Home | Bibliography | Page 2 of 2 | Outline of Page 1

The Philosophy of Science of Arthur Eddington - Second Page

Background: In philosophy we are trying to learn things in logic, ethics and metaphysics, but if our learning is to have a foundation in logic, then it must begin with logic's first principle, namely the distinction between sense and nonsense in language. How is that distinction preserved or made when thinking about philosophical problems? That is the subject of what I have called "Wittgenstein's logic of language". (How quotation marks are used in philosophy and why.)

The physicist Arthur Eddington wrote about the philosophy of science, and in this page I continue trying to understand what he has said about it, about what Eddington means and whether he gives a true account of "what physics is" or does not.

Outline of this page ...

Absolute versus Relative Space

Our life is lived in "absolute space"; the concept 'relative space' or 'space frame' has no application in our life of every day. It is like the concept 'relative motion', except that we do find a few uses for that. Actually we do have an application for 'relative space': Eddington imagines a man in a free-falling elevator (Arthur Eddington, The Nature of the Physical World [1927 Gifford Lectures], Chapter VI, [Ann Arbor: 1958] p. 110). But if we take the elevator out of the office-building and drop it from an airplane, we create a situation familiar to us. A sky-diver releases an object from his hand and the object does not move; the object is at rest relative to the sky-diver's space frame (as is the sky-diver; the earth is in motion). Seen from our space frame (where the earth is at rest), both the diver and the object are in motion.

In what way science is useful to philosophy

Is scientific progress useful to philosophy? Certainly. The realities that are discovered lighten the philosopher's task, [i.e.] imagining possibilities. (Wittgenstein, Last Writings on the Philosophy of Psychology [Note 3])

This is related to imagining language-games in fictitious ethnologies. And Eddington's book is good for this: even if we learn no science, we are introduced to many new pictures, concepts. Like fairy tales, these push back the limits of our imagination. [The importance to philosophy of being able to imagine alternative pictures.]

New Conceptions of the Universe, in History

Whenever we talk about "radical new conceptions of reality", we should remind ourselves of the Greeks: for, it is often said, they already had these "new" conceptions (e.g. Schroedinger's wave theory looks like Plotinus). Look and see, is the only reply: you would need to find a picture that the Greeks were unaware of. According to Etienne Gilson there were many:

There are times when the universe undergoes abrupt changes in the minds of men.

The thirteenth century was one of them. [Europeans] then experienced the sudden revelation of what had been the science and philosophy of Greece ... the eternal and uncreated cosmos of Aristotle, peopled with species immutably fixed under their present appearances, and completely alien to history both in its origin and its duration ...

The sixteenth century was another such time. Tycho Brahe saw with his own eyes a comet going through one of the supposedly solid celestial spheres, thus for the first time in history proving their non-existence and at the same time wrecking the cosmos of the Greeks.

The nineteenth century saw still another one of these revolutions ... Charles Darwin established ... that living species have not always existed with exactly the same characteristics that they possess today. At that very moment the vegetable and the animal kingdoms, such as Aristotle had conceived them, joined his cosmography in the world of lost illusions.

The world of modern science differs from that of the Greeks in this, that it has a history.

The unexpected confluence of nature and history, two orders that in the past had always been separated, not to say opposed, will remain ... the distinctive feature of the late-nineteenth-century view of the world. (The Philosopher and Theology [Note 4])

If we imagine there to be laws of nature (rather than laws of physics), it is suggested that even these have evolved since the so-called Big Bang. What was "the real" for the Greeks was what is permanent, not subject to change. This was a feature, not only of their metaphysics (e.g. Plato's Ideas), but also, according to Gilson, of their "science". The question is: is the contemporary view a question of facts or of outlook? Has the discovery of facts justified new theories -- because these facts are anomalies that the old "theories" cannot account for? And why shouldn't new facts suggest a new point of view -- isn't that what happens?

We only give up an hypothesis for an even higher gain. (Wittgenstein's Philosophical Remarks [1929-1930], tr. Hargreaves and White (1975), p. 284)

Isn't being able to account for all the evidence just such a gain? Ideally, that is how science works. But is that also the practice?

What are the Rules for Reasoning in Science? (Philosophy of Science)

A philosophy of science should give an account of (make clear) the rules for reasoning in science as derived from the actual practices of scientists. Those rules, if there are such rules, would tell us what kind of discipline science is.

Statistical Averages versus Strict Causality - Science's New Model of Explanation

And a philosophy of science should also state what the standards of scientific explanation are: 'A scientific explanation looks like this': e.g.

(1) If A happens, then B will always happen. This is strict causality.

(2) If A happens, then B will happen X% of the time, C will happen Y% of the time, etc. This is probability, statistical averages.

As I understand Eddington, (2) has replaced (1) as the model or goal of scientific explanation. But Drury quotes Claude Bernard:

Claude Bernard - Conjecture versus Truth

In every science we must recognize two classes of phenomena, those whose cause is already defined; next those whose cause is still undefined ... we compile statistics [only in the latter case, i.e.] only when we cannot possibly help it; for in my opinion statistics can never yield scientific truth ...

