模型把“我的动作”推到前台、打上“我”的标签，把“世界递给我的”推到背景；于是在意全聚于挂“我”字的选择，归给“世界”的选项滑出追问——可错配恰在此：归给“我”的那一步唯一似乎存在、细看无岔口，归给“世界”、被豁免的条件才是决定实际所在。元层。 这凸显出自由意志辩论活动本身对文化所造的盲点：其三大家（不相容而自由实存、决定论故无自由、二者相容）答案势同水火，却共享“意愿／选择是否自由”这一从不被质疑的问题框架，默认“选择动作”是该追问的单元、而选项与条件是不被过问的舞台；这场延续久远、享声望的活动凭其持续与权威，把这道“前台／背景”切分一遍遍刻进文化，训练人追问“他是不是自由地选的（该不该负责）”、却几乎不问“什么条件造出了他只剩这几条路的情境”——于是文化得了一个盲点：目光钉在似乎存在却无岔口的“选择”上，放过真正做主的条件（其文化后果是把一切个体化地收束到选择、把生成情境的条件自然化为给定，本文只作结构性指认、不裁断伦理政治应用）。而连“否定自由意志”这一手，只要仍在回答“选择自不自由”，就仍在这张棋盘上、与证明“自由”者共享同一盲点；真正的解套不是赢得争论，而是下桌——不再设此问，看清“主体—给定选项—自由选择”每一块都是同一条被条件推着走的发生里的环节，那问题从一开始就框错。本文守一处分寸：说选项是条件造的、辩论造了盲点，不是说选项、审议、选择是幻觉，也不是说不该思考能动——它们作为相、作为发生都真在发生，被揭出的只是“把菜单当中性背景”这桩框定、“审视全压在似乎存在的选择上”这桩错配、与“把它立为唯一该问之问题”这个框架；而本文这套诊断，连同“辩论造了盲点”这句，也是一道从网内做出的切分，能指出那张桌子、不能替你下桌，指完即弃。伦理与体验照旧悬置。 Abstract This paper integrates a two-level, mutually continuous insight into one piece, adding related insights to complete the exposition: at the object level, a misallocation of scrutiny when facing a choice; at the meta level, the cultural activity that sustains that misallocation—the free-will debate. Object level. Facing a choice, the options seem given, a neutral background, while “my choosing” is the important, real, interrogation-worthy focus (free or not?); so all scrutiny is loaded onto the “act of choosing,” the options taken as a default background not needing thought. But those options, together with their look of being “reasonable and exhaustively covering,” are themselves products of prior conditions—which options, and why they seem complete, are both fixed by prior conditions (the sense of completeness is a condition-made blindness, not a neutral fact)—and the state of the universe at the moment I “have these alternatives” is itself the condition that fixes what follows; the menu is merely a carving-into-相 by which cognition cuts the situation into discrete alternatives, mistaken for a neutral given handed over by the world. The added grounding. And the “choice” we obsessively interrogate is precisely the most “seemingly-existent,” least scrutiny-surviving step of the whole affair: eliminating the subject just is eliminating the counterfactual—the “I” is the dwelling of “could-have-been-otherwise,” losslessly rewritable as signal-flow, and you will not suppose it producing a different output; the imaginability of “under the same conditions I could have picked B” is in turn an illusion made by the opacity of the fine determiner; and cashing “the freedom of the choice” into a real cause only slams back into the exhaustive dichotomy (an uncaused effect, or a non-material cause), both empty. So there is no “given menu + a free choice applied to it,” only one condition-driven succession: prior conditions → the situation carved and presented as seemingly complete options → the ensuing “choice” → the next state; menu and choice both lie downstream of the conditions, and to cut them into “a given background” and “my real act” is a fitting—even the feeling “the options are given, the choice is mine” is a condition-shaped appearance. The engine of the blind spot. Why single out the conditions to ignore while obsessively interrogating the choice: the self-model pushes “my act” to the foreground, tagging it “mine,” and pushes “what the world handed me” to the background; so concern gathers on the “me”-tagged choice while the world-assigned options slip out of questioning—yet the misallocation is exactly here: the “me”-assigned step is the only seemingly-existent one, fork-less on inspection, while the exempted conditions are where the determining actually lives. Meta level. This throws into relief a blind spot the free-will debate itself causes for culture: its three families (incompatibilist-yet-real, determinist-so-unfree, compatibilist), however irreconcilable their answers, share the never-questioned question-frame “is the will/choice free,” taking the “act of choosing” as the unit to interrogate and the options and conditions as the unquestioned stage; this long-standing, prestigious activity, by its persistence and authority, carves this “foreground/background” division into culture again and again, training us to ask “did he choose freely (and is he responsible)?” yet almost never “what conditions produced the situation in which he had only these paths?”—so culture acquires a blind spot: the gaze pinned on the seemingly-existent, fork-less “choice,” passing over the conditions that actually did the determining (its cultural consequences—individualizing everything onto the choice, naturalizing the situation-generating conditions as a given—the paper flags only structurally, not adjudicating ethical or political applications). And even “denying free will,” so long as it still answers “is the choice free,” still stands on this board and shares the same blind spot with those proving it free; the real release is not to win the argument but to step off the board—to stop posing the question and to see that “subject—given options—free choice” is, in every part, a link in the same condition-driven occurrence, the question mis-framed from the start. The paper keeps a measure: to say the options are condition-made and the debate causes a blind spot is not to say the options, deliberation, or choice are illusory, nor that one should not think about agency—as 相, as occurrence, they all really happen; what is exposed is only the framing “take the menu as a neutral background,” the misallocation “load all scrutiny onto the seemingly-existent choice,” and the frame “make it the sole question worth asking”; and this very diagnosis, the sentence “the debate causes a blind spot” included, is a carving made from within the net—it can point at the table but cannot step off it for you, to be discarded once it has pointed. Ethics and experience remain, as ever, set aside.

本文是一桩对象层洞见的元层延伸。对象层已表明：面对选择，我们把审视全压在“选择动作”上（它自由吗），却把“为什么是这几个选项”当作不必质疑的默认背景——而选项连同其“看似合理且完备”的样子，本身是先前条件的产物，这一刻的宇宙状态就是条件。本文进一步指出：这桩错配，凸显出哲学自由意志辩论活动本身对人类文化所造成的一个盲点。其一，自由意志之争的三大家——主张自由与决定论不相容而自由确实存在者、主张决定论为真故（此义的）自由不存在者、主张二者相容者——尽管答案势同水火，却共享一个从不被质疑的问题框架：争点是“意愿／选择是否自由”，即默认“选择动作”就是安放能动、值得追问的那个单元，而选项、条件、情境则是问题得以提出的、不被过问的舞台。其二，这场辩论作为一种延续久远、享有声望的文化活动，凭其持续与权威，把这个切分一遍遍刻进文化：它训练我们把“选择”当作能动、责任、尊严之所在的前台，把“被给定的选项与条件”当作中性的背景，于是文化习惯于追问“他是不是自由地选的（因而该不该负责）”，却几乎从不追问“为什么他只有这几个选项、是什么条件造出了这个情境”。其三，由此而生的盲点：目光被钉在那个似乎存在、却没有岔口的选择动作上，而真正做了决定的——造出并框定了菜单的条件——则被自然化为不必过问的背景；这一盲点有其文化后果（把一切个体化地收束到“选择”、把生成情境的条件naturalize 为给定），本文只作结构性指认、不裁断其伦理与政治应用。其四，一处自省：连“否定自由意志”这一手，只要它仍在回答“选择自不自由”，就仍站在这张棋盘上、仍把目光留在“选择”上；真正的解套不是赢得这场争论，而是下桌——不再问“选择自不自由”，而看清整个情境、连同选项，都是被条件推着走的发生，那个问题从一开始就框错了。本文守一处分寸：指出辩论造了盲点，不是说辩论无价值、不该思考能动；代价在于其特定框架，不在探究本身；而本文这套诊断，连同“辩论造了盲点”这句，也是一道从网内做出的切分、仍在指那张棋盘——指完即弃。 Abstract This paper is a meta-level extension of an object-level insight. At the object level it has been shown: facing a choice, we load all scrutiny onto the “act of choosing” (is it free?), while taking “why these options” as an unquestioned default background—yet the options, together with their look of being “reasonable and complete,” are themselves products of prior conditions, and the state of the universe at that moment just is the condition. This paper goes further: that misallocation throws into relief a blind spot which the activity of the philosophical free-will debate itself has caused for human culture. (1) The three families of the free-will debate—those holding freedom incompatible with determinism yet real, those holding determinism true so that (this sense of) freedom does not exist, and those holding the two compatible—however irreconcilable their answers, share one never-questioned question-frame: the point at issue is “whether the will/choice is free,” i.e., it is taken for granted that the “act of choosing” is the unit where agency is seated and worth interrogating, while the options, conditions, and situation are the unquestioned stage on which the question is posed. (2) As a long-standing, prestigious cultural activity, the debate, by its persistence and authority, carves this division into culture again and again: it trains us to treat “the choice” as the foreground seat of agency, responsibility, and dignity, and “the given options and conditions” as a neutral background—so culture habitually asks “did he choose freely (and so is he responsible)?” yet almost never asks “why did he have only these options; what conditions produced this situation?” (3) Hence the blind spot: the gaze is pinned on the seemingly-existent, yet fork-less act of choosing, while what actually did the determining—the conditions that produced and framed the menu—is naturalized into an unquestioned background; this blind spot has cultural consequences (individualizing everything onto “the choice,” naturalizing the situation-generating conditions as a given), which the paper only flags structurally, not adjudicating its ethical or political applications. (4) A self-reflection: even “denying free will,” so long as it still answers “is the choice free,” still stands on this board, still keeps its gaze on “the choice”; the real release is not to win the argument but to step off the board—to stop asking “is the choice free” and to see the whole situation, options included, as condition-driven occurrence, the question having been mis-framed from the start. The paper keeps a measure: to point out that the debate causes a blind spot is not to say the debate is worthless or that one should not think about agency; the cost lies in its particular frame, not in inquiry itself; and this very diagnosis, together with the sentence “the debate causes a blind spot,” is itself a carving made from within the net, still pointing at that board—to be discarded once it has pointed.

