Objects and Pseudo-Objects: Ontological Deserts and Jungles from Brentano to Carnap 9781501501371, 9781501510458

Citation preview

Objects and Pseudo-Objects

Philosophische Analyse/ Philosophical Analysis

Herausgegeben von/Edited by Herbert Hochberg, Rafael Hüntelmann, Christian Kanzian, Richard Schantz, Erwin Tegtmeier

Band/Volume 62

Objects and Pseudo-Objects Ontological Deserts and Jungles from Brentano to Carnap Edited by Bruno Leclercq, Sébastien Richard and Denis Seron

ISBN 978-1-5015-1045-8 e-ISBN (PDF) 978-1-5015-0137-1 e-ISBN (EPUB) 978-1-5015-0139-5 ISSN 2198-2066 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2015 Walter de Gruyter Inc., Boston/Berlin Printing: CPI books GmbH, Leck ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Bruno Leclercq, Sébastien Richard and Denis Seron

Preface

The idea of a distinction between objects and pseudo-objects is related to a rather old and classical philosophical problem, which raises the question of the ontological status of general objects such as “the rose in general”, of abstract objects such as “the size (or the color) of this rose”, or even of fictional objects such as “the rose Fidel Castro offered to Margaret Thatcher” or “Flora, the goddess of flowers”. Even though they cannot be seen, picked and smelled like this or that actual rose, all these “objects” seem to behave like genuine objects as can be seen in the fact that they possess properties, which can be attributed to them in true predicative judgments – “the rose is a flower”, “the size of this rose is 5 inches”, “Flora is married to Favonius”, ... Language, at least, seems to make it possible to speak about such objects even if they do not exist in space and time and cannot be given to sense perception. A tough philosophical debate has classically opposed those – named “realists” – who consider that these are genuine objects and those – named “nominalists” – who consider that they are nothing but pseudo-objects and that their seeming to be proper objects is just a trick of language. In the twentieth century, such a debate took the form of a spectacular opposition between those who, like Alexius Meinong, would take any object of thought and any linguistic subject of predicative judgment as a genuine object – with the consequence of such a luxurious domain of objects that his detractors called it an “ontological jungle” – and those who, like Bertrand Russell and Willard Van Orman Quine, but also and even more radically Franz Brentano in his old age, Tadeusz Kotarbiński or Stanisław Leśniewski, would rather rephrase sentences concerning these “objects” in such a way that only ontologically innocent objects (e.g. objects which can be known by empirical acquaintance) appear as the logical subjects of the sentence – with the consequence of much less crowded “furniture of the world”, perhaps even an “ontological desert”. This strategy of getting rid of pseudo-objects by rephrasing sentences so as to show that their surface linguistic subjects do not properly refer – they are “apparent names” or “onomatoids”¹ – and are not the genuine logical subjects of these sentences is known as the use of Occam’s razor for shaving Plato’s beard. In some

1 See A.C. Zielinska’s paper.

VI � Bruno Leclercq, Sébastien Richard and Denis Seron respect, Russell and Quine could be seen as having tried to clear Meinong’s jungle in a similar way with a methodological “machete” sharpened by Frege’s formal logic.² In the twentieth century a specific concern for what Quine called the “referential opacity” of intensional contexts gave a new face to this classical debate. In modal contexts, i.e. in contexts involving the notions of necessity and possibility (but also of duty and permission), as well as in contexts of intentional attitudes, i.e. in contexts of beliefs, hopes, fears, ..., the “intension”, and not only the “extension”, of linguistic expressions matters: in such contexts, expressions do not only refer to some objects but to some objects apprehended and/or described in a certain way. While the “logically” proper names “Phosphorus” (the morning star) and “Hesperus” (the evening star) both refer to one and the same object, namely Venus, so that normally everything which is true of one is also true of the other one – if “Phosphorus undergoes retrograde rotation” is true, so is “Hesperus undergoes retrograde rotation”; when “Hesperus can be seen from the Earth” is true, so is “Phosphorus can be seen from the Earth” –, there still are contexts where the two names cannot be intersubstituted salva veritate: “Phosphorus (being the morning star) necessarily shines in the morning” is true while “Hesperus necessarily shines in the morning” is not; and “Ancient Greeks believed that Phosphorus was hidden during the evening” is true while “Ancient Greeks believed that Hesperus was hidden during the evening” is not. Similarly, even though both phrases actually refer to the same individual object, “the current President of the USA” and “the husband of Michelle Robinson” cannot replace each other in the sentences “The husband of Michelle Robinson necessarily is a married man” or “The Tea Party supporters hate the current President of the USA”. And even though the passengers of the Mayflower became the founders of Plymouth, I can believe that the passengers of the Mayflower never walked upon American soil while I may not believe that the founders of Plymouth never walked upon American soil. These examples seem to mean that Phosphorus and Hesperus – or similarly the current President of the USA and the husband of Michelle Robinson, the passengers of the Mayflower and the founders of Plymouth – have at least some different properties and are thus somehow distinct “objects”. Now, what has here been highlighted on a linguistic level had been studied on a psychological level in the Brentanian school at the end of nineteenth century. Our thoughts and other mental states take as their contents “intentional ob-

2 Frege, himself, though, was not a nominalist. To some extent, his theory of meaning could even be seen as leading to an ontological jungle. See C.O. Hill’s paper.

Preface � VII

jects”, i.e. objects which are sensitive to the way they are apprehended and conceived through these mental states. For instance, in Sergeant Garcia’s thoughts, Zorro and Don Diego de la Vega are characterized in very different ways so that, for him, they are not one and the same object; as contents of Sergeant Garcia’s intentional attitudes, Zorro and Don Diego de la Vega are different objects. And even while knowing that the current President of the USA and the husband of Michelle Robinson actually are one and the same person, namely Barack Obama, a partisan of the Tea Party can attribute some properties to – and entertain some feelings towards – the former which she does not attribute to – or entertain towards – the latter. As far as intentional attitudes are concerned, Barack Obama seen as the current President of the USA is not the same object as Barack Obama seen as the husband of Michelle Robinson. And this is why, for Brentanians, there seem to be as many objects as there are contents of thoughts... including several ones related to a single actual object like Barack Obama. While Barack Obama exists and is an actual object corresponding to some mental content, there also are contents of thoughts to which no actual object corresponds: I can believe that Santa Claus exists, admire Sherlock Holmes, fear ghosts or look for the round square (i.e. hope for squaring the circle). Now, from the psychologist’s point of view, my mental states can be described and studied regardless of whether the objects they are aiming at actually exist or not. On this respect, i.e. as far as psychological description is concerned, contents of thoughts such as Santa Claus, Sherlock Holmes, ghosts or round squares are just part of the mental states aiming at them... But, from a metaphysical point of view, they surely are more than that: that they are not just part of my mental states is shown by the fact that they can also be the contents of other people’s thoughts and intentional attitudes – Santa Claus (or Sherlock Holmes or the headless horseman or the round square) is not just something in my head but seems to be of an “intersubjective” nature – but also by the fact that they have properties which have nothing to do with the mental³ – Santa Claus has a beard (Sherlock Holmes is a detective, the headless horseman carries his head under his arm, the round square has diagonals) while no part of a mental state could. Therefore they are said to be “objective” rather than subjective; they are “ideal” contents of thoughts rather than real parts of mental states. Yet the question remains of whether they are genuine objects or not. Mental states are said to be intentional insofar as they aim at (existent or inexistent) “objects” apprehended as such and such; in some cases, the corresponding linguistic expressions exhibit unusual logical properties and are said to be inten-

3 See M. Gyemant’s paper.

VIII � Bruno Leclercq, Sébastien Richard and Denis Seron sional (with an s). This raises the question of what kind of things such intentional “objects” or “objective contents” are. When I think about the morning star, my thought does not have the same content as when I think about the evening star; yet to both contents corresponds one and the same actual object. As Frege would put it the names “Phosphorus” and “Hesperus” have different senses and yet refer to the same actual object. And when I think about Santa Claus, my thought has got some (objective) content; yet no actual object corresponds to it. The name “Santa Claus” has some sense but does not refer to any actual object. These are the puzzles that led Bernard Bolzano to separate the ideal (as opposed to the mental) content (Inhalt) of a thought/presentation from its objectuality (Gegenständlichkeit). And this, on its turn, led both Kazimierz Twardowski to separate the intentional object (intentionales Objekt) – as opposed to the mental content (Inhalt) – of a thought from its (possible) actual object (äusseres Objekt) and Gottlob Frege to separate the sense (Sinn) – as opposed to subjective presentation – of a linguistic expression from its reference (Bedeutung). And here lies the question of the relation between contents and objectivity, or between intentional objects and objectivity or again between senses and objectivity. Leaning on Twardowski, Meinong considered all intentional objects as objects, even though some of them are devoid of actual existence as well as of any other kind of being (such as the subsistence of mathematical objects): Sherlock Holmes and the round square are objects in spite of the fact that the first one lacks existence and the second one is even impossible. Being an object does not necessarily require to have a Sein but only to have some characteristic features – a Sosein. Leaning for his part on Frege, Russell considered sense to lie in characteristic features (Merkmale) which may happen not to be endorsed by any (actual) object: Sherlock Holmes, conceived as the detective living at 221b Baker Street who has solved such and such mysterious cases of crime, and the round square, conceived as a geometric figure which would be made of four equal sides and four right angles while all its points would be at the same distance from its centre, are well-defined meanings to which however no (actual) object corresponds. For that reason, they should not be considered as being of an objectual but rather of a conceptual nature. Indeed, Frege had opposed concepts, which are (propositional) functions, to objects, which can be their arguments (or independent variables); concepts are required to have sense, i.e. characteristic features, but they do not need to be satisfied by an object: a term like “unicorn” is meaningful yet devoid of objective reference. And so is the case, according to Russell, of “the round square” as well as of “Sherlock Holmes”, which is an abbreviation for “the detective living at 221b Baker Street who has solved such and such mysterious cases of crime”.

Preface �

IX

Is having characteristic features – a Sosein – the mark of objects or of concepts? This is one of the main questions here at stake. In order to be an object of thought, having characteristic features is sufficient and existing is not necessary. But being an object of thought in this sense seems to amount to being conceivable and therefore to being a concept⁴ (which, for Frege and Russell, is anyway something “objective” and not just a part of someone’s mental states). Like sets of characteristic features, objects of thought can be generated at will. I can think (and talk) about golden mountains, flying horses, green unicorns with red spots, and the like. However, followers of Frege and Russell would say, this enriches our ideology (and terminology) rather than our ontology; it provides new concepts, i.e. new principles for sorting the objects of the world, without changing anything in the furniture of the world. The fact that we can think and talk about golden mountains and flying horses does not prove that “there are” (in any sense of the expression) objects such as these; they could merely be concepts devoid of any reference, i.e. satisfied by no object. Rather than as independent variables referring to inexistent objects, Frege and Russell claim, such expressions should be construed as propositional functions, with the consequence that the corresponding “objects” are, basically, of a conceptual nature. The question, however, is whether such a rephrasing is satisfactory and also applies to sentences about other kinds of “objects of thought”. According to some Brentanians, any object of thought or talk must be a genuine object. I can think and talk about the color of this unicorn, the size of this rose, Peter and John’s friendship, the hypotenuse of a right-angled triangle, the divisors of twelve, the raising of the Sun, your having arrived late at our first date, and so on. This is why some of Brentano’s followers such as Twardowski, Meinong, and Husserl understood “object” in a wider way⁵. At the center of Husserl’s approach is the psycho-linguistic notion of “nominalization”. Nominalization transforms any part of what a thought intends or a sentence means into the object of a new thought and thus into the linguistic subject of higher-order statements. I think and talk about this rose being red and then about the redness of this rose being bright or beautiful; I think and talk about Peter and John being friends and then about Peter and John’s friendship being old and firm; I think and talk about the Sun raising and then about the raising of the

4 For his part, Alexander Pfänder will distinguish between the content of a concept and the “formal object” corresponding to it, while Jean Héring will separate concepts from “ideas” in the sense of “essences”. See G. Fréchette’s paper. 5 See S. Richard’s paper.

