Tadeusz Kotarbinski

First published in Mysl Wspolczesna, 1949, No. 10(41). Translated from Polish by Olgierd Wojyasiewicz in Tadeusz Kotarbinski, Gnosiology: The Scientific Approach to the Theory of Knowledge edited by G. Bidwell and C. Pinder (Pergamon Press, 1966). Transcribed into hypertext by Andrew Chrucky, March 22, 2003.

The responsibilities of the secondary school teacher undoubtedly include the striving to make the pupils understand meanings of words as clearly and as distinctly as possible. This task acquires particular importance in the teaching of those disciplines which in library catalogues are called philosophical. This is so because the principal shortcomings of those disciplines, which also account for protracted controversies -- for example, in epistemology, ontology, general theory of value, etc., do not consist in defective observation or experimentation, or in using wrong forms of inference, but are mainly reducible to the habit of thinking, and correspondingly speaking vaguely. Hence, whether he so desires or not, the teacher must build a system of verbal explanations and, as it were, compile a dictionary of those terms which sow confusion. And since every subject taught leads to some philosophical issues, every teacher must try to contribute to such a philosophical dictionary, and the professional teacher of philosophical subjects must help him in that, working out the dictionary not only for his own use, but also for use by teachers of other disciplines. For instance, the controversy is revived from time to time as to whether mathematics is an empirical or a purely deductive science. I have no intention of solving that problem here. My point is only that the controversy would not be chronic if the participants would distinctly realize the ambiguity of the term "empirical". In the genetic sense, such a statement is empirical which can be understood only if the person concerned has ever previously observed something, or if he has observed at least one of the objects denoted by one of the terms involved in that statement. In the methodological sense, only such a statement is empirical which for its founding requires at least one observation statement as a premise.

We are thus struggling with confusion due to unnoticed ambiguity, which becomes particularly vicious in the case of similar meanings example above n other cases, trouble arises from the vagueness of a term about which we do not know whether it covers by its extension a given limiting object. Very often there are differences in the appraisal of the scientific character of a given theorem or its foundation; for instance -- are observations on animal behaviour, made in the natural environment without laboratory control, scientific if they are fairly(?) careful and made by an expert (?), or are they not scientific even if both these conditions are satisfied? In other cases, we correctly apply a given term to appropriate objects, but we cannot specify the properties which combine to form its meaning and are helpless in the face of obscure theories in which that term plays an important role. By way of example we may mention here the term "time" in the general theory of events or the term ''ttuth" in the foundations of epistemology. In still other cases, we witness delusions concerning the existence of alleged entities, delusions which are largely explained by the presence in our language of nouns and adjectives of various kinds -- for example, general and abstract, such as "equality", "law", etc.

The best method of eradicating such defects in thinking, which originate from wrong suggestions of language, would be to avoid all such stigmatized words. But such a principle, not qualified by additional explanations and limitations, would result in a most detrimental sterility of research. We must know how to render innocuous the various words, while preserving the wealth of valuable content which they carry. In some cases, we can save the full content, essential in given investigations, by discarding an unnecessary auxiliary problem which requires the use of a hopelessly vague term. Thus, in order to group all the elements of a given set into two classes according to size, it is not necessary to consider which are "large" and which are "small" (these being extremely vague terms), but it suffices to see which in a given pair is larger than the other, and to group in one class all those which are larger than those grouped in the other class. We use here the term "greater than", which is incomparably less vague than are the terms "great" and "small". An example of a similar procedure can be found in making the analysis of a deductive system independent of the terms "truth", "true", etc. These are not, we admit, classical examples of vagueness, but they are controversial enough to make their elimination, whenever this can be done without losing the sense of a given inquiry, a justified endeavour. In the theory of deductive systems this is achieved by interpreting axioms as the primitive statements of a given system, and interpreting theorems as theses which are deducible from axioms in conformity with rules of inference, neither the interpretation of the primitive character of the axioms nor the interpretation of the rules having any reference to "truth". Sometimes in the case of a confusing word, it suffices to accept a thesis or theses containing that word, and the controversy is resolved. For instance, when we witness a controversy as to whether man is a machine, it suffices, in order to take a stand in the controversy, to accept the statement that machines do not move of their own will and do not perform actions. But in most cases we have to find definitional equivalents of given terms. We then proceed either synthetically, by imparting to a given word some arbitrary meaning (as was done by logicians with the sentential connective "if... then...", whose function was radically modified as compared with the colloquial usage), or we proceed analytically, by trying to grasp the meaning of a given word in a definition which specifies those properties which combine to form its intuitive meaning. The controversy about the existence of universals revives from time to time like a many-headed hydra. The philosopher who has to take sides in that controversy will first define the term universale by stating, for instance, that a universale is any such object, associated with a given general term, as has only those properties which are common to all the designata of that term -- that is, common to all the individuals connoted by that general term. When this definition is adopted, the thesis about the existence of universals can easily be reduced ad absurdum if it is assumed that:

