一条主张管到哪里
How Far Does a Claim Reach?
真或假以前,先问地址
我们习惯把一项主张放进真假两格。可很多哲学争论并不是结论在原处为假,而是它被搬到了没有着落的地方。一条在特定模型、对象或逻辑分支里成立的话,离开这些条件以后仍以同样口气说下去,问题首先出在地址。
《第一批判》把这种审查叫作辖域。辖域不是主题标签,也不只是“适用于某些情况”的松散提醒。它要求写明一项判断的效力在哪里取得,越过哪条线以后该判作无所立,而不是仓促改判为假。
这一区分很重要。没有着落,表示现有凭证不够把判词送到那里;它给新的公设、新的经验输入或新的分支留下位置。若一律判假,批判自己反而会冒领跨域裁决权。
一刀留下的四类账
一次划界把“这”从“非这”里切出来,事情却没有干净地分成两个盒子。《第一批判》登记四种欠账方式。第一是收纳之亏:被划入这一侧的对象仍没有被构完整收下。第二是错置:某项被放到不属于它的一侧。第三是两属:同一项同时命中两边。第四是判外:判准在这个对象上没有合法使用。
这四类不是四种真值,也不自称穷尽一切可能。它们是审计开放清单,告诉我们一刀可能在哪些地方欠账。尤其“判外”不能被挤回真假两格;判准若够不到对象,强行给出真假,只是在没有辖权的地方模拟判决。
四栏也不能混成一句“总有余项”。收纳之亏说的是这一构没有收尽;错置与两属说的是当前分类发生了内部问题;判外说的是工具本身没有进入资格。它们的修复方式不同:补充构、重画边界、承认重叠,或者撤回判词。
反面不是余项
划出“这”时,反面是这一刀所规定的“非这”。它与构同生,仍属于当前划界的产物。余项则是这一刀未能穷尽之物,首先表现为被划入一侧却未被收纳的剩余。两者若混在一起,理论会误以为只要说出了反面,就已经保存了未知。
例如把动物分成“会飞”与“不会飞”,不会飞是反面;一只受伤而暂时不能飞的鸟在这套分类中如何处理,可能产生错置或两属;而“飞行”这条标准能否说明动物的生态角色,则可能整个判外。未知不是一个名叫“不会飞”的现成抽屉。
物自体在这里也不是躲在现象背后的单一实体,而是每一个“这”收纳不尽之名。对象不同,极限名随之复数;作业改变,边界也改写。这个名字负责守界,不负责偷偷给边界之外填入内容。
四栏地址
一项主张至少要登记四个坐标。对象栏问它在说什么;层级栏问它使用的工具在发生序列哪一层可用;分支栏问它采取哪套逻辑或理论选择;公设栏问它依赖哪些可替换承担。
“所有判断非真即假”在经典逻辑的某些场域可以成为工作规则,但若把排中律从分支公设伪装成无条件背景,就会静默扩大辖域。“主体必须继续认识”也许可以在一套明定认知驱力的模型里成立,却不能由“非不能非”直接推出。地址一写,争论便从口号回到依赖。
四栏不是行政表格。它们让替换可以传导:若某条公设撤回,依赖它的条目失去原地址,被送往待重审;若只更换分支逻辑,不依赖该分支的主张不必一起倒下。没有地址,理论只能整块接受或整块拒绝。
实得而非趋近
辖域也改变了“进步”的含义。若不存在一个固定、可测距离的全知终点,就不能说每次修订使我们离终点更近。但把一项越界主张送回合法地址,本身是实得;把判外从“假”里分出来,也是实得。
实得不是趋近,因为它不借未来的完整答案为现在担保。一个局部结论可以很稳,同时明确承认它没有覆盖整体。恰恰是边界写得清楚,使这个局部结果能够被可靠地调用。
所以问“一条主张管到哪里”,不是缩小哲学的雄心。它把雄心从占有全部答案,改为建立可复核、可替换、可继续的判断。主张不必因为有限就失效;它只需停止把有限的凭证出示成无限的通行证。
Ask for the Address Before the Verdict
We are trained to place claims into two boxes, true and false. Yet many philosophical failures do not begin with a false conclusion at its proper location. They begin when a statement valid within a model, for an object, or under a logical branch is carried elsewhere without changing its tone. The first defect is an address error.
