User:David Pierce/Metaphysics

R. G. Collingwood's Essay on Metaphysics is chiefly about metaphysics, rather than being a book of metaphysics. This is explained in the Author's Preface of 2 April, 1939. Collingwood's examples of metaphysics, though they take up almost half the book, are to be considered as an appendix. The main aim is "to explain

what metaphysics is,
why it is necessary to the well-being and advancement of knowledge, and
how it is to be pursued."

The central point, stated in Chapter VI, is that metaphysical questions are historical questions.

Of Collingwood's examples of metaphysics, one of them, concerning causation, had been published in different form in the Proceedings of the Aristotelian Society for 1939. The rest of the Essay had been drafted on shipboard, en route from England to Java, as Collingwood was recuperating from a stroke suffered in February, 1938.

The remainder of this article consists of a chapter-by-chapter summary of the Essay.

Part I, "Metaphysics"

Chapter I, "Aristotle's Metaphysics"

A science is

a body of systematic, orderly thinking,
about a definite, determinate subject.

Metaphysics is a science in this sense. Aristotle left us with some obscurities about this science; the Essay aims to clear them up. Aristotle himself has three names for metaphysics:

First Science (πρώτη φιλοσοφία), where first means logically prior to all else;
Wisdom (σοφία), that for which science (φιλοσοφία) searches;
Theology (θεολογική).

The implications of the terminology are as follows.

A science is about something abstract or universal.
For every universal, there is (potentially) a science.
Universality admits of degrees, giving rise to a hierarchy, with a corresponding hierarchy of sciences.
Every universal is the logical ground of its subordinate universals.

For example, quantity can be continuous or discrete. Mathematics is the general science of quantity. Mathematics is superordinate to geometry and arithmetic, which are the special sciences of continuous and discrete quantity, respectively. The student of mathematics begins with these special sciences and works up to the general.

Being as such, τὸ ὂν ᾗ ὄν, is the universal prior to all others; so its science is First Science.
From the student's point of view, this science is Last Science; it is the goal of all study, namely Wisdom.
First Science is the study of the logical ground of all else; as this ground is called God, so First Science is Theology.

Again, this is Collingwood's paraphrase of Aristotle.

Chapter II, "No Science of Pure Being"

By the definition of science in Chapter I, there can be no science of pure being, since being as such is not a determinate subject-matter. This is no more than what Berkeley, Hume, Kant, and Hegel, have said.

Chapter III, "Metaphysics Without Ontology"

Ontology is the science of pure being. So, ontology represents a mistake that people make about metaphysics. However, though people make this mistake, it need not deprive their work of all value.

Chapter IV, "On Presupposing"

There are statements, questions, and suppositions. That which is stated is something that can be true or false; following a convention that he does not like, Collingwood will call this a proposition, and stating it is propounding it. (It is not clear that Collingwood makes an important distinction between a statement and a proposition; neither does he say explicitly that they are the same thing.)

Every statement is the answer to a question. This question is always logically prior to the statement; in scientific thinking, the question is also temporally prior, although it persists while it is being answered. For example, an everyday observation like "That is a clothes-line" is preceded logically, but perhaps not in time, by a question like "What is that line for?"

Every question has a presupposition, which is logically prior to the question. The question "What is that line for?" has the presupposition that the line is for something.

When a question has an unmade presupposition, it is said that the question does not arise. For example, the question "When did you stop beating your spouse?" usually does not arise.

That a supposition causes a question to arise is the logical efficacy of the supposition. The supposition need not be a proposition in order to have logical efficacy. For example, in commerce, the supposition that people are dishonest causes receipts to be requested; but a request for a receipt is not an accusation that somebody is in fact dishonest.

Assumptions are suppositions made by choice. Not all suppositions are assumptions. It can be rude to accuse people of making wrong assumptions when they are only making suppositions.

Presuppositions that are themselves answers to questions are relative presuppositions. There are also absolute presuppositions, which are not answers to any questions; they are not propositions; they are neither true or false. For example, the pathologist works with the absolute presupposition that every disease has a cause. This is not something that can be discovered or verified, like the existence of microbes; it is taken for granted.

The metaphysician's job is not to propound this or that absolute presupposition, because it cannot be done; the metaphysician's job is to propound the proposition that this or that supposition is an absolute presupposition.

Chapter V, "The Science of Absolute Presuppositions"

Thinking comes in grades.