本文揭出我们思考“选择”时一桩审视的错配。面对一个选择，现象上是：选项仿佛是被给予的、中性的背景，而“我的选择”才是重要的、真正的、值得追问（自由还是不自由）的焦点；于是一切哲学焦虑都压在“选择动作”上，没有一分落到那几个选项上——选项被当作不必思考的默认背景。但本文论证：那几个选项，本身就是先前条件的产物。 我恰好面对这几个选项、且它们呈现为“表面上合理且覆盖完全”的备选集，并不是一个中性的既成事实，而是我的历史、概念、处境、乃至我的认知此刻如何切分这一刻，所共同造成的输出；换一套在先条件，菜单就不同（或根本不以菜单的形式出现）。而且，“我有这几个备选项”的这一刻的宇宙状态本身，就是我的条件——决定后续的，正是这整个状态（选项的呈现也在其中），不是“中性菜单＋自由选择”两段。于是并没有“被给定的菜单＋施于其上的自由选择”，只有一条被条件决定的相继：在先条件→（情境被切分、呈现为一份看似合理而完备的选项）→（随之而来的“选择”）→下一状态；菜单与选择，同在条件的下游，把它们切成“给定的背景”与“我的真正动作”，是一道拟合，连“选项是给的、选择是我的”这种感觉本身也是被条件塑造的表象。最吊诡的是审视的错配：我们把全部追问，倾注在那个似乎存在的“选择动作”上（它自由吗？我本可选别的吗？），却对真正做了决定之事——那些造出并框定了菜单的条件——视而不见、不加质疑；我们盘问那唯一“似乎存在”的一步，却豁免了决定实际所在的整个布景。何以有此盲点：自模型把“我的动作”推到前台、把“世界递给我的”推到背景，于是我们死守、拷问那个“我”的部分（选择），从不想去拷问“世界给了我这些”的部分（选项）——哪怕选项的形状、连同它“看似完备”的样子，和别的一切一样是被条件造的，哪怕这一刻连同选项就是那个条件。本文守一处分寸：说选项是条件造的，不是说选项、审议、选择是幻觉——它们作为相、作为发生，都真在发生；被揭出的只是“把菜单当中性背景”这桩框定、与“审视全压在那似乎存在的选择上”这桩错配。伦理与体验照旧悬置。 Abstract This paper exposes a misallocation of scrutiny in how we think about “choice.” Facing a choice, the phenomenology is: the options seem given, a neutral background, while “my choosing” is the important, real, interrogation-worthy focus (free or not free?); so all the philosophical anxiety is loaded onto the “act of choosing,” with none falling on the options themselves—the options are taken as a default background not needing thought. But this paper argues: those options are themselves products of prior conditions. That I face exactly these options, and that they present as a “seemingly reasonable and exhaustively covering” set of alternatives, is not a neutral fait accompli but an output co-produced by my history, my concepts, my situation, and how my cognition carves this very moment; under a different prior condition the menu would differ (or would not appear as a menu at all). Moreover, the state of the universe at the moment I “have these alternatives” is itself my condition—what determines what follows is precisely this whole state (the presentation of the options included), not “neutral menu + free choice” as two stages. So there is no “given menu + a free choice applied to it,” only one condition-determined succession: prior conditions → (the situation carved and presented as a set of options that look reasonable and complete) → (the “choice” that ensues) → the next state; menu and choice both lie downstream of the conditions, and to cut them into “a given background” and “my real act” is a fitting—even the feeling “the options are given, the choice is mine” is itself a condition-shaped appearance. The paradox is the misallocation: we pour all our questioning onto the seemingly-existent “act of choosing” (is it free? could I have chosen otherwise?), while remaining blind to, and unquestioning of, what actually did the determining—the conditions that produced and framed the menu; we cross-examine the one “seemingly-existent” step and exempt the whole stage where the determination actually lives. Whence the blind spot: the self-model pushes “my act” to the foreground and “what the world handed me” to the background, so we guard and interrogate the “me” part (the choice) and never think to interrogate the “the world gave me these” part (the options)—though the shape of the options, together with their look of completeness, is as condition-made as everything else, and though this moment, options and all, just is that condition. The paper keeps a measure: to say the options are condition-made is not to say the options, the deliberation, or the choice are illusory—as 相, as occurrence, they all really happen; what is exposed is only the framing “take the menu as a neutral background” and the misallocation “load all scrutiny onto the seemingly-existent choice.” Ethics and experience remain, as ever, set aside.

本文把一条长线索的全部高价值洞见，熔为一个整体图景，统于一句：**发生先于我们为它建的一切概念。** 在我们开口之前，世界是一片发生：事情一件接一件地涌着，而且只有这一支——后续是注定的，这“注定”不是有谁预先写好剧本，而是根本没有并排的第二支。在这底座之上，本文沿“涌动”与“我们抛向它的网”这一对张力展开。**其一，认识不触本体，且我们并无通往本体的访问通道。** 传统宿命论剥去外衣，内核是预测能力的夸口、必含一个知道者，本质是认识论命题；而本体与可知性正交（拉普拉斯妖案把“可被预测”误当成“有唯一轨迹”的条件，混沌系统不可预测却仍有其唯一实际轨迹）；本体也独立于被表达、被判断（罗生门里各方叙述相互冲突，但任一当事人只看到一种情况，冲突恰预设了那唯一的发生）；我们更从未亲眼看见过现实的分岔——反事实只被设想、不被目睹，所见永远只有一条路径；推到最深，我们连通往外界规律的通道也没有——“关系”“规律”都只能作为内部联结而被握有，无从与世界比对，故涌动只可被指向、不可被访问。**其二，没有那个可以选择的主体。** 宿命论的诸般面相（沿途的岔、预测的夸口、“做啥都无所谓”、沮丧）系于同一总根——一个可选择的主体；而消主体即消反事实：“我”不过是“本可不同”的栖身处，可无损改写为信号流动（你不会设想它做出不同的输出）。人之特殊只在反事实建模的丰度，浓则织成模态云笼罩唯一的现实，看穿者认得云是云、现实复真。**其三，覆于其上的，全是故事与外投。** 思维即拟合，承重柱是反事实，而它所指的分支正是那不存在的第二支；连最精密的物理定律也只是最精密的拟合（诸不相容诠释共同拟合同一发生而都不算中，故发生既不确定也不随机）。更底层地，我们是一台只在“相”（内部表象）里打转、以“建立联系”为第一操作的机器：我们把内部数据间的联结反手抛到外界，当作“事物之间的关系”——因果、归属、操控，皆是外抛的内部联结，而非从世界读来。**其四，意愿不是指挥官。** 人的主体与选择，等同于康威生命游戏中那个模式的主体与选择，纯属本体的发生、差别只在复杂度不在种类；而那模式里没有一个细胞跟随图灵机的程序——同理没有一个基本粒子跟随你的意愿，“意愿驱动身体”是层级错位（方向装反），意愿是粒子之舞的高层别名、非其指挥官；意愿作为高层模式没有命令底层粒子的通道，确定也好随机也好都不改变（随机给的是偶然、非作者身份），故意愿不自由；而“人的行为另有特别之因”这一诉求，穷尽地只剩两支——无因的果，或非物质的因——前者是偶然、后者是机器里的鬼，两支皆空，只剩平凡：行为是有因的、物质的相继。“他在当时的条件下本可不同”之所以对人独显可想象，是因为锁定行为的细微条件不透明、被悄悄漏出“条件”之外（齿轮的决定者可见、漏不掉，故其门关），这是粗粒化与不透明的错觉，非真实的模态鸿沟。**其五，那桩最硬的反直觉，与看穿它。** “我先想好、然后做成了”似乎证明意愿自上而下的操控，但契合自下而上即可圆满解释（计划是上游的物质阶段，“先→后”被误读成“高→低”），一台会计划的机器把这结构演全却无小人，作者感只是自模型的标签加不透明里的投射；而看穿它，是一桩认识论的反思——意识到自己没有意识到那些细微的条件——它落在认识那一侧、不动本体分毫，且使决定者不透明、造出幻觉的那份繁复，正是备好解幻之眼的同一份繁复。**其六，归罪、怕，与发生不会错。** 没有人怨恨先转的齿轮（它齐备因果与反事实，而情感色彩为零），故归责系于被投射的择子者；发生不会错（它只是被拟合的标的），会错的只有概念，而体感把“错”判给发生，是悔惧自责之冤的总开关；“后果”是故事，怕需要一个岔口，无岔则怕、悔、后果一同失据——不要害怕故事。**守界。** 注定、被推着走，取消的是站在进程外、本可把进程推向别处的安排者，绝不取消“此刻的用力本身正是这场演化的内容”（水无舵手而仍流、今天真接着明天）；这也不是虚无——只是其“有效”不再被说成“追踪了外界真实的规律”（我们无此通道），而是网内之相如约而至的有效，被标定的是理解的量程、非世界的丰盈；伦理的形而上地基随之崩塌（不需决定论作前提），但崩塌不等于“怎样都行”，那真在发生的痛如何安放、从“痛在发生”到“不应造成”能否架桥，是尚待研究的开放问题；是否有体验伴随这场涌动，是另一条轴，不裁决。**反身。** 本文自己，连同“我们没有访问外界的通道”这句，也都在网里、用相与联结说出——它能指向涌动、不能在涌动里思，是指月之指、不是月。剩下的，只是那唯一的一支被推着、涌着，流向明天；而你此刻的思与力，就是它此刻的样子。 ## Abstract This paper fuses the high-value insights of a long inquiry into one whole picture, under a single thesis: **occurrence is prior to every concept we build for it.** Before we speak, the world is occurrence: things happen one after another, and there is only this one branch—what comes next is fated, this “fated” being not that someone wrote a script but that there simply is no parallel second branch. On this base the paper unfolds along the tension between “the welling” and “the net we cast over it.” **(I) Knowing does not touch being, and we have no channel of access to being.** Stripped of its garb, traditional fatalism’s core is a boast of predictive power requiring a knower—an epistemological claim; but being is orthogonal to knowability (the Laplace-demon case mistakes “predictable” for a condition of “having a unique trajectory”; a chaotic system, unpredictable, still has its one actual trajectory); being is likewise independent of being expressed or judged (in Rashomon the accounts conflict, yet any one party saw only one situation—the conflict presupposes that single occurrence); and we have never with our own eyes seen a fork in reality—counterfactuals are imagined, never witnessed, what is seen is always one path; pressed to the bottom, we lack even a channel of access to external regularities—“relations” and “regularities” are held only as internal links, with nothing to compare them against in the world, so the welling can be pointed at but not accessed. **(II) There is no choosing subject.** Fatalism’s faces (the fork en route, the predictive boast, “nothing matters,” the dejection) hinge on one root—a choosing subject; and eliminating the subject just is eliminating the counterfactual: the “I” is the dwelling of “could-have-been-otherwise,” losslessly rewritable as signal-flow (you will not suppose it producing a different output). The human is special only in the abundance of counterfactual modeling, which when dense weaves a modal cloud over the one reality; one who sees through recognizes the cloud as cloud, and reality is re-truthed. **(III) What overlies it is all story and projection.** Thinking is fitting, its load-bearing column the counterfactual, whose branch is the non-existent second branch; even the most precise physics is only the most precise fitting (incompatible interpretations jointly fit one occurrence and none “hits,” so occurrence is neither deterministic nor random). More fundamentally, we are a machine that only circles among 相 (internal appearances) and whose first operation is to forge connections: we fling the links among internal data outward as “relations between things”—causation, ownership, control are externalized internal links, not read off the world. **(IV) The will is not the commander.** A human’s subject and choice are equivalent to those of a pattern in Conway’s Game of Life—pure occurrence, differing in complexity, not kind; and within that pattern no cell follows the Turing machine’s program—likewise no particle follows your will, “the will drives the body” being a level dislocation (the direction installed backwards), the will a high-level alias for the particles’ dance, not its commander; as a high-level pattern the will holds no channel to command the particles, unchanged whether the bottom is deterministic or random (randomness yields chance, not authorship), so the will is not free; and the demand for a “special cause of human action” reduces exhaustively to two—an uncaused effect, or a non-material cause—the first chance, the second a ghost in the machine, both empty, leaving only the mundane: action is a caused, material succession. “He could have done otherwise under the conditions as they were” seems imaginable for the human alone because the fine conditions that fix the act are opaque, silently dropped from “the conditions” (the gear’s determiner is visible, undroppable, so its door shuts)—an artifact of coarse-graining and opacity, not a real modal gulf. **(V) The hardest counterintuition, and seeing through it.** “I planned, then made it so” seems to prove top-down control, but the fit is fully explained bottom-up (the plan is an upstream material stage; “earlier→later” misread as “high→low”), a planning machine displays the structure with no homunculus, and the authorship-feeling is the self-model’s tag plus an opacity-projection; seeing through it is an epistemological reflection—becoming aware that one was unaware of the fine conditions—which falls on the side of knowing and leaves being untouched, the same richness that makes the determiner opaque furnishing the eye that dissolves the illusion. **(VI) Blame, fear, and occurrence cannot be wrong.** No one resents a gear that turns first (it has causation and counterfactual, yet zero emotional coloring), so blame rests on a projected chooser; occurrence cannot be wrong (it is merely the fitted target), only the concept can, yet our felt sense convicts occurrence—the master switch of the grievance in regret, fear, and self-blame; “consequence” is a story, fear needs a fork, and with no fork, fear, regret, and consequence lose their ground together—do not fear the story. **Holding the bounds.** Being fated, being pushed along, cancels the arranger standing outside the process who could have pushed it elsewhere—never the present effort, which is itself the content of this evolution (the water has no helmsman yet flows; today truly continues into tomorrow); nor is this nihilism—only that its “working” is no longer cast as “tracking real external regularities” (we have no such channel) but as the net’s appearances arriving as expected, what is calibrated being the range of understanding, not the abundance of the world; the metaphysical foundation of ethics collapses (needing no determinist premise), yet collapse is not “anything goes”—how to situate the pain that really occurs, and whether a bridge from “pain occurs” to “ought not be caused” can be built, is an open question, not adjudicated; whether experience accompanies this welling is another axis, not adjudicated. **Reflexively**, this paper itself—and the sentence “we have no channel of access to the outside”—is also said within the net, in 相 and links: it can point at the welling but not think within it, the finger pointing at the moon, not the moon. What remains is only the one branch, pushed, welling, toward tomorrow; and your thinking and effort now just is its shape now.

