← Critique of Pure Reason
第 4 章 · 24 分钟
Schematism and the analogies of experience
Categories acquire cognitive use only in time; the second analogy answers Hume head-on.
概念地图
Schema and analogies
Categories acquire cognitive use only in time; the second analogy answers Hume head-on.
深色为本章概念，浅色为其他章
陪练模式
不想只是读？让它带你走一遍。
导师把这一章拆成小步，每一步先问你，再讲；你用自己的话答，它先认下对的部分再纠偏。用你自己的 AI，对话只存在这台设备上。
Locus01
延伸阅读
Hume · on succession and necessary connection
Points out that what we observe is one thing following another, never the tie that binds them. He argued against reading necessity into observation; the cost is that the distinction between the order of events and the order of my perceptions loses its footing as well.
Foundational
Kant · the Schematism · the schemata of the pure concepts
Categories are pure concepts, intuitions are sensible, and the two are heterogeneous: they cannot be laid on each other without a mediator, and the mediator is a determination of time. He argued against applying categories straight to the material; the cost is a chapter so compressed that Kant himself calls it an art concealed in the depths of the human soul.
Turn
Kant · the analogies of experience · above all the second
Answers Hume head-on: causal rules are not induced from repeated succession, they are the condition under which a subjective sequence of perceptions is judged to be an objective sequence of events. The book's one frontal engagement with Hume.
Turn
The strong and weak readings of the second analogy · twentieth-century debate
Whether the argument establishes 'every event has its particular cause' or only 'there must be some rule by which order is fixed' has never been settled. Which reading holds decides how much of this chapter survives modern physics.
Revision
机制02
Heterogeneous things need a third
Chapter 3 secured the categories' entitlement and left a technical problem: a pure concept carries no sensible content, an intuition is nothing but sensible content, and the two do not meet. Kant's fix is to insert something that touches both sides, which he calls a schema. Its place in the argument is not a patch but a seal: once a category can land only by way of a determination of time, its range of use is closed off with time's — whatever is not in time gets no cognitive use out of the categories. The boundary work of chapters 5 and 6 rests entirely on this step.
机制Two heterogeneous things, through a mediator that touches both, are joined and thereby confined to the mediator's range.
可迁移性测试
Move it to systems integration: two interfaces that know nothing of each other meet through an adapter that speaks both formats. The isomorphism breaks in that an adapter can be written many ways, while Kant's mediator has only one material available to it, namely time.
机制03
Fixing order by rule
The second analogy takes up something that does not look like a problem: how do I know that the order I perceive is the order in which things happened. Kant groups three principles about existence in time under the heading analogies of experience, and the second one's job is to supply the criterion for 'objective order'. What it does in the argument is convert causation from a regularity read off the world into an instrument required to make a judgement at all.
机制Objective order, through being fixable only by rule and never readable off perception, makes causation a condition of experience.
可迁移性测试
Move it to a distributed system: the order in which nodes receive messages is not the order in which events happened, and the real order can be fixed only by causal dependency, never by arrival. The isomorphism breaks in that a system can add a synchronised clock, while here there is no second source.
Derivation04
The shape of the second analogy
What is being reconstructed is the shape of the second analogy, and the range its conclusion is sealed into.
Time itself cannot be perceived
A premise Kant chose. There is no object called time sitting there to check against.
Whether the object is coexistent or successive, my perceptions come one after another
A premise Kant chose; the pair of examples, the house and the ship, exists to serve it.
The difference between the order of events and the order of my perceptions must have some criterion
A premise Kant chose. Give it up and the chapter's problem disappears — which is also why it is the least questioned line.
推导 · 0 / 4
Worked derivation05
The house and the ship
Seeing a house (roof first, then foundation) and seeing a ship drift downstream (upstream first, then downstream) are both successive perceptions. On what ground is the second an objective event and the first not?
First write down how the two perceptions differ as material, then check whether the difference is in the material at all.
裂缝06
Cracks in this reading
争议地形07
本书主张
Causal rules are the condition under which a subjective sequence of perceptions is judged an objective sequence of events, and so hold necessarily of all possible experience.
另一种看法
One line holds that objective order is in practice fixed by cross-checking several observers, instrument records and theory — fallible empirical work; and that modern physics keeps constraints on causal order while dropping any deterministic principle of causation.
分歧扎在
The disagreement is rooted in whether the criterion of objective order is an a priori condition or empirical work, not in whether causation is useful. Both sides grant that perceptual order is not the order of events.
什么证据能裁决
What would settle it is the actual practice of physics: if in some domains objective order is fixed without a deterministic causal rule, that principle is not a necessary condition of any experience. Relativity's causal structure and quantum statistics are public, recheckable cases.
The balance tips to the second: modern physics keeps the ordering constraint and drops determinism. Kant's weak reading — there must be some rule — survives; the strong one does not. What is missing is a formulation that states the weak reading clearly and does not slip back into the strong one in the next sentence.
Textual test08
The 'criterion argument, not induction' reading as a testable specification
The reading under test: the second analogy treats the causal rule as the condition for judging objective order, not as a regularity induced from repeated succession.
Edition: A 1781 and B 1787 side by side, the analogies cited by standard A/B pagination with a standard translation alongside
Fixed in advance; the two editions differ in this section, so both must be checked.
Decisive passages: the second analogy in full, including the house and the ship
Specified in advance, not picked after reading.
What counts as support: the key step is 'without judging thus, objective order could not be distinguished', with no appeal to the number of observations or to repetition
Both must hold; the examples alone, with no criterion-style inference, do not count.
What the rival reading predicts: if this were an inductive argument, the key step should appeal to how often the succession occurs or to the generality of experience
The two readings differ observably within the same section.
Pre-registered failure condition: the decisive passages rest on repeated observation, or the step 'otherwise objective order could not be distinguished' is absent
Either counts as the reading being refuted.
推导 · 0 / 3
接口09
Which slot it hangs on
挂在哪个槽位The conclusion you already hold is probably 'correlation is not causation' — a caution about inference. This chapter goes at the slot before it: where the 'before and after' in your hands came from.
Does this chapter (a) replace 'correlation is not causation', or (b) add a slot: order is not raw material either, and objective sequence has to be fixed by a rule?
慢变量Register two slow variables: how many conclusions you hold depend on some 'before and after', and in how many of those you have checked what rule fixed it. Look at the second first.
小结10
本章小结
01The schema mediates between category and intuition and at the same time seals the categories' use inside time.
02The second analogy asks how objective order can be fixed, not how causal laws were discovered.
03'Every change follows a causal law' solves the criterion problem; it is not a discovery about nature.
04Kant does not write the strong and weak conclusions separately, and modern physics keeps only the weak one.
05'Time itself cannot be perceived' is consistent with any evidence, so this course books it as unadjudicable.
提取练习 · 合上书，先自己答一遍。
?Is 'the perceptual sequence alone cannot distinguish objective order' (a) the first two premises merged or (b) a textual claim refutable by the original?
?Does Kant's answer to Hume address (a) where the idea of necessary connection comes from or (b) how objective order can be fixed?
Forced choice11
A · Identify the mechanism
Which of these is a mechanism?
二选一
Forced choice12
D · Judge the interface
Which one hangs on a slot in your existing structure?
二选一
读到这里，把它记为已读。