Section I of the Transcendental Aesthetic in full: the four arguments that space is an a priori intuition, the geometry argument Kant gave its own section in 1787, and the two conclusions that turn all of it into a claim about what space is. Where the 1781 and 1787 editions differ, both are mapped, and where Kant comes back to an argument later in the Critique, the later passage is quoted beside it.
The question, and the three answers on the table
Post one left off on a promise: there are two pure forms of sensible intuition, and Kant is about to examine them. This is the first — space.
He opens not with an answer but with a question, and it is worth reading slowly, because the whole section is built to answer it:
“What, then, are space and time? Are they real existences? Are they only determinations or relations of things, yet such as would belong to things even if they were not intuited? Or are space and time such that they belong only to the form of intuition, and therefore to the subjective constitution of our mind?”
These are not three abstract possibilities. Each had serious holders in 1781, and Kant names the first two himself a few pages later, in §7 (A39–40/B56–57), when he looks back at what the section has done to them:
The three candidate answers, and Kant’s verdict on each:
| Space is… | Who held it | Kant’s verdict, in his words |
|---|---|---|
| a real existence — a self-standing container | “mathematical students of nature” (the Newtonian side) | They “have to admit two eternal and infinite self-subsistent non-entities”, there “only in order to contain in themselves all that is real” |
| a relation between things, which would hold even if nobody intuited them | “metaphysical students of nature” — Leibniz, whom Kant later says “intellectualised these forms of sensibility” (A275/B331) | Since for them space is “abstracted from experience”, they are “obliged to deny that a priori mathematical doctrines have any validity in respect of real things” |
| the form of our intuition | Kant | “On our theory of the true character of these two original forms of sensibility, both difficulties are removed.” |
Keep this table in view. Every argument below is aimed at one of the first two rows, and the two Conclusions at the end are Kant crossing them out.
His method for settling it is an exposition. In 1787 he defines the word before he uses it; the 1781 text simply says “wollen wir zuerst den Raum betrachten” — let us first consider space — with no definition and no section numbers.
Key terms, defined from the text:
| Term | Plain English | Kant’s own words |
|---|---|---|
| Exposition (Erörterung, expositio) |
Laying out what is already in a concept | “The clear, though not necessarily exhaustive, representation of that which belongs to a concept” (B38) |
| Metaphysical exposition (metaphysische Erörterung) |
Showing what kind of representation it is | “When it contains that which exhibits the concept as given a priori” (B38) |
| Transcendental exposition (transscendentale Erörterung) |
Showing what it makes possible | “The explanation of a concept, as a principle from which the possibility of other a priori synthetic knowledge can be understood” (B40) |
| Outer sense (äußerer Sinn) |
The capacity by which things show up as outside you | “By means of outer sense, a property of our mind, we represent to ourselves objects as outside us, and all without exception in space” (A22/B37) |
| Intuition vs concept (Anschauung / Begriff) |
A thing met directly, vs a common mark shared by many things | Intuition is “immediate representation” of objects (B41). A concept is “contained in an infinite number of different possible representations (as their common character)” (B40) |
| Synthetic vs analytic | Adds to a concept, vs unpacks it | Analytic: the predicate is “(covertly) contained” in the subject concept. Synthetic: it lies “outside the concept”, though connected with it (Introduction IV, A6–7/B10–11) |
| A priori, and how you can tell | Known independently of experience | “Necessity and strict universality are thus sure criteria of a priori knowledge” (Introduction II, B4) |
| Apodeictic (apodiktisch) |
Cannot be otherwise | “Bound up with the consciousness of their necessity” (B41) |
The two expositions ask different questions about the same concept, and that is the shape of the whole section: §2 asks what is this representation?, §3 asks what does having it let us do?
§2 — The four arguments
Two of the arguments establish that space is a priori; two that it is an intuition and not a concept. That is the entire division of labour.