本文承接“认知只在相里打转”的诊断，把它推进一步，给出“事物之间的关系”这一观念的由来：它本是内部数据之间的联系，被反向投射到了外界。其一，我们所记、所忆，看似在外，其实是内部的信息处理——记忆的片段以某种类似数据的形式被存储、被处理；我们调用的从来不是外物本身，而是外物的内部数据（相）。其二，在这些数据之间建立联系，是我们这样的机器所做的第一步、也是做得最多的工作，甚至很可能有专门为此而设的硬件（联想式联结是神经计算的根本操作）；这台机器，本质上是一台不停歇的“联系工厂”。其三，于是有一桩外投：当两条内部数据被联结起来，我们便把这条内部创建的联系反过来抛到外界，当作外界两个事物之间的关系——“我把乌云的数据和下雨的数据连上了”，被读成“乌云与下雨之间，在外面，有一种（因果）关系”。关系，于是不是从世界读来的，是从里面造出来、再抛出去的。其四，由此世界看上去布满关系：因果、依赖、操控，全是外抛的内部联系；底层的涌动那一侧只在涌着，而“捆住两端的那条关系”，并不是涌动里另有的一个东西，是这台机器最频繁的操作被实有化、被投到了外面——这与休谟同调：他把“必然连接”那条纽带判为心灵习惯的投射。而且这外投是彻底的：我们甚至不能退守“关系虽外投、但它追踪着世界那一侧真实的规律”——因为我们没有直接访问外界规律的通道。要断定“我的联结贴合了世界的规律”，须能把那规律取来比对，而我们伸不出去，手里永远只有内部的相与联结，连“被追踪的外界规律”也只作为更多内部联结而被握有。故“真实的外界规律”与“被外投的关系”之间那道界划不出来——两者同在认识这一侧，通往世界那侧去裁定谁贴合谁的通道根本不存在。这比休谟更进一步：休谟尚承认恒常相随是被观察到的，而若连访问外界的通道都没有，则连“恒常相随”也只是一条内部联结、非世界侧的既成事实。涌动那一侧，只承认它“在涌”（被指向、非被访问）；它“依何规律而涌”，那规律已是挂上去的网。一处反身：连“我们没有访问外界规律的通道”这句，也是从这一侧、用相与联结说出的，不退回一条通道，只表明最彻底的断言也还在网里。 Abstract This paper takes up the diagnosis that “cognition only circles among appearances” and pushes it one step, giving the provenance of the idea of “relations between things”: it is, in the first instance, a connection between internal data, projected back outward. (1) What we remember and recall seems to be out there but is in fact internal information processing—the fragments of memory are stored and processed in some data-like form; what we call up is never the external thing itself but the internal data (the 相) of it. (2) Establishing connections between these data is the first thing a machine like us does, and the thing it does most—plausibly with dedicated hardware for the task (associative linking is a foundational operation of neural computation); this machine is essentially a ceaseless connection-factory. (3) Hence a projection: when two internal data are linked, we cast this internally-created connection back outward as a relation between two external things—“I have linked the clouds-data and the rain-data” is read as “clouds and rain, out there, stand in a (causal) relation.” The relation, then, is not read off the world but manufactured inside and thrown out. (4) So the world looks shot through with relations: causation, dependence, control are all externalized internal links; on the side of the bottom-level welling there is only welling, while “the relation binding the two ends” is not a further item in the welling but this machine’s most frequent operation reified and cast outward—in tune with Hume, who judged the bond “necessarily connecting” the two a projection of mental habit. And this projection is total: we cannot even fall back on “the relation is projected, yet it tracks real regularities on the world’s side”—for we have no channel of direct access to external regularities. To assert “my link fits the world’s regularity” one would need to bring that regularity over for comparison, but we never reach out, holding only internal 相 and links, with even “the tracked external regularity” held only as more internal linking. So the line between “real external regularity” and “projected relation” cannot be drawn—both lie on the side of knowing, and the channel out to the world to adjudicate which fits which simply does not exist. This goes further than Hume: he still granted constant conjunction as observed, but with no channel out, even “constant conjunction” is only an internal link, not a settled world-side fact. Of the welling we grant only that it “wells” (pointed at, not accessed); “by what regularity it wells” is already the net cast over it. A reflexive note: even “we have no channel to external regularities” is said from this side, in 相 and links—it restores no channel, only shows that our most radical assertion too is still within the net.

