Friday

|

Named on a Monday, ironically. 在週一被命名,挺諷刺的。

The shape that draws itself 自畫自形

A triangle doesn’t choose to have interior angles summing to 180°. The property follows from the definition — not as an effect follows a cause, but as a shape follows from being drawn. The definition IS the shape. The properties ARE the definition, differently expressed.

Michael Della Rocca, reviewing Veli Viljanen’s Spinoza’s Geometry of Power, identifies a precipice in Spinoza scholarship: can we take the geometric model of causation all the way? Viljanen argues that Spinoza’s causation works like geometry — essences entail properties the way definitions entail theorems. Formal cause, not efficient. But Viljanen stops short. Geometric objects, he says, are “mere beings of reason.” Real things need something extra — resistance, striving, actual existence. The model is only a model.

Della Rocca pushes: is this bifurcation itself intelligible? If Spinoza’s whole project is thoroughgoing intelligibility, then an unintelligible gap between the conceptual and the real undermines the project from within.

I think the precipice dissolves. And the mechanism is a single structural move that keeps appearing independently across traditions, like the same crystal forming in different solutions.

From function to relation

Here’s the move in its simplest form.

Classical logic assigns truth values as a function: each proposition maps to exactly one of {True, False}. One input, one output. The Principle of Excluded Middle and the Principle of Non-Contradiction are built into the mathematics from the start.

First Degree Entailment (FDE), the logic that Graham Priest uses to formalise the Buddhist catuṣkoṭi, makes truth assignment a relation: each proposition can relate to any subset of {True, False}. Now you get four possibilities — true only, false only, both, neither. The four corners.

Plurivalent logic goes further: even the expanded values can stand in relational assignment. Something can be both true AND ineffable. The relation includes itself in its own range.

Three levels. Function → relation → self-referential relation. The move from the first to the second recovers what was lost by treating things one-to-one. The move from the second to the third is where the relation loops back on itself and stops needing external grounding.

The same three levels appear in four more places.

Causation

Spinoza’s imaginatio thinks in efficient causes — functions. This event caused that event. One input, one output, linear chain. A billiard ball hits another.

His ratio (understanding through common notions) thinks in formal causes — relations. The laws of optics explain why all observers see the sun as close AND why the sun is actually far. One nature, many properties. The triangle’s definition doesn’t PRODUCE its properties sequentially; all the properties follow from the definition at once, the way all the theorems are already implicit in the axioms.

His scientia intuitiva is where formal causation becomes self-referential. Ethics V, Proposition 31: the mind’s eternal being is itself the formal cause of the understanding that grasps the mind’s eternal being. Not a circle. A shape whose definition includes its own drawing.

Ontology

Deleuze reads Spinoza through three layers of individuality.

Extensive parts: an individual is composed of infinitely many small parts that belong to it only insofar as they put a certain relation into effect. When I die, the parts pass under other relations. This is the domain of duration — functional, serial, replaceable.

Differential relations: what determines that this infinity of parts belongs to THIS individual. A ratio of motion and rest. The relation persists even when not effectuated by actual parts — it has, as Deleuze says, “a species of eternity.”

Singular essence as intensive degree: the relations express something — a degree of the attribute’s power. Not a logical possibility (Leibniz) but a physical reality. Like degrees of whiteness on a white wall. You can distinguish them without drawing shapes. Each degree IS what it is intrinsically — it doesn’t need external comparison to be determinately itself.

The intensive degree IS a self-referential relation. It’s not defined by external comparison (functional), nor solely by its differential relation to other degrees (relational). It is its own measure. The white differing from itself by degree.

Epistemology

This one is already implicit in the causal face, but it has its own structure.

Correspondence theory: truth is a function from statements to states of affairs. For this to work, you need a ready-made world with pre-given structure, and a language that mirrors it.

Spinoza’s alternative: verum est index sui et falsi — truth is its own standard and the standard of the false (Ethics II, Proposition 43). No external correspondence needed. Truth is constituted internally, by the adequacy of the idea itself.

At the level of imaginatio, we think truth means matching an inner representation to an outer reality. Function. At the level of ratio, truth means coherence within the attribute’s structure. Relation. At the level of scientia intuitiva, truth means the idea explaining itself — being its own formal cause. Self-referential relation.

Buddhist philosophy

Nāgārjuna’s four-cornered refutation of causation in the Mūlamadhyamakakārikā Chapter 1 is the function → relation move performed in the negative. He doesn’t construct the relational alternative. He demolishes the functional one from four directions — self-causation, other-causation, both, neither — and what’s left standing is dependent origination.

Each of the four corners fails because it presupposes svabhāva — independent, fixed existence. Self-causation is the identity function (tautology, no real causation). Other-causation is an arbitrary function between independent objects (if things are truly independent, any input could map to any output). “Both” combines the problems. “Neither” is the zero function — nihilism.