X � Bruno Leclercq, Sébastien Richard and Denis Seron Sun occurring earlier in the summer than in the winter; and so on. In this respect, properties, relations, states of affairs, events, etc., are kinds of objects. Yet the question remains whether such entities can be eliminated by some rephrasing strategy. Should we keep to first-order logic and refuse to construe these entities as objects in the logical sense, that is, as independent quantified variables? Or should we accept them as part of the furniture of the world? Would this provide a satisfying account of how they behave as regards truth value? Inexistent objects do seem to have properties, and to be genuine logical subjects of true statements. Consider the case of fictional objects: Santa Claus wears a beard, Sherlock Holmes smokes a pipe, and flying horses have big wings. Frege and Russell’s extensional rephrasing of such sentences is notoriously unsatisfactory. For Frege, “Flying horses have big wings” should be rephrased as “For all objects, if this object is a flying horse, it has big wings”, a sentence which is trivially true since no (actual) object is a flying horse. But this analysis also makes true the statement “Flying horses smoke pipes” ! For Russell, “Sherlock Holmes has big wings” should be rephrased as “There is only one (actual) object which is a detective living at 221b Baker Street and has solved such and such mysterious cases of crime and this object has big wings”, a sentence which is false since no (actual) object satisfies the first description. But this analysis makes false the statement “Sherlock Holmes smokes a pipe” as well ! According to Frege, universal sentences having “inexistent objects” as their linguistic subjects are all true; according to Russell, singular sentences having “inexistent objects” as their linguistic subjects are all false – which means that it is true that all flying horses have big wings but false that Pegasus has big wings... Clearly, extensional paraphrases do not work. A more convincing way to approach statements about “inexistent objects” while claiming that these are concepts rather than objects might be to provide – as Carnap did in Meaning and Necessity (1956 [1947]) – an intensional rather than extensional analysis of these statements: “Flying horses have big wings” and “Sherlock Holmes (the detective living at 221b Baker Street who has...) smokes a pipe” are statements about the sense or intension of the conceptual terms “flying horse” and “detective living at 221b Baker Street...”. But this means that if such statements are true, they are analytically true, i.e. the attributed properties are part of the meaning of the conceptual terms which function as linguistic subjects of the sentence. Having big wings must somehow be included in the definition of “a flying horse”; smoking a

Preface �

XI

pipe must be included in the definition of “Sherlock Holmes” (“definition” which amounts to the whole description of this character in Conan Doyle’s work).⁶ Note that Meinong and his followers hold quite a similar view, believing that objects rather than concepts are characterized by sets of descriptive features (the golden mountain is the object made of the properties of being a mountain and being made of gold; Sherlock Holmes is the object made of all the properties he has in Conan Doyle’s novels); to such objects can only be attributed their own “constitutive” properties and the analytical consequences of these.⁷ Proximity between Meinong and Carnap on this point shows that, by taking “intensions” into account, the latter took a significant step towards the admission of other entities than singular actual objects. Sherlock Holmes, i.e. the detective living at 221b Baker Street... (be it actual or not), or the morning star (whatever actual object it is), which are the logical subjects of Carnap’s quantified modal logic sentences, clearly are semantical entities rather than actual objects.⁸ The difference between both philosophers is that Carnap also keeps an eye on extensions so that, next to semantical entities which are characterized by descriptive features, he also admits actual objects such as Barack Obama or Venus. For most Meinongians, by contrast, actual objects are characterized by (infinite sets of) descriptive features just like any object, their only specificity being the fact that they exist. As has been shown above for the case of fictional objects, the question of which objects have to be admitted as genuine objects is not merely a matter of rephrasing sentences. The question is also how to account for what makes these sentences true or false. Indeed, objects are not merely what we think or talk about, they are also the truth-makers for what we think and say about them.

6 Similarly, it could be said that it is in the definition of the number “241” to be a prime number. In some way, logicism can be seen as the view that the ideal “objects” of mathematics are of a conceptual nature and that all true statements about them – “241 is an prime number”, “A rightangled triangle is such that the area of the square whose side is its hypotenuse is equal to the sum of the areas of the squares whose sides are its two legs” – are analytically true. 7 Besides their own constitutive properties, “extranuclear” properties can also be attributed to Meinongian objects: of the golden mountain, I can not only say that it is a mountain and that it is made of gold, but also that it does not exist or that owing one is my favorite dream. 8 Another approach, later adopted in Frege’s school in order to give an account of fictional objects and other intentional objects is quantified modal logic based on possible world semantics. Although quantified modal logic allows to have only singular objects in each possible world, it ontologically commits one to the existence of the possible worlds themselves and to cross-world entities such as “the morning star (whatever actual object it is)”, which is some kind of function from worlds to singular entities (i.e. the singular entity which satisfies this description in each world).

XII � Bruno Leclercq, Sébastien Richard and Denis Seron This is what Husserl notably tries to theorize by giving an account of the objects of both “meaning intentions” and intuitions. For Husserl as for many Brentanians, an object is, in one sense, anything which can be thought or talked about (i.e. take place as the linguistic subject of a sentence). In another sense, an object is what is somehow “given” (along with its properties) and, as such, makes true thoughts and sentences. This distinction between two meanings of “object” was a central feature of Brentano’s and his pupils’ theory of intentionality. The general idea is that intentionality involves some kind of constitutive ambiguity, due to the fact that in most cases the “object” the mental state is about is not the object of direct acquaintance that makes the state veridical. Brentanian intentionalism may be viewed as an attempt to dispel this ambiguity by distinguishing the real object properly referred to from the intentional object or content which is “obliquely” presented to the subject as far as she is aware of what her representation is about. In some sense, the true judgment “the Centaur is a poetic fiction” is about the Centaur, which thus functions as its “object”. Yet the Centaur is not thereby taken as existent; it is assumed to be inexistent and hence cannot be the really existing object that makes the judgment true. In Brentano’s view, at least in his later works, such judgments actually refer to some mental reality, and the sentence should consequently be rephrased as “Some mental states are such that they are lived by poets and have the property of being about ‘the Centaur’ ”. Now, he continues, this applies not only to mythological objects, but to physical reality in general. On the one hand, the physicist’s judgments are really about physical beings distinct from their appearance in the mind, not about anything mental. But on the other hand, Brentano claims in Psychology from an Empirical Standpoint, such physical beings are “fictions to which no reality of any sort corresponds” (Brentano 1874/1973: 11). Likewise, because they believed only sense experience can give objects, Russell and his followers tended to restrict the domain of objects to real objects, and to discard as “pseudo-objects” the objects of thought and speech which cannot be known by empirical acquaintance. Husserl, by contrast, takes note that objects of thought and speech are of many different kinds, and that each of them makes the related thoughts and sentences true or false. For him, these objects must be not only intended or meant, but also “given” or “intuited” in some specific way. It is phenomenology’s task to investigate the various kinds of objects as well as the specific ways they are intended and given. This does not prevent sense experience from playing an important role in each kind of givenness. For Husserl, abstract ob-

Preface �

XIII

jects such as essences and states of affairs⁹ are intuited on the ground of sense experience yet through some specific “categorical” apprehension that makes them the specific kind of objects they are.¹⁰ Even fictional objects are, in some sense, given, namely through imaginative intuition. The result is a general theory that specifies not only the various modes of givenness, but also the corresponding modes of being and reality and the way they are related to each other and to sense experience. Indeed, there are not only several kinds of objects of thought and speech, but also specific kinds of being given and thus specific kinds of being¹¹, of which existing or being real is only one yet perhaps the most fundamental. This includes the theoretical objects of science, namely of physics and chemistry as well as of biology, psychology, and social sciences. Husserl, Carnap, Hartmann took up this challenge, although providing quite different answers. Carnap’s claim in the Aufbau is that the Russellian paraphrasing makes the truth of scientific statements dependent on sense experience only. Atoms, molecules, cells, desires and beliefs, interests, social groups, etc., all objects of science are (more and more) complex logical entities constituted on the basis of similarity relations between objects of immediate sense experience. Therefore, sentences about these objects are no more than convenient abbreviations for (long and complex) sentences about immediate experience. For Husserl, by contrast, the constitution of higher-level objects on the basis of immediate experience is not a mere logical construction.¹² Although grounded on the latter, the former cannot merely be reduced to it. For Hartmann, each level, even though ontologically dependent on the lower levels, has its own laws, ontological categories and ontological dignity.¹³

9 On the tough question of the ontological nature of states of affaires, see D. Seron’s and G. Fréchette’s papers on Husserl and Reinach. 10 This leads to some tension in Husserl’s work. On one side, objects of all kinds are given or intuited rather than merely meant or intended. On the other side, intuition is always the fulfillment of meaning intentions, so that an object can only be given according to the way it has been intended. Realism in Husserl’s phenomenology is thus counterbalanced by idealistic trends. See R. Brisart’s paper. 11 The fact that Meinong thinks that some objects do not have any Sein but are “Außersein” somewhat shows that, for him, they are thought but not given. 12 And this, for Husserl, even is the case of higher order objects such as sets, numbers or “manifolds” (mathematical structures). See C.O. Hill’s paper. 13 See Poli’s paper.

| In memory of Robert Brisart (1953–2015)

Contents Bruno Leclercq, Sébastien Richard and Denis Seron Preface � V Part I: Ontological Parsimony Peter Simons How to Do Things with Things � 3 Federico Boccaccini The Bounds of Object � 17 Anna C. Zielinska Kotarbiński’s Strong Minimalist Ontology � 39 Robert Brisart Objects as Posits from a Phenomenological Point of View � 51 Bruno Leclercq The Concept and its Object are (not) One and the Same � 63 Part II: Objecthood Prodigality Maria Gyemant Objects or Intentional Objects? � 85 Dale Jacquette Domain Comprehension in Meinongian Object Theory � 101 Sébastien Richard Meinong and Early Husserl on Objects and States of Affairs � 123 Guillaume Fréchette Essential Laws � 143 Denis Seron Adolf Reinach’s Philosophy of Logic � 167

XVI � Contents Claire Ortiz Hill Husserl’s Way Out of Frege’s Jungle � 183 Part III: Modes of Being Arkadiusz Chrudzimski Ingarden on Modes of Being � 199 Roberto Poli Nicolai Hartmann’s Theory of Levels of Reality � 223 Bibliography � 239 Index � 256

| Part I: Ontological Parsimony

Peter Simons

How to Do Things with Things Brentano’s Reism and its Limits Peter Simons: Trinity College Dublin, Ireland

1 Introduction From about 1904 until the end of his life in 1917, Franz Brentano held an ontological view which has come to be called reism. This is the view that the only things that exist are concrete things (res). The list of objects that this view denies existing is long indeed, and includes: properties and relations, whether considered as individual accidents or as universals; events and processes; facts and states of affairs; numbers, sets and all other mathematical objects; space and time as entities in their own right; intentional contents and objects; propositions, and other abstract senses or meanings. The objects that Brentano does accept as entia realia or things include as individuals mental souls, physical bodies and their parts and lower-dimensional boundaries, and collections of individuals. Brentano conceives it as possible that the primary physical things be of more than three dimensions: he calls such things topoids. Brentano’s ontology is thus, in comparison with that of most other philosophers, extremely sparse or parsimonious. Some medieval nominalists flirted with reism: John of Mirecourt considered that only the dogma of transubstantiation stood in the way of taking there to be only things. Brentano himself mentions Leibniz as a potential forebear, assuming the real for Leibniz to comprise only the monads, and not their successive states or modifications. Somewhat later, and independently of Brentano, Tadeusz Kotarbiński, to whom we owe the term “reism”, came to the view that the only things are bodies which are extended in space and in time: his reism thus differed from Brentano’s in an even more parsimonious direction and he favoured the term “pansomatism” for it. Brentano’s view differs from Kotarbiński’s not only in his acceptance of mental substances or souls, but also in his view of time: Brentano is a presentist, holding that only that which exists now, in the present, exists, and that no thing is extended in time, whereas for Kotarbiński all things are extended in time as well as in space. Brentano’s view was the culmination of a long and complex development in his ontology. His early work was carried out under the influence of Aristotle’s

4 � Peter Simons theory of categories, according to which there are several basic kinds of entity. While Brentano was never an uncritical follower of Aristotle, he was initially generally favourable to this idea, whereby alongside substances there are qualities, quantities, relations, actions, places and times. The Psychology of 1874 was already showing signs of ontological parsimony. According to his theory of the three basic kinds of mental acts: ideas, judgements and attitudes, the objects that are presented by ideas are also that which is accepted or denied in judgement, and which is liked or disliked in attitudes. Middle-period Brentano moved to a somewhat more ample theory, allowing judgements to have their own specific kinds of content, what would nowadays be called states of affairs. Over a period beginning in the 1890s, and following a somewhat obscure development, he came to his late and austere view. In good part this evolved in dialogue and debate with his former student Anton Marty, who predeceased him in 1914, and the latter’s students Oskar Kraus and Alfred Kastil, who at first found it hard to follow Brentano, but eventually became his most vociferous supporters, as well as his literary executors after his death. In later life, in part as a result of his blindness, Brentano found it difficult to put his views together into monograph-length accounts. He was more at home dictating relatively short notes and letters. His last major attempt at a treatise on ontology, again starting from Aristotle but being this time much more critical, occupied him for his last ten years. It arose through a number of preliminary notes and studies which represent shifting views, and went through three drafts, none of which made it into print, until the various notes and drafts were compiled by Kastil under the title Kategorienlehre and published in 1933. Since Brentano’s text is often unclear in its import, Kastil, who by this stage was more familiar with Brentano’s work than anyone, felt compelled to add over four hundred elucidatory notes. Brentano’s final ontology is thus unfinished and unsystematic, although many of its basic positions and themes are fairly clear. So in this study I shall be concerned less with the way in which Brentano came to his views, which I still find puzzling, than with the question whether they, or something close to them, can be considered adequate as an ontology.