Statistics can bring to birth only conjectural sciences ... we get a conjecture of greater or less probability about a given case, but never any certainty, never any absolute determinism ... I condemn not trying to get beyond them. (Drury, The Danger of Words (1973) p. 9)

Now, that was the position of Einstein too, I think.

Eddington, however, reports that:

It is a consequence of the advent of the quantum theory that physics is no longer pledged to a scheme of deterministic law. Determinism has dropped out altogether in the latest formulations of theoretical physics and it is at least open to doubt whether it will ever be brought back. (Eddington, Chapter XIV, p. 294-5)

Eddington's confusion about the distinction between science and philosophy

Meanwhile we may note that science thereby withdraws its moral opposition to free-will. Those who maintain a deterministic theory of mental activity must do so ... not with the idea that they are thereby making it more conformable with our experimental knowledge of the laws of inorganic nature. (ibid. Chapter XIV, p. 294-5)

Jacob Bronowski pointed to Heisenberg's "Principle of Indeterminacy" (or uncertainty) as scientific proof that there is no absolute certainty, and therefore that science shows us that we should to be tolerant of other people's ways of life and thinking.

Every discipline has its limits, but rather than recognize those Eddington and Bronowski try to extend science into the subject matter of philosophy. The notion 'strict causality' was taken by scientists from philosophers, not vice versa; in philosophy strict causality is a metaphysical, not a scientific, notion -- i.e. in science strict causality is a method; in philosophy it is a speculation, a picture. The relation of the concept 'determinism' to the concept 'free-will' is a philosophical, not a scientific, question, and the question of whether or not there is free-will is neither. As to "absolute certainty", the question is whether the combination of words 'absolute certainty' (cf. 'absolute grounds', 'absolute knowledge') has any use in our language, which is a question belonging to logic, not science.

The religious reader may well be content that I have not offered him a God revealed by the quantum theory, and therefore liable to be swept away in the next scientific revolution. (ibid. Chapter XV, p. 353)

This is the megalomania of a mad scientist. Wittgenstein said to Drury (ca. 1930):

There is a tendency nowadays for scientists when they reach middle age to become bored with their real work, and launch out into absurd popular semi-philosophical speculations. Eddington is an example of this. (Recollections p. 117; cf. CV p. 42 [MS 125 21r: 1942 § 1])

The Metaphysical Determinism formerly fostered by Scientists

However, Eddington's "we may note that science thereby withdraws its moral opposition to free-will" is not without historical significance, according to Gilson:

Most scientists are unable to discern ... the exact point beyond which demonstrated certitude gives way to the views of imagination.... The illusion is ... natural, for science would never progress if imagination did not anticipate today the scientific truth of tomorrow. The danger appears at the point where the scientist who follows his reasonable anticipations ascribes to the undemonstrable a probability akin to the certitude of the demonstrated. He does not distinguish between the free play of his imagination creating a world of its own and [his imagination] inventing within the boundaries of given reality.

The most remarkable extrapolation of this type, since the universal mathematicism of Descartes [see Gilson's The Unity of Philosophical Experience (1937), Chapter V], was seen among the scientists of the nineteenth century who decided that everything in the world obeyed the laws of a universal mechanicism and determinism. To justify this attitude they began by recognizing only quantitative relations among things ... This was a natural step. When he follows his imagination the scientist is naturally inclined to picture to himself the universe such as it has to be in order to be explainable in terms of his own science.... This time, however, normal as it was, the case was somewhat different. The method of physico-mathematics itself, quite apart from its concrete applications, was being turned into the universal law of nature. In other words, it was then decreed that what could be known of reality was in itself such as it had to be in order to be scientifically intelligible. Since, they said, the universe is not entirely knowable unless it consists exclusively of quantitative relations ruled by mechanical laws, this was obviously what it really was.

The operation was fantastically arbitrary, but it nonetheless took hold of innumerable minds with the force of a religious faith. This is in effect what it was, but it mistook itself for science ...

... what pretended to be science actually was a very questionable metaphysics. (The Philosopher and Theology p. 225-7)

Gilson says that this is the error Aristotle denounced of "conceiving being qua being under the form of one of its particular modes" (ibid. p. 226) -- i.e. of attributing to the whole of reality what is characteristic of but one of its parts. E.g. if part of reality is quantifiable, then the whole of reality must be quantifiable; and, therefore, what is not quantifiable is not real. Eddington does just what the Victorian physicists did; he retains their "quantificationism", simply replacing their "mechanicism and determinism" with pointer-readings and indeterminism. Eddington is speaking as a metaphysician rather than a physicist.

The Relation of Metaphysics to Physics

According to Gilson (if I understand him correctly), metaphysics concerns itself with the "reality that underlies" what physics reveals; however, this reality is not scientifically demonstrable, only metaphysically demonstrable.