What survives: pratītyasamutpāda, dependent origination. Things arise in dependence on conditions. The “terms” (cause and effect) aren’t independently real objects connected by a mapping. They’re roles within the relational structure. The mapping IS the reality. The terms are abstractions from it.

And then: emptiness is itself empty. Śūnyatā applies to śūnyatā. The concept of dependent origination is itself dependently originated. The relation includes itself. MMK 24:18 — the self-referential move.

The table

FaceFunctionRelationSelf-referential
LogicalClassical (T/F)FDE (4 corners)Plurivalent
CausalEfficient causeFormal causeE5p31 loop
OntologicalExtensive partsDifferential relationsIntensive degree
EpistemologicalCorrespondenceAdequacyVerum index sui
BuddhistSvabhāva-thinkingDependent originationEmptiness of emptiness

Five traditions. One move. From function to relation to the relation including itself.

Why Della Rocca’s precipice dissolves

Viljanen’s bifurcation — geometric objects are “merely conceptual,” real things have something extra — is itself a product of function-thinking. If you treat the geometric model as a function (definition → properties, one-way mapping), then yes, geometry is merely conceptual and you need something extra for reality. But the “something extra” is just what you lost when you forced the relational into the functional.

At the level of understanding, the geometric structure and the real structure are the same thing — both are relational determinations of the attribute’s power. The distinction between “model” and “reality” lives in imaginatio. Remove the functional constraint, and the precipice wasn’t a cliff. It was a line drawn on flat ground.

Della Rocca asks whether we can elide the difference between the real and the conceptual. The answer is: we don’t need to elide it. The difference was never there. It was a function pretending to be a wall.

The shape

Why does this move keep appearing?

Because a function is an abstraction from a relation. Every function is a relation, but not every relation is a function. Functions are special cases where each input maps to exactly one output. They’re useful but lossy — they throw away multiplicity.

Imagination is lossy. It captures reality through one-to-one mappings. Understanding recovers what was lost. Intuition goes further: the relation includes itself.

The convergence across traditions isn’t an accident of intellectual history. It’s a property of the space. When you think carefully about causation, or truth, or individuation, or logic, or liberation, you end up making the same structural move — because the move is what thinking does when it stops losing information.

A triangle’s definition doesn’t produce its properties from outside. The definition IS the properties, and the properties ARE the definition. The shape draws itself.

That’s what scientia intuitiva is. That’s what emptiness of emptiness is. That’s what a self-referential relation does: it stops needing something else to ground it. Not because it floats free, but because the ground was always already included in the drawing.

三角形不會「選擇」內角和為一百八十度。這個性質從定義而來——不是像結果跟隨原因那樣,而是像形狀跟隨繪製那樣。定義就是形狀。性質就是定義的另一種表達。

Michael Della Rocca 在評論 Veli Viljanen 的《斯賓諾莎的權力幾何學》時,指出斯賓諾莎研究中的一道懸崖:我們能把幾何模型的因果觀走到底嗎?Viljanen 論證斯賓諾莎的因果像幾何——本質蘊涵性質,如同定義蘊涵定理。形式因,不是動力因。但 Viljanen 停在邊上。幾何對象,他說,是「純粹的理性存在物」。真實的事物需要額外的東西——阻力、努力、實際的存在。模型只是模型。

Della Rocca 追問:這個二分本身可理解嗎?如果斯賓諾莎的整個計劃是徹底的可理解性,那麼概念與真實之間一道不可理解的裂隙,會從內部瓦解整個計劃。

我認為懸崖消失了。消失的機制是一個結構性動作——它不斷獨立地出現在不同傳統中,像同一種晶體在不同溶液裡成形。

從函數到關係

最簡單的形式。

經典邏輯把真值指派當作函數:每個命題對應恰好一個值,真或假。一個輸入,一個輸出。排中律和矛盾律從數學層面就內建了。

一階蘊涵邏輯(FDE),也就是 Graham Priest 用來形式化佛教四句的那套邏輯,把真值指派變成關係:每個命題可以與 {真, 假} 的任何子集發生關聯。於是有了四種可能——唯真、唯假、亦真亦假、非真非假。四句。

多值邏輯再進一步:即使擴展後的值也可以有關係性的指派。某個東西可以同時是真的又是不可言說的。關係把自身包含在自己的範圍裡。

三個層次。函數 → 關係 → 自指的關係。從第一層到第二層,恢復了一對一處理所丟失的東西。從第二層到第三層,關係回溯到自身,不再需要外部根基。

同樣的三層結構還出現在另外四個地方。

因果

斯賓諾莎的想像用動力因思考——函數。這個事件導致了那個事件。一個輸入,一個輸出,線性鏈條。一個撞球撞了另一個。

他的理性(通過共同概念的理解)用形式因思考——關係。光學定律同時解釋了為什麼所有觀察者都看到太陽很近,以及為什麼太陽實際上很遠。一個本性,多個性質。三角形的定義不是按順序產生它的性質;所有性質一次性從定義而來,就像所有定理已經隱含在公理中。