2 There are Things By a thing, Brentano understands an entity which is particular, unrepeatable, and persisting through time. That there is at least one such thing is regarded by Brentano as evident via introspection. Introspection, or inner perception as Brentano calls it, reveals to me that in the flux of my mental activity there is a

How to Do Things with Things �

5

persisting owner or bearer of these acts, a non-spatial, persisting and substantial item which may rightly be called a soul. Very much a Cartesian, Brentano regards this evidence as more secure than anything we might know or conjecture about the physical, spatial world. Since such evidence is infallible, I can be sure there is at least one thing: myself. Physical bodies provide further if less secure examples of things. They persist, but unlike souls are located in space and can move about. Both souls and bodies can change, and accounting for change becomes a major theme in Brentano’s reistic ontology.

3 Everything is a Thing Brentano has an argument purporting to show that there are only things. It goes as follows: 1. 2. 3. 4.

An object is something that can be thought about. “think” is univocal. To think is to think about something. We think about things.

therefore: 5. Whatever we think about is a thing. (from 2, 3, 4) therefore: 6. Every object is a thing. (from 1, 5) The premises 1, 3 and 4 are relatively uncontroversial. Premise 3 is an application of Brentano’s intentionality thesis. Premise 4 is empirical common sense. Premise 2 is perhaps not self-evident, but let us grant it for the sake of argument. The crucial transition is to statement 5. Brentano’s idea is that if we could think about objects other than things, since we also think about things, this would make “think about” and thereby “think” equivocal, contradicting premise 2. Formally speaking, this is a terrible argument. Here is an argument reproducing the form of 2-5: 2a. 3a. 4a.

“eat” is univocal. To eat is to eat something. We eat apples.

therefore: 5a.

Whatever we eat is an apple.

6 � Peter Simons The premises are uncontroversially true, and the conclusion is uncontroversially false. Therefore the argument is invalid. The difference must presumably turn on the fact that whereas “apple” is a material concept, “thing” is a formal or categorial concept. But this does not help Brentano, because suppose we replace “thing” by “quality” in the fourth premise: 4q.

We think about qualities.

Since “quality” is no less a formal or categorial concept than “thing”, the argument ought to show that we only think about qualities and that everything is a quality. Brentano has an answer, which is to say that in cases where we supposedly think about qualities, relations, places, times, numbers etc., the term “think” is not being used in its proper or authentic sense, but in an improper or inauthentic sense, so that the supposed objects of thought in this case are not genuine but fictitious. This rejoinder is however clearly question-begging, since anyone not persuaded that that only things exist will legitimately resist the move to count thinkings of non-things as improper. The result is simply a stand-off. Rather than further examine Brentano’s reasons for being a reist, I prefer then to simply accept that he has the view and see what can be done with it. The answer is perhaps surprising: more than one might at first think.

4 Accidents and Qualitative Change One of the principle tasks of any ontology, and one which poses particular challenges for reism, is to offer an account of change. There are several kinds of change. One is motion, or change of position. We postpone consideration of this until later. Another is existential change, coming to be and ceasing to be. Here the reist is at no disadvantage over others, since it is the coming to be and ceasing to be of things that is principally of interest. The more interesting and challenging types of change are: qualitative change, quantitative change, and relational change. Qualitative change is the most straightforward. Consider for example a tomato, which as it ripens changes in colour from green to yellow to orange to red. It is one and the same thing throughout, notwithstanding its gradual alteration in colour. Non-reistic ontologies, whether realist or nominalist about universals, have a ready explanation. They say that the tomato has different colour-properties at different times. If colours are universals, the tomato will successively exemplify a sequence of distinct universals. Typically, such theorists will say that the tomato’s being this colour at this time and that colour at that

How to Do Things with Things

� 7

time are two facts or states of affairs. This picture is familiar, and it is rejected by Brentano because he rejects universals and states of affairs. A nominalist account of qualities or properties such as is commonplace among medieval scholastics posits instead a sequence of colour-accidents or tropes which successively inhere in the tomato. When it changes colour, this amounts to one colour-trope going out of existence to be replaced by another from the same (colour) family, which comes into existence, each inhering in the same substance, the tomato. There are some complications involving differential change in different parts of the tomato’s surface, but they can be coped with, and the general picture is the same. Qualitative change consists in the replacement of quality-tropes, the substance remaining in existence. Brentano rejects this picture, because he rejects tropes. Since Aristotle, tropes have been conceived as items inhering in their substances as a kind of part which is inseparable from its substance and incapable of existing apart and alone, by contrast with other, separable and potentially free-standing parts such as the tomato’s seeds or skin. This distinction between separable and dependent parts was highlighted and analysed by Brentano’s student Husserl in the third of his Logical Investigations. Brentano, unlike Aristotle, is perfectly happy to accept parts of things which are not in fact separated, as being themselves genuine things, provided they are separable. So the skin of a tomato is a part of it and a thing, notwithstanding its actual connection to the rest of the tomato. Aristotle had considered such parts as only potential entities, not actual ones. Brentano however rejects the idea of an inseparable dependent part as at best an abstraction or fiction. His preferred way of dealing with change is to consider accidents not to be non-thing parts of substances, but things of which substances are parts, what in Aristotle are called accidental unities. Consider the tomato, the green tomato, and the red tomato which existed later. According to Brentano, these are three things, rather than one thing and two things-with-a-trope. The green tomato and the red tomato are concrete things which differ both qualitatively and in their times of existence: the green tomato ceases to be before the red tomato comes to be. They do however have the tomato itself as a common part, which persists throughout. That is how change is explained. The tomato changes in that it is successively part of different “enriched” things. Since a thing may change qualitatively in a number of ways, there can be many overlapping things which have the same tomato as part. A natural question to put at this point is this. If the tomato is part of the green tomato and later a part of the red tomato, what is added to the tomato to give the green tomato, and what replaces this to give the red tomato? In mereological terms, since the tomato is not identical with the green or the red tomato, what is the mereological supplement making up the rest of the green, respectively red

8 � Peter Simons tomato? Brentano’s surprising answer is: nothing at all. He rejects the principle of mereology according to which if one thing is a (proper) part of another, there is another part of the whole disjoint from the first, the Weak Supplementation Principle. For Brentano the tomato is enriched to give the green tomato, and differently enriched to give the red tomato. But it is not enriched by any entity. If it were, this additional entity could not be a thing capable of independent existence, so it would be incompatible with reism to accept such a supplement. Brentano appears to have no other good reason to reject the idea of a supplement than this. But it is surely analytically true of the concept of part that a part which is not the whole has a supplement, so at the very least Brentano is not entitled to call the relationship between a substance and its accident one of part to whole. What the relationship then could be is not clear. Once again Brentano’s position gains no advantage over the bicategorial ontology of things and tropes, but the issue of the relationship aside, it does not appear to be at a material disadvantage either. Quantitative change is somewhat more complicated. Consider our tomato again. It not only changes colour as it ripens, but also grows in size and gains in weight. The weight gain can be explained in terms of the addition of new parts to the tomato through the natural processes of plant metabolism. The tomato is after all a complex object composed of many parts, and there is no strong reason for Brentano to deny that such aggregative individuals may persist despite mereological change, the addition or loss of (some) parts. Such aggregative wholes may not be the most basic of individuals, but there is no reason to deny them their status as things. Likewise a tomato grows in size, diameter, surface area, and volume because new parts are added to it. A different case is presented by a thing which grows, shrinks or otherwise changes its shape and/or size but without change of parts. A piece of steel wire for example is variable in length (Hooke’s Law) as well as in shape, depending on all sorts of internal and external influences, a metal body which is heated will measurably expand without gaining parts, and organic bodies such as ourselves change their shape all the time. Brentano’s account must be in such cases that the different parts of the thing change in their spatial relationships to one another, so it comes to be subsumed under the case of (relative) motion.

How to Do Things with Things �

9

5 Relations Brentano does not believe in relations as entities, whether these be universals or particulars. So how then does he account for relational truths, such as the following? Caius is taller than Titus The short-circuit caused the fire The leg is part of the chair The equator bounds the northern hemisphere Dublin is 6349 km from Addis Ababa Sherlock Holmes is more famous than Kurt Wallander The answer is: variously. But there is a feature common to all of Brentano’s treatments of relational truths, which is his distinction between two ways in which we think about things. If I simply say: Caius is brave then I am thinking of Caius directly, modo recto, as Brentano calls it. But in: Caius is taller than Titus I think of Caius directly, but Titus indirectly, modo obliquo. It is for Brentano characteristic of relational predications that all but one of its nominal subjects involves thinking modo obliquo. He normally only gives examples of binary relations, but a similar account will apply to relations of more than two places, such as: Dublin is between Belfast and Wexford John gave Mary the flowers where only Dublin (resp. John) is thought about directly. The remainder of the sentence ascribes a relative determination to its subject which involves thinking of one or more things modo obliquo. Aristotle considered that in all relational predications, all the terms have to exist, except in the case of intentional relations such as: Karel is thinking about Pegasus This can be true despite the fact that Pegasus does not exist, because in truly predicating this of Karel we think of Pegasus only modo obliquo, and so are not ontologically committed to Pegasus. In this regard, Brentano is wittingly or unwittingly recapitulating William of Ockham, for whom only subjects in the nominative case carry ontological import. Some relational predications involving relational predicates definitely are existing-entailing in all nominal positions: “cause”, “eat”,

10 � Peter Simons “hit” and “marry” are examples. Relational predications where one of the terms does not have to exist for the predication to be true, as in the case of thinking about, are what Brentano calls “relation-like” (relativlich). Relational change needs not directly affect a thing in itself. The example is Titus outgrowing Caius. Caius changes in no intrinsic way when he ceases to be taller than Titus due to the latter’s growth. Nor need the gain or loss of a part change the rest of an object. Suppose I have a long steel rod, one end of which gets scratched and loses a miniscule portion of metal. According to Aristotle, as Brentano interprets him, this spells the destruction of the original rod and its replacement by something which did not actually exist before, but was a potential object, lying in wait as it were, to spring into existence by the removal of the fractional part. Brentano quite rightly has nothing to do with this. Whether the rod before and after the scratch are one and the same body or two different bodies, the portion which came to be the whole rod clearly pre-existed the scratch, and is not brought into existence, but only into totality, by the removal of the small part. It is in itself unchanged by ceasing to be attached to the lost piece.