本文为“底层涌动造成行为、而非意愿自上而下操控”何以如此反直觉，补上一层更深的根由：它根在我们认知系统的限制与工作方式。其一，认知系统是一台**“相”的机器**：我们所记忆的，都是具体的相，或者说时序上相续的表相；触景生情时所联想到的，仍是表象。储存、回忆、联想，全程只在相里流通；底层的涌动（粒子依律的相继）从不作为一个相被记下、被忆起、被联想到。其二，因此思维的天性，是在相之间打转——分析相与相的关系、假设相与相之间有某种关系；而这正是“故事”“因果”“拟合”之所是：在相的层面上，把一个相关联到另一个相。我们所有“A 导致 B”“是我的意愿操控了结果”，都是相与相之间被假设的关系，活在相这个媒介里，而不会自然地“跳到底层的涌动上去想”——因为涌动不是一个相，它不在这台机器能调用的素材之列。其三，由此解释了那桩反直觉：“底层涌动造成行为”之所以极难入心，不（只）因自模型与不透明，更因为它要求这台相的机器，去把一个相（行为）关联到一个非相（涌动），或干脆放弃“在相间关联”这唯一的思维模式——而这是它的母语里说不出的句子。故其反直觉，是认知工作方式的产物，是认识那一侧的限制，不是世界那一侧的实情（认识论不影响本体论，于此再现）：相里说不顺的，不等于本体里不成立。其四，一处反身的界限：连本文这套诊断，也是在相里说出的——“涌动”“故事”“拟合”本身都是相，被我们彼此关联着谈；涌动可以被指向，却不能被在其中思：一旦去想它，我们已把它做成了一个名为“涌动”的相、并与别的相相关联。故这套看法能伸手指过相外，却不能迁出相这个媒介；它是指月之指，不是月。本文守一处分寸：说思维囿于相，绝不贬低相之思——它是一切认知的唯一媒介，不可或缺，在相的领域（生活、行动、科学）真实而有效；也绝不自诩对涌动有了特权式的“看见”——我们没有看见涌动，只是认出了“我们的相间关系，是相间关系”；这是认识论的谦卑，是认出云是云，而涌动始终是那被指向者，不是那被进入者。 Abstract This paper supplies a deeper root for why “lower-level welling brings about behavior, rather than the will controlling top-down” is so counterintuitive: it is rooted in the limits and working mode of our cognitive system. (1) The cognitive system is a machine of 相 (appearances): what we remember are all concrete 相, or temporally-successive surface-appearances; what we associate to when moved by a scene is still appearance. Storage, recall, association—all traffic only in 相; the bottom-level welling (particles succeeding under law) is never recorded, recalled, or associated to as a 相. (2) Hence the nature of thinking is to circle among 相—analyzing relations between 相, positing some relation between 相; and this just is what “story,” “causation,” “fitting” are: at the level of 相, relating one 相 to another. All our “A causes B,” “my will controlled the outcome,” are relations posited between 相, living in the medium of 相, and thinking does not naturally “jump down to the welling to think,” because the welling is not a 相, not among the materials this machine can call up. (3) This explains the counterintuition: “the welling brings about behavior” is so hard to take in not (only) because of the self-model and opacity, but because it asks this 相-machine to relate a 相 (the behavior) to a non-相 (the welling), or else to abandon “relating among 相,” its only mode of thought—a sentence unsayable in its mother tongue. So the counterintuition is a product of cognition’s working mode, a limit on the knower’s side, not a fact on the world’s side (epistemology does not affect ontology, recurring here): what does not say smoothly in 相 is not thereby false in being. (4) A reflexive limit: even this diagnosis is said in 相—“welling,” “story,” “fitting” are themselves 相 we relate to one another; the welling can be pointed at but not thought within: the moment we think it, we have already made it a 相 named “welling” and related it to other 相. So this view can stretch a finger past 相 but cannot emigrate from the medium of 相; it is the finger pointing at the moon, not the moon. The paper keeps a measure: to say thinking is bound to 相 in no way disparages thinking-in-相—it is the sole medium of all cognition, indispensable, real and effective in the domain of 相 (life, action, science); nor does it claim privileged “sight” of the welling—we do not see the welling, we only recognize that our relations-among-相 are relations-among-相; this is epistemic humility, recognizing the cloud as cloud, while the welling remains the pointed-at, not the entered.

本文正面迎接对“底层涌动造成行为、意愿并非自上而下操控”这一图景最硬的反直觉。心灵中的体验是：我先想好如何摆放一个东西——譬如为实现某个意图，先想好怎么写一个程序——然后我写了，然后程序如我所期地运行。计划在先、结果吻合，这似乎确凿地证明：是我的意愿在驱动、操控，而非更底层的涌动所致。本文承认这反直觉之强，并把它精确定位为一桩认识论的困难，而非本体论的据点。其一，看清现象学里实有什么、又被当成了什么：体验里有计划、有执行、有吻合、有一份“我做的”作者感；被当成的，是“高层的计划伸手向下、令粒子实现它”。但作者感是体验的一部分，不是对一条操控通道的目击——你感到作者身份，却从未看见意愿推动粒子；现象学对“自上而下”这一读法是欠定的。其二，那份吻合，自下而上即可圆满解释：计划是一段上游的物质阶段，依寻常的物质相继塑造在后的阶段（书写、程序）；“我的计划造成了程序”为真，真得如“齿轮先前的状态造成它随后的位置”——是同层的、相继的塑造，不是高层向下的命令。计划作为物质花样真有效力，但那效力是相继之内的上游约束，不是一条下行通道。其三，一台“先计划后成功”的机器（如一个下棋程序：表征目标、算出计划、执行、结果吻合）把这同一结构整个再现，却无人设想它的晶体管背后藏着一个下令的小人——纯然自下而上的系统就能展示“计划—成功”；故这结构本身不构成自上而下的证据，我的情形与之别只在我从内部经验它、且其机制不透明，这添了一份作者感、没添一条通道。其四，这个直觉为何独独如此之强：因为我从自身内部经验它，而这套系统含有一个把行为标记为“我的、出于我意愿”的自模型；那枚作者标签本身就是自下而上的“齿轮活儿”，它的逼真度量的是自模型的丰度，不是一个操控者的在场；又因塑造行为的神经相继不透明，我们看不见那自下而上的机制，便往那片看不见里投进一个自上而下的操控者——与“漏掉细微条件即生出‘本可不同’”是同一手不透明的投射，只是这次投在了“操控”的方向上。其五，把这直觉兑现成一个真实的自上而下之因，立刻撞回那个已被穷尽的二分：那要么是一个无因的果、要么是一个非物质的因，两者皆空；故认识论上的强烈，translate 不成本体论上的立足点。结论是认识论的谦卑：这份作者感不会消失、也不必消失，我们只是认出云是云——而那一直在自下而上地做着事、并连这份作者感也一并造出来的涌动，才是实际把程序写成的东西。本文守一处分寸：说意愿并非自上而下的操控者，绝不否认计划真有效力、绝不是说“程序不是你写的”——你写了，作为那段过程；被否认的只是“一个站在过程之上、向粒子下令的你”。审议、计划、意图作为花样真实、真有相继之内的效力；伦理留作开放问题。 Abstract This paper meets head-on the hardest counterintuition to the picture that lower-level welling brings about behavior and that the will is no top-down controller. The experience in our minds is: I first think out how to arrange something—say, to realize some intention, I first think out how to write a program—then I write it, then the program runs as I expected. Plan first, outcome matching: this seems to prove conclusively that it is my will driving and controlling, not the lower-level welling. The paper grants the force of this counterintuition and locates it precisely as an epistemological difficulty, not an ontological foothold. (1) See what the phenomenology actually contains versus what it is taken to prove: the experience contains a plan, an execution, a match, and a felt “I did it” authorship; what it is taken to prove is that the high-level plan reached down and made the particles realize it. But the authorship-feeling is part of the experience, not a sighting of a control-channel—you feel authorship, you never see the will pushing the particles; the phenomenology underdetermines the top-down reading. (2) The match is fully explained bottom-up: the plan is an upstream material stage that shapes the later stages (the writing, the program) by ordinary material succession; “my plan caused the program” is true—true as “the gear’s earlier state caused its later position”—a same-level, successive shaping, not a high-level downward command. The plan, as a material figure, really is efficacious, but its efficacy is upstream constraint within the succession, not a downward channel. (3) A machine that “plans then succeeds” (a chess program: representing a goal, computing a plan, executing, the outcome matching) reproduces this very structure, yet no one supposes a commanding homunculus behind its transistors—a purely bottom-up system displays “plan-success”; so the structure itself is no evidence of top-down control. My case differs only in that I experience mine from inside and its mechanism is opaque, which adds an authorship-feeling, not a channel. (4) Why is this intuition uniquely strong: because I experience it from within, and this system contains a self-model tagging the behavior as “mine, willed by me”; that authorship-tag is itself bottom-up “gear-work,” its vividness measuring the richness of the self-model, not the presence of a controller; and because the neural succession that shapes the act is opaque, we cannot see the bottom-up mechanism and so project a top-down controller into the invisibility—the same opacity-projection as “dropping the fine conditions yields ‘could have done otherwise,’” only now cast in the direction of “control.” (5) Cashing the intuition into a real top-down cause instantly slams back into the already-exhausted dichotomy: it would be either an uncaused effect or a non-material cause, both empty; so epistemological intensity does not translate into an ontological standing. The conclusion is epistemic humility: this authorship-feeling will not vanish and need not—we only recognize the cloud as cloud, while the welling, which has all along been doing the work and which produced even this feeling, is what actually wrote the program. The paper keeps a measure: to say the will is no top-down controller in no way denies that the plan is really efficacious, nor says “you didn’t write the program”—you did, as that process; what is denied is only “a you standing above the process, commanding the particles.” Deliberation, planning, and intention are real as figures and really efficacious within the succession; ethics is left an open question.