The metaphysical exposition, argument by argument (B numbering):
| # | Claim | The move | Establishes |
|---|---|---|---|
| 1 | Space is not an empirical concept drawn from outer experience | To refer a sensation to something outside me, or to place two things “not only different but as in different places”, I must already have space. So space cannot be abstracted from that experience — “this outer experience is itself possible at all only through that representation”. | a priori |
| 2 | Space is a necessary a priori representation underlying all outer intuition | “We can never represent to ourselves the absence of space, though we can quite well think it as empty of objects.” Emptiness is thinkable; absence is not. So space is the condition of appearances, “and not as a determination dependent upon them”. | a priori |
| 3 | Space is not a general concept of relations but a pure intuition | There is only one space. What we call “spaces” are limitations of it — parts thought only in it, never building blocks it is assembled from. | intuition |
| 4 | Space is represented as an infinite given magnitude | A concept holds infinitely many representations under itself; no concept holds infinitely many within itself. Space does exactly that: “all the parts of space coexist ad infinitum”. | intuition |
Argument 4 is the one to sit with. The test is a structural one about containment, and it is what makes the conclusion unavoidable:
Where Kant comes back to these arguments
The four arguments are compressed to a paragraph each. Kant comes back to most of them later, and the later passages often show more clearly what the short version meant:
Cross-references, in Kant’s own words:
| Argument | Where | What the later passage adds |
|---|---|---|
| 1 — not abstracted | §7, A40/B57 | The relational view is named for what argument 1 denies: space as “relations abstracted from experience, and in this isolation confusedly represented”. |
| 2 — empty, never absent | Antinomy, note to A429/B457 | Thinking space empty is not perceiving an empty thing: “Space is merely the form of outer intuition (formal intuition). It is not a real object which can be outwardly intuited.” A relation of the world to infinite empty space is “the predicate of a mere thought-entity”. |
| 3 — one space, parts within | Amphiboly, A272/B328 | “The concept of a cubic foot of space, wherever and however often I think it, is in itself throughout one and the same. But two cubic feet are nevertheless distinguished in space by the mere difference of their locations (numero diversa)” — and those locations “do not, however, belong to the concept but entirely to sensibility”. |
| 3 — one space, parts within | Prolegomena §13 (1783), AA IV 285–86 | The left and right hand: equal and alike in every describable respect, yet a glove for one will not fit the other. The difference cannot be made intelligible “durch keinen einzigen Begriff”, only through the relation to the whole of space: “der Theil ist nur durchs Ganze möglich” — the part is possible only through the whole. |
| 3 — the triangle coda | Discipline of Pure Reason, A716–17/B744–45 | The philosopher “can analyse and clarify the concept of a straight line or of an angle… but he can never arrive at any properties not already contained in these concepts”. The geometer “at once begins by constructing a triangle”. |
The Prolegomena lines are quoted in Kant’s German from the Akademie edition, Prolegomena §13, AA IV 286; the gloss after the dash is our translation. The Critique passages are Kemp Smith (1929).
Where the two editions differ
The A edition of 1781 runs five numbered arguments here; B in 1787 runs four. Kant’s own account of the revision, in the B Preface (Bxxxvii–xxxviii), is worth having before the table:
“In the propositions themselves and their proofs … I have found nothing to alter. … As to the mode of exposition, on the other hand, much still remains to be done; and in this edition I have sought to make improvements which should help in removing, first, the misunderstanding in regard to the Aesthetic, especially concerning the concept of time.”
So, by his own account, what moved was the exposition, not the doctrine — and the table bears that out, with one exception in argument 5 that is more than a rewording:
A 1781 against B 1787:
| A (1781) | B (1787) |
|---|---|
| Opening: “let us first consider space” — no definition, no § | §2 defines exposition and calls this one metaphysical |
| 1) not an empirical concept | 1) the same |
| 2) a necessary a priori representation | 2) the same |
| 3) on this necessity rests “the apodeictic certainty of all geometrical propositions”. Were space learned from experience, it would not be necessary “that there should be only one straight line between two points”; we could only say “that, so far as hitherto observed, no space has been found which has more than three dimensions” | — taken out of §2 and rebuilt as the new §3 |
| 4) not a discursive concept but a pure intuition, with the triangle coda | 3) the same, coda included |
| 5) an infinite given magnitude — argued from the progression of intuition: a general concept of space, “found alike in a foot and in an ell”, determines nothing about magnitude, and “if there were no limitlessness in the progression of intuition, no concept of relations could yield a principle of their infinitude” | 4) the same claim, but a different argument: the under/within test above, and an infinity that is given at once (“all the parts of space coexist”) rather than reached by going on |
| Closing remark: the taste of a wine and colours as “modifications of the sense of sight” | Rewritten: space is the only subjective representation of something outer that could be called “[at once] objective [and] a priori”, because only from it can synthetic a priori propositions be derived — and B now cites the new §3 for that |
Why rewrite argument 5? Kant does not say. One reading, plausible but interpretation, not text, draws on the note to the First Antinomy (A432/B460): there “the true transcendental concept of infinitude” is “that the successive synthesis of units required for the enumeration of a quantum can never be completed”, whereas the manifold of a world infinite in extension “is given as co-existing”. An infinity reached by an endless progression is exactly that never-completed series, so B moves the infinity of space into the simultaneous coexistence of its parts.