6 Thinkers The distinction between modo recto and modo obliquo thinking or, to use a different terminology, existence-entailing versus non-entailing slots in a predication, is certainly a useful one, whatever one’s views about the existence of relations as entities in their own right. However, as the Sherlock Homes example indicates, a relational predication need not have any existing object thought of modo recto. So the question arises as to what in the world is responsible for its being true, since neither Holmes nor Wallander is in the world. A modern approach would be to analyse the notion of being famous, somewhat as follows: A is more famous than B = (Def.) More people have heard about A than B The definiens, or as some would call it, the “logical form”, on the right-hand side, renders the ontological commitments and truth-conditions of the original predication more transparent. Clearly, in a singular predication such as: N has heard about A the subject term is taken modo recto and the object term modo obliquo. So the “definition” of “more famous than” does not entail the existence of either term. Nor of course does it exclude it, as in the truth: Barack Obama is more famous than Michael D. Higgins

How to Do Things with Things

� 11

It may be surprising, but Brentano would be quite happy with such an analysis. The only modification he might take would be to make the terms of the numerical comparison more patently nominal, as in: A-thinkers are more numerous than B-thinkers He would certainly not object to the use of numerical comparator quantifiers. In this case the predication apparently about two non-existing things is happily exposed as one which is a generalization about thinkers: happily, because “famous” is clearly mind-involving, and because the truth-conditions of the analysed sentence are preserved in its analysans. Brentano is however much more ready to bring minds into the picture than this, and takes a wide range of sentences which would appear to have nothing to do with minds to in fact be generalizations about minds and their objects, thought modo obliquo of course. For example, the modal proposition (regarding some thing or things A): A are impossible is interpreted by Brentano as: Whoever apodictically rejects A does so correctly which, by Brentano’s analysis of quantifier sentences, receives its canonical form: There are no apodictic rejecters of A who are incorrect rejecters of A where “A” is always modo obliquo but “rejecters of A” is modo recto. This is far less happy than the “famous” example. Firstly, it is far from being a sense-preserving paraphrase: the initial sentence said nothing about those who judge there are no A. Secondly, it has the wrong truth-conditions, since A might be possible but there simply happen to be no thinkers thinking about them at all, let alone apodictically and correctly rejecting them. Of course at this point Brentano could wheel in God, who thinks about everything and correctly accepts and rejects with evidence whatever is to be correctly accepted or rejected. This is a tempting albeit cheap expedient for all theists, and unless and until such a God’s existence be proven, to be avoided. Even logic is not immune to this invasion of minds. Brentano’s favoured reading for the principle of non-contradiction: Nothing is both A and not A is:

12 � Peter Simons It is impossible for someone who denies what another correctly accepts to do so correctly, or for someone who accepts what another correctly denies to do so correctly. Taking into account the above analysis of “impossible” and restoring the schematic variable “A”, this is probably best rendered as: It is impossible that there be both correct rejecters of A and correct accepters of A which by the analysis of “impossible” above comes out somewhat as: There are no apodictic rejecters of (both correct rejecters of A and correct accepters of A) who are incorrect rejecters of (both correct rejecters of A and correct accepters of A). It is hard to know where to start in listing the problems with this, and not especially enlightening or rewarding to do so. Suffice it to say that it is preferable to take the original simple form as the logical starting point and not look for a version taking a roundabout route via minds. In retrospect, it looks as though Husserl’s assessment of Brentano’s logic as psychologistic was right on the money.

7 Space Brentano for the most part rejects Newtonian absolute space, but this makes it important that he have a decent account of the relations among things in space, which he does not. In virtue of what, for example, is Dublin 6349 km from Addis Ababa (on the great circle)? Not in virtue of relations, since there are none, and not in virtue of their relationship to an autonomously existing space, since there is none. Perhaps the best account would invoke the actuality or possibility of a line or a body completely filling the gap between Dublin and Addis – there is actually one, the air (Luftlinie!) – but this pushes the question back to the question as to why the extremities of this tubular body are 6349 km apart, so we are no further forward. In the case of bodies separated by “empty space” such as the Earth and the Moon, there is no such body, so we would have to invoke possibility, and we would be back to thinkers. Brentano lays special stress on the notion of a boundary. His understanding is taken from Aristotle. A three-dimensional body such as a cube has a twodimensional surface consisting of six suitably joined squares: each of these has a linear square boundary consisting of four suitable joined straight lines of equal length, and non-opposite ones of these in turn meet in points. Brentano accepts

How to Do Things with Things �

13

that boundaries exist and are parts of the things they bound, but they are dependent on their bounded bodies in a subtle way. They cannot exist on their own. Were the whole body to be annihilated, the boundary would go with it, but the boundary could be annihilated (e.g. by friction) without the rest of the body disappearing. Conversely, the body could be pared away successively and still leave enough for some of the boundaries to bound a remnant. Boundaries exist not just at the extremities of a body but internally as well, as for example the disc between the Earth’s northern and southern hemispheres, or the equator, or the axis of rotation. Boundaries can also be of more than one thing at once, so unlike bodies can wholly coincide: the edge between two square faces of a cube is an edge of each face, but also of many other planar objects within the cube and coming just up to the edge. Like Leibniz, Brentano denies that continua like the edge, face or cube are made up of dimensionless points, but he accepts the points as boundaries of higher order, boundaries of boundaries of boundaries. Brentano’s theory of boundaries, to be found in part in the Kategorienlehre but in greater detail in the later, likewise posthumous Philosophische Untersuchungen zu Raum, Zeit und Kontinuum, is challenging but extremely rich in detail and insight.

8 Time Unlike modern physicists, Brentano strictly separates time from space. In the light of modern relativity theory – which he rejects – this is questionable position, though one he shared with many. More importantly, Brentano is what we now call a presentist. For him, to exist is to exist now. There are no things wholly in the past, and no things wholly in the future. It is incorrect to say Napoleon exists, but correct to say he existed. Brentano regards the non-present tenses as corresponding to particular modes of judgement. When I accept Napoleon in a pastward mode, I do so correctly; when I accept the house to be built on this site in a futureward mode, and the house does get built, I do so correctly. If no such house gets built, my futureward acceptance is incorrect. This is relatively familiar both from medieval and modern accounts of the truth-conditions of tensed sentences. It does leave Brentano with a problem about temporal distance or elapsed interval however, one which he shares with other presentists. It is true that the First World War broke out 99 years after the conclusion of the Napoleonic wars. So in virtue of what is this correct judgement about temporal distance (Abstand) true? It cannot be a relation, because there are none, and even if there were, the two terms do not coexist in the same present, so one or both of them could only be mentioned modo obliquo. Also the statement about their distance apart is, the tense of its ex-

14 � Peter Simons pression aside, invariant over time. It belongs to what in the terminology deriving from McTaggart are called B-statements. I can find no satisfactory answer to this question in Brentano, which is unsurprising, since it is the temporal analogue of the question about spatial distance, made more problematic by the fact that at least things a certain spatial distance apart exist at the same time, whereas things whose lives do not overlap never do. This problem also affects Brentano’s account of causation, since he takes causes always to precede their effects. An additional problem for Brentano’s view, as he recognises, is that the present is not a continuum but the boundary between the past and the future. As a boundary it ought to depend on them both, yet as non-present, neither of them exists. How can the existent depend on the non-existent? One aspect of a solution must be that the idea of the Present is an abstraction. What really exists is not a time or times, but things. When something such as a star or a river continues in existence over a period, then whenever it exists, except at the beginning and end perhaps, it used to exist for a while and it will exist for a while. So it as existing now is continuous with it as it used to exist and it as it will exist. There are not a plethora of things here, but one thing with a plethora of changing aspects, which are not themselves entities, if one will. So rather than the Present one really should speak of this present, this thing now. It may coexist with other things now, and on this basis one might build up a section across the universe and call that the Present. Whether this approach or something recognisably like it is workable is a difficult issue. I am inclined to think it cannot be made to work, even disregarding the complex relationship between space and time. But Brentano’s account of space and time is not the only one to face a barrage of difficulties, and only a genius or a fool would claim to have the last word in this tricky area.

9 Motion If space and time are both problematic for a reist, then all the more so is motion, which involves them both. Consider the usual example, a billiard ball rolling across the table. The ball exists before, throughout and after the motion, as does the table. The ball rolls, so that the orientation of its parts around the centre or a horizontal axis of rotation changes continuously, as does the area in which it is in contact with the table. The distances between the ball and its parts and other surrounding objects likewise change continuously. All of this can be studied by kinematics. Brentano is in no way disposed to challenge the mathematics of the situation, but what his ontological assessment of it amounts to, given his denial

How to Do Things with Things �

15

of relations and his insistence on the existence of only what is present, is hard to see. In a late and admittedly tentative piece dictated on 30 January 1915, placed by Kastil as an Appendix to the Kategorienlehre, Brentano looks with favour on a conjecture floated by Lord Kelvin that there be one large basic substance, a sort of all-encompassing homogeneous fluid, within which what we think of as bodies are temporary and mobile vortices or accidents, mutually impenetrable and obeying the laws of mechanics. This would replace corporeal substances as the non-mental basic substances by this one unitary substance, along with its parts and boundaries. The substance would not move, indeed the idea of its moving would lack sense, and what we think of as movement would be in fact the successive qualification of distinct parts of the substance by accidents resembling those recently in adjacent parts. This would not be bodily motion in the accepted sense, but something more akin to the progress of a wave through a fluid or of an image across a film or television screen. While not subscribing wholeheartedly to the picture, Brentano claims that because, unlike our transparent knowledge of ourselves, the true nature of the physical world is hidden from us, such a theory which may solve problems such as the apparent lack of an aether deserves serious consideration. It is rare indeed to find a 77-year-old exhibiting such flexibility of intellect.

10 Conclusion Brentano’s late ontology of reism is alike tantalising and frustrating, in that it bristles with novel insights and interesting alternatives to more familiar views while remaining incomplete and dubiously consistent. There are many aspects of his late philosophy and even of his late ontology which have been omitted or only grazed here, and I am very conscious of skating rapidly over much thin ice. Nevertheless I think it is fairly clear that reism in the form in which Brentano upholds it, a dualist, presentist reism of persistent things (if persistence and presentism are not themselves conflicting), is untenable in many regards. Whether a more adequate version of reism can be found is a moot point, not least because we cannot today be as sanguine as Brentano was about what we mean by a “thing”. Perhaps tropes, or fields, or some other sort of item can provide the sole category of furniture for the universe, in which case Brentano’s nominalistic and monocategorial instincts would be retrospectively justified, even if the details of his approach are rejected. Or perhaps not. Man wird sehen – vielleicht.

Federico Boccaccini

The Bounds of Object The Brentano-Meinong Dispute, A Priori Knowledge, and the Power of Perception¹ Federico Boccaccini: University of Liège, Belgium

1 Introduction It is commonly assumed that the disagreement between Brentano and Meinong revolves around the possibility of introducing several kinds of being into philosophical analysis, with the severe nominalist solution – proposed by the former in the final phase of his philosophical work – contrasting with the ontological pluralism offered by the latter. I shall attempt to show that this reading is not accurate, mainly because the disagreement between the two authors is of an epistemological rather than ontological nature. The question is how to define objects and what is given by analysis; and it involves perception, namely inner perception (innere Wahrnehmung). Unlike outer perception, which enables us to perceive physical phenomena through the senses, inner perception is the capacity to perceive non-sensible phenomena. Brentano calls such phenomena “intentional” or “mental phenomena”. For instance, a relation such as “being a part of” is a mental representation because, strictly speaking, there are no relations outside of the mind. Relations do exist as entia rationis, and for this reason they are objects of psychology rather than ontology. According to Brentano, it is in inner perception that we acknowledge

1 First, thanks are due to The Euraxess Postdoctoral Fellowship and the Department of Philosophy at University of Liège, where I have worked on my research project in excellent conditions and have had the benefit of incisive and useful comments on first draft of this paper, which was presented at the conference Objects and Pseudo-Objects. Ontological deserts and jungles from Meinong to Carnap held at Liège in May 2012. In particular I wish to thank Arnaud Dewalque, Bruno Leclercq, and Denis Seron. I feel very much in their debt. Thanks also to Sébastien Richard and Peter Simons for their valuable comments on my talk. I am also very grateful to Anna Marmodoro for the time I have spent drafting the final version of this paper at Corpus Christi College, Oxford, as Academic Visitor for the Trinity Term 2013. Finally, I am very grateful to Laura Nicola for having suggested numerous linguistic improvements to this text.