  1. "A exists" means the same as "some object is A";
  2. every object has either a certain property or its negation;
  3. every designatum of a general term has a specific property.

This concludes the lengthy introduction, the task of which was to persuade the reader that it would be well to think of some general recommendations outlining the methods of working out all the explanations to be used in a philosophical dictionary and serving to eliminate all that Francis Bacon called idola fori.

The reader's attention will now be drawn to only one of such recommendations, to which the present author attaches special importance -- namely, to the principle of reism. What is the requirement formulated by that doctrine? This only -- that all ultimate formulations, and hence also all ultimate explanations of words, should include no nouns and adjectives other than concrete nouns and concrete adjectives. The point is not, of course, that such ultimate explanations should consist of concrete nouns and/or concrete adjectives alone, to the exclusion of connective negation particles, copulas, etc. The essential point is that these ultimate explanations should include no nouns and adjectives other than concrete ones. Instead of "noun or adjective" let us say briefly "term". The principle of reism can then be formulated as fol1ows: let us see to it that we know how to reduce every statement to a form which contains no other terms than concrete ones.

The question now arises which terms are to be accepted as concrete. These are terms of three kinds. The first kind consists of singular names of persons or things, terms which we use as grammatical subjects in true singular sentences about individual persons or things (of the type "A is B") -- that is, proper names such as "Plato", "Rome", etc. The second kind includes general names of persons or things -- that is, terms which we use as grammatical subjects in true general sentences (of the type "every A is B") about persons or things, such terms as "man", "city", etc. As for the terms of the second kind there is no doubt that they can be meaningfully used as subjective complements both in general and in singular sentences. Some people doubt, on the other hand, whether the terms of the first can be used meaningfully as subjective complements. Our opinion is that these two kinds must be treated on an equal footing in that respect. But this controversy is not essential for the principal issue here under consideration. What is important is the discussion of the third kind of terms -- that is, concrete empty terms. These are terms which cannot be subjects of any true singular or general sentence about persons or things, but are by definition reducible to certain combinations of singular or general terms which are names of persons or things; these combinations are such that if some other singular or general names of persons or things are combined in this way, then the whole becomes a singular or general name of persons or things. This clumsy formula becomes simpler if we introduce abbreviations: "denoting term" for "singular or general name of persons or things", and "elementary sentence" for "true singular or general sentence about persons or things". This yields the following formulation: an empty term is a term which cannot be the subject of an elementary sentence, but is by definition reducible to such a combination of denoting terms that other terms, combined in this way, yield a whole which is a denotins term. Let the word "chimera" serve as an example. It is defines in ancient mythology as lion's head with goat's body amd serpent's tail. No such monster has ever existed, and hence it is an empty term. But if its denoting elements (that is, "lion's head", "goat's body", "serpent's tail") are replaced by other denoting terms, "head", "thorax", "trunk", we obtain the whole "head with thorax and trunk", which is a term and denotes any (mature) insect. A similar operation cannot be performed on those terms which are not empty concrete terms, although they also do not denote persons or things. They are such words as "smoothness", "relationship", "tune", "shift", and, generally, what are called names of properties, relations, contents, events, etc. All these nouns and adjectives which have the appearances of concrete terms but are not concrete terms, we take the liberty of calling apparent names or onomatoids. Hence the world of terms is divided into concrete terms (or just terms) and apparent terms (or onomatoids), and the principle of reism requires elimination from all ultiInate formulations of all onomatoids, so that only concrete terms -- singular, general and empty -- are left.