All metaphysics become outdated because of their physics: that of Saint Thomas (and of Aristotle himself) because the Aristotelian physics became obsolete, that of Descartes because of the Cartesian mechanics, that of Kant because of the Newtonian physics, and that of the last born, Bergson, who lived just long enough to realize that he could not catch up with the physics of relativity. (ibid. p. 232)

If this is so -- i.e. putting aside the question of whether metaphysics is other than "the free play of ... imagination creating a world of its own" --, this gives further support to Drury's remark:

Considering the vast complexity of the matrix of nature, isn't it certain that there is still much evidence, lying before our eyes and beneath our hands, which we have failed to notice as yet? And may not such evidence in the future transform our idea of nature ...? [There is a universal] tendency to take transitory concepts as final and absolute. [When some thinkers] make the idea of evolution the basis of their philosophy, even of their religion, [they] are making just [this] error. (cf. DW p. 110)

Eddington's notion that science could ever possibly be in "moral opposition to free-will" mistakes physics for metaphysics (in Gilson's sense).

The nineteenth-century determinism Gilson charitably characterized as taking "hold of innumerable minds with the force of a religious faith", Wittgenstein characterized more harshly:

Belief in the causal nexus is superstition. (TLP 5.1361, quoted in DW p. 55)

The Water in my Teapot - The Uniformity of Nature

The so-called principle of the uniformity of nature is an induced principle: our experience leads us to believe in it; though as a principle we apply it to things outside our experience (i.e. future events).

I believe the water in my teapot will boil and not turn to ice -- only because it has done this countless times. The certainty is similar to that about my having hands and about fire burning me (PI § 472; Philosophical Remarks p. 285). If I were desperate to have ice, if e.g. I needed it to save a life, I would not as a last resort place the teapot on the fire in the hope that the water might turn to ice. I would no more do that than I would use fire in the desperate hope that it might have the same effect as ice.

And this has nothing whatever to do with science: naive man did not require philosophers to teach him induction; our belief in certain regularities of nature is not scientific; it predates science, is taken for granted, is maybe instinct (or at the very least it is logically possible for it to exist in the absence of science).

If we are rational we do not base our actions on improbable beliefs such as that water in a teapot set over a fire will freeze. (Is that a definition of 'rational'? If by 'definition' we mean 'any explanation of the use of words in the language (meaning)'.)

[Someone] can fight, hope and even believe[,] without believing scientifically. (CV p. 60 [MS 134 141: 13.4.1947 § 1b])

Or philosophically. The point is that we do not appreciate just how removed from the life of everyday science is; just how specialized parts of it are as a way of thinking: e.g. so-called Laws of Nature have nothing to do with our normal life, and so we should not assume that we know what scientists mean when they utter the expression 'Law of Nature'.

The law of conservation of momentum and energy results from the overlapping of the different aspects in which the "non-emptiness" of space presents itself to our practical experience ... We can measure certain forms of energy with a thermometer, momentum with a ballistic pendulum, stress with a manometer. Commonly we picture these as separate physical entities whose behavior towards each other is controlled by a law. But now the theory is that the three instruments measure different but slightly overlapping aspects of a single physical condition, and a law connecting their measurements is of the same tautological type as a "law" connecting measurements with a metre-rule and a foot-rule.

The law itself will endure as long as two and two make four; but its practical importance depends on our knowing that which obeys it [i.e. to what it applies]. (Eddington, Chapter XI, p. 238-9)

"Aspects of the non-emptiness of space" has no relation to our naive concepts. (Here Eddington speaks as if there were both laws of nature and laws of physics; but he portrays the former as being rather like Plato's Ideas: they're always out of reach ...)

We consider all sorts of ideas and doubts, carry out investigations, in philosophy which have no connection to our life as it is lived. In philosophy it is logically possible that the water in the teapot will turn to ice; in science it is theoretically possible that it will.

Theories are "Transitory Affairs"

Is the work of Charles Darwin to be regarded as likely to be of more permanence than the work of Isaac Newton? Are not Darwin's theories, his way of looking at things, just as likely to be "transitory affairs"?

Hearing the certainty with which certain scientists talk about evolution, you would think that they had been standing there watching the whole thing happen, which of course they were not: they have made deductions from the evidence (and deductions (Drury's "models, pictures, maps") should never be confused with the evidence itself). Newton's Rule II requires us to assume that e.g. the radiocarbon-dating method works when applied to what we imagine to be very long times indeed; but it is only a theory that it does and can never be anything else.

Science does not finally understand things ("We have finally understood how ..."); it comes to an understanding of things. An understanding that is always subject to change.

Scientists lay down their own fossil record -- namely their own work. Perhaps someday we shall find in our dictionaries this definition of the word 'brontosaurus': mythological beast invented in the 19th Century; a plastic toy figure children once played with. (Perhaps ours will be known as the Age of Scientific Credulity.)

It may be claimed that I am making too much of the evolution of science; that much is retained even while much is changed. But I do not see that Einstein's gravity has anything more in common with Newton's gravity than the word 'gravity'; to redefine 'gravity' as a property of space rather than as an attraction between bodies (Eddington, Chapter VI, p. 119) can hardly be called the simple evolution of a concept. We could say that Wittgenstein's work is no more than an evolution of the work of the Greeks -- but that would be misleading, to say the least. 'Everything resembles everything else -- in some way or another' is a rule of grammar.