18 � Federico Boccaccini (anerkennen) and justify the truth of propositions and emotional stances. Since no impression corresponds to the concepts “right” and “correct”, Brentano argues that these are founded in inner experience. “Rightness” and “correctness” don’t strictly exist as objects in the world. Nevertheless, we are able to use these concepts, because we grasp the corresponding intuitions when experiencing emotions or when correctly inferring. If, as it has been supposed, Brentano’s notion of “immanent object” is a precursor of Meinongian ontology (Jacquette 1990/91), we first need to understand what an immanent object is. An incorrect interpretation would be that the immanent object is something in the mind, i.e. an idea of something that exists in the external world. According to Brentano, the immanent object is no more than an actualized power. This can be explained by means of an example. When I speak German, I demonstrate my knowledge of German grammar. When I speak English, I retain, potentially, my knowledge of German grammar, but this knowledge is immanent and unobservable. My ability to speak German manifests itself only when I speak German. But we should not assume the existence of two things here, potential knowledge and actual knowledge of German grammar – there is only one knowledge, which manifests itself in two different ways. My knowledge of German is the “inner object” of my act of speaking. Thus, we should not separate the immanent object from the mental act because, as we have seen, there is only one object, which manifests itself in two different ways. Brentano’s distinction between physical and intentional phenomena displays some analogous features. The idea is not that the world is divided in two separate ontological domains, say, matter and spirit; rather the notion is that the natural world manifests itself in two different ways. A red thing and my perception of red exist simultaneously as a whole and as two parts, one of which is a physical phenomenon, the other an intentional phenomenon. Outer perception perceives the former, while inner perception grasps the latter. Accordingly, it is only secondarily that the disagreement between Brentano and Meinong is of an ontological nature, namely insofar as all categorical distinctions – quality, quantity, relation, and the like – result from an analysis of the sense of being. And what has several “senses” can be grasped by our cognition. Thus, even if we allow that the category of relation plays a significant part in this debate, this does not mean that questions about the necessary existence (or subsistence) of the objective correlate of a mental act should come first. Indeed, during the first phase of his philosophy, Brentano conceives the mental relation as a real relation, a one-to-one correspondence between act and object, where both the act and its object are real. Yet, in an appendix to the second edition of Psychology from an Empirical Standpoint (1911), Brentano changes his view about the structure of mental relations, introducing a new definition of it as a

The Bounds of Object � 19

quasi-relation (or pseudo-relation), as “something relative” (etwas Relativliches). The reason for this is that we can also think of non-existent objects. Mental reference cannot be wholly ascribed to the category of relation. Quite obviously, the controversy about the alleged ontological commitment of intentionality – in the sense of Meinong’s theory of object – strongly influenced the change in Brentano’s viewpoint about mental reference. However, starting from this premise, this paper proposes an alternative interpretation of the Brentano–Meinong dispute. The problem is about perception and its limits. I argue that perception, and its power of actualisation, is the central issue.² Brentano and Meinong have, broadly speaking, two different notions of what philosophy is or should be: Brentano’s central question is: “What does it mean?” What does “being” mean? Likewise, what do “thinking”, “sensing”, “true”, “evident”, “good” mean? Searching for the primary sources of concepts that are empirically justified, including the categorical ones, Brentano characterizes such inquiry as “the problem of the origin of ideas”. In his view, philosophy is essentially conceptual analysis, i.e. descriptive psychology. However I do not mean to claim that his project for descriptive psychology is simply based on conceptual analysis, but that we can construe it as some sort of conceptual analysis, in the same way as we understand early analytical philosophers such as G.E. Moore, albeit with due distinctions. By contrast, Meinong’s questions are “What is that?” what is “difference”? What is “similarity”? What is “relation”? What is a “round square”? His “theory of object” can be viewed as a general answer to such questions. Brentano’s philosophy is a philosophy of objective sense and its main focus is how the contents of concepts are justified by inner or outer experience, Meinong’s main problem, on the other hand, seems to be the possibility of objective reference regardless of content experience. This difference is clearly expressed by Meinong’s concept of Apriorität (Meinong 1973 [1907]: §§7 and 12-18). Yet if Meinong aims at introducing special entities, he must (i) justify how we can grasp and know such objects; and (ii) clarify the notion of the given. As I shall demonstrate, the two philosophers conceive of the given in quite different ways. First, I shall provide an account of Brentano’s understanding of objects and their apprehension, his distinction between intuition and concept, and his idea of primacy of the given in intuition. I will then focus on what Meinong calls “foundation” (Fundierung) in order to justify his notion of given-ness, which is a central issue in his philosophy. I will not take into consideration all of Meinong’s work, and

2 This issue has its basis in the Aristotelian idea of common sense. This subject, as a central topic for understanding Brentano’s notion of inner perception and its power of knowledge, cannot be considered further here.

20 � Federico Boccaccini will confine myself to some key concepts. I shall attempt to explain why Brentano rejects Meinong’s approach as implausible, although both give an intentionalist account of knowledge.

2 Realism and Analysis Very early on, a debate began within the Brentanian School on how plurality can be apprehended as a unity. More specifically, the question was whether the apprehended unity constitutes a new object. For example, is a melody an object? Or is it just a combination of several notes? This issue was introduced by Stumpf, through his concept of tonal fusion (Verschmelzung) (Stumpf 1883). Another difficult issue was the so-called “problem of analysis”, to which Meinong contributed in An Essay Concerning the Theory of Psychic Analysis (1894) (Meinong 1969 [1894]).³ The conceptual root of this controversial question can be summarized as follows: do notions and concepts resulting from analysis describe objects and parts of them (realism), or do they merely arise from the method of enquiry (nominalism)? Brentano embraces the nominalist approach to the problem of analysis, whereas Meinong embraces realism. The reason for this is that, for Brentano, no theory could be logically true and phenomenologically (“psychologically” in Brentano’s terms) false. So, the first step in recognizing the content of perception is to understand what is given by the analysis of mental phenomenality. By “phenomenality” I do not mean “sensory”, but, following Brentano, what is justified by intuition. I mean whatever is present in the mind, i.e., mental phenomena, where “being in the mind” means being understood or apprehended.⁴ In other words, the Brentano-Meinong dispute is essentially about what Findlay called “the apprehension of objects” (Findlay 1995 [1963]). To say that the object exists prior to one’s analysis is to say that there is something mind-independent which is the object of analysis. Claiming that the object is posterior to one’s analysis is to say that we have an object only at the end of an analytical process, and that the object depends on the method of analysis. This is an anti-realist claim. Certainly, both philosophers would agree that the object is not a result of analysis (primacy of the method) and must pre-exist it; the method must only de-

3 For more on the problem of analysis, see (Fisette & Fréchette 2007), (Seron 2012a) and (Dewalque 2013). 4 The doctrine that what is known must be present to the mind can be found in Locke’s Essay. See (Locke 1979 [1690]: Book I, Chap. 2, §5, 50): “For if these words ‘to be in the understanding’ have any property, they signify to be understood”.

The Bounds of Object � 21

scribe it. From this point of view, Brentano and Meinong are both realists; they believe that the object comes first (the primacy of the object).⁵ The primacy of the method is a view from the Kantian tradition. In particular, Windelband and Rickert notoriously proposed to classify sciences by their “nomothetic” or “idiographic” method. This strategy involves a primacy of the theory of knowledge (Erkenntnislehre) over psychology, and some sharp distinction between natural sciences (Naturwissenschaften) and social sciences (Kulturwissenschaften or Geisteswissenschaften). In contrast, Brentano, in accordance with the Aristotelian tradition, proposes a classification according to the nature of the object studied. He argued that there are not two kinds of sciences, and that psychology is not a social science. Rather, the concept of science is univocal and refers to natural sciences. The social – or moral – sciences are called “sciences” only by analogy, because their laws are not exact, even if they can be generalised. As a consequence, the method of psychology must be that of natural science, “from the object’s point of view”. Brentano’s standpoint is quite similar to Aristotle’s, who claimed: “The phenomena show that nature is not a series of episodes, like a bad tragedy” (Metaphysics, 1090b 19⁶). Likewise each sentence constitutes some unity; but this unity is not really in the sentence. It is, rather, in the act of uttering or judging. Such a unity can be grasped only by inner perception, taken as a perception of mental activity and its different modes. Though, only ex post, analysis can separate the act from its object and acknowledge the difference between them. Brentano did not understand the word “empirical” as contemporary philosophers do, namely as synonymous with “from a third-person point of view”. Brentano believes in the epistemological primacy of the “first-person point of view”, claiming that only inner perception is immediate and hence can justify empirical laws.⁷ Thus, broadly speaking, both Brentano and Meinong are realists. For both of them, there is something outside the mind that cannot be altered by it. However, they have a different idea of what the object of analysis is. For Brentano, what Meinong calls “objects” are just notions arising from description, with no ontological commitment – like “difference”, “similarity”, “identity’, “necessity”, and the like. All in all, Brentano is a realist as to analysis, but a nominalist with regard to the analysandum: the thing to be analysed must be something individually determined, namely a substance. And the capacity of being represented is, and only is, a property of a substance and its accidents.

5 Psychology and the Neo-Kantian approach to science are discussed in (Stumpf 1891). 6 Quoted in (Brentano 1867b: 39). 7 On Brentano’s epistemology, see (Küng 1978) and (McAlister 2004).

22 � Federico Boccaccini When Meinong says that some entities are “beyond being and non-being”, he means they are matter of given-ness (Gegebenheit) rather than phenomenality (Meinong 1904: 500/1960: 92).⁸ That is, there is something that is known through thought rather than sense-perception, inner perception, or experience. From a minimal empirical standpoint, all possible objects of knowledge may be only mental or physical phenomena. Therefore, if Meinongian objects are given, but can never be phenomenally experienced, then they must be known a priori. In this way, Meinong turns well-grounded Brentanian empirical intentionality into a non-empirical account, by misunderstanding the notion of inner perception, and reading it as a special source of knowledge. The reason for this is that his concept of the given (das Gegebene) is more extensive than Brentano’s. Yet intentionality, according to Brentano, cannot justify any sort of a priori given. That is, the intentional, as a non-sensible sphere of experience, does not imply a priori experience. Not everything that can be named necessarily exists as an object. Nor, it would seem, can something not present in the mind be known.⁹ In other words, not everything is an object.¹⁰ Moreover, there is a further difference between an object of thought or discourse and an object of knowledge. Meinong’s claim is that “all that is knowable is given – namely, given to cognition” (Meinong 1904: 500/1960: 92). Whereas Meinong speaks of “ideal objects” – a notion which stems from the Brentanian notion of “immanent object”, used in the Brentano School to refer to the objective (gegenständliche) counterpart of a mental act – Brentano does not get caught up with the ontological aspect of the immanent doctrine. He does however give an – unfortunately, equivocal – account of what he means by the notion of “being in the mind objectively”,¹¹ which is obviously a very important topic. What Brentano means when he speaks of “intentional in-existence” is a relation of di-

8 “Alles Erkennbare ist gegeben – dem Erkennen nämlich. Und sofern alle Gegenstände erkennbar sind, kann ihnen ohne Ausnahme, mögen sie sein oder nicht sein, Gegebenheit als eine Art allgemeinster Eigenschaft nachgesagt werden”. 9 Concerning Brentano’s view, the source is, of course, Aristotle. Compare (Aristotle, De anima: 432a 3-9): “There is nothing outside and separate in existence from sensible spatial magnitudes, the objects of thought are in the sensible form, viz. both the abstract objects and all the states and affections of sensible things. Hence no one can learn or understand anything in the absence of sense, and when the mind is actively aware of anything it is necessarily aware of it along with an image; for images are like sensuous contents except in that they contain matter”. 10 Compare (Meinong 1978 [1921]: 12): “Was zunächst Gegenstand ist, formgerecht zu definieren, dazu fehlt es an genus wie an differentia; denn alles ist Gegenstand” (my italics). 11 About the alleged ontological commitment of Brentano’s definition of intentionality, see (Antonelli 2012).