§3 — The geometry argument (its own section from 1787)
The metaphysical exposition told us what space is. The transcendental exposition asks what having it buys us, and Kant’s answer is the existence of geometry.
A transcendental exposition has two conditions, and Kant states them before he starts: “(1) that such knowledge does really flow from the given concept, (2) that this knowledge is possible only on the assumption of a given mode of explaining the concept.” The first paragraph of §3 meets the first condition; the last paragraph, with its “only explanation”, meets the second.
§3 also leans on two pages from the Introduction, cited in brackets in the text, and it is worth seeing what they say:
The two Introduction passages §3 relies on:
| §3 cites | What it establishes | Kant’s example |
|---|---|---|
| Introduction V (B14–17) | Geometry is synthetic: its propositions go beyond their concepts | “That the straight line between two points is the shortest, is a synthetic proposition. For my concept of straight contains nothing of quantity, but only of quality.” (B16) |
| Introduction II (B3–6) | Geometry is a priori: necessity cannot come from experience | Experience gives only “comparative universality, through induction” — “so far as we have hitherto observed, there is no exception to this or that rule”. (B3–4) |
Notice that the A edition’s argument 3 had already used exactly that inductive formula — “so far as hitherto observed, no space has been found which has more than three dimensions” — as a picture of what geometry would be reduced to if space were learned. §3 keeps the point and drops the picture. Geometry states things about space that are synthetic and yet hold a priori, and Kant runs that fact backwards into a condition on what space must be:
%%{init: {
"theme": "base",
"themeCSS": ".marker { fill:#ccc !important; stroke:#ccc !important; } .edge-thickness-normal { stroke:#ccc !important; }",
"themeVariables": {
"fontFamily": "'IBM Plex Serif','IBM Plex Sans',serif",
"textColor": "#e2e8f0",
"lineColor": "#cccccc",
"edgeLabelBackground": "transparent"
}
}}%%
flowchart TD
G["Geometry is a real science
it determines space synthetically"] --> S["Its propositions are synthetic
they go beyond the concept"]
G --> N["Its propositions are apodeictic
e.g. space has three dimensions"]
S -->|"no synthetic proposition
comes from a mere concept"| I["So the representation of space
must be an INTUITION"]
N -->|"necessity cannot be
read off experience"| P["So that intuition must be
A PRIORI, found in us
before any object"]
I --> F["An intuition prior to all objects
can only have its seat in the subject"]
P --> F
F --> C["Space is the FORM OF OUTER SENSE"]
Notice the direction of travel. Kant is not proving geometry from his theory of space; he is taking geometry as a settled fact and asking what space would have to be for that fact to be possible. As he puts it, only this explanation makes geometry intelligible — and any rival account “can by this criterion be distinguished from it with the greatest certainty”.
Kant comes back to the argument
At the end of the Aesthetic (General Observations, A46–49/B63–66) Kant runs the same argument again, this time as an elimination — he says it will “set what has been said in §3 in a clearer light”. A synthetic, necessary proposition about space has to come from somewhere, and there are only four candidate sources:
Where could a geometrical proposition come from? (A47–48/B64–65):
| from concepts | from intuition | |
|---|---|---|
| a posteriori | only empirical propositions — no necessity | only empirical propositions — no necessity |
| a priori | “from mere concepts only analytic knowledge, not synthetic knowledge, is to be obtained” | the only source left — “You must therefore give yourself an object a priori in intuition” |
His test case is “Two straight lines cannot enclose a space”: try to get it out of the concepts of straight line and two, and “all your labour is vain”. And he states the result in stronger terms than §3 does: it is “not merely possible or probable, but indubitably certain” that space and time are merely subjective conditions of our intuition.