To posit the uniformity of nature over billions of years is indeed Newtonian. But the distant past is not within "the reach of our experiments"; indeed we have little knowledge over about 2500 years old, and much of that history is doubtful. I mean human testimony about natural events: our belief in the uniformity of nature is based on human experience; and we would not have the theories we do have if we did not have that belief -- what would be the point? We have some evidence, on the basis of which we speculate about the distant past: but we cannot by definition have knowledge of this speculated about past; because all knowledge ultimately rests on human experience, human testimony.

The Limit of Scientific Explanation

Why do you demand explanations? If they are given you, you will once more be facing a terminus. They cannot get you any further than you are at present. (Z § 315)

At some point you are going to have to "just accept" something. And so why not just accept the present fact rather than try to explain it? That is the point of what Drury calls "clarification by full-stop".

Scientific explanations lead us on indefinitely from one inexplicable to another ... and we never find a real resting place. Philosophical clarity puts a full-stop to our enquiry and restlessness by showing that our quest is in one sense mistaken. (DW p. xii)

That is the first fact to accept, that explanations and speculation have no logically necessary terminus. What is important is to see the stopping point that is most useful to the understanding. As for example, when Wittgenstein wanted to understand what gives meaning to our language, he saw that if language-meaning is to be objective, it must be given its meaning by what is public to all users of the language. And so it is the facts in plain view that logic must describe. They are the "terminus".

Scientific explanation versus Speculation

The "theory of abstraction" wants to explain language meaning but it is metaphysical speculation, not science. And that is important. Why? Verification (neither necessarily naive nor simple) is the difference between science and metaphysics. Speculation is not verifiable: it is, instead, explanation floating free of that tether. For example, if we explain the red shift with the hypothesis of the Doppler effect, then we are not justified to base another hypothesis on that hypothesis and yet another on that -- or at least we are not justified to say that these "hypotheses" are science rather than speculation: they are not an understanding of reality, but a picture show of possibilities logical, but unreal.

Goethe thought it a mistake to replace the facts in plain view with a theory about them. Untethered speculation is not even a scientific theory, but it is fantasy.

Everyday Language versus Scientific Jargon - 'flat space' and 'curved space'

Enthusiasts, popularizers of science, advocates of scientism, whichever they should be called, utter the words 'flat space' and 'curved space' as if the meaning of those expressions required no explanation, as if this were a question of fact rather than of grammar. And people say that they 'cannot understand' when they ought to say 'Nonsense!' In our common language 'space has physical properties' is nonsense, which is why 'flat space' and 'curved space' are meaningless combinations of words there, except as they are used in the ways we are all familiar with (e.g. "Look up there; do you see the curved space in the rock? That's where I saw the bird's nest last week.")

The Rationalism of Descartes, and of Einstein

Descartes: 'There can be no extension which is extension of nothing'; i.e. 'reality is a plenum'. This Descartes "deduced from his clear and distinct ideas". But Pascal, in the spirit of Newton's "Experiments are the true masters in physics", demonstrated through his experiments with vacuums that there is indeed "extension which is extension of nothing", or, in other words that there is "void space" -- i.e. empty space, space wherein there is nothing perceptible -- i.e. wherein there is nothing.

Einstein returned to Descartes view: he posited an imperceptible "ether" to fill Pascal's void space. Einstein redefined the word 'space', saying that 'space is endowed with physical properties' -- i.e. that the combination of words 'empty space' is without meaning. [Note 5]

"Empty Space is filled by the Ether" - There is no void space in Physics

We must rid our minds of the idea that the word space in science has anything to do with void ... the physicist does not conceive of space as void. Where it is empty of all else there is still the ether. (Eddington, Chapter VI, p. 137)

[James Jean's account of the status of the ether is different.]

Both Descartes and Einstein's 'There is no empty space' were rules of grammar, rules not in agreement with our language of every day (Pascal used the language of every day to talk about space). Descartes' rule was idle; i.e. he did not stipulate how it was to be applied, verified e.g. (PG i § 82, p. 127). However, Einstein's new use of the word 'space' was justified by a theory (CV p. 44 [MS 127 36v: 27.2.1944 § 1], quoted in Philosophy of Mind): e.g. light requires a medium; there can be no light in empty space. Nor can there be measuring-rods or clocks (Einstein op. cit.), so that there is no way to quantify empty space; according to Einstein, this makes 'empty space' meaningless.

In Einstein's language 'space can be curved' (by definition, since 'space has physical properties'). In the language of every day, there can indeed be a curved space (e.g. the space defined by the rim of a coffee cup or a Roman arch); but in the language of every day 'Space is curved' is meaningless, since 'space has no shape, except the shape of the objects that limit it' is a rule of grammar -- i.e. 'space' is not a name-of-object (even though the use of that word is learned ostensively).