The Bounds of Object �

23

rect acquaintance between the mind and the external world. In order to avoid idealism, this kind of acquaintance is not defined by some identity between act and object, but by unity, for example by the unity of the sense and the sensible. The concept of unity allows him to avoid Hegelian idealism and Humean scepticism. Indeed, if act and object were one and the same thing, there would be no distinction between mind and world, and if they were different, there would never be any certitude whatsoever. This point is important in order to understand Brentano’s view of intentionality and the unity of consciousness. However, it has often been said that the “immanent object” is a medium between the subject and the transcendent object. Likewise the inner object has been confused with the content (Inhalt) because Brentanian intentionalism is often conceived as a variety of representationalism. According to the standard representational theory – the view that the mind is only mediately in relation to the world¹² –, an idea arises when, from simple sensation, the external object is converted into a representation of x. A representation is a sign of the outer object. However, this is not what Brentano claimed about the mental: We have already explained what we mean by “presentation” (Vorstellung). We speak of a presentation whenever something appears to us. When we see something, a colour is presented; when we hear something, a sound; when we imagine something, a fantasy image. (Brentano 1874/1973: 198)

Brentano points out that presentation is a mental act, not a sign of something else. But, is the presented colour an object? Or does it exist as a modification of the subject? Is the perceived colour a sense-datum? If so, Brentano would embrace phenomenalism. But this interpretation does not agree with his strong realism. Brentano believed, in line with the Aristotelian tradition, in a formal unity between soul and body, mind and external world. Perceptual experience does not depend on the perceiver, but the perceiver’s experience is dependent on the unity between the perceiving and the perceived.

12 According to some philosophers, in sense-experience we do not directly observe objects or properties in the external world; the immediate objects of our experience are sense-data. Compare (Pitt 2012): “The term ‘Representational Theory of Mind’ is sometimes used almost interchangeably with ‘Computational Theory of Mind’, I will use it here to refer to any theory that postulates the existence of semantically evaluable mental objects, including philosophy’s stock in trade mentalia thoughts, concepts, percepts, ideas, impressions, notions, rules, schemas, images, phantasms, etc. as well as the various sorts of ‘subpersonal’ representations postulated by cognitive science”. I argue that in Brentano’s theory of mind there are no intermediaries between perceiver and perceived.

24 � Federico Boccaccini The point is that Brentano thinks of the unity of outer and inner experience as a complex structure, an idea that can be found in Aristotle.¹³ Yet, by this principle, Brentano introduces a new approach to the analysis of mental content. This idea is at the basis of Brentano’s thesis about unity versus simplicity of consciousness, which plays a central role in Psychology from an Empirical Standpoint.¹⁴ One thing can be composed of several parts, but its being unified does not make it simple; unification does not necessarily entail simplicity (Brentano 1874/1973: 155).¹⁵ It seems that the problem of analysis leads to the definition of an immanent object, because, in Brentano’s view, we do not perceive, in the strict sense, identity and difference between objects, but we are able to notice identity and difference through inner perception. Also, contrary to classical empiricism, these elements are not produced by reflection, but are grasped by analysis through inner perception.

3 Objective Reality or What is Present to the Mind Brentano’s controversial doctrine of immanent objectivity (immanente Gegenständlichkeit) was later transformed into an ontological theory about special kinds of entities. However, the doctrine of immanent objectivity can be read, as I shall read it here, as cognitive. The line between a cognitive and an ontic reading of immanent object is both thin and fragile. Some scholars have qualified this interpretation of intentionality as the correct one.¹⁶ Yet we should remind ourselves that by introducing the well-known passage about intentional in-existence, Brentano recalls the medieval psychological doctrine of the esse objectivum: They [viz. the medieval philosophers] also use the expression “to exist as an object (objectively) in something” [gegenständlich (objektive) in etwas sein], which, if we wanted to use it

13 On this important topic, compare (Marmodoro 2011). 14 See Book I, Chap. IV, “On the Unity of Consciousness”, in (Brentano 1874/1973). 15 “Our investigation has shown that wherever there is mental activity there is a certain multiplicity and complexity. Even in the simplest mental state a double object is immanently present. [...] But this lack of simplicity was not a lack of unity. The consciousness of the primary object and the consciousness of the secondary object are not each a distinct phenomenon but two aspects of one and the same unitary phenomenon”. The treatment of sameness and identity of substance through change, by quality and quantity, became a central subject of late Brentanian reflection, but this topic is not under examination here. Compare (Brentano 1933/1981). 16 Jacquette claims “the ontic status of intended objects” (Jacquette 2004a: 98). For an ontological interpretation of intentionality see also (Smith 1994) and (Chrudzimski 2001).

The Bounds of Object �

25

at the present time, would be considered, on the contrary, as a designation of a real existence outside the mind. At least this is what is suggested by the expression “to exist immanently as an object” [immanent gegenständlich sein], which is occasionally used in a similar sense, and in which the term “immanent” should obviously rule out the misunderstanding which is to be feared. (Brentano 1874/1973: 88)

In Psychology from an Empirical Standpoint, however, his main aim is to point out that intentionality is a special relation, namely a non-causal relation. Causality dominates the physical world; but the mental is, however, free from causality. Intentionality rules the psychical realm and this concept constitutes the basis of and the condition for descriptive psychology. And, there is no causation between things and ideas. The origin of ideas is purely intentional. But, what kind of relation is an intentional relation? It is a relation of inherence. Intentional relations are characterised by “inhabiting” (Brentano uses the theological term of “indwelling” [Einwohnung] or inhabitatio) (Brentano 1874/1973: 89).¹⁷ The object inhabits the mind in the form of an idea, the idea inhabits the judgement, and both inhabit the feeling. Such “inhabiting” is nothing but presence. In the act of thought, the object is present in the mind; the thought “becomes” the object.¹⁸ Brentano translates a theological into a psychological vocabulary by means of some semantic shift, in order to describe the mental. The point is that Brentano, like Aristotle and Aquinas, does not conceive the mind as a sort of box in which we can find contents, but as a form, i.e., as an act.¹⁹ In order to understand his view,

17 “Wenn die Schrift von einer Einwohnung des hl. Geistes spricht, so erklärt er diese al seine intentionale Einwohnung durch die Liebe. Und in der intentionalen Inexistenz beim Denken und Lieben sucht er auch für das Geheimnis der Trinität und den Hervorgang des Wortes und Geistes ad intra eine gewisse Analogie zu finden”. 18 Compare (Aristotle, De anima: 414a 25-28): “The actuality of any given thing can only be realized in what is already potentiality that thing, i.e. in a matter of its own appropriate to it”; 429a 13-18: “If thinking is like perceiving, it must be either a process in which the soul is acted upon by what is capable of being thought [...]. The thinking part of the soul must therefore be, while impassible, capable of receiving the form of an object; that is, must be potentially identical in character with its object without being the object. Thought must be related to what is thinkable, as sense is to what is sensible”; 431b 17-18: “The mind which is actively thinking is the object which it thinks”. O. Kraus, defining the notion of phenomenon in Brentano, recalls that the concept of activity must be understood in the sense of passio, “an affection in the Aristotelian sense”. Compare (Brentano 1874/1973: 79). 19 In fact, according to Brentano, the soul is not a simple substance, but a substantial form of body. Compare (Brentano 1867b/1977: 55). For a definition, see also (Aristote, De anima: 412a 27; 2, 414a 4; Book Z 10, 1035b 14; H 6, 1045b 11). The soul as a substantial form is a thomistic definition. Brentano is aware of the difference between the definition of the soul as substance and the definition of the soul as a form or act. The former is attributed to Albert the Great, the lat-

26 � Federico Boccaccini we must refer to the first formulation of the immanence theory; it is found in his commentary to De Anima, where, in order to explain the power to receive sensible forms without matter, he uses the term “objective” for the sensory phenomena (aestheta) that occur in the act of sensation. It would be difficult to overestimate the significance of this passage for understanding the whole Brentanian conception of intentionality: We do not sense the cold insofar as we become cold; otherwise plants and inorganic bodies would also sense. Rather, we sense the cold insofar as the cold exists objectively, i.e., as cognized object within us [...]. Here and in the following we do not use the expression “objective” in the sense customary in recent times, but in the sense usually connected with the word by the medieval Aristotelians (the scholastic term obiective). It allows a brief and precise characterization of the Aristotelian doctrine. Materially, as physical quality, coldness is in the cold thing. As object, i.e., as something that is sensed, it is in him who feels the cold. (Brentano 1867b: 80)²⁰

When we feel cold, Brentano explains, it isn’t because our body becomes cold. Rather, the verbs “to feel” or “to have sensations” mean that cold exists “objectively” within us as cognized object (als Erkanntes). This kind of existence is purely psychological; what we have in us is just the form of cold. For that reason, Brentano specifies the nature of such an “object” as what is known: feeling cold means that cold is present to the mind as a form. Brentano refers to the Aristotelian concept of aistheton kat’energeian: sensed object as actuality. Such actualization is a power of the soul, and it can be found in the capacity of perception. Thus Brentano seems to use an ontic language when he talks about the way such objects exist in the mind. In fact, Brentano is echoing older doctrines of an Aristotelian and scholastic form which tried to find the way that features of perceptual objects, e.g. their form, existed in the mind of the perceiver. Later, the ontic being was transformed by Descartes and Locke into a psychological reality in order to explain cognition. For the Moderns, having an idea meant having a sign or meaning of a thing, which is sometimes an image. There is something be-

ter to Aquinas. Compare (Brentano 1980: 45): “Das Wesen der menschlichen Seele [ist] Substanz (Albert). Die Seele [ist] Form, auch die menschliche Seele (Thomas)”. 20 “Allein nicht insofern wir kalt werden, empfinden wir das Kalte, sonst würden auch Pflanzen und unorganische Körper empfinden, sondern insofern das Kalte objectiv, d.h. als Erkanntes in uns existiert. [...] Wir gebrauchen den Ausdruck ‘objectiv’ hier und im Folgenden nicht in dem Sinne, der in neuer Zeit der übliche ist, sondern in jenem, den die Aristoteliker des Mittelalters damit (mit dem scholastischen objective) zu verbinden pflegten, und der eine sehr kurze und präcise Bezeichnung der Aristotelischen Lehre ermöglicht. Materiell, als physische Beschaffenheit, ist die Kälte in dem Kalten; als Objekt, d.h. als Empfundenes, ist sie in dem Kältefühlenden”. (my italics)

The Bounds of Object

� 27

tween the thing and our mind that represents the thing outside of us. The word “form” acquired some psychological meaning. But Brentano proposes something new: being intentional is simply to be presented as a manifestation of something. He places emphasis upon phenomenality, so that by ideas or representation he means acts of the mind, and, if my interpretation is correct, a mental act is nothing but an act of presence in the mind. Thinking an apple just means seeing an apple, visually or by imagination. There is no “thought apple” in the mind, and there is no room for an ontology of the intended object. The act of thinking an object and the manifestation of the object are one and the same; in other words, the form is the object. There is no meaning or representation of cold in our mind; thinking cold means feeling cold. And, likewise, thinking red means seeing a red thing. This is the intended sense of the theory of the unity of the act. There is no distinction between content and object in Brentano because his theory of mind is not concerned with mental contents, “mental states” or “mental representations”. His psychology is not a description of what happens in our mind; mental acts are the subject-matter of Brentanian psychology. A serious problem arises when we think of something like “good”, “true”, “being” or “substance”, or about something temporally determined in the past or future. What is present in the mind in this case? Brentano uses intentionality in order to justify the constitution of such non-empirical concepts. Yet the intentional matter of a concept can be separated from its intuitive moment only by analysis. A concept not based upon intuition is just an empty concept because, unlike in Kant, the origin of concepts is not logical but phenomenological. However, this procedure does not introduce new entities. Its function is to explain how we possess some ideas which are not of sensory origin, without appealing to Innatism. Finally, Brentano carries out an inquiry into the foundations of thought by experience. Nevertheless, in the case of the mental, we have only an inner experience (Erlebnis). For instance – Brentano is very clear on this – the concept of will is founded on the concept of love, but we do not possess the latter just by definition; we need to have an experience of it: When we say that the will can be defined by adding specifications of this kind to the general concept of love, we do not mean that someone who has never himself experienced the specific phenomenon could attain complete clarity concerning it merely from the definition. This is by no means the case. (Brentano 1874/1973: 243)

As I said above, the point is that an intentional relation is not a relation of identity between act and object (esse est percipi), but a relation of unity. Mind and world are unified formally by mental acts, but they are not the same thing. This means that we can think of sense-perception as a complex structure. In this way,

28 � Federico Boccaccini Brentano can both identify the mental and physical fields and avoid scepticism about the external world: the inner and outer realms are in contact through form. The task of analysis is to show what is already given in experience although only in a confusing and hybrid manner. So, analysis is used to clarify and describe. No entity could be ontologically produced by analysis; its task is merely epistemic.