The reader will certainly ask for the foundation of this principle. First of all, we shall refer to the teacher's experience. Is this not the way to explain words in a natural manner, natural at least from the didactic point of view? Should we wish to explain to a child what the word "similarity" means, should we not show him in turn several pairs of objects which look alike, and say: "Look, here are two sparrows: this one is grey and that one is grey, this one is hopping and that one is hopping, this one has a short beak and that one has a short beak. These two birds are similar one to the other. And here are two windows: both are rectangular and both have rectangular panes, separated by thin pieces of wood. These two things are similar one to the other. Do you understand now what is similarity?" Or suppose that in the class we encounter in the text the word "recovery", which the pupils do not understand. We shall tell them: "Whenever a person has been ill, and later was better, and now is still better, we speak of recovery".

Suppose then that the principle of concretism, understood as a guiding line for teachers, has been founded sufficiently by reference to the psychological naturality of such a guiding line. But the reader will ask in turn -- How can that psychological naturality be explained? Reference, quite correct, to the course of development of language in the life of both individuals and ethnic groups does not provide an exhaustive answer. It is a fact that names of persons and things appear earlier in a language than do other nouns, but this fact in turn calls for an explanation. Here the reist will risk a thesis which is no longer founded by didactic considerations, but is of an ontological nature. That thesis states that every object is a thing or a person, and that this justifies everything which we have tried to explain above. Let us subject that ohtological statement to closer scrutiny.

Since every object is a thing or a person, and "entitv" and "individual" are extensional equivalents of "object", then to put it briefly, there exists nothing else but things and persons -- of course, if the word "exists" is understood in its fundamental sense. In that fundamental sense: A exists, is the same as: a certain object is A, which is equivalent to the formulations: some entity is A, some (at least one) individuals are A, etc. Each of these formulations can either be shortened into: something is A, or expanded pedantically: for some x, x is A. Now the reist deems that only things and persons exist, since it is true only of things and persons that certain objects are things or persons. Should a person further ask about the definition of the term "object", we should have to refer to the meaning of the copula "is" in singular empirical statements (such as "this is green", with an indication of a leaf; or "the Earth is spherical"; or "I am gay"; or "Peter is a carpenter") and say that only that, and all that, is an object about which we may meaningfully formulate a singular sentence (of the type "A is B") with the copula so understood.

From that point of view it is true that Vesuvius exists, and that the dinosaurus exists (of course not necessarily now, but among past objects); it will be false that a centaur or a cyclops exists; but it will be nonsensical to say "the eruption of Vesuvius exists" or "the simultaneity of action and reaction exists". Consequently, we may say correctly that centaurs and cyclops do not exist, but we may not say correctly "the eruption of Vesuvius does not exist" or "the simultaneity of action and reaction does not exist", for the formulations which are here rejected are not subject to a meaningful negation, but must be eliminated as nonsenses. The negation of a false statement is a true statement, the negation of nonsense is itself nonsense. It would be nonsense to say "over whereby exists", but to say "over whereby does not exist" would be nonsense, too. The reservation must be made that we do not mean the words "over whereby" in their possible material supposition, but we take them here in their usual role. The other reservation may also prove advisable. It can meaningfully, and even correctly, be said that -- for example -- simultaneity of action and reaction exists, if the word "exists" is comprehended in its secondary sense, for instance, in a sense in which that statement would mean the same as "a body presses another body when it is itself pressed by that other body".

To recapitulate. The point is that the world of objects is identical with the world of things and persons, and therefore only things and persons exist in the essential sense of that word, and hence the conclusion that only names of things and persons may occur in such a meaningful, and correspondingly in such a true, statement, from which results the assertion of existence of designata of such terms. The concretist has nothing against metaphorical formulations, against using words in their secondary senses. Without sentences with nouns and adjectives other than the names of things and persons we could not communicate briefly and quickly. In lectures, let us not renounce apparent names of properties, relations, events. etc. The point is only that we should know in each case how to eliminate any such apparent terms, for only then shall we interpret correctly the reality, which consists exclusively of things and persons.