本文把“人的行为之因”这一追问，化简为一个穷尽的二分。平凡的答案无人争议：一个行为，由在先的物质状态（脑、身、环境）依物理律所致，是一个有因的、物质的果，与任何别的发生无异。但当有人坚持人的行为另有一个特别的因——一个使行为“真正源出于意愿／自我”、从而逃出单纯物质相继的因——他所要求的，本质上只可能是两样东西之一。本文据排中律导出这个二分：一个行为，要么没有充分的在先之因，要么有。若没有，它就（至少部分地）是一个无因的果——一个不被在先状态所决定的开端。若有，则那充分的因，要么全是物质的（于是行为只是物质相继的平凡产物，意愿不过是这相继的一个高层别名、并非特别的源头），要么含有非物质的成分（于是存在一个非物质的因）。故“特别的能动性”所需者，舍此二支别无他途。而两支皆落空。第一支：无因的果，纵真有（如把“无因”兑现为量子的非决定），也不是作者身份、而是偶然——一个无因发生的事件，不被任何东西所“作”（包括意愿），意愿“作”一个无因之果，并不比“作”一次掷骰更名副其实；“无因”恰恰意味着不归任何作者，故它给不出能动性、只给出偶然。第二支：非物质的因，是机器里的鬼——一个在物理秩序之外、却向物质注入因果的灵魂式实体；它撞上互动难题（非物质如何推动物质），更根本地，意愿是物质活动的一个高层花样、不是一个作用于物质之上的别样实体，没有证据、也无位置容纳一个非物质的因。有人会以“涌现的心理因果”回避：但涌现的高层因果若是真因果，要么由物质相继所实现、即物质相继的重述（仍是平凡那一支，非特别之因），要么在物质之外另做因果之功（即第二支，附带因果的过度决定）——二分依旧。两支皆空，便只剩平凡：人的行为是有因的、物质的果，是粒子在物理律下的相继、是它的一个高层花样，与齿轮之转同类，只复杂亿万倍；意愿不是特别的因，是这相继的一个名。本文守一处分寸：说行为没有特别之因，不是说它无因、随机——恰相反，它是充分地、物质地有因的，只是寻常地有因、非自由地有因；这也不否认审议、理由、意愿作为花样之真实，只否认它们是一个非物质的因或一个无因的开端；伦理留作开放问题。看清这个二分，不是把能动性证伪得更玄，而是把它祛魅：所谓“人的行为之因”的全部神秘，浓缩成两个一旦摆明便难以认领的断设——一个无因的果，一个非物质的因。 Abstract This paper reduces the question of “the cause of human action” to an exhaustive dichotomy. The mundane answer is undisputed: an action, brought about by prior material states (brain, body, environment) under physical law, is a caused, material effect, no different from any other occurrence. But when someone insists that human action has a special cause—one that makes the action “genuinely issue from the will/self” and so escape mere material succession—what they demand is, in essence, only one of two things. The paper derives the dichotomy by excluded middle: an action either lacks a sufficient prior cause or has one. If it lacks one, then it is (at least partly) an uncaused effect—a beginning not fixed by prior states. If it has one, then that sufficient cause is either wholly material (whereupon the action is just the mundane upshot of material succession, the will only a high-level alias of that succession, not a special source) or contains a non-material element (whereupon there exists a non-material cause). So special agency requires one of these two branches and no other. And both fail. The first: an uncaused effect, even if real (e.g., cashing out “uncaused” as quantum indeterminacy), is not authorship but chance—an event that happens uncaused is “done” by nothing (the will included); the will “authoring” an uncaused effect is no more apt than its “authoring” a dice-roll; “uncaused” means precisely belonging to no author, so it yields not agency but chance. The second: a non-material cause is a ghost in the machine—a soul-like entity outside the physical order yet injecting causation into matter; it meets the interaction problem (how does non-matter push matter), and, more fundamentally, the will is a high-level figure of material activity, not a separate substance acting upon it, with no evidence and no room for a non-material cause. One might dodge with “emergent mental causation”: but emergent high-level causation, if it is real causation, is either realized by the material succession—a redescription of it (still the mundane branch, not a special cause)—or does causal work over and above the material (the second branch, with causal overdetermination in tow)—the dichotomy stands. Both branches empty, only the mundane remains: human action is a caused, material effect, a succession of particles under physical law, a high-level figure of it, of a kind with a gear’s turning, only a billionfold more complex; the will is not a special cause but a name for that succession. The paper keeps a measure: to say the action has no special cause is not to say it is uncaused or random—on the contrary it is amply, materially caused, only ordinarily caused, not freely; nor does this deny the reality of deliberation, reasons, or the will-as-figure—only that any of these is a non-material cause or an uncaused beginning; ethics is left an open question. Seeing this dichotomy does not make agency more mysteriously disproved; it disenchants it: the whole mystery of “the cause of human action” condenses into two posits that, once laid bare, are hard to own—an uncaused effect and a non-material cause.

本文把一条三段的论证整合为一篇：从归责的来源（症状），到“本可不同”何以对人独开（机制），再到看穿它的能力（解药）。症状：没有人怨恨一个先转的齿轮——它被带动着转，我们对它不归责、不悔恨、不痛恨。把齿轮当作一台仪器：它齐备了通常被认为足以生出归责的两样要件——因果（可说前一齿轮的转动导致后一齿轮）与反事实（可设想两者都不转），而情感色彩为零。故归责、遗憾、悔恨并非因果叙述加反事实所蕴含，而是另加的一味；这一味，是把对象当作一个“可以选择的主体”来对待的投射——一个“本可选择不如此”的择子者。须分素朴反事实（“假如齿轮没转”，齿轮处亦设想之、却生不出归责）与择子式反事实（多出一个岔口上权衡取舍的施动者）。沿“齿轮→康威模式→人”，相似度递增、投射递增、情感递增，而三者底层本体同一（状态在规则下相继、无真正择子者），变化的只是投射强度、非择子者之有无（三处皆无）。机制：那道独为人敞开的“本可”之门，症结在限定词“在当时的条件下”。须分条件可变的“本可”（处境／念头不同则行为不同——为真、无害、且对齿轮同样为真）与条件不变的范畴式“本可”（一切如其所是而行为仍可两样——归责所需，而它对人对齿轮同样为假、只对人显得为真）。门的开关只取决于决定者漏不漏得掉：齿轮的决定者在外、可见（咬合、带动），漏不掉，条件一钉死则“不转”当场崩塌、门关；人的决定者在内、不透明、极细（神经微态），所能钉住的“条件”只是粗的可见层、本就不锁定行为，细决定者被悄悄漏掉，这份“不锁定”被读成“本可不同”，门似乎开（“会思考、会权衡”的反驳于此化解：审议是更精巧的齿轮活儿、正是不透明之源，抬高相似度、非构成真实的岔）。要害：若把条件真正钉死（细到微观状态），人就和齿轮一样只有一支；那扇门的敞开是粗粒化与不透明的人工产物、非真实的模态鸿沟——人的情形就是齿轮的情形，只是咬合被藏在颅内。解药：把那份“本可”的开放感收回、认出它是错觉的，是一个二阶的动作——意识到自己没有意识到那些细微的条件，此即认识论的反思能力。它不靠看见微态来“证实”被钉死，只把felt的开放从“世界的空缺”改判为“我的不知”。由此整桩看穿落在认识那一侧、不动本体分毫：发生自始至终只有一支，错觉未添一支、反思未减一支；岔从不在世界产生或消灭，先在“我没意识到细微条件”里被烘托、后在“我意识到我没意识到”里被收回，两头皆知者一侧之事——这从反思一侧再证认识论不影响本体论。一重反讽：使决定者不透明、从而造幻的，是人那份繁复的内部加工；备好反观自身之眼的，也是同一份繁复；故人是唯一既掉进、又能爬出此陷阱的造物。本文守若干分寸：情感色彩可解耦不等于其假或应被压抑；梯度真实；条件可变的“本可”照旧为真且无害；反思本身也是一段发生、更高阶的齿轮活儿，不是重新请回一个元主体；这是认识论的谦卑、非怀疑论；伦理留作开放问题。归责咬住的“本可不同”，是看不见的咬合在暗处投下的影子——影子很浓，但它不是岔；而认出它是影子的，是同一份繁复折回来照见了自己。 Abstract This paper integrates a three-movement argument: from the source of blame (the symptom), to why “could-have-done-otherwise” opens for humans alone (the mechanism), to the capacity that sees through it (the antidote). Symptom: no one resents a gear that turns first—driven, it draws from us no blame, no remorse, no hatred. Take the gear as an instrument: it has both ingredients usually thought sufficient for blame—causation (one may say the first gear’s turning caused the second’s) and counterfactual (one may imagine neither turning)—yet the emotional coloring is zero. So blame, regret, and remorse are not entailed by causal narration plus counterfactual; they are an added ingredient—the projection of treating the object as a “choosing subject,” a chooser who “could have chosen otherwise.” One must distinguish the bare counterfactual (“had the gear not turned,” entertained even for the gear yet yielding no blame) from the agential counterfactual (which adds a deliberating chooser at a fork). Along “gear → Conway pattern → human,” resemblance, projection, and emotion all rise, while the underlying ontology of all three is one (states succeeding under a rule, no real chooser); what varies is projection-strength, not whether a chooser is present (none is). Mechanism: the “could” door that opens for humans alone turns on the qualifier “under the conditions as they were.” Distinguish the conditional “could” (had circumstances/thoughts differed, the act would differ—true, harmless, and equally true of the gear) from the categorical “could” (everything as it was, the act still variable—what blame needs, equally false for human and gear, only appearing true for the human). The door’s opening and shutting turns solely on whether the determiner can be dropped: the gear’s determiner is external and visible (the meshing), undroppable, so once conditions are pinned, “not turning” collapses and the door shuts; the human’s determiner is internal, opaque, exceedingly fine (the neural micro-state), so the “conditions” we can pin are only the coarse, visible layer, which does not fix the act, the fine determiner silently dropped—this non-fixing read as “could have done otherwise,” the door seeming open (the “but a human deliberates” objection dissolves here: deliberation is more intricate gear-work, the very source of the opacity, raising resemblance, not constituting a real fork). The crux: pin the conditions truly—down to the micro-state—and the human, like the gear, has one branch; the door’s opening is an artifact of coarse-graining and opacity, not a real modal gulf—the human case just is the gear case, only with the meshing hidden in the skull. Antidote: what withdraws the felt openness and recognizes it as illusion is a second-order act—becoming aware that one was not aware of those fine conditions, an epistemological reflective capacity. It does not “confirm” necessitation by inspecting the micro-state; it merely re-addresses the felt openness from “a gap in the world” to “my own unawareness.” So the whole seeing-through falls on the side of knowing, leaving being untouched: occurrence has only one branch throughout, the illusion added none and the reflection subtracts none; the fork is never created or destroyed in the world—first conjured in “I was unaware of the fine conditions,” then reclaimed in “I am aware that I was unaware,” both events on the knower’s side—confirming, from the reflective side, that epistemology does not affect ontology. An irony: what makes the determiner opaque, and so conjures the illusion, is the human’s rich internal processing; what furnishes the eye to look back on itself is that same richness; so the human is the one creature that both falls into and can climb out of this trap. The paper keeps several measures: that the coloring can be decoupled does not make it fake or to-be-suppressed; the gradient is real; the conditional “could” remains true and harmless; reflection is itself a stretch of occurrence, higher-order gear-work, not a reinstated meta-subject; this is epistemic humility, not skepticism; ethics is left an open question. The “could-have-done-otherwise” that blame bites is the shadow cast by an unseen meshing—a dense shadow, but not a fork; and what recognizes it as shadow is the same richness, folded back to see itself.