In a way those remarks make something clearer, but they are hardly an explanation of the meaning of the expression 'curved space' as it is used by physicists.

"Space is Finite but Unbounded"

If the universe is a globe, then how do we know that the stars we see in the northern sky aren't the same stars we see in the southern sky; -- "in spherical space" beyond "brings us back towards the earth from the opposite direction" (Eddington, Chapter VIII, p. 167). A line on the Euclidean plane extends indefinitely in opposite directions; but a line on the surface of a globe joins itself to form a circle. [Note 6]

The modern view is that space is ... finite though unbounded [as is the surface of a globe]. In such a space light which has travelled an appreciable part of the way "round the world" is slowed down in its vibrations, with the result that all spectral lines are displaced towards the red. Ordinarily we interpret such a red displacement as signifying receding velocity in the line of sight.

The phenomenon is intelligible if what has really been observed is the slowing down of vibrations consequent on the light from these objects having travelled part of the way round the world. (ibid. p. 166-7)

This is an apparent alternative to the Doppler effect explanation of the red shift.

'Curved space' versus 'non-Euclidean space'

Eddington says that the expressions 'curved space' and 'non-Euclidean space' are "used practically synonymously; but they suggest rather different points of view". 'Curved space' is used to refer to the two-dimensional picture, the surface of a globe e.g., that gives us an idea of what we are talking about; whereas 'non-Euclidean' refers to "the dropping of all relations to an external (and imaginary) scaffolding [i.e. the surface of the globe] and holding on to those relations which exist within the space itself" (Eddington, Chapter VII, p. 157). The latter says nothing to me except: don't try to picture non-Euclidean space. I wrote above that if 'space' is defined to have physical properties, then a sense can be given to 'curved space'; but that way of applying the word may not be what physicists mean at all.

In general, there is nothing that explains the meaning of words as well as a picture ... (LC p. 63)

If we cannot (whether by definition or by lack of imagination; it doesn't matter which in this context) picture three-dimensional non-Euclidean space, then we lose something important to our ideal of scientific explanation. A reality presented to us in nothing more than equations relating pointer-readings is not a reality that most of us can understand. And I think, perhaps naively, that it is not the understanding of reality that physicists really want. So that, I suspect that when Eddington say that the rainbow we see is an "illusion", he is expressing his desire to see the "really real behind" the spectroscope readings.

If picturing is our normal way of understanding, it is also our normal way of misunderstanding.

"Looking at objects without any preconceived idea"

Apropos of "seeing is concept-laden" (percepts without concepts are blind) or "we see in our language".

There is a way of looking at electrical machines ... which sees these objects as arrangements of copper ... in space, without any preliminary understanding ... It is quite analogous to looking at a mathematical proposition as an ornament. -- It is of course an absolutely strict and correct conception; and the characteristic and difficult thing about it is that it looks at an object without any preconceived idea (as it were from a Martian point of view) ... (Z § 711)

But if we "look at an object without any preconceived idea", e.g. at some electrical machine -- do we see anything? We see nothing except that general class of objects. (It is the difference between seeing trees -- or seeing oaks, maples, birches, etc.) Even seeing "arrangements of copper" is already a preconception. We say we see "something" -- but we see nothing coherent, and so what sense does it make to speak of 'seeing' at all? (Percepts are seen without concepts -- i.e. without being conceived as this or that -- as if the windows of the soul were open. 'Is that possible?' is a grammatical-definitional question, i.e. of describing a logical possibility.)

What the philosopher would like is to see the world through the eyes of a newborn baby, before conceptualization, in the time "when meadow, grove and stream, the earth and every common sight" does seem ... The question is -- does a newborn, a creature without concepts, see anything at all (beyond perhaps contours and colors). The presumption [presumptions can be refuted] of Wittgenstein's remark is that Martians would have concepts, just not the same concepts as ours (a concept is a way of arranging things -- of seeing arrangements rather than unintelligible scenes); but cf. a visitor who, unlike us, has not read the guide book -- what does he see when he sees the old city? The same as us (he is not blind), but not the same as us (he is concept-blind).

There is a purpose to looking at things "as it were from a Martian point of view". By trying to look at a thing in a new way -- e.g. at an electrical machine as a functionless piece of art, or by imagining a people who use mathematical equations (or musical notation) only to ornament the walls of their rooms (WLFM p. 39), we become aware of the way we habitually look at the thing, of what our "preconceived idea" of it is. What we want from philosophy is a fresh view of things.

North American robins take off horizontally. Mocking-birds almost leap into the air. Is seeing a robin seeing that it will take off horizontally? Is it knowledge -- i.e. is the ability to predict events knowledge? Seeing a class of objects {robins}and the consistency of nature {take-off patterns}, induction leads to generalization: North American robins take off horizontally.

The concepts 'seeing' and 'knowing' are related through verification: 'How can you know such and such?' -- 'By observing it.' And 'knowing' and 'expecting' are related concepts: 'Why do you expect it to happen?' -- 'I have seen it happen many times before.' Is seeing, therefore, as N.R. Hanson says, knowledge? Does talking that way make anything clearer? And is Hanson stating a fact or making a rule of grammar? (Even that isn't clear.)

