Aquí, el registro de las cartas que existió entre los dos y que muestran la colaboración de Einstein con Los Andes y la visión de Mario Laserna.
Bogotá, 17 September 1953
Institute for Advanced Study
Princeton, New Jersey
Dear Professor Einstein,
Last year, while at Princeton, I asked your permission to write down some of the discussions we had on questions dealing with the philosophical and methodological foundations of Physics. You generously granted your permission and suggested that I send you the manuscript so that you could go over it and make some corrections and observations. I have now completed what I consider to be a dialogue containing most of the problems we discussed. Some points I have put in as systematic a way as possible but without great detail. On the whole I consider these Prolegomena to Physical Science as a general discussion on the non-inductive character of the method of Physics. I feel a great deal more could be said in order to clarify matters; however this is beyond my possibilities at present and I believe it would take me several years of intensive work.
My intention in sending you the manuscript is to find out whether if you approve it, then it could be published as a dialogue between us. To say it briefly, I could not imagine a greater honor than to have your name associated with this paper. Not only do I consider that your opinions on these matters have unequaled value but in this particular case it so happens that whatever I have said has been either a paraphrase of our conversations or inspired by your writings. Therefore, once you have made your observations and corrections, I would beg you to let me know if it can be published with a brief commentary by you or with co-authorship. If you wish to change the form of the paper or do not care to have it published at all associated with your name, please inform me, I also would appreciate any suggestion as to the journal where publication should be done.
A second paper on the question of external reality dealing with a definition of primary and secondary qualities so as to preserve as objective most of the data we consider so in daily life is under way. It is also based on problems and opinions dealt with in conversations with you.
At present I am acting President of Universidad de los Andes. Dr. Zuleta Angel was appointed our Ambassador to Washington three months ago.
It is my great desire to return to Princeton for a few days. Perhaps if some questions arise in relation to the paper, I might discuss them with you on this occasion.
I thank you again for your generous support of Universidad de los Andes and for your never failing encouragement and personal interest in my intellectual preoccupations.
With best personal regards, I am, respectfully yours,
September 22, 1953
Dr. Mario Laserna
Universidad de los Andes
Calle 18-A Carrera 1-E
Apartado Aéreo 4976
Dear Mr. Laserna:
I gather from your dialogue that a considerable difference of opinion exists between us with regard to this problem. Because I am not of the opinion that there exists an essential difference between concepts and methods in the fields of common sense and science.
Every linguistic utterance is wholly confined to the conceptual sphere. Concepts, as far as they have any basis, are judged logically free inventions of the mind (together with propositions connecting them). But those concepts and propositions receive their value and justification exclusively through their only intuitively given connection with perceptions (Erlebnissen). There is no logical way to deduce concepts and propositions from our crude experiences (induction). This is equally true for concepts like red, tree, as for concepts like distance, atom, etc. The difference lies in the fact that scientific concepts and propositions are mostly brought into connection with sense-perceptions in a more indirect and complicated way. Also the use of numbers does not involve a difference in essence between scientific and common sense methods.
Apart from these differences of opinion I must confess that I do not want to appear in this field as a responsible partner. I have not studied epistemology thoroughly enough and I am not sufficiently acquainted with the tremendously extended literature in that field. It is enough if you mention that we had several discussions about these questions.
With kind regards,
December 14, 1954
Dear Professor Einstein:
After a year and a half of being Acting President of Universidad de los Andes, I have been relieved of these administration duties. The new President, doctor Alberto Lleras - Camargo was until recently the Secretary General of the organization of American States and he also was president of Colombia for one year (1945). All this has given him great administrative experience and, in my opinion, gives him a prestige of valuable use for a private and young University like ours. I am certain popular support will increase substantially and that we will be able to undertake many activities that lack of sufficient funds kept in the realm of the impossible.
As a consequence of my not being any longer engaged in the worldly activities of administration I am at present attempting to reassume the more quiet and exciting ones of intelectual speculation. At present the question of the nature of the data of physical science and of the process of constructing theories out of them interests me greatly. On several occasions I had the privilege of discussing with you some of these questions; I consider these discussions of great value to me since not only I was able to hear your opinions on these matters but through them I also clarified my views a great deal. Of course I shall always keep in mind your remark that it is very difficult to convince anyone of the correctness of opinions held on these epistemological questions. I myself now feel that things are vastly more complex and difficult than what I initially thought. My intentions at present are limited to gather opinions as to the nature of those perceptions that permit the construction of a physical theory. Would it be taxing your kindness too much if I asked you some questions? I would greatly prefer to be able to call on you as I did in past occasions; I do not know, however when I may be able to visit Princeton again.