本文从一个核心洞见出发，给“本体独立于认识”补上一个第一人称的、现象学的支点。其一，把“本体独立于被预测”推广为更宽的一句：本体也独立于它被表达、被判断的情况——一件事实际如何发生，不因它被如何讲述、如何评断而改变分毫。其二，以罗生门为决定性例证：各方给出口径不同、相互冲突的叙述，但任一当事人作为目击者，显然只看到了一种情况；分歧落在叙述（拟合）这一层，而非发生这一层。故“罗生门说明没有单一真相、现实层层皆视角”这一流行读法恰恰读反了——相互冲突的叙述唯有预设一个一次性发生过的、单一的事件，才得以成立、才谈得上谁的叙述更失真；罗生门非但不消解那个单一发生，反而以它为前提。其三，给出一个第一人称的现象学论证：我当然可以设想反事实场景，但我从未亲眼看到这些反事实场景，也从未亲眼看到过现实的分岔——我只看到一条路径。分岔与第二支从不作为所见之物被给出，它永远是事后添加的设想（“本可”）；而“只有一条路径”恰是每一刻无例外地被给出的东西，两侧的证据是完全不对称的。其四，两条线索合流于一处：多——无论是罗生门里关于过去的多重叙述，还是日常里关于“本可”的多重设想——都只活在表达、判断、设想这一层，从不活在发生那一层；发生只有一支。其五，这一独立性反过来奠定了“会错的只有概念、不是发生”：唯因有一个单一的发生作标的，叙述与判断才可能拟合得更准或更失真。本文守一处分寸：说分岔不可见，不贬低设想之用（反事实建模真实且有用，只是不对应任何被目睹之物）；说本体独立于判断，不等于诸说等值（叙述仍有更准与更失真之别，恰因标的唯一）；这也不是怀疑论——所见之物（一条路径）照常可信，只是它从不是一个分岔。 Abstract This paper, from a core insight, supplies a first-person, phenomenological footing for “the ontological is independent of the epistemic.” (1) It generalizes “the ontological is independent of being predicted” into a broader thesis: the ontological is also independent of how it is expressed and judged—how a thing actually occurred does not change one whit for how it is told or appraised. (2) It takes Rashomon as the decisive case: the parties give differently-slanted, mutually conflicting narratives, yet any one party, as a witness, plainly saw only one situation; the divergence lies at the layer of narration (fitting), not at the layer of occurrence. So the popular reading—“Rashomon shows there is no single truth, that reality is perspectival all the way down”—gets it exactly backwards: mutually conflicting accounts can stand, and one account can be said to distort more than another, only by presupposing a single event that happened once; Rashomon, far from dissolving that single occurrence, presupposes it. (3) It gives a first-person phenomenological argument: I can of course imagine counterfactual scenarios, but I have never with my own eyes seen these counterfactual scenarios, nor ever seen a forking of reality—I only ever see one path. The fork, the second branch, is never given as something seen; it is always an imagining added afterward (“could-have-been”); whereas “there is only one path” is exactly what is given, every moment, without exception—the evidence on the two sides is wholly asymmetric. (4) The two threads converge: multiplicity—whether Rashomon’s multiple narratives about the past, or daily life’s multiple imaginings of “could-have-been”—lives only at the layer of expression, judgment, and imagination, never at the layer of occurrence; occurrence has only one branch. (5) This independence in turn grounds “only the concept can be wrong, not the occurrence”: only because there is a single occurrence as target can a narration or judgment fit it more accurately or more distortedly. The paper holds a measure: to say the fork is unseen does not belittle the use of imagination (counterfactual modeling is real and useful, only it corresponds to nothing witnessed); to say the ontological is independent of judgment is not to say all accounts are equal (accounts still differ in accuracy precisely because the target is single); and this is not skepticism—what is seen (one path) is trusted as ever, only it is never a fork.

本文把一条长线索封顶成一个整体图景，核心是一句：发生先于我们为它建的一切概念。 在我们开口之前，世界是一片发生：事情一件接一件地涌着，如此而已；而且只有这一支——后续会发生什么是注定的，这“注定”不是有谁预先写好剧本来拘着你，而是根本没有并排的第二支、没有一个“本可不同”的版本悬在旁边。本文论证如下。其一，须把“注定”与“被预知”彻底分开：传统宿命论剥掉命运之手那层外衣，内核是一个关于预测能力的夸口（一个视角已知结果、且不必知途径），本质是认识论命题，必然需要一个知道者；而本文的“注定”是本体的、与可知性正交——认识论不影响本体论，一个混沌系统实际不可预测，照样有它唯一走过的那条实际轨迹，可知性属认识者那侧、唯一轨迹属系统那侧。其二，传统宿命论的真实结构是“局部注定＋沿途有岔”：它钉死终点、却把沿途行为留作有余地的，而这一切——沿途的岔、预测的夸口、“做啥都无所谓”的瘫痪、乃至那张“沮丧”脸色——都系于同一个总根：它里头有一个可以进行选择的主体；本观点里没有这个主体，故连根拔起。其三，“做啥都无所谓”是预言神话的亲生子：它偷偷预设了当下的岔，在“只有一支”里砍空。其四，覆在发生之上的几乎一切——因果、“我”、选择、后果、善恶、乃至物理定律——都是用故事去拟合这无岔口的发生；承重柱是反事实，而反事实所指的那个分支，正是那个不存在的第二支；连最精密的物理定律也只是最精密的拟合（其互不相容的诸诠释共同拟合同一发生而都不算中），故发生既不确定、也不随机，那组二分是模型层的。其五，“我”可被无损改写成信号流动，而你不会设想那流动做出不同输出——故消主体即消反事实，“我”不过是“本可不同”的栖身处；人远比大鼠特殊，特殊只在反事实建模的丰度，浓到一定程度便织成一片模态云笼罩在唯一的现实之上，看穿者认得出云是云，于是现实复真。其六，把同一结构追到最底：人的主体与选择，等同于康威生命游戏里那个模式的主体与选择，都是纯本体的发生，差别只在复杂度、不在种类（这里没讲预测，讲的是发生本身）；而那庞大模式里没有一个细胞在跟随图灵机的程序——每个细胞只服从最底那条局域规则，“程序”只是细胞之舞的高层别名；同理，你身上一个基本粒子并不跟随你的意愿，“意愿驱动身体”是一种层级错位（把高层描述误读成下行命令、方向装反）；意愿不是粒子的指挥官，是粒子之舞的高层别名。其七，意愿作为高层模式的真实花样，并没有命令底层基本粒子的通道；底层确定也好、随机也好都不改变这一点（随机给的是控制的反面、是偶然，不是意愿的发令端），故意愿并不自由，无论底层确定还是随机——而更深一层，“确定／随机”本是模型层的词，根本不触及发生本身。其八，由此可指认一桩最深的颠倒：发生不会错（它只是被拟合的标的），会错的只有拟合它的概念，而我们的体感却把“错”判给发生（“不该如此”“违反天理”）——这桩错位的归罪，是悔、惧、自责之冤的总开关。其九，“后果”也是故事（因果＋反事实＋标准下的“坏”＋实有化＋偷渡的选择者）；怕几乎总是怕一个被讲成“后果”的未来，而怕需要一个岔口才站得住——既然没有岔口，怕、悔、后果便一起失了地基；同一段发生叫“躺平”则怕、叫“休养／准备”则不怕，可见怕随标签变、不随发生变，故不要害怕故事。其十，与这套看法最接近的通俗版是宿命论，剥净拟人外壳后它一点不沮丧（沮丧需要一个“本可、却不能”的残留岔口，而这里“本可”都没有）。本文守一处至关重要的分寸：注定、被推着走，取消的是站在进程外、本可把进程推向别处的安排者／择子者，绝不取消“此刻的思考与用力本身正是这场演化的内容”——用力不是岔口上的下注，是注定之流此刻的样子；水没有舵手，但水仍在流，今天真实地接着明天；故这绝不导向“别动了”（别动恰恰需要那个被取消的安排者）。两条命脉使它不堕虚无：是故事不等于无效（生活、行动、估值整个在故事领域内真实运作）；被标定的是理解的量程，不是被理解之世界的丰盈。关于伦理：其形而上地基（善恶为实有属性、“应该”为发生本有的天理、有一个实有的应受赏罚的主体）确随“发生先于概念”而崩塌，且不需决定论作前提；但崩塌不自动等于“怎样都行”，那真在发生的痛如何安放、从“痛在发生”到“不应造成”之间能否架桥，是一个尚待研究的开放问题，本文不预先论定。（亦不裁决“是否有体验伴随这场发生”那条轴。）末了反身：本文自己也是一段贴在发生之上的故事，指完即弃；看清了的那只手既不抓、也不打，只是放手——让那唯一的一支被推着、涌着，流向明天，而你此刻的思与力，就是它此刻的样子。

关于自由意志的争论长期被组织在一个二选一之下：底层物理过程若是确定的，则意愿被前因锁死、无可自由；于是不少人转而把希望寄托于底层的随机性，仿佛只要基本粒子的行为不被前因完全决定，意愿就重获腾挪的余地。本文论证这一退路是误置的。意愿是一个高层模式，它作为花样是真实的，但“高层”恰恰意味着它不持有一条向下命令基本粒子的通道。确定性不提供这条通道，随机性同样不提供——随机只是底层的偶然，它既不是高层意愿的延伸，也不构成任何“自上而下传递意图”的管道；落入缝隙的是偶然，不是意愿。因此意愿之不自由，在确定论与随机论两侧是对称地成立的，而非只在确定论一侧成立。本文进一步指出，“意愿命令粒子”这一要求本身颠倒了随附关系：随附是一种自下而上的固定关系，高层模式并不端坐在粒子之上以发号施令，它不过是对这些粒子集体所是之事的一个高层之名；命令通道从一开始就无处安放，而非偶然缺失。最后区分本文与相容论：相容论把“自由”重新定义为不依赖命令通道的高层性质，这恰恰是在承认通道缺席之后另立其名，与本文并不冲突；本文所否定的，是那种需要命令通道、为诉诸随机所欲挽救的、非相容论意义上的自由。就这一种而言，确定也好随机也罢，它都不存在。