Eddington's Metaphysics - "Secondary qualities are not real"

Eddington takes it for granted that color forms no part of our concept 'substance': "color is no part of the essential nature of substance." (Eddington, Chapter XIII, p. 273) If we were asked to define the word 'substance' we might reply: that is a white powder given to children in chemistry classes with the order that it be heated in a crucible (and beyond that I don't remember). That is, the word 'substance' in every day language only occurs in the expression 'an unknown substance', if it occurs at all. That is, Eddington is using a word from Aristotle's Metaphysics.

... the advances in natural science have been due to a wise and deliberate selection of certain aspects of the total given whole, and the ignoring of others. The division of qualities into primary and secondary was a great discovery in methodology, not a metaphysical discovery. (DW p. 78)

Suppose that instead of the word 'substance' Eddington had used the word 'object'. Then the reply would be that we don't use the word the way Eddington suggests: it makes no sense to ask what is essential to objects as such. We do not define the word that way, but rather: ostensively; objects are pointed to, and that is as far as our definition of the word 'object' goes. We learn to use color-words when we learn to use object-words (as we learn to use size-words and location-words, weight-words, and so on).

"Physics cannot explain our mental conceptions" - Why do we perceive an illusory table rather than the "arrangement of atoms that is really there"?

No physical theory is expected to explain why there is a particular kind of image in our minds associated with light, nor why a conception of substance has arisen in our minds in connection with those parts of the world containing mass. (Eddington, Chapter VII, p. 156)

According to Eddington, physics cannot discover the facts that can explain concept formation. This 'cannot' is not theoretical; it is metaphysical -- i.e. it has its source in Eddington's confused "use" of the words 'mind' and 'consciousness'. The facts cannot be discovered because the "conceptions" are "in our minds".

Light waves are propagated from the table to the eye; chemical changes occur in the retina; propagation of some kind occurs in the optic nerves; atomic changes follow in the brain. Just where the final leap into consciousness occurs is not clear. We do not know the last stage of the message in the physical world before it became a sensation in consciousness. (ibid. Chapter XII, p. 267-8)

[Color is] a mental image, which is in our minds and not in the external world. (ibid. p. 254)

Eddington asks why we perceive a table (as in table and chairs) rather than an arrangement of atoms -- why don't we perceive what is "really" there (i.e. the atoms)? The naive table, as opposed to the "scientific table", is a "mental illusion". Why do we perceive colors (an illusion) rather than electromagnetic waves (the really real)? What is wrong with our minds?

Whether or not a scientific question can be derived from this confused form of expression, I don't know. But if there is a question here that cannot be answered, that impossibility has nothing to do with "a leap into consciousness" [Drury's criticism of that notion] or "in our mind".

Is it that we have a better understanding of the mind than Eddington had? Logic does not say anything about the mind, only about our usage of the word 'mind'. Eddington says "no physical theory" -- Why? is there some other kind?

Ever since Einstein's theory showed the importance of securing that the physical qualities which we talk about are actually connected to our experience, we have been on our guard to some extent against meaningless terms. Thus distance is defined by certain operations of measurement ... (ibid. Chapter X, p. 222)

But Eddington is deceived by words whose "connection with our experience" he thinks he knows perfectly well.

Eddington's Metaphysical Idealism

Note.--Better than the word 'idealism' would be 'idea-ism'. The "new way of ideas" began with the proposition that "the direct object of perception is an idea in the mind". Idealism removed the word 'direct': the object of perception is an idea in the mind -- and that is all (There is no as it were indirect object -- i.e. "physical object" -- "behind" the "perception").

... the mental activity of the part of the world constituting ourselves occasions no surprise; it is known to us by direct self-knowledge, and we do not explain it away as something other than we know it to be -- or, rather, it knows itself to be ... Our bodies are more mysterious than our minds ... (ibid. Chapter XIII, p. 276-8)

... no one can deny that mind is the first and most direct thing in our experience, and all else is remote inference ... (ibid. p. 281)

Eddington leaps into Idealism, but being a scientist, rather than a philosopher, he does not want to accept all the consequences of that leap. He asks what we "should regard as a complete description of the scientific world".

The description should include nothing that is unobservable but a great deal that is not observed. Virtually we postulate an infinite army of watchers and measurers. From moment to moment they survey everything that can be measured by methods which we ourselves might conceivably employ. Everything they measure goes down as part of the complete description of the scientific world.

By employing the known physical laws expressing the uniformities of Nature we can to a large extent dispense with this army of watchers. We can afford to let the moon out of sight for an hour or two and deduce where it has been in the meantime. But when I assert that the moon (which I last saw in the west an hour ago) is now setting, I assert this not as my deduction but as a true fact of the scientific world. I am still postulating the imaginary watcher; I do not consult him, but I retain him to corroborate my statement if it is challenged. (ibid. Chapter X, p. 226-7)

In writing about the philosophy of science, I am always afraid of behaving like Wittgenstein's savages (PI § 194). But in straying into philosophy, the Professor of Astronomy does just that.