You once expressed the view that it would be possible to constructor all physical theory through the data of any one unique sense field.
Would there exist a set of theoretical conditions that characterize the perceptions belonging to such a sense field? (for instance some type of topological properties). Would the given perceptions be unique only until they reach some apparatus which transforms them or would you say that if man possessed only one sense he could the theoretically construct all physical science?
In a letter (sept. 22, 1953) to me, you asserted concepts as far as they have any basis are, judged logically, free invention of the mind (together with propositions connecting them). In case we suppress given sense fields would our concepts be altered? i.e, though there is no logical connection between percepts and concepts, yet some other relation is conceivable. What is it? I was somewhat puzzled last spring when you expressed your opinion that there exists no basis for distinguishing between primary and secondary qualities. Today this opinion seems to me implies and is implied by the idea that physical science is the systematization of our perceptions so that given a certain pattern we can correlate with it past and future patterns of sense perceptions. But if one were to consider physics as describing and relating entities belonging to a word of definable objective structures, i.e, if one were to consider that certain concepts like shape or distance are deductible from our experience in a direct and logical way, one could not dispense with a distinction between primary and secondary qualities. Do you agree to this distinguishing between two types of physical theories? What would be the objections against the second type of interpretation of physical theory?. Since the concept of object (tree, chair) demands that we consider form as a property of objects it would follow that a physical science of the first type (correlating sense perceptions) would not belong to the world of objects. How are then, our theories verified?
I would deeply appreciate if at least to some of these questions you would give me your points of view. It is clear that if at any moment I plan to publish any material on these topics I will request yours authorization if I am to quote yours views. As of today I have no intentions of becoming an author in this very complicated field.
Please accept my best wishes for your personal well being.
January 8th 1955
Dear Mr Laserna:
thank you for your letter of December 14th.
I still believe that one cannot distinguish, in principle, between primary and secondary qualities; it is basic for all physics that one assumes a real world existing independently from any act of perception. But this we do not know. We take it only as a programme in our scientific endeavours. This programme is, of course, pre-scientific and our ordinary Language is already based on it.
The concepts body - object and shape are not given to us directly by our sense- impressions but are a result of a mental construct.
That this is not so easy to see is only produced by the fact that those steps made by everyone of us in early childhood seem to us logically necessary. But this is not so.
With kind regards,
'Prolegomena to Physical Science
Nota del editor: Después de numerosas conversaciones con el profesor Einstein y después de consultar sus principales escritos sobre la filosofía de la física, Mario Laserna resolvió elaborar el proyecto de diálogo adjunto como una forma de resumir sus intercambios verbales y lo sometió a consideración del eminente científico. Como se menciona en el texto, el profesor Einstein no lo firmó pero autorizó a Mario Laserna mencionar que habían tenido diversas conversaciones sobre estos temas. Este texto indica el tipo de temas que fueron discutidos y Mario Laserna ha afirmado que esta interpretación del pensamiento de Einstein está basada en sus propias afirmaciones verbales y en sus escritos. Sin embargo, debemos respetar integralmente el deseo del profesor Einstein de no aparecer como coautor de este diálogo como el mismo lo señaló en la carta de 1953.
L. The great question which seems to me has not been discussed sufficiently is the difference between experimental science and inductive science. In my opinion, physics is experimental but it is not inductive.
E. This opinion seems to me correct; Physics is not based on induction. It has never been so. Physical concepts are not facts found in experience but are creations of the mind. Their usefulness lies in their capacity to connect facts, observed by the senses, by means of laws.
L. How is it then that all text books on scientific method declare that physics is an inductive science and that the validity of its laws depends on the use of induction?
E. This is because it is assumed that scientific concepts are facts which can be observed i.e. that they are realities. Newton believed that he had found in nature, as existing so that anybody could observe them, things like force, mass and acceleration. With them he explained all known facts of the planetary system, as a first approximation, of course. But Newton did not realize he had not observed force and mass; they were, in reality, not his observations but his inventions. Of course they are not arbitrary inventions; they are created by the mind in order to establish relations between facts; they are afterwards tested as adequate or inadequate by experiments.
- L.From what you say one must necessarily draw the conclusion that the language used in Physical Science does not contain words which deal with objects of the world of sense perception. In other words, Physical science does not contain statements dealing with chairs, cats, colours or other perceived objects.