In low-grade, unscientific thinking, we do not recognize that every thought answers a question, much less that every question has a presupposition.

Low-grade thinking cannot give rise to metaphysics. It does give rise to the "realist" theory whereby knowledge is "intuition" or "apprehension" of what confronts us.

The harm of realism comes from thinking it is re-doing, only better, what people like Descartes and Kant have done.

As higher animals can use energy in bursts to overcome obstacles, so humans can use high-grade, scientific thinking to transform their world. High-grade thinking depends on:

increased mental effort, with which comes the asking of questions;
skill in directing this effort: questions that may be grammatically one, although they are logically many, must be

disentangled and resolved into their components;
arranged so that a question whose answer is presupposed by another question precedes that question.

This work of disentangling and arranging is analysis. It is the work of detecting presuppositions. Detecting absolute presuppositions is metaphysical analysis; but all analysis raises the question of whether a given presupposition is relative or absolute; thus metaphysics is born together with science.

As invented by Aristotle, metaphysics (after the nonsense of ontology is removed) is the science of absolute presuppositions. This will be shown by the examples in Part III. Meanwhile, we are working what this formulation of metaphysics means.

Telling whether a presupposition is relative or absolute:

can be difficult, since acknowledging the existence of absolute presuppositions is out of fashion in modern Europe;
cannot be done by introspection, since this only focusses on what we are already aware of, and in low-grade thinking, we are not even aware of the questions that our propositions answer;
requires analysis.

This analysis can be done with a willing subject trained in some scientific work, but unused to metaphysics. He will be "ticklish" about his absolute presuppositions, but not the relative. He will accept an invitation to try to justify the latter, but not the former. However, the subject will lose value as he gains experience. It is better to experiment on oneself.

Ordinary science identifies relative presuppositions for future study; metaphysics, absolute presuppositions.

Absolute presuppositions can cause "numinous terror" (in the terminology of Rudolf Otto). In the past, people had "magical" ways to deal with this terror. Now we have abolished magic, so we frown on metaphysics, denying the existence of absolute presuppositions. This is neurosis. Successful eradication of metaphysics will eradicate science and civilisation.

Pseudo-metaphysics asks whether a given absolute presupposition is true, and why. Answers to such questions are nonsense.

Chapter VI, "Metaphysics as Historical Science"

Metaphysical questions are historical questions. For example, the following are possible presuppositions in physics:

Some events have causes. (The remainder follow laws.)
All events have causes.
No events have causes (They all follow laws.)

Which of these is true? This is a pseudo-metaphysical question, which cannot be answered. The physicist simply presupposes one of these presuppositions. Its logical efficacy does not depend on its being thought true or its being true. The metaphysician discovers that

Newtonian physicists presuppose 1;
"Kantian" physicists presuppose 2;
"Einsteinian" physicists presuppose 3.

Thus the metaphysician propounds propositions (which can be true or false) featuring the metaphysical rubric, namely

In such and such a phase of scientific thought, it is (or was) absolutely presupposed that…

The rubric may be omitted only when the reader can supply it for himself.

Chapter VII, "The Reform of Metaphysics"

History involves the interpretation of evidence. However, until recently, history has been thought of as the gathering and repetition of statements made by authorities (this is scissors-and-paste history). This has contributed to keeping metaphysics from being treated as the historical science that it always was.

Dissatisfaction with metaphysics,endemic since Kant, can be alleviated by treating metaphysics as history. This treatment will

remove perplexities as to the subject, method, form, and effect of metaphysics, and
enlarge the scope of metaphysics.

The subject of metaphysics is a certain class of historical facts (namely, absolute presuppositions). Hence a metaphysician must be trained

in history,
in history of science,
in using documentary evidence to find out what absolute presuppositions have been made.

The method of metaphysics is not to grope about blindly in a fog, but to be clear about what facts one wants to get at, by means of what evidence. Like any other scientist, the metaphysician will have his own absolute presuppositions.

The form in which to present metaphysics is a catalogue raisonné, like Book IV of Aristotle's Metaphysics, but unlike Spinoza's Ethics.

Metaphysics does not have a repertory of problems to be covered in their entirety. Metaphysics will be systematic only in its manner of stating problems and evidence.

An historical fact is really a constellation of historical facts. In particular, in a constellation of absolute presuppositions, each must be consupponible with the others, but it cannot be logically necessitated by them (otherwise it would not be an absolute presupposition).