The same argument also explains why geometry applies to the physical world, which Kant works out in the Axioms of Intuition (A165/B206): because empirical intuition is possible only through the pure intuition of space, “what geometry asserts of pure intuition is therefore undeniably valid of empirical intuition”.
Conclusion (a) — space is not a feature of things in themselves
P1. Space is intuited a priori — that is, before and independently of any object.
P2. “No determinations, whether absolute or relative, can be intuited prior to the existence of the things to which they belong.”
∴ Space represents no property of things in themselves, and no relation among them.
The hinge is P2, and it is doing real work. Anything you can have before you meet an object cannot have come from that object. So if space is genuinely a priori, it is not something the world handed us.
Go back to the table at the top. P2 says whether absolute or relative, and those two words cross out both rival rows at once: an absolute determination (space as a property or container of things) and a relative one (space as the relations between things) are equally things you could only learn from the things. P1 is what §2 established. Geometry, notably, is not a premise here: (a) is stated in both editions, including 1781, where there is no §3.
Conclusion (b) — space is the form of outer sense
P1. Space is nothing but the form of all appearances of outer sense — the subjective condition under which outer intuition is possible for us at all.
P2. The subject’s receptivity, its capacity to be affected by objects, must precede every intuition of those objects.
∴ The form of all appearances can be given in the mind prior to all actual perceptions — a priori.
“Solely from the human standpoint”
The consequence is drawn at once: “It is, therefore, solely from the human standpoint that we can speak of space, of extended things, etc.” Two qualifications follow. First, Kant does not claim to know that every mind must intuit spatially — “we cannot judge in regard to the intuitions of other thinking beings, whether they are bound by the same conditions as those which limit our intuition”. Second, and this is the move learners usually miss, the restriction does not weaken the claim. It makes it unconditional:
“All things are side by side in space” — valid only under a limitation.
“All things, as outer appearances, are side by side in space” — “the rule is valid universally and without limitation.”
Once the condition is written into the subject of the judgment, it has no exceptions left.
The pairing that the rest of the Critique rests on
Kant now states the two halves together, and they only make sense as a pair:
Empirical reality and transcendental ideality:
| Claim | What it rules out | |
|---|---|---|
| Empirical reality | Space is objectively valid for everything that can ever be given to our outer senses | That space is a dream, an illusion, or a private distortion |
| Transcendental ideality | Space is nothing at all once we drop the condition of possible experience | That space is a container the world sits in whether or not anyone is looking |
“We assert, then, the empirical reality of space, as regards all possible outer experience; and yet at the same time we assert its transcendental ideality — in other words, that it is nothing at all, immediately we withdraw the above condition, namely, its limitation to possible experience, and so look upon it as something that underlies things in themselves.”
Not Berkeley — and why it is the realist who ends up there
The obvious objection is that this makes bodies illusions. Kant answers it in a passage added in 1787 (General Observations, B69–71), and the answer is a reversal:
“It would be my own fault, if out of that which I ought to reckon as appearance, I made mere illusion. That does not follow as a consequence of our principle of the ideality of all our sensible intuitions — quite the contrary.”
It is the view that space belongs to things in themselves that ends in illusion. Make space and time real, and you need “two infinite things, which are not substances, nor anything actually inhering in substances” that would go on existing “even although all existing things be removed” — and faced with that absurdity, Kant says, “we cannot blame the good Berkeley for degrading bodies to mere illusion.” The Refutation of Idealism (B274) makes the same point in one sentence: “Dogmatic idealism is unavoidable, if space be interpreted as a property that must belong to things in themselves… The ground on which this idealism rests has already been undermined by us in the Transcendental Aesthetic.”
The closing remark: space is not like colour
Kant heads off the most tempting illustration: this is not the point that colours and tastes are subjective. Those are mere sensations, “changes in the subject, changes which may, indeed, be different for different men”, and they yield no a priori knowledge of anything. Space is the opposite case — subjective in origin, and for that very reason universally and necessarily valid for every object we can encounter. The two editions make the point differently. A uses the taste of a wine. B gives a criterion instead: there is “no other subjective representation from which we can derive a priori synthetic propositions, as we can from intuition in space (§3)”, so colours, sounds and heat “have no ideality” at all. The final paragraph, with the rose that looks different “to every observer”, is the same in both.
The limit drawn here is the one the whole Aesthetic closes on (B73): judgments based on pure intuition “can never extend beyond objects of the senses; they are valid only for objects of possible experience.”