This view evokes appraisals which differ greatly from one another. Some people see in it an excessive and over-bold simplification of the picture of reality, while others think that these are self-evident consequences of conventions about meanings of words. It would be a hopeless task to try to bring reism into agreement with all the appraisals by its opponents who themselves differ so much in their opinions. But we make it a point to be on good intellectual terms with logicians, who are past-masters ins exact notation. What appraisal of concretism can we expect of them? I think that rather a favourable one. Using his own language, a semasiologist would classify what are called the names of properties, relations contents, events, etc., in the various semantic categories, and say that these are semantic categories other than either the category of subjects or the category of subjective complements of inherential and subsumptive statements. An inherential statement says that a given individual belongs to a given class, while a subsumptive statement says that the totalitv of individuals of a given class belongs to some other class. The scheme of an inherential statement may be symbolized x e M (x is an element of the class of M's, or x is one of the M's, or x is an M), and the scheme of a subsumptive statement, M C N (the class of M's is included in the class of N's, or whatever is an M, is an N, or any M is N). Now a logician will simply never think of substituting for x anything but a singular name of an individual, and for M or N, a general name of the individuals belonging to a given class, in doing which he will never identify any individual with any property, relation or event. Hence, if he wants to write down the statement saying that the relation of equality holds between x and y, he will resort to another scheme and write x = y, which is a special case of the general scheme xRy, which can be included neither in the scheme of inherential statements nor in the scheme of subsumptive statements. A logician will moreover conclude that the reists identify the totality of individuals with the totality of things and persons, against which he will not raise any objections, but will only say that deciding this issue is not a task of formal logic. It is perhaps superfluous to add that the objection against reism that in logic there are statements which do not include any terms -- for instance, any theorem of the sentential calculus, be it the principle of transposition: Pp, q [(p --> q) --> (~q --> ~ p)], where p and q are sentential variables, would be out of place. The reists do not assert that sentences without terms are impossible. They only claim that if a sentence in its ultimate formulation, without any abbreviations and substitutions, includes some nouns or adjectives, then these are names of concretes, whether singular or general or empty.

On the other hand, the reists will find it more difficult to reach an agreement with those ontologists who accuse them of an excessive simplification of reality. We mean here those who accept the ontological interpretation of Aristotelian categories or of any similar list. All entities are there classified as substances, properties, relations, events, etc. Now the reists accept only one ontological category -- namely, that which has traditionally been called the category of substances. And please note, even within that category they eliminate Aristotelian substances in the secondary sense -- that is, general substances, in other words the universals. What remain are only the Aristotelian substances in the primary sense: Socrateses, horses, stones -- in a word, individual things or persons; the term "thing" becomes modernized and covers everything which is temporal and spatial and physically defined -- for instance, physically influencing something else. Guessing at the existence of anything else is considered by the concretists as committing a hypostasis -- that is, imagining the existence of some entities -- caused by a delusion due to the existence of certain nouns. This gives rise to a number of special problems -- for instance, whether the term "class" is a genuine term or an onomatoid, that is, whether certain objects are classes or not. The answer will depend on which of the two uses of that term, distrutive or collective, is involved. If it is said that the class of M's is included in the class of N's, whereby it is meant that whatever is an M is an N, then the term "class" is used distributively, and the term itself is an onomatoid which vanishes in the ultimate formulation. If, however, the term "class" is used to denote a group of school pupils, the term "class" occurs as a name of a team, of an object of which the individual pupils are component parts.

Some of these hypostases enjoy support not from physicists themselves, but from many philosophical interpreters of physics. In that field, any discussion is extremely difficult, since one of the most topical but unperformed tasks of concretism is to work out a dictionary of mathematics and physics in the reistic interpretation. In this connection the present author will raise only one point, where the issue seems to him clear enough. Now it is not possible to accept such an interpretation of experimental data in which a particle of a physical body, and hence a small object, would have, under certain conditions, to be identical with a wave. The noun "wave" is a special case of the general term "process" or "event", and as such, from the reistic point of view, yields nonsense when substituted for a name of a thing in an ultimate interpretation. To say about a, thing that it is a wave is equally nonsensical as to say about a thing that it is a mode or equality. A similar attitude must be adopted by the concretist with reference to the opinion that reality is a structure of changes which are not changes of anything. The term "change", being also a special case of the term "event", is an onomatoid.To think that the world consists of changes is to build reality out of hypostases, and if it is moreover thought that those changes are as it were, subject-less, this is as if a person were to think that a hospital consists of diseases, recoveries, and deaths, without patients. For the reist, reality is a fabric composed of changing things. Stress is laid on "fabric" and "changing", in order to avert a possible objection that for the reist the world is a "statis conglomerate" ("a mere sum") of "rigid and chageless solids".