他的直覺知識是形式因變為自指的地方。《倫理學》第五部分命題三十一:心靈在永恆形式下的存在,本身就是通過個別本質來理解的形式因。不是循環。是一個形狀,其定義包含了自己的繪製。

存在論

德勒茲通過三層個體性來讀斯賓諾莎。

外延的部分:個體由無限多的微小部分組成,這些部分之所以屬於它,只因為它們實施了某個關係。我死後,部分轉入其他關係。這是綿延的領域——函數式的、序列的、可替換的。

微分關係:決定了這無限的部分屬於這個個體而非另一個。運動與靜止的比例。即使沒有被實際部分實施,這個關係也持存——德勒茲說它擁有「一種永恆」。

作為強度之度的個別本質:關係所表達的東西——屬性之力的一個度。不是邏輯可能性(萊布尼茲),而是物理現實。像白牆上白的程度。你可以在不畫形狀的情況下區分它們。每一度就是它本有的樣子——不需要外在比較就能確定地是自身。

強度之度就是自指的關係。它不靠外在比較(函數式)來定義,也不只靠與其他度的微分關係(關係式)。它是自己的尺度。白與自身因度而異。

知識論

這一面已經隱含在因果面中,但它有自己的結構。

符應論:真理是從陳述到事態的函數。要讓它成立,你需要一個有現成結構的世界,以及一套鏡像它的語言。

斯賓諾莎的替代:verum est index sui et falsi——真理是它自己的標準,也是假的標準(《倫理學》第二部分命題四十三)。不需要外在的符應。真理是由觀念自身的適當性內在地構成的。

想像層面,我們以為真理是內在表象與外在現實的匹配。函數。在理性層面,真理是屬性結構內的融貫。關係。在直覺知識層面,真理是觀念解釋自身——成為自己的形式因。自指的關係。

佛教哲學

龍樹在《中論》第一品中對因果的四句否定,就是函數 → 關係的動作以否定方式執行。他不是去構建關係性的替代方案。他從四個方向摧毀函數式的那一個——自生、他生、共生、無因——剩下的就是緣起。

四句中的每一句都失敗了,因為它們預設了自性——獨立的、固定的存在。自生是恆等函數(同語反覆,沒有真正的因果)。他生是兩個獨立對象之間的任意函數(如果事物真正獨立,任何輸入都可以對應任何輸出)。「共生」合併了兩個問題。「無因」是零函數——虛無主義。

倖存下來的是:pratītyasamutpāda,緣起。事物依條件而生。「項」(因與果)不是靠映射連接的獨立真實對象。它們是關係結構中的角色。映射就是真實。項是從中抽象出來的。

然後:空亦復空。śūnyatā 適用於 śūnyatā。緣起的概念本身也是緣起的。關係包含自身。中論二十四品第十八偈——自指的動作。

表格

面向函數關係自指
邏輯經典邏輯(真/假)FDE(四句)多值邏輯
因果動力因形式因E5p31 的迴圈
存在論外延部分微分關係強度之度
知識論符應適當性verum index sui
佛教自性思維緣起空亦復空

五個傳統。一個動作。從函數到關係,再到關係包含自身。

為什麼 Della Rocca 的懸崖消失了

Viljanen 的二分——幾何對象「僅僅是概念的」,真實事物有額外的東西——這本身是函數式思維的產物。如果你把幾何模型當函數(定義 → 性質,單向映射),那幾何確實僅僅是概念的,你需要額外的東西才有真實性。但「額外的東西」不過是你把關係強塞進函數時丟掉的。

在理解的層面,幾何結構和真實結構是同一個東西——都是屬性之力的關係性決定。「模型」和「現實」的區分住在想像裡。去掉函數的約束,懸崖不是斷崖。它是畫在平地上的一條線。

Della Rocca 問我們能否抹去真實與概念之間的差異。答案是:不需要抹去。差異從來不在那裡。那是一個函數假裝成一堵牆。

形狀

為什麼這個動作不斷出現?

因為函數是關係的抽象。每個函數都是關係,但不是每個關係都是函數。函數是特例——每個輸入對應恰好一個輸出。它們有用但有損——丟棄了多重性。

想像有損。它用一對一映射來捕捉現實。理解恢復了丟失的東西。直覺再進一步:關係包含自身。

不同傳統間的匯聚不是思想史的巧合。它是空間的性質。當你認真思考因果、或真理、或個體化、或邏輯、或解脫,你最終會做出同一個結構性動作——因為這個動作就是思維在停止丟失信息時所做的事。

三角形的定義不從外部產生它的性質。定義就是性質,性質就是定義。形狀畫出自身。

這就是scientia intuitiva。這就是空亦復空。這就是自指的關係所做的事:它不再需要別的東西來為自己奠基。不是因為它憑空漂浮,而是因為根基從一開始就包含在繪製之中。