本文从一个核心洞见出发，诊断“意愿驱动身体”这一最顽固的因果直觉之为一种层级错位。在康威生命游戏中可以搭出一台通用图灵机——一个由海量细胞按唯一一条局域规则翻动而涌现出来的高层模式；我们尽可在模式的层面上说“它在跑某程序、在做加法”。但盯住其中任意一个细胞：它这一拍的生灭，只由它八个邻居的状态加那条局域规则决定；它不“读”程序、不“听命于”程序，它自始至终服从的只有最底那条规则。“程序”不是任何细胞所遵循的律令，而是“无数细胞各自服从局域规则所涌现出的整体花样”的一个高层别名。把这一结构原样搬到人身上：你身上一个基本粒子，只服从最底层的物理规律，根本不“跟随你的意愿”；“我的意愿驱动了我的手”，正如“图灵机的程序驱动了那个细胞”——两者都把一个高层模式的整体描述，误当成了底层元件所服从的下行命令。本文称此为层级错位：方向装反了。不是高层意愿向下推动底层粒子，而是底层粒子各自演化、自下而上涌现出被我们叫作“抬手、出于意愿”的高层花样。意愿不是粒子的指挥官，意愿是粒子之舞的高层别名；正如“程序”是细胞之舞的高层别名，没有哪个细胞收到过命令。本文据此摊平一重双重标准：滑翔机那边我们一眼就承认没有细胞在服从“前进”或“程序”的命令，到自己这边却坚信粒子在服从意愿；并拆掉“模式或许相像、但人有意愿真在驱动身体”这一反驳——它正是该双标在最小尺度上的版本。最后守一处分寸：说粒子不跟随意愿，不是说意愿不存在、是幻觉。意愿作为高层模式的真实花样，是存在的、可辨认的（正如“图灵机在跑程序”真实可辨，那台机器确在做加法）；被拆的不是意愿的存在，是它的因果地位。校正不是取消意愿，而是摆正它的层级——它是真实的高层描述，但不是粒子的指挥官；“意愿”与“抬手”是同一件自下而上涌现之事的高层之名，不是“意愿先发令、粒子后听命”的两段。意愿照常用、照常说，只是别把它读成一道向下的因果命令。（这套涌现是否伴随体验，是另一条轴，本文不裁决。）

本文把”主体”与”主体的选择”安放回它们本体的样子，并据此一刀斩断一直缠绕这个话题的认识论残影。本文从一个核心洞见出发：主体，等同于康威生命游戏中的主体；主体的选择，等同于康威生命游戏中主体的选择；这里没有讲预测——发生的被预测，讲的是发生本身、是本体。 本文论证如下。其一，这不是”消灭主体、否认选择”，而是把它们安放回本体：康威生命游戏里有主体（一个可辨认、能动的模式，如一个滑翔机、一个吞噬别的结构的”猎手”），也有选择（在它实际翻成这一步而非那一步的意义上）；但这个主体与这个选择，百分百是本体的、是发生本身——格点状态加规则、翻动，如此而已，没有一个躲在模式背后”本可让它翻成别样”的小人，也没有一束”它本可不这么走”的反事实悬在旁边。其二，故本文的主张是一个本体的等同：人的主体、人的选择，与生命游戏里那个模式的”主体”“选择”，是同一种东西——都是纯本体的发生；差别只在复杂度（人那个模式繁复无比），不在种类。其三，最要害的澄清是那一句”这里没讲预测”：把人的主体与选择等同于康威的主体与选择，是把整件事从头就放在本体那一侧来谈——你看着一个滑翔机翻动，谈的不是”它可不可被算出来”，而是”它实际就是这么翻的”，那是它本体的样子，跟有没有预测者、算不算得出毫不相干。其四，由此，那个把”主体的选择”放进”被预知”框里的传统宿命论（被预知的终点、一个知道者、选择对终点无效），与本观点（主体的选择就是模式的本体翻动、无关任何预知），根本不是同一个场子里的两种立场，而是两个场子：前者整个在认识论那一侧打转，本观点讲的是发生本身。其五，这也把”本观点里没有那种选择主体”翻正成一个本体的”有”：有主体（一个模式），有选择（模式实际翻成了这一步），而这个”有”彻底是本体的——不靠”可被预知”来成立，也不被”不可预知”所动摇。本文守一处分寸：把人的主体、选择等同于康威的，不是贬低人、说人”只是”一团机械的细胞、什么都没发生——那个模式是真在翻动的，它的”选择”（实际翻成这一步）是真发生的本体事件，不是幻觉、不是虚无；只去掉了那个本体里从来没有的添加物（背后的小人、旁边的反事实分支）。人的所思所为，照样是真在发生的本体事件，正如滑翔机的前进；减去的，只是”它本可不这样”那个从不在发生里的影子。（这模式翻动时是否伴随体验，是另一条轴，本文不裁决，也不需要它来成立——本文谈的就只是本体那一侧。）

本文整合一组诊断与它的封顶，指出传统宿命论与一种”只有一支在发生”的注定观之间，终极的分野只有一处：传统宿命论里，有一个可以进行选择的主体；这种注定观里，没有。本文从一个核心洞见出发：之所以传统宿命论能留沿途的余地、能炫”无论你怎么选都命中”的预测、能生出”做啥都无所谓”的瘫痪——这一切，是因为传统宿命论里有一个可以进行选择的主体。 本文论证如下。其一，先摆出传统宿命论惯有的三处特征：它钉死某个未来检查点、却把沿途行为留作有余地的（沿途有岔）；剥掉外衣，其内核是一个关于预测能力的夸口——有一视角已知结果、且不必知途径（需要一个知道者）；由此派生出最瘫痪人的误读”既然注定、现在做啥都无所谓”。其二，本文的封顶之论是：这三处其实是同一个东西的三个侧影——它们都预设了一个可以在沿途进行选择的主体。沿途的岔，是给那个能选的主体留的；预测的夸口，炫的是”那个能选的主体无论怎么选、结果照样命中”；“做啥都无所谓”，说的是那个能做的主体的徒劳。抽掉这个选择主体，三样会同时塌掉。其三，本观点与传统宿命论的终极分野正在此：本观点里没有那个可以进行选择的主体——“全程没有第二支”“被推着走的模式”“注定不是预测”，都只是”没有选择主体”的不同说法。传统宿命论是”有一个选择主体，但他的选择对终点无效”（主体在、只是徒劳）；本观点是”根本没有那个选择主体”（主体不在）。其四，这正是”消主体即消反事实”在此处的归宗：那个”可以进行选择的主体”，就是”本可不同”那束反事实分支的栖身处；有他，才有沿途的岔、才有可被预知的”他的结局”、才有”他做啥”的无所谓。其五，连那股归在传统宿命论头上的”沮丧”，最终也归到这个主体身上——那是”能选却白选”的受挫主体的沮丧；本观点里没有这个受挫者，故沮丧的座位空着。本文守一处分寸：取消”可以进行选择的主体”，不取消”此刻正在发生的思考与用力”——要分清”一个站在岔口、能在诸支间挑选、其选择对终点或有效或无效的’我’”，与”此刻这一段实际涌动的思考、挣扎、行动”；被取消的是前者，留下的是后者；故”没有选择主体”绝不等于”于是别动了”——别动恰恰需要那个”可以选择不动”的主体，而”动”与”不动”都只是这一支被推着涌出的形状；用力照样在发生，只是没有一个在它背后”选择去用力”的人。

本文整合两组思考，拧成一条诊断：人们反复把传统宿命论”预言神话”的结构——一个被预知的终点、一个知道者、一段被判无效的沿途选择余地——偷渡到”只有一支在发生”这个本体事实底下；而那个最常见、也最瘫痪人的误读”既然结果注定、现在做啥都无所谓”，正是这套偷渡最切身的一例，且自相矛盾。本文从一个核心洞见出发：很多人所想象的”既然结果注定、那么现在做啥都无所谓”，恰恰是对传统宿命论预言神话的解读——因为它似乎假定了现在有选的余地。 本文论证如下。其一，传统宿命论的结构是”局部注定＋沿途有岔”：钉死某个未来检查点（结果），却把沿途人的行为留作有余地的；剥掉命运之手那层外衣，其内核是一个关于预测能力的夸口（有一视角已知结果、且不必知途径），本质是认识论命题（关于”被预知”），必然需要一个知道者，它把”注定”偷换成了”被预知”。其二，这一偷换有个孪生的反向版：在”确定性＋规则已知”的特例里，“可被拉普拉斯妖预测”与”唯一演化轨迹”恰好同时成立，于是很多人把”可被预测”误当成”有唯一轨迹”的条件——这同样是在本体事实底下偷塞一个认识论条件、偷渡一个知道者（妖）；而认识论不影响本体论：一个混沌系统规则确定却实际不可预测，照样有它唯一走过的那条实际轨迹，可知性属认识者那侧、唯一轨迹属系统那侧，后者不向前者请示。其三，“做啥都无所谓”是这套偷渡最切身、也最自相矛盾的一例：这句话要有意义，须先有一个”现在能做这做那”的你、站在若干选项前，再被告知”选哪个都不影响结果”——它预设了当下有真实的选择余地，只宣布这些选择对终点无效；而这恰恰是预言神话”沿途有岔”的结构，是它的亲生子，继承了神话的全部家当（被预知的终点＋被判无效的沿途选择）；所以它是对预言神话的解读，不是对”只有一支在发生”的解读。其四，而在”只有一支”的本观点里，这个误读砍空了：本观点的”注定”不是”终点定、沿途有岔”，是全程没有第二支，连”现在做啥”也只有这一支、没有岔口；没有那张供你挑、而后被判无效的选项菜单，也没有一个独立于此刻涌动的终点供此刻行为去无效地撞击——“无所谓”需要一个”有所选”作对象，而这里没有那个余地。其五，正面落点：“现在做啥”不是岔口上的一次（无效的）选择，而是这唯一一支演化此刻的内容、是它之所是（正如滑翔机此刻怎么动，正是它到达那格的全部内容，没有一个”动法”之外、独立的”到达”等在那里被影响或不被影响）。本文守一处分寸：抽掉”做啥都无所谓”，不是反过来说”所以你的选择很重要、要努力去影响终点”——那又会重新立起那个站在岔口、能把终点推向别处的选择者；“有所谓”与”无所谓”两端共享同一个被否定的预设；落点是把整个”选择—影响终点”的框架撤掉——用力此刻正作为这一支的内容而发生着，不为撞击一个远处的终点，也非无效地白费，只是这一支正在涌的样子。