After all that has been said above, it would not be astonishing if the reader wondered at a certain dualism to be seen in the formulations used by the concretist. Why is reference always made to things and persons? Does he accept the Cartesian distinction as between two kinds of substances, not reducible to one another, called res extensae and res cogitantes? Can they not all be reduced to one of the two kinds, either the extensive entities, or physical objects, or the sentient entities, that is, psychic beings? In the present author's opinion, this can be done, and for that purpose he chooses the first solution: for him, any psychic being is a physical object, and hence John who experiences emotions, hears, thinks, and makes decisions, is the same John who talks, writes, and in general moves in a purposive manner. This is not only reism, but also somatism, which is a special case of the former. In principle, however, there might be reists who would reject sornatist monism, and accept dualism or spiritualist monism. Somatism is certainly a form of materialism, but it does not imply mechanism. The assumption that every object is a physical object does not in the least mean that the laws of mechanics sufficiently explain everything that happens to objects, for it seems that the laws of mechanics do not even explain all that happens to objects in those respects in which physics is concerned with objects. A fortiori, they do not explain those changes with which psychology is concerned. Thus, somatism faces the immense task of formulating the conceptual structure of psychology from the reistic point of view. That standpoint excludes the description of basic psychologieal statements, such as statements about smells, tunes, tastes, immanent coloured patches, senses of thoughts, contents of perceptions, images of concepts, or acts of perception, of imagining, or of conceptual comprehension, etc. All these nouns are onomatoids. Elsewhere, the present author has taken the liberty of calling radical realism that view which accepts what are called names of contents to be onomatoids and hence prohibits their use as subjects in elementary sentences. From the standpoint of reistic somatism, efforts are made to interpret bbasic psychological statements as statements about persons (the latter being, of course, identical with certain physical objects). There would be, for instance, such statements as: "X experiences as follows: A is B". As applied to a given case, this would yield: "John sees so: this is black". Of course, the sentence of the type "A is B" is a special form of sentence, and in the general formula of the psychological statement its place may be occupied by a sentence of any form. If we take p to stand for an arbitrary sentence, we obtain the following general formula of the psychological statement: "x experiences as follows: p", or "x is experiencing as follows: p". Thus the only specific name element of the psychological statements would be the term "experiencing" or one of its special forrns: "seeing", "wishing", etc. All these terms are predicable about persons, are names of persons (the latter being, from the somatist standpoint, identical with certain physical objects). And here are some other examples of psychological statements:

"John thinks so: 2+2 = 4",
"John feels so: they are playing sadly",
"John doubts so: do angels exist?",
"John desires so: be happy".

Perhaps it would even not be too bold to generalize this formula so that not only a sentence, but any phrase referring to how a given person experiences, could be substituted for "p". It might even be an inarticulate exclamation, so that a given psychological statement would be: "John experiences so: Oh!"

Having worked out the principles of reism in its somatist and radical realist form, the present author for some time had an illusion of having invented something new. But soon his attention was drawn to the fact that such a view was represented by Franz Brentano in the last years of his life. In the writings dictated by Brentano, who was then blind, and appended to a new edition of his Psychologie vom empirischen Standpunkt in Meiner's Philosophical Library (cf. No. 193 of that publication, Chapter entitled "Von den Gegenstanden des Denkens", and passim, Leipzig 1915), we find explicit warnings against hypostatising those nouns which are not names of things and persons, and an explicit tendency to accept things and persons as the only entities. Brentano was thus undoubtedly a reist, a concretist, although he did not use such terms to describe his standpoint. As a former priest, he stopped at the threshold of somatism and never crossed it. To the end of his life he remained a dualist reist. He was not the first reist: he himself found that even in Leibniz's works appears what is probably the first formulation of that principle. In Nouveaux essais sur l'entendement humain Leibniz said that all thorny philosophical issues would vanish immediately if we made sure that in formulating theorems we should use only names of concrete objects. With this, for him somewhat unexpected, formulation Leibniz initiated the brief series of the forerunners of reism, which in a very vague form may be traced back even to the Stoics. He was, however, a spiritualist reist. The present author knows of no consistent and conscious somatist reists. Is that so because the doctrine is erroneous? This is not known, but the claim may be hazarded that it is a true doctrine, and that is why the reader's attention was occupied with its analysis.