E. Yes, this is so. And it is with this type of objects that inductive laws are generally established. All dogs bark at the full moon. This is an inductive law. It may be true or false. At any rate it deals with objects. And physical science does not. It is mainly due to the desire of the positivists who wanted to establish science as mainly a classificatory activity, that the pre-eminence of the inductive method was proclaimed.
L. The objects one classifies in physical theory have no individuality i.e. pressure or temperature are not objects like cats or chairs. It seems to me that behind the use of the inductive method there lies hidden some non-physical or non perceptible reality, that is, there is some meta-physical assumption.
E. This is not very clear. Perhaps, you can explain with more precision. First you must define what you mean by inductive law.
L. Well, an inductive law is a general proposition, i.e. which refers to a certain class of individuals, the truth of which has been established by examining a certain number of them in order to verity that they possess a common property or fact.
E. And what is meta-physical in this?
L. It seems to me this method is based on the motion of essence and accident. Take for example the proposition God is yellow.
E. Such a statement is not made in the language of physical science.
L. That is precisely the problem. But since it is an statement which describes reality an ordinary person would think it is a statement found in physical science. What this statement presupposes is that there exists a certain essence called gold and that in some way or another our experience, a) tells us when we have a piece of gold in front of us, b) this identification is possible without using yellowness as a condition for identification, c) we are able after identifying this essence as gold to add that it is also yellow.
E. We can put it thus: to establish the proposition if x is gold then x is yellow means that we must be able to assert x is gold by itself, i.e. without considering the property yellowness. Then if we say that for all x if x is gold x is yellow we have an inductive law. The question is then, how do we establish the proposition x is gold? X is yellow has a physical meaning; but if we must declare something to be gold independently of its measurable, and verifiable properties, such a proposition has no physical meaning whatsoever.
L. But it may have meanings outside of Physical science. At any rate this is another problem; perhaps a very important one, but we would not deal with it now. What seems to me proved by our analysis is that this type of propositions in which in order to assert x is gold we exclude physical qualities, have no place in physical science. And since in order to establish an inductive law, we have to give meaning to these propositions, this type of induction is not used in physical science.
E. On the other hand, if physical qualities like yellowness are part or the definitions of gold , the preposition gold is yellow is analytical and not empirical but we most examine other meanings of induction and inductive laws.
L. For example a proposition like all pennies have Lincons image is this an inductive law? Has the truth or falsity of it been obtained by empirical methods? It seems to me the answer depends on the definition of pennie.
E. Yes, in this case it would be a matter of definition; but a property like having Lincons image is not a property which can be formulated in the language of physical science, that is, the laws that we speak of in physics do not establish that type of relations.
L. Why? is not such a proposition the description of something physical?
E. Yes, but a law of physics is not a description of something but a description of a relation i.e. of the way in which different physical concepts are related to one another. In other words a law of physical science is a function in which the related variables are quantifiable; if one gives to a set of variables a given value, within a physical formula, the other variables take on definite values. Lincons image can not be considered a variable; it can not take any numerical value.
L. The matter is quite clear now. Inductive law do not express relations between well defined entities like in the gas formula PV=MRT. In inductive laws, or at least in a common type of inductive laws like Ripe apples are sweet one does not establish a functional relation between ripe apples and sweetness. What one rather does is to predict that a certain cluster of qualities (those wich define ripe apple), will be followed by the perception of another quality; in this sense inductive laws are the formulation of certain sense expectations. But they do no Express quantifiable relations between concepts.
E. It could also be said that in inductive laws an object is identified by a certain group of qualities, and the laws say that there is another quality which will always appear as co-existing with the initial ones but which is not part of the definition of the object. Such is not the method of physical science. Here we have no initial group of qualities identifying an object. As a matter of fact what we call objects in the ordinary sense, have no rights of citizenship in the World of physical science. Physical concepts, as I already said, are never perceived.
L. Then how are objects, or what corresponds to our experience of objects, identified in physical science? If it is not through their sensible qualities, how do we know an object is the same in two different occassions? Is the concept of identity a concept of physical science?
E. I believe that identity is not a concept of physical science in the sense that it is either used or perceived within our experience of the objects dealt with in physical science. Identity is either a metaphysical concept or one derived from the world of he senses. Of course we could describe a situation contemplated in physical science which would correspond to our ordinary meaning of identity.
L. I suppose one could say that given a certain process described adequately by physical laws if the energy exchange relations of a certain region of space satisfy given conditions, then that entity contained in the region of space observed remains identical.
E. Some of these formulations give me the impression of swallowing without having anything in my mouth. But perhaps this is unavoidable.