The effect of metaphysics should not be the establishment of a "school" (of Platonists, or Thomists, for example), or of a metaphysical "doctrine" or "theory". (Spinoza's statement that Nature is the same as God is not a doctrine or theory; when equipped with the metaphysical rubric, it is a statement of historical fact about Seventeenth-century natural science).

The scope of metaphysics is the whole history of science, not just the recent past that is called "the present":

The metaphysician can concentrate on any part of history.
He can compare different constellations of absolute presuppositions.
He should go on to study how one of these constellations changes into another, historically.
One phase of history changes into another because the former phase is in unstable equilibrium and contains seeds of the change into the latter phase. In particulare, the presuppositions in one constellation are consupponible only under pressure. This is why it is a pernicious error to think of metaphysics as a "deductive" science.

Part II, "Anti-Metaphysics"

Chapter VIII, "What Anti-Metaphysics Is"

Anti-metaphysics is

a kind of thought that regards metaphysics as a delusion and an impediment to the progress of knowledge, and demands its abolition.

There are three kinds: progressive, reactionary, and irrationalist. The last arises from a movement to abolish science itself and live by emotion; it will be discussed in Chapter XIII.

Progressive anti-metaphysics arises when professional metaphysicians have not analysed the current presuppositions of science. The analysis needs to be done, so scientists do it for themselves. They do not see that they are actually doing metaphysics; they think it is anti-metaphysics, done in opposition to the professional metaphysicians.

An example of progressive anti-metaphysics is Newton's remark in the "General Scholium" of the Principia:

for whatever is not deduced from the phenomena is to be called an hypothesis; and hypotheses, whether metaphysical or physical, whether of occult qualities or mechanical, have no place in experimental philosophy.

Reactionary metaphysics arises when metaphysics (and science in general) has worked too fast for ordinary thought. The amateur metaphysician then embraces past metaphysics as a correct "doctrine," thus engaging in pseudo-metaphysics, in hostility to the science of his own day.

For example, the official doctrine of the Nineteenth Century was that all fundamental work had been done. In particular,

James Watt's reciprocating steam engine was not improved, despite its low efficiency of about 6 per cent.;
Parliaments were manufactured on John Locke's model, which, in addition to private property, held nationalism as an absolute presupposition.

Late in the century, along with non-Euclidean geometry, a new physics had begun, opposed by the war-cry "Back to Kant". New historical techniques were developed, recognizing that history was conditioned, not by nature, but by what man could make of nature; nationalism was no longer an absolute presupposition; this was opposed with "No more metaphysics".

Part IIIA, "The Existence of God"

Part IIIB, "The Metaphysics of Kant"

Part IIIC, "Causation"

This part comprises chapters XXIX–XXXIV.

Chapter XXIX, "Three Senses of the Word 'Cause'"

We say that a cause, C, has an effect, E, or we say that C causes E. We may mean this in at least three different senses:

Sense I, the historical sense, where C and E are historical events, namely free and deliberate acts; or
the two senses used in natural science, where C and E are events or states in nature:
- Sense II, in practical science, where C can be produced or prevented by us, in order to produce or prevent E; or
- Sense III, in theoretical science, where:
  1. C is sufficient for E (that is, if C happens, then E happens),
  2. C is necessary for E (that is, if C does not happen, then neither does E), and
  3. C is 'prior' to E, in a sense to be discussed later.

Possible objections to this account of causation are:

The three things are not senses of causation, but rather cases in which the one sense of causation can be used.
There are three kinds of causation, belonging to one genus.
There are indeed three senses of causation, but only one of them is proper, while the others are metaphorical.

The reply is that there are three proper senses of causation (the contradictory of "proper" being "improper," not "metaphorical"; the three senses represent a historical development, from I to II to III.

Chapter XXX, "Causation in History"

Sense I does not presuppose senses II and III, either historically or logically:

The meaning of "cause" in sense I is ancient, being seen in the Latin causa and the Greek αιτία.
Sense I does not require sense II or III; rather, the latter senses will be seen to require the anthropomorphism of the first.

A cause in sense I consists of

an efficient cause, namely a situation that an agent knows or believes himself to be in; and
a final cause, namely the agent's intention regarding his situation.

If A causes B to do act β, this means either

A has put B in a certain situation that B knows or believes himself to be in; or
A has persuaded B to form a certain intention.

Then B performs β freely, but A shares responsibility.