There is a doctrine well know to philosophers that the moon ceases to exist when no one is looking at it. (ibid. p. 226)

Berkeley did not play Hide and Seek

Eddington imagines that Bishop Berkeley played hide and seek with objects: we see the moon; we turn away (the moon "goes out of existence"); then we quickly look over our shoulder to see if the we have caught the moon out; -- perhaps it has not "come back into existence" fast enough!

Hide and seek has nothing to do with Idealism. If "mind is the first and most direct thing in our experience, and all else is remote inference" (Eddington), then my direct object of perception is the "ideas" in my mind.

John Locke's "ideas"

Whatsoever the mind perceives [with-]in itself, or is the immediate object of perception, thought or understanding, that I call idea. (John Locke, An Essay Concerning Human Understanding [1690] ii, 8, 8)

I presume it will be easily granted me, that there are such ideas in men's minds: everyone is conscious of them in himself ... (Introduction, 8)

But then what do I know other than my ideas? Nothing. Why do I say that the moon exists? Because I perceive "it" -- i.e. have an idea of it. What the moon is in itself -- i.e. apart from my idea of it -- if anything, I have no way of knowing. Indeed, I have no reason to believe that it is anything more than an idea that I perceive in my consciousness. (But it is involuntary? So are the gryphons in my nightmares.) If I "look away" from the moon, I do not perceive "it" -- I have no idea of "it" and so no reason to assert that "it" exists. To exist is to be the content of an idea (which is what a "perception" is: the content of an idea).

That is what "to be is to be perceived" means: it is not a statement about objects, things in themselves -- because I know nothing of them; it is a statement about me (as are all other statements), a definition of 'to exist'. This is where Descartes' program in philosophy leads: if I begin in my mind (with my ideas), I can never get out of my mind (my ideas). The consequence of Idealism is solipsism -- mute solipsism, since our psychological language is meaningless if I am the only one who speaks it; and all language is psychological in Idealism.

Beginning as he does, Eddington has no defense against this criticism. Eddington's "watchers" -- what are these but so many "other minds". But that is nonsense: how can Eddington have an "idea" of a consciousness that is not his own? Eddington says he knows his consciousness as a thing in itself, but then that is the only thing he knows in itself: to exist is to be an object of Eddington's consciousness. He has no right to say anything more than that; unless he wishes to argue that his "army of watchers" is the God of Descartes.

Samuel Johnson's "refutation of Berkeley" by kicking an unyielding stone (Eddington, XV, p. 326-7) is not to the point -- i.e. it is not a counter-argument, not a disproof of the thesis Berkeley argues for -- if Johnson is seen as trying to counter one thesis with another. But regarded as a reminder of how we normally use our language, it is correct criticism. And similar to this is the story about Diogenes the Cynic, that "when someone declared that there is no such thing as motion, he got up and walked about" (Diog. L. vi, 39), i.e. that is an ostensive definition of what we normally mean by the word 'motion' (To say that there is no motion is in effect to say that the words 'mobile' and 'immobile' are without meaning, because those words require their antitheses to give them meaning).

"I make no hypotheses"

In private Isaac Newton speculated voluminously about the Book of Revelation. -- But he did not publish his speculations as part of his Principia. He did not publish his speculations at all. His "I make no hypotheses" belonged to his project in philosophy, but I think that not speculating (at least in public) was also an ethical demand Newton made of himself. And so the scientist who uses his authority as a scientist to publish his metaphysical speculations -- i.e. speculations about the ultimate but hidden-from-man nature and causes of things -- is not a disciple of Newton.

(But does it do any real harm? Or is there even a benefit to it? "... letting one's thoughts develop by expressing them." There is a lot to learn from criticizing bad philosophizing (one's own and other's) -- a lot one won't learn from silence or keeping silent.)

Examining the Extraneous ("Know thyself")

Now, what reason have I to speculate above about the red shift? I have none; I am only allowing myself to be charmed by what are for me no more than idle pictures, because I don't know on the evidence they are based and so I cannot treat them as scientific hypotheses; I do nothing more than play with them conceptually.

And I think, either become a serious student of astrophysics or leave astronomy alone ... But there is for me something far deeper here. "Know thyself" -- what does this precept mean for me as an individual? I think that any interest I myself take in natural science or pure mathematics (as e.g. axiomatic geometry), other than what is needed in order to understand the Philosophy of Science and the Philosophy of Maths, is no more than running away from "Know thyself" -- i.e. it is a flight from life ("we are discussing no small matter, but how to live") into oblivion (as with wine or sleep). The natural sciences, mathematics, even history, and every other concern except the understanding and ethical conduct of my life are, for someone of my nature, oblivion seeking. Wittgenstein to a student: "Take life seriously!"

Plato, Phaedrus 229e-230a

"The North Wind fell in love with a girl and carried her off to his realm in the north" becomes "A strong wind gust blew the girl onto the rocks where she died and the river carried off her body" (Plato, Phaedrus 229c-d). Thus, seemingly, the myth is made plausible, the supernatural event made natural, the myth explained.