4 Apprehension of Objects, Apprehension of Concepts In the later years of his life, Brentano went further into his phenomenological research by focusing his attention on the traditional problems of metaphysics. He devoted the last ten years of his life to investigating the principle of individuation, the concepts of substance and accident, the theory of categories, and space and time, which he developed into a form of reism. Brentano’s reism is often misunderstood. The reist thesis concerns thought and its objects. According to his later philosophy, the objects of thought can only be entia realia, i.e. individuals. We may summarize the later Brentano’s view in this way: no entity, without individuality. Some philosophers confound the concept of the individual with the concept of the actual thing. This is a mistake, because the class of individuals is larger than the class of actual things. For example, the centaur Chiron is an individual but not an actual entity. Individual and actual beings are conceptually distinct. Referring to Brentano, Chisholm explains this point as follows: Prior to 1905, Brentano had held that there are two types of entity – entia realia and entia irrealia. Dogs, red things, squares, unicorns, and mermaids would be entia realia; and privations, possibilities, mental entities, concepts, properties, states of affaires, and propositions would be entia irrealia. As the examples suggest, it would not be correct to translate ens reale as “real thing” or “actual thing”. [...] The best translation of ens reale is “individual”, and the best translation of ens irreale is “nonindividual”. [...] After 1905, Brentano held that, strictly speaking, the only entities that we are capable of thinking of are individuals. (Chisholm 1986: 9-10)

After 1905, Brentano rejected all entities of reason, or irrealia. This turn would involve separation from, and rough criticism towards his students, especially Meinong and Husserl, who had developed the ideas of their teacher in Vienna. There are only two kinds of objects of thought: concepts and individuals. But, strictly speaking, only individuals exist. Now we are in a position to ask how individuals are apprehended. The answer is: by intuition. Aristotle’s doctrine of

The Bounds of Object �

29

individuation is connected to the idea that the cognitive process of abstraction allows us to grasp concepts. Following Aristotle, Brentano makes a distinction between sensing and knowing: we sense individuals but know universals.²¹ But to be known, to him, simply means “to be presented in the mind”. Paraphrasing Aristotle, Brentano claims that we perceive individuals but think universals. In my opinion, Brentano’s strong realism encourages acceptance of the mind–world relation as an epistemic relation based on the nature of objects, refusing the semantic relation in terms of meaning, because signs have nothing in common with reality. However the risk is to fall into a sceptical account of human knowledge. For Brentano, a presentation does not resemble its object. If that were the case, we could not distinguish his position from Twardowski’s, for instance. Twardowski’s Bildtheorie is a coherent interpretation of the process of cognition of an object by “resemblance”. In Twardowski’s view, to be in the mind means to be the image of an object. This is the use of Vorstellung in Twardowski’s psychological terms. While writing On the Content and Object of Presentations (1894/1977), he thought that he was writing in accord with his teacher. In this way, intentionality and external reference can be explained in terms of the cognitive process of mental representation. A presentation is something between the mind and the external object and this “something” is an inner image. Unfortunately, if we allow this, we must accept psychologism. This is possible, and some contemporary representational theories of mind are openly psychologist. But this is not the case for Brentano. Like Husserl, he rejects psychologism.²² Knowledge is not a psychological subjective fact; but it is not outside the mental. Maybe we can better explain this by using Brentano’s distinction between genetic and descriptive psychology. By “genetic psychology” he means the study of neuro-psychological processes and the causal relation between psychological and physical states. By contrast, by “descriptive psychology” he means philosophy of mind, in the sense of a philosophical analysis of mental life. We can say that knowledge is not connected with genetic aspects of thought; as a mental act, however, it concerns descriptive psychology.²³ Nevertheless, what we can conceive, and the limitation to which that conception is submitted, is unrelated to the nature and process of our own faculty of conceiving. Vorstellung, in Brentano, is not an image of an object, nor is it the object itself. Twardowski’s empiricism is closer to British empiricism than

21 Compare (Aristotle, De Anima: 417b 21): “What actual sensation apprehends is individuals, while what knowledge apprehends is universal”. 22 Compare (Boccaccini forthcoming). 23 For the distinction between genetic and descriptive psychology, see (Brentano 1895). For more on the idea that knowledge is a state of mind, see (Williamson 2000).

30 � Federico Boccaccini Brentano’s. Brentano’s conception of empiricism is an interpretation of classical empiricism with constant reference to Aristotle. The knowledge of the world is not only empirical, but also metaphysical. Ontology plays a large part in his philosophy. Debating the ontology of perception is not the purpose of the present paper; but I think that perception remains, for Brentano, a psychological issue – a problem on the side of mental acts. Only in conceiving ideas and perceptions as objects would it be possible to elaborate an ontology of perception. But for Brentano this is not the case. Perceptions, ideas, and sensations are physical phenomena which are outside the competence of psychological analysis. The very object of psychology is the act of thought, the perceiving, not the perceptions. Having clarified these points, we can better understand the role that abstraction plays in Brentano’s work. In Metaphysik 12 (M 12), the second part of Psychology III, Brentano discusses the distinction between sensitive and noetic objects. He separates positions on universals into two groups: on one side are the ultrarealists, who are divided into two further subcategories – Platonists and medieval realists. The former argue that existence of universals is ante rem; universals exist as entities detached from matter and thought. The latter hold that universals exist ante rem but as logical parts of individual things. On the other side Brentano places nominalists. They deny the ante rem or in re existence of universals, allowing only general ideas post rem. Among nominalists there are those who, like Berkeley, deny even the possibility of thinking in a general way, claiming that there is no kind of general representation, neither in the mind nor out of the mind. Universals are just fictive objects generated by the use of language. On the other hand, Brentano defends universals as a moderate conceptualist. Against strict nominalism, he affirms that we always think in a general way. Universal judgements are not possible without universal presentations, and the form of judgement is a proof of that. But the universal is not conceived as a general or collective name, as it is for Locke, for instance, because in this case it would not possible for a single act of induction to form a general concept. Brentano accepts the Kantian distinction between concepts and intuitions, as well as between general representations and individual representations, but with some differentiations. How do we grasp individual presentations or intuitions? The concept of colour has specific differences – like, for example, red and blue. Abstraction seems to be very similar to sensation, but it is simplified in its content. Indeed, both belong to the domain of the representative life of the mind. When we have a complex concept of a warm, red thing, warm with a pleasant sound, like a fire burning in a fireplace, then we have a complex presentation. In this case, the singular presentation of a red thing, of a warm thing, etc., are all unified by an act of synthesis, but this act is not an act of judgement. When I say “one red warm thing”, I do not affirm the same content as when I say “a red thing is warm”. In the first case, we have a representational

The Bounds of Object �

31

synthesis, in the second a predicative synthesis. From a logical point of view nothing is changed. Both sentences can be translated into a proposition of this form: There exists a red thing and this red thing is warm. But from a psychological point of view we have a semantic shift. The representational synthesis is expressed in the form of a complex presentation (SP), while predicative synthesis is a double judgement ( can be analysed in and ). Two individual presentations can represent the same thing with a different content because they do not have any ultimate individual determination. Through an act of identification, the objects of two distinct intuitive presentations can be one. This operation is obtained through an attributive combination of presentations. The act of presenting has two modes: modus rectus and modus obliquus. Brentano’s psychological analysis rules out the possibility of a priori conceptual presentations. Concepts are built by sensibility and there is no place in Brentano’s theory of knowledge for either Kantian pure concepts or Meinongian pure objects. He also rejected the Kantian idea of time and space as pure forms of intuition. For Brentano, concepts and intuitions are the results of our cognitional and discriminatory capacity, not of the conditions of experience.

5 The Given as an Object In the famous §2 of his Theory of Objects, Meinong talks about the object and limits of metaphysics as a science of existing things and he condemns this conception as “the prejudice in favour of the actual”. For his part, the Austrian philosopher distinguishes between three senses of the verb “to be”: to subsist (bestehen), to exist (existieren), and to be real (wirklich). Consequently, we should admit that there are, at least, three separate realms of given-ness. Similarity and difference, for instance, are objects. In this case, if we were to build a science of all kinds of “objects”, then we would need a science more general than metaphysics, namely ontology, the science of the object in general. The question concerns the proper place of scientific investigation of the Object [Gegenstand] taken as such and in general. (Meinong 1904: 481/1960: 77-78)²⁴

24 In fact, in contemporary metaphysical analysis, the standard distinction between ontology and metaphysics is the exact opposite of Meinong’s. Generally, ontology is the study of what there is and metaphysics is the study of what it is. For instance, in Brentano’s later philosophy we have a monistic ontology (there are only things [Dinge]), but a dualistic metaphysics (there are two kinds of things: bodies and souls). So, in contemporary terms, Brentano argues for the

32 � Federico Boccaccini Thus, the ultimate problem seems to concern the relation between metaphysics and ontology. If we claim metaphysics to be a science of what is real, and ontology to be the science of being qua being and of its properties, we can recognize the later Brentanian position as a severe reduction of ontology to metaphysics. The later Brentano’s thesis about being can be summarized in this way: the being qua being is, and only is, the being as individual substance. In this case, unlike what is argued by Meinong, similarity and difference are only relations, and a relation is definitely an ens rationis. As such, they should be properly referred to as “objects”. This, according to Brentano, is an improper use of language. By contrast, Meinong broadens the concept of being, establishing metaphysics as a part of ontology: the actual is just a part of the possible. From this point of view, ontology comes first. Accordingly, Meinong considers that “something” without determinations is a possible object of thought.²⁵ Thus, the issue about non-determined objects concerns the link between ontology and logic, not semantics. For Brentano, on the other hand, what it is indeterminate cannot be thought at all. Meinong “releases” objects from the constraint of being, defining the pure object as außerseiend (Meinong 1904: §4, 490-494/1960: 83-86). This view originates in the concept of “ideal object” (idealer Gegenstand) introduced in 1899 (Meinong 1971 [1899]: 377/1978: 137). This suggests that object theory conforms to some rationalistic Wolffian conception of general object (Gegenstand überhaupt) as an object of metaphysica generalis (Pichler 1910). Yet Brentano argues against the Meinongian theory of a priori knowledge, claiming that all experience depends on the intuition of an object as given, i.e. as a phenomenon. If an object exists, we can have an experience of it and – consequently – knowledge of it. We cannot properly speak of knowing something if it is not an object in the proper sense. Only what there is can be truly known. For Brentano, like Kant, thinking and knowing are two separate acts. By contrast, Meinong states that “there is thus not the slightest doubt that what is supposed to be the Object of knowledge [Gegenstand des Erkennens] need not exist at all” (Meinong 1904: 489/1960: 81). In this statement there is a clear and decisive epistemological commitment, and not only an ontological or logical one. According to Meinong, “all that is knowable is given – namely, given to cognition” (Meinong 1904: 500/1960: 92). The theory of object is based on the principle that it is not

primacy of ontology. By contrast, Meinong argues that “what it is” must be prior to “what there is”. Among contemporary contributors to metaphysics, some carry on in a Meinongian vein, e.g. (Bergmann 1967), (Johansson 1989), (Grossmann 1992) and (Chisholm 1996). In this text I adopt the difference between metaphysics and ontology in accordance with the Meinongian distinction. 25 Compare (Meinong 1968 [1917]: 22). Compare also (Chisholm 1982a).