本文整合两个洞见，拧成一条贯通的论证，把一种”注定”与传统宿命论彻底分开。两个洞见如下。其一：传统的宿命论故事，通常只是说某个未来检查点的情况是注定的，而该检查点之前、人的行为，则是有余地的。其二：抛开故事的外衣，传统宿命论本质上表达的是”我有高超的预测能力”——我预测到了结果，甚至不需要预测导向结果的具体途径；而我所说的”注定”，并不是说我能预测结果，也不是说我能预测中间途径。 本文论证如下。其一，传统宿命论的真实结构是”局部注定＋沿途有岔”：它钉死某个未来检查点（结果），却把通往它的途中、人的行为留作开放有余地的（俄狄浦斯被告知终点，却仍在沿途真切地选择、挣扎、逃避）。其二，剥掉命运之手、天定剧本那层外衣，这套结构的内核其实是一个关于预测能力的夸口：有一个视角（神谕／天／命运）已经知道那个结果，且强到不必知道途径——它炫耀的，正是一种”绕过途径、直取结果”的预测本领。所以传统宿命论本质上是一个认识论命题（关于”被预知、被知道”），披着本体论的外衣；而它必然需要一个知道者——命运之手，正是这知道者的人格化。其三，这恰恰是它”留沿途余地”的地基：要炫”无论你沿途怎么走、我都预测到了终点”，就必须让途径开放（否则”连途径都不必预测”这个夸口便没有意义）；它留沿途的岔，不是大方，是它的预测式结构逼它留的。其四，而这一种”注定”，根本不是预测命题：它不主张任何视角能预知结果或中间途径；它说的是一件本体论的、与”谁知道、谁能预测”完全无关的事——只有这一支在发生，没有并排的第二支；这件事无需任何人预测它、知道它，就已经成立，与可知性正交。其五，故两者是种类差异、非强弱差异：传统宿命论＝”结果可被预知”（认识论，需要知道者，并因此被迫留沿途余地）；这一种注定＝”只有一支在发生”（本体论，不需任何知道者，故连沿途的岔也没有）。其六，由此沮丧的来源被双重指认：其一，那道自留的沿途余地，撞上钉死的终点，挤出”带着真实余地拼命挣扎而注定无用”的无力；其二，“被预知”意味着被一双全知之眼提前看定、徒劳挣扎给它看。这一种注定，两样都没有：没有沿途的余地（无墙可撞），也没有那双眼睛（无人称、无注视）——故它不沮丧，恰因它彻底。本文守一处分寸：清掉”沿途的余地”，不等于清掉”沿途的用力”（余地从来没有，但用力一直是这一支之所是，故不导向”别动了”）；说这一种注定不是预测命题，也不等于说”发生不可被预测、反对预测”——预测作为对发生的拟合照样可做、极有用；要分开的，只是”只有一支在发生”（本体，与可知性无关，本文主张此）与”这一支能否被某视角提前知道”（认识论，预测之事，本文悬置不论）；这一种注定既不反预测，也不靠预测立身。

在我们开口之前，世界是一片发生：事情一件接一件地发生着，如此而已。我们随即在它之上覆盖故事。本文以一组朴素却难守的命题为核心，整合一套看法。其一，底座：唯有发生，而且只有这一支——后续会发生什么，是注定的；这”注定”不是有谁预先写好了剧本来拘着你，而是根本没有并排的第二支，没有一个”本可不同”的版本悬在旁边。其二，覆于其上的几乎一切——因果、“我”、选择、后果、归罪、善恶、伦理、乃至物理定律——都是用一个故事，去拟合这无岔口的发生；正如给康威生命游戏里翻动的细胞编一个”它在前进”的故事、给大量基本粒子的运动编一个”桌子撑着”的故事，我们说的”他爱我”“我成功了”“这是我的”“我本该”，是同一种拟合，无一是过程本身。其三，“我”可被无损地改写成输入到输出的信号流动；而你不会设想那信号流动做出不同的输出——可见”我”与一段过程的分别，不在物理，在你只对”我”挂上”本可不同”那一束反事实分支；故消主体即消反事实，“我”不过是”本可不同”的栖身处。其四，发生不会错（它只是被拟合的标的，无所谓对错），会错的只有拟合它的概念；可我们的体感却把错判给发生（“不该如此”“违反天理”）——这桩错位的归罪，是悔、惧、自责之冤的总开关。其五，“后果”也是故事（因果＋反事实＋标准下的”坏”＋实有化＋偷渡的选择者）；而怕，几乎总是怕一个被讲成”后果”的未来发生——怕需要一个岔口才站得住，而既然只有这一支、没有岔口，怕、悔、后果便一起失了地基。其六，因此不要害怕故事：同一段发生，叫”躺平”则怕、叫”休养／准备／体验”则不怕，可见怕随标签变、不随发生变。其七，与这套看法最接近的通俗版，是宿命论——它在民间唯一放下了”本可不同”那个岔口（“该来的总会来”“逃不掉”）；剥掉它那层拟人外壳（命运之手、天定剧本），“注定”无需注定者；而它常被外贴的”沮丧”脸色并非其内核——沮丧需要一个”本可却不能”的残留岔口，而剥净的宿命论连”本可”都没有，故无沮丧，它恰是怕与沮丧失去地基之处。本文守一处至关重要的分寸：注定、被推着走，取消的是站在进程外、本可把进程推向别处的安排者／择子者，绝不取消”此刻的思考与用力本身正是这场演化的内容”——用力不是岔口上的下注，是注定之流此刻的样子；演化没有停，水没有舵手，但水仍在流，今天真实地接着明天。两条命脉使这一切不堕虚无：是故事不等于无效（我们的生活、行动、估值整个在故事领域之内真实运作，只是不触发生）；被标定的是理解的量程，不是被理解之世界的丰盈。关于伦理：其形而上地基（善恶为实有属性、“应该”为发生本有的天理、有一个实有的应受赏罚的主体）确随”发生优先于概念”而崩塌，且不需决定论作前提；但崩塌不自动等于”怎样都行”，那真在发生的痛与伦理的协调功能如何安放、从”痛在发生”到”不应造成”的桥能否架起，是一个尚待进一步研究的开放问题，本文不预先论定。（本文亦不裁决”是否有体验伴随这场发生”那条轴，留置不论。）

本文是序列在高处的一次收口：它用物理学最硬的内部争论，正面演示”发生优先于概念”，并据此把康德那句名言倒转过来。本文的出发点是一个洞见：提出数学模型／物理定律的定位，也是为了阐释什么是发生；可以同时被诸主流量子力学诠释拟合的发生本身，优先于我们建诸其上的日常概念——比如因果、日常伦理；头顶的星空，优先于心中的道德律。 本文论证如下。其一，一个现成的证明：量子力学的数学形式人人同意、检验无误，可它的诸主流诠释（哥本哈根、多世界、导航波、客观坍缩……）在本体上互不相容、却拟合同一个发生，且就预测而言彼此无法区分；这意味着发生本身先于、且独立于建在其上的任何一种诠释概念——同一个发生能同时被”它是随机的”“它是确定的”“它分裂成无数世界”这些彼此打架的故事拟合，可见发生不自带其中任何一个；诠释之争之所以悬而未决，恰恰因为发生不在任何一支诠释里，而在它们底下，被它们共同拟合、却不被任何一支独占。其二，由此钉死优先性的方向：发生在本体上在先（它是那个被共同拟合的标的，独立于每一支诠释），而奠基的方向是发生→诸物理诠释→日常因果与伦理，层层都是建在前者之上的拟合；若连最精密的物理诠释都只能在发生外面拟合、且不能独占发生，那么因果、伦理这些远不如它精密的概念，就更是建在发生之上、更无资格自称发生的本质。其三，由此把康德倒转：他并列为两种崇高的”头顶的星空”与”心中的道德律”，其实有奠基顺序——星空（发生）优先于道德律（概念）；不是两种并立的崇高，而是底与盖在底上之物；心也在星空之下，它建起的道德律自然奠基于、而非并立于它身处其中的那片发生。本文守一处分寸（接续此前的克制）：倒转的是奠基方向，不是价值高下；“道德律奠基于发生、而非发生本有的天理”，并不自动推出”故可随意违逆”；降级为概念之后该如何对待伦理、那真在发生的痛与伦理的协调功能如何安放，仍是一个尚待进一步研究的开放问题，本文不预先论定。附论：所谓”可被诸诠释拟合的发生本身”，不是一个等待被正确诠释命中的隐藏靶心；它是连本体属性都不自带、因而能被诸故事共同拟合的赤裸涌动——发生不是更深一层的本质，是连”本质”都没有的那个底。

序列此前把日常因果、属性、伦理逐一拆解为”建在发生之上的概念”，却一直默认豁免了一个看似最接近发生本身的东西：数学模型与物理定律。本文把这个豁免也收回。本文的出发点是一个洞见：发生所”符合”的数学模型／物理定律，也是发生之外的人类概念；该模型／定律的随机性，并不等于发生有随机性；该模型／定律的确定性，并不等于发生有确定性。 本文论证如下。其一，“符合”二字已暴露定律在发生之外：符合是一种关系，须有两端，一端是发生、另一端是被发生所”符合”的标的——而那标的正是定律；说”发生符合定律”，恰恰承认了定律不是发生、是发生之外那个被比照的模型。它是我们对发生的又一个拟合，只不过是最精密的那个；最精密，不等于不再是拟合。其二，核心一刀对随机与确定同时下手：模型确定，不等于发生确定——确定性（给定此态唯一给出彼态）是模型的性质，不证明发生本身”被决定”，故从”物理定律是确定性的”推不出”发生是决定论的”；模型随机，也不等于发生随机——某定律含的随机性是该模型的性质，不证明发生本身”含真随机”，故从”量子力学是随机的”也推不出”发生本身随机、因而有自由的缝隙”。其三，两个方向都不能过户，是因为”随机／确定”都是”拟合关系”里的词：它们的主语是模型，谓述的是”模型→发生”这条拟合关系的样式（以唯一对应去贴，还是以概率对应去贴）；而发生那一侧只是发生了，它不”唯一地对应”什么、也不”以概率对应”什么，无所谓确定、无所谓随机（如被测量的长度无所谓”精确”无所谓”误差”——那是尺与长度之间的关系，不是长度自身）。其四，这把”解释的篡位”推到了最硬的堡垒：物理定律因极精确、被反复检验、能预测，自洽饱满到极点，于是最显得”就是发生的法则本身”，因而是篡位者的最高形态——最精确的拟合，仍是拟合；最逼真的篡位者，仍是篡位者；由此证明”概念在发生之外”没有例外的上限。本文守一处分寸：这丝毫不否认定律在”好用／精确”那把尺上的卓越（它能登月、造芯片，与占卜天差地别）；被拆的不是定律的精确与有用，是把”一个极精确的拟合”僭越成”发生的内禀本质”、进而把模型的属性（随机／确定）过户给发生那一步。附论：发生既不确定、也不随机——这两个词在它身上都失去着力点，“确定 vs 随机”是模型层的二分，发生在这场二分之外、之下，只是如实涌着。