(Another example, from Herodotus' History 2.54-57: Zeus' oracle at Dodona in Epirus. And see as well Ionian skepticism and the birth of history. But it may happen that naturalized myths are far shallower than their religious understanding.)

If our skeptic, with his somewhat crude science, means to reduce every one of [the myths] to the standard of probability, he'll need a deal of time for it. I myself have certainly no time for the business, and I'll tell you why, my friend [Phaedrus]. I can't as yet "know myself", as the inscription at Delphi enjoins, and so long as that ignorance remains it seems to me ridiculous to inquire into extraneous matters. Consequently I don't bother about such things, but accept the current beliefs about them, and direct my inquiries, as I have just said, rather to myself, to discover whether I really am a more complex creature and more puffed up with pride than Typhon, or a simpler, gentler being whom heaven has blessed with a quiet, un-Typhonic nature. (Phaedrus 229e-230a, tr. Hackforth)

Now I am quite acknowledge that these allegories are very nice, but he is not to be envied who has to invent them; much labor and ingenuity will be required of him; and when he has once begun he must go on And if he is must go on and rehabilitate Hippocentaurs and chimeras dire, Gorgons and winged steeds flow apace, and numberless other inconceivable and portentous natures. And if he is skeptical about them, and would fain reduce them one after another to the rules of probability, this sort of crude philosophy will take up a great of time. Now I have no leisure for such enquires; shall I tell you why? I must first know myself, as the Delphic inscription says; to be curious about the which is not my own concern, while I am still in ignorance of my own self, would be ridiculous. (Phaedrus 229d-230a, tr. Jowett)

That was Socrates's view of how a "lover of learning" or "lover of wisdom" (ibid. 230d, 278d), that is, a philosopher, lives. But Epictetus says that even among those men who do try to obey the Delphic precept "Know thyself" -- by seeking to discover both what the specific excellence that is proper to man is and what one's own limits (both highest and lowest) as an individual are, in order to know how one should live one's life -- most seem more concerned with the argument than with doing the deeds that follows from it.

Ethics exists for the act of living

The first and most necessary department of philosophy deals with the application of principles; for instance, "not to lie". The second deals with demonstrations; for instance, "How comes it that one ought not to lie?" The third is concerned with establishing and analyzing these processes; for instance, "How comes it that this is a demonstration, what is consequence, what is contradiction, what is true, what is false?"

It follows then that the third department is necessary because of the second, and the second because of the first. The first is the most necessary part, and that in which we must rest. But we reverse the order: we occupy ourselves with the third, and make that our whole concern, and the first we completely neglect. Wherefore we lie, but we are ready enough with the demonstration that lying is wrong. (Manual, tr. Matheson)

Closing Remarks

Again and again comes the reminder, here in Wittgenstein's words, that expresses the entire ethics of philosophy.

The difficulty in philosophy is to say no more than we know. (BB p. 45) [Note 7]

That comes, as do Augustine's words -- "He only errs who thinks he knows what he does not know" -- from the same Socrates who made ethics a subject of philosophy. Now, what do I know about "the Doppler effect explanation of the red shift"? Astrophysically, nothing.

Note 3: Ludwig Wittgenstein, Last Writings on the Philosophy of Psychology [1948-9] i, § 807; tr. Luckhardt and Aue. Oxford: 1982. [BACK]

Note 4: Etienne Gilson, The Philosopher and Theology, tr. Cecile Gilson (New York: 1962), p. 128-9. [BACK]

Note 5: Albert Einstein, "Ether and Relativity" in Sidelights on Relativity; tr. Jeffery and Perrett (orig. New York: 1922; repr. New York: 1983), p. 23. [BACK]

Note 6: See Einstein's "Geometry and Experience" in Sidelights on Relativity, page 47-8. [BACK]

Note 7: The words that should come easily to the educated man are often the hardest for the uneducated man to say, namely the words 'I don't know'.

Socrates and the Greek investigation of pride (hubris, pride as a disfiguring vanity) was absorbed by Christian theologians with the command "Know thyself"; in their language 'self-knowledge' = 'humility' (as well as 'meekness'). This is found in The Imitation of Christ's command "Confess thine ignorance". Because when the questions are at the heart of life and understanding, the words 'I don't know' are not easy for a proud man to say. (Cf. CV p. 26.)

This is knowledge. To know of a thing that you know, that you know it. And to know of a thing that you do not know, that you do not know it. (The Analects of Confucius, Book ii, Chapter xvii) [But that is Confucius read through the eyes of Greek philosophy: the Chinese Confucius was exercising his wit.]

The latter -- to know that you don't know -- is the far more important kind of knowledge for philosophy. [BACK]

Site copyright © September 1998. Send Internet mail to Robert [Wesley] Angelo. Last revised: 1 April 2006 : 2006-04-01

The URL of this Web page:

Back to top of page

Wittgenstein's Logic of Language - Introduction and Table of Contents | Bibliography | Site Search | Site Map