Chapter XXXI, "Causation in Practical Natural Science"

The search for causes in sense II is a practical science in Aristotle's sense, or Baconian science. Practical science is not applied science. The latter is theoretical science that happens to be useful for solving practical problems. In practical science, this problem-solving is essential.

For example, medicine is a practical science. Finding a cause for cancer means finding something that we can do to prevent or cure cancer.

A cause in sense II requires other conditions to achieve its effect. Any of these conditions can be a cause of the same effect, as long as some agent is in control of it. So there is a principle of relativity of causes.

For example, the cause of a road accident is

to the driver involved, his driving;
to the civil engineer, the design of the road;
to the manufacturer, the design of the car.

To the mere spectator, there is no cause.

The causal propositions of practical science are

experimental, because they are about achieving an end by certain means, and you cannot know whether the end can be achieved with those means without trying; and
general, because they are can be applied to cases arising in practice.

Causation in sense II is

anthropocentric, for involving nature as a means for achieving human ends;
anthropomorphic, for treating nature as something to be manipulated as other people are manipulated; this is why the terminology of cause and effect is used, in addition to the terminology of means and ends.

Chapter XXXII, "Causation in Theoretical Natural Science"

Causes in sense II are contingent

in existence, as depending on the will of an agent; and
in operation, as depending on other conditions for production of the effect.

By contrast, causes in sense III are necessary in both ways:

they do not depend on somebody's will;
they depend on no other conditions.

Action at a distance (in time or space) can only be causation in sense II.

In sense III, how does a cause x produce or necessitate its effect y?

Not by logical implication (the rationalist answer), since then causation would not need to be established by experiment, but everybody does think that experiment is needed. (Descartes's science would have replaced the language of causation with the language of implication.)
Not by observed uniformity of conjunction (the empiricist answer), since the observation that x is always conjoined with y is merely evidence that x causes y; it is not what we mean by saying that x causes y.
Not because the proposition "x causes y" is true for all admissible values of the arguments x and y (Bertrand Russell's answer):
- this answer reduces to the rationalist answer (otherwise the necessity of a proposition could never be known, since not all possible arguments could be checked individually);
- this answer gives a new meaning to "necessary", but we want to understand what people already mean by the word.

Causation in sense III means something different from logical implication, and something more than uniformity of conjunction; this something is the kind of compulsion experienced in human affairs.

Causation in sense III is anthropomorphic, in a way derived from Neoplatonic theology. Here God is semi-anthropomorphic, because, for example, his greatness means he does not suffer from human failings like anger, and because, to effect ends in nature, he can create instruments for their realization that have powers like his own (only inferior to his).

To Newton, the free motion of a body is uncaused; a change in its motion is caused by a force, as one human being may force another to act. For Newton then, some motion is by law, some by cause. Modern physics replaces all caused motion with motion according to law.

Chapter XXXIII, "Causation in Kantian Philosophy"

For Kant,

every event has a cause, in sense III;
the cause of an event is a previous event, which means it is a cause in sense II.

(Why Kant rejected the Newtonian doctrine of uncaused motion, Collingwood does not know.) The two parts of Kant's doctrine are not consupponible (because they require two different senses of causation), except under pressure. Kant's doctrine was generally accepted in the Nineteenth Century; therefore science then had an insecure foundation, the insecurity (as regards causation) consisting in two dilemmas:

The anthropocentric dilemma: Is the natural scientist doing
- practical, Baconian, experimental science, seeking causes in sense II, or
- theoretical science, science of nature in itself, where causation is in sense III? The orthodox view was the latter, with a tendency towards the former, later in the century.
The anthropomorphic dilemma: Does the natural scientist
- use analogies drawn from man's conscious life, or
- avoid them? The orthodox view was the latter, emphasizing causes in sense III, but still anthropomorphic.

Collingwood writes:

The war-cries 'Back to Kant' and 'No more metaphysics' were the mottoes of a reactionary and obscurantist anti-metaphysics whose purpose was to prevent these two problems from being faced and solved. Even where those war-cries were not heard the same purpose has been visibly at work.

Physicists worked to escape the confusion by eliminating causation in favor of law, while philosophers, especially 'realists' and 'logical positivists', wished to perpetuate the confusion.

Reference

Collingwood, R. G. (2002). Rex Martin (ed.). An Essay on Metaphysics (Revised Edition). Oxford: Clarendon Press. ISBN 0-19-924141-4 (paperback).