The First Critique calls the relevant examination jurisdiction. Jurisdiction is not a topic label or a vague warning that a claim applies only “in some cases.” It records where a judgment acquired standing and where it must become no-verdict because its warrant no longer reaches.
No-verdict matters. It says that the present credential is insufficient to deliver a judgment at this address. It leaves room for another postulate, new empirical input, or a different branch. If critique automatically calls everything outside a jurisdiction false, critique itself falsely claims authority over the excluded field.
Four Kinds of Debt Left by a Cut
A demarcation separates “this” from “not-this,” but reality does not fall cleanly into two boxes. The First Critique registers four ways in which an operation can leave debt. The intake-deficit occurs when an object admitted to a side is not exhausted by the construct. Misplacement assigns an item to the wrong side. A double-hit occurs when the same item meets both criteria. Out-of-field means that the criterion has no licensed use for the object in question.
These are not four truth values and they do not claim to exhaust every possibility. They form an open audit list describing where a cut may owe an account. Out-of-field, especially, must not be forced back into true or false. When a criterion cannot reach an object, issuing a verdict merely simulates judgment where no jurisdiction exists.
Nor should the four entries collapse into the slogan that “there is always a remainder.” Intake-deficit concerns what a construct has not taken in. Misplacement and double-hit concern internal trouble in the current classification. Out-of-field concerns the qualification of the instrument itself. Their remedies differ: enrich the construct, redraw the boundary, admit overlap, or withdraw judgment.
The Reverse Side Is Not the Remainder
When “this” is drawn, its reverse side is the “not-this” specified by the same operation. It is produced with the construct and remains internal to the present demarcation. The remainder is what the operation has failed to exhaust, beginning with what remains untaken even on the admitted side. Confusing them allows a theory to believe that naming an opposite has already preserved the unknown.
Suppose animals are divided into those that fly and those that do not. Non-flying is the reverse side. A wounded bird that cannot currently fly may expose misplacement or a double-hit depending on the rule. Whether flight can classify an animal's ecological role may be out-of-field altogether. The unknown is not a ready-made drawer labeled “does not fly.”
The thing-in-itself likewise becomes not a single entity concealed behind all appearances but a limit-name for what each “this” fails to take in. Different objects yield plural limit-names, and new operations rewrite their boundaries. The term guards a limit. It does not fill the other side with secret content.
Four Coordinates
A claim receives at least four coordinates. The object coordinate asks what it concerns. The level coordinate asks where its instruments become available in the developmental sequence. The branch coordinate names the chosen logical or theoretical branch. The postulate coordinate lists the replaceable undertakings on which it depends.
“Every proposition is either true or false” may function within a classical domain, but excluded middle expands its jurisdiction silently when presented as unconditional background rather than a branch postulate. “A subject must continue cognizing” may hold inside a model that explicitly assumes a cognitive drive, but it cannot be derived directly from the performative axiom. Once addresses are explicit, disagreement returns from slogans to dependencies.
The coordinates are not bureaucracy. They make propagation possible. If a postulate is withdrawn, dependent entries lose their old address and return for review. If only a logical branch changes, claims independent of that branch need not collapse with it. Without addresses, a philosophy can be accepted or rejected only as a block.
Gain Without Approach
Jurisdiction also changes the meaning of progress. If there is no fixed, measurable standpoint of complete knowledge, revision cannot be described as reducing our distance from that destination. Yet returning an overextended claim to a valid address is a real gain. Distinguishing out-of-field from false is a gain as well.
Gain is not approach because it does not borrow authority from a future complete answer. A local result can be robust while openly declining to cover the whole. Indeed, a clearly marked boundary is what makes a local result dependable enough to be used elsewhere.
Asking how far a claim reaches does not diminish philosophical ambition. It replaces the ambition to possess every answer with the work of producing judgments that can be reviewed, replaced, and continued. A claim does not fail because it is finite. It fails when a finite credential is presented as an unlimited passport.