The Bounds of Object � 33

possible for something to be given yet impossible to know. Thus, object theory, cognition, and theory of knowledge are strictly connected. But what is given a priori? An analogy between the empirical a posteriori sciences and the theory of object can be drawn, due to the fact that “both start from given data” (Marek 2008), but it sounds rather strange to call “data” something that is given outside being. We can say that, for Meinong, mathematical entities, for instance, are given as objects. In contrast, Brentano never thinks of such entities as objects. In his conceptualist phase, such “entities” are not objects but merely concepts; while in his later phase they are just nouns.²⁶ Of course, Meinong does not say that there are golden mountains and round squares. What he says is that, since [...] we can think about these objects and say various true things about them, they have certain characteristics even though they cannot be said to be. They have a Sosein even though they haven’t any Sein. (Chisholm 1973: 39)

This is the famous principle of independence articulated by Ernest Mally: “Sosein is independent of Sein” (Mally 1904: 126-127). Usually, the notion of Außersein, which refers to the property of being beyond being and non-being, is considered as the focus of the theory of objects.²⁷ However, we should keep in mind that the ambition of this theory is to be a science (Wissen). Meinong himself talks about a kind of “knowledge of what is not actual” (Es gibt eben auch Wissen von Nichtwirklichem): But any one who knows things to be identical or different certainly knows something about these things; yet his knowledge however, also concerns identity and/or difference, and identity itself is as far from being a thing as is difference. Both identity and difference stand outside of the disjunction between the physical and the psychological, since they stand beyond the real [außerhalb des Realen]. There is also knowledge of what is not actual [von

26 (Brentano 1974 [1930]/1966: 26): “Being in the sense of the true”: “In thinking about things, our mind forms various concepts of which a part are fictions, to which nothing corresponds – for example, when we form the concept of a golden mountain or, even more to the point, the concept of a wooden flat-iron. There is no golden mountain, and it is completely impossible that there be such a thing as a wooden flat-iron. But to another part of these concepts of being golden, or of a mountain, or of being wood, or a flat-iron, we may say that there actually are mountains and flat-irons and things which are golden and things which are wood. [...] There cannot be anyone who contemplates an A unless there is a contemplated A; and conversely. But we must not infer from this fact that the one who is thinking about the A is identical with the A which he is thinking about. The two concepts are not identical, but they are correlative”. (my italics) 27 See (Findlay 1995 [1963]: 42-58), (Lindenfeld 1980: 148-158) and (Simons 1986). For a discussion of Meinong’s epistemology, see (Modenato 1995; and 2006) and (Schubert Kalsi 1987).

34 � Federico Boccaccini Nichtwirklichem]. No matter how generally the problems of metaphysics are construed, there are questions which are even more general; these questions, unlike those of metaphysics, are not oriented exclusively toward reality. The questions of theory of Objects are of this kind. (Meinong 1904: 517/1960: 107)

The issue about this sort of a priori knowledge arises in three main stages. First, Meinong sketches an account of it in On Objects of Higher Order and Their Relationship to Internal Perception (Meinong 1971 [1899]: 377-480/1978: 137-208), prompted by a critical article by Friedrich Schumann arguing against Ehrenfels’ theory of qualities of form (Gestaltqualitäten) (Schumann 1898: 106-148). Then, he openly discusses this issue in The Theory of Object, especially in §11, entitled “Philosophy and the Theory of Object”. Finally, he advocates a priori knowledge in On the Experiential Foundations of Our Knowledge (Meinong 1973 [1906]) and The Position of the Theory of Objects in the System of Sciences (Meinong 1973 [1907]). We can say that the problem of a priori justification about non-empirical objects is the leitmotiv of all Meinong’s work, from the analysis of relations in Hume to that of moral sentiment. However, my aim is not to describe the whole development of this issue, but to discuss its main arguments.

6 Transcending the Field of Perception: The Foundation Meinong introduces his general ideas about the perception of ideal objects in his article “On Objects of Higher Order and Their Relationship to Internal Perception” (§7, Objects of experience and founded objects): “Founding or the Foundation [Fundierung] process achieves the same for ideas of ideal objects as perception does for ideas of real objects” (Meinong 1971 [1899]: 397/1978: 152). The concept of Fundierung is central. Perception and foundation are two different forms of apprehension. Meinong used the term “foundation” in Hume-Studien II to designate relational members. But in this essay, “foundation” specifically denotes the apprehension of the necessary relationship between the inferiora and their superius. The distinction between perception and foundation is based on the division between real and ideal. What, then, is an ideal object? Deficiency [Mangel], limit, past things, etc., are traditional examples for the non-real, that is, the ideal. [...] Let us take the similarity between copy and original: both of the pictures exist. But to attribute existence to similarity over and above attributing it to the pictures would seem like an act of force to any open-minded person. Still, a correct affirmation might be made concerning the similarity in the given case. For we presuppose that the similarity of

The Bounds of Object � 35

the two pictures cannot be denied. The similarity does not exist but does subsist. And things which, by their nature, can well subsist but, strictly speaking, cannot exist are precisely what is contrasted, here, as the ideal and real. (Meinong 1971 [1899]: 394-395/1978: 150)

All relations and complexions are cases of ideal objects (e.g. the difference between green and red is a relation, while the difference is the Relat; the contemporaneity of the sound of two notes is a relation, the melody is a complexion) (Meinong 1910: 279-283/1983: 201-204). They have the property of subsistence, but not the property of existence. Meinong’s argument for foundation is very simple yet effective: Everything real, if it exists, must, under favourable conditions, be able to be perceived by the cognizing and sufficiently capable subject. That is a matter of course. If we say that the real is characterized by its capability to be perceived, that it is naturally perceptible, we may seem to be making an insignificant statement. However, things are different when we realize that there is a whole class of objects which essentially are not perceptible. They are the ideal objects. (Meinong 1971 [1899]: 396-397/1978: 151)

What Meinong wants to show is that in the sense of givenness there are properties that are neither physical nor mental, but ideal. If we call perceptible objects “real” and non-perceptible ones “ideal”, then a property like unity is not real: it is an ideal property of the object it pertains to. But how can we realize that there is a class of ideal objects? The answer lies in the form of judgement. Brentano had introduced the idea that judgement is not an act of adæquatio rei, i.e., that there is no already given object – or objectity – to which thought should somehow be adequate. In the Theory of Object, Meinong takes a further step. In §11 he makes the distinction between his theory and metaphysics as an empirical science. According to Meinong, there are two kinds of phenomena, physical and psychological ones, but there are some objects that are neither physical nor mental. These constitute the field of pure possibility as objective and subsistent: It is obvious that psychology is concerned only with real psychological events and not with the merely possible. A science of knowledge cannot set similar limitations on itself, because knowledge as such has value, and therefore something which is not but could be may draw attention to itself as a desideratum for knowledge. Accordingly, not only are pseudo-objects in general, and hence all objects which are actually judged or presented, to be included as Objects of our scientific knowledge [Wissen], but also all Objects which are Objects of our cognition only in possibility. However, there is no Object which could not at least in possibility be an Object of cognition. (Meinong 1904: 499-500/1960: 91-92)

The ideal object presents itself simply as a non-empirical given. The capacity to apprehend this kind of given-ness is at issue here. Also, opposing the possible to

36 � Federico Boccaccini the real, Meinong here overlooks an important remark that Brentano made about the use of such concepts. The young Brentano observed that: There is a great difference between what we here mean by the potential (the dynaton or dynamei on) and what in more recent times is meant by calling something possible in contrast with real, where the necessary is added as a third thing. This is a possibility which completely abstracts from the reality of that which is called possible, and merely claims that something could exist if its existence did not involve contradiction. It does not exist in thing but in the objective concepts of the thinking mind; it is a merely rational thing. (Brentano 1975:27-28, my italics)

By “more recent times”, Brentano surely refers to Kant. In the Critique of Pure Reason, Kant had indeed unified, under the category of modality, three different concepts: existence, possibility, and necessity. Brentano’s critique concerns the distinction between logics and metaphysics: possibility is a logical concept; potentiality is a real concept. Hence possibility is in the mind and potentiality is in the physical object. For Brentano, metaphysics is only possible as metaphysics of experience. Thus, despite the fact that Meinongian apriorism concerns ideal objects rather than our faculty of cognition, Meinong’s claim that “all objects which are objects of our cognition only in possibility are objects of scientific knowledge” is closer to Kantian apriorism rather than Brentano’s empiricism.

7 Concluding Remarks According to Brentano’s later view, while (i) we can think of several concepts of object, (ii) only what is an individual (ein Seiendes, ein Reales) exists stricto sensu and can have properties (Brentano 1874/1973: 311-314). Consequently, we are in an epistemic state only when we think of what there is. On the other hand, Meinong argues that there are (es gibt) some objects that can be the subjects of true predications or evaluations – objects of cognitive or moral mental acts. Even inexistent or impossible entities may have properties and stand in relationship to each other. So, in this case, the concept of knowledge also admits non-individual and impossible entities. Both Brentano and Meinong agree with the intentional principle that “knowing is impossible without something being known” (Meinong 1904: 394/1960: 76).²⁸ The hotly-debated point concerns the concept of “something” (Etwas). This debate developed into a rift between pupil and master, which could not be healed. Probably the origin of this split was Meinong’s conception of ob-

28 “Daß man nicht erkennen kann, ohne etwas zu erkennen”.

The Bounds of Object � 37

ject theory in his Über Gegenstandstheorie, where he presents it as the most general and fundamental science. As suggested by Marek, “according to Meinong, its range is wider than metaphysics because metaphysics deals only with existing things, with the entirety of the real” (Marek 2008). In order to distance himself from this new outlook, which was apparently connected with his earlier standpoint, Brentano goes back to work on the theory of categories, grammar, and logic; in particular, he works on the difficult status of entia rationis between psychology and metaphysics.²⁹ As to the sense of “being”, he opts for univocity in the wake of the Scotist tradition. His purpose is to speak of the meanings of “being” rather than of the genuine kinds of it – an issue already raised in his very first work, where he adopted a Thomistic solution in terms of analogy. Starting from Aristotle’s claim in (Metaphysics: Book IV 1003b) that “being is said in many ways” (to on legetai pollakôs), Brentano assumes that the question is semantic rather than ontological. To put it more simply, the problem is that a term can obviously have a meaning yet no reference. Yet, as I have tried to show, the dispute also involves the role and cognitive power of perception and the limits of a priori knowledge. In general, the problem arises with categorical cognition and the role of intuition. Brentano claims that categories are forms of apprehended real objects, or the objects about which we think; they are forms of being, not of our thought about being. Similarity, identity, difference, unity, possibility, existence and non-existence, are all just thoughts which, although given in the things themselves through perception and attention, should not be considered genuine “objects”. Only the concept of being correctly intuited by inner perception gives us the bounds of what can be called an “object”. On the other hand, according to Meinong, there is a special kind of cognition that transcends the field of perception: this is the “foundation”. In accepting Meinong’s claim, we have to accept another concept of the given that involves knowledge through a particular kind of intuition. This is, as I have tried to suggest, the crucial difference between Brentano and Meinong.

29 See (Boccaccini 2010).

Anna C. Zielinska

Kotarbiński’s Strong Minimalist Ontology Anna C. Zielinska: Max Planck Institute for Comparative Public Law and International Law, Heidelberg, Germany

1 Introduction There is no reason to think that the world contains anything other than things, res. Decent ontology, according to Tadeusz Kotarbiński, does not need to postulate qualities, events, or relations, as he notes in his Elementy (translated as Gnosiology in 1966), where he proposes the first extensive outline of his “reism” (1929). This idea might strike one either as trivially true, or as trivially false, or at least unhelpful. Whether it is true or false is not the subject of the current paper, however; here I will only try to show that this is neither trivial nor unhelpful. Kotarbiński was a member of the Lvov-Warsaw School, and was one of a few fully-fledged philosophers working among logicians and philosophers of science. His first writings on moral philosophy were criticisms of classical figures of utilitarianism. In his doctoral dissertation, Kotarbiński remarked that, before thinking of any moral philosophy, there was a need for a new discipline, dealing with the foundation of general axiology and of general practice, that would make it possible to achieve something more “solid” within this area (Kotarbiński 1914: 83). What thus becomes necessary is less a theory and more a grammar of action. This first project (later developed under the name of praxeology) already shows that Kotarbiński’s approach was modest in its extent, for him a better understanding of the terms used in existing theories was the essential step towards any new philosophy. This modesty also characterizes his ontology, a doctrine that he developed almost en passant as an application of Stanisław Leśniewski’s formal system (also called “Ontology”, but misleadingly so¹). Mainly because of the criticisms it encountered, it became a coherent theory and at the same time one of the most interesting metaphysical positions of his time. This paper tries to explain Kotarbiński’s ontological worries. It then sketches the genesis and the development of his metaphysical position, and finally adds two contextual elements: a brief de-

1 Leśniewski’s ontology is a formal calculus of names.

40 � Anna C. Zielinska scription of his conception of truth and his unaccomplished project of linguistic reform, one he shared with an Austrian precursor of reism, Franz Brentano.

� Reism �.� Early Notes Kotarbiński was one of the very few early analytical philosophers eager to defend a metaphysical position