L. How do we know that given experiences have put us in contact with parts of the same type of reality? How do we know two different pieces of matter are both pine wood. And I put this question not in the realm of practical daily affairs, but in the rigorous world of scientific knowledge. It is quite clear that sense experience cannot be the complete answer.
E. Sense experience is involved in the process. But the answer to your question of how two atoms of experience may be said to belong to the same reality, in your example, pine wood, is that, scientifically pine wood or iron or gold or oxygen is identified by the laws that it fulfills, i.e. by the relations which it satisfies. This is how the scientist proceeds. He takes a piece of a material A. He invents concepts like weight, color, heat of fusion, and the discovers what relations expressed in a functional way these concepts satisfy when present in A. Then whenever another piece of matter B satisfies those same relations, we say that A and B are two portions of the same substance or physical reality. Therefore gold or oxygen are names for the class of all portions of matter that satisfy the relations that define gold or oxygen. Normally if two pieces of matter are portions of the same substance, the sense counterparts to the physical properties i.e. the color, hardness, specific weight etc. appear the same. Identity of sense perception is, to say it roughly, a first approximation to the physical science counterpart of being two portions of the same physical substance.
L. I am inclined to believe that this method of defining a given substance or entity by the relations it satisfies is not as unfamiliar as its theoretical sophistication and its serious epistemological consequences would make it appear. For instance, in chemistry it seems to me such a method has always been used. A given chemical substance is identified as being, let us say, alcohol of a certain type, if it fulfills certain properties i.e. if it exhibits some reactive qualities. Once the substance has been tested through a process of reactions with other substances, one is able to identify it. It is true that the properties exhibited in a reaction are usually described in the language of the senses; color, precipitation, heat, etc. but one could always reduce these sense manifestations to concepts definable within physical science. This would mean that ultimately chemical processes should be reducible to the language proper to physical science. At present the relations satisfied by chemical reactions are not functional or numerical relations, but relations between concepts belonging to the language of the senses.
E. What is very important is that within a given stage of development of scientific theory, i.e. systematized experience, if we want to develop a theory T and express within it relations t1, t2, ts, we have to identify those concepts or entities that are related by the ti; how do we identify them in order to know whether the ti hold or not? Of course the answer is operationalism; but how do we know that the scales and apparatus used in carrying out such an operation are the same physical objects? The answer is that a less developed theory than T1, call it To, has provided us with relations tio ?¡¡¡¡¡¡¡¡¡¡ such that the objects we use in our apparatus satisfy them. i.e. we say something is a dry piece of pine wood if it satisfies some relations we consider a part of the definition of pine wood. What is then, if we carry on a regression into more elementary, immediately given or experienceable theories, the basic experenceable and verifiable relations we start from? This theoretically prior experience must be something inherent in the field of some sense; it seems to me the answer is that the basic experiences given in a sense field are the familiar propositions of Euclidean Geometry as know through a visual or tactual model.
L. The validity of this last assertion depends on the impossibility of the existence of more than one visual or tactile model of Euclidean Geometry.
E. Let us draw our conclusions.
L. Yes; and I believe we have some important ones to draw.
E. It is quite clear that induction is not the method of proving the validity of the laws that constitute physical science. Neither is induction used for establishing the law, nor once the law has been established can there appear an instance that contradicts it. This is son because on cannot decide that something is an instance to which the law should apply without first identifying that something. And the identification is carried on by imposing as a condition for classifying it under a given name that it fulfills certain laws. In practice what happens is that we classify things by using sense cues, and not by the abstract way of investigating hat functional relations are satisfied.
L. Induction then is a method for establishing laws within the world of sense experience, but not in the world of quantifiable concepts. It may also be said to consist of laws relating not so much to the external world as to our neurological structure. One can make the following assertion: Physical science will not be altered by discoveries in the field of neurology, but the science dealing with sense perceptions and not with abstract quantified concepts, that is, the science of inductive laws, may change with our knowledge in the field of neurology.
E. One should not fall into the mistake of assuming that because sense experience by itself does not constitute scientific knowledge, then sense experience has no place in the formation of scientific theories.
L. Yes; such mistakes must be avoided. But it would be also of great interest to know which part of sense experience is used for forming scientific theories. Obviously we could possess all our physical science If the visible spectrum did not include the wave lengths corresponding to 7000 8500 A?¡¡¡¡. Therefore the experience f red is not indispensable.
E. Undoubtedly the question of what is objective experience is a very old and difficult problem. It is the ancient question of primary and secondary qualities. It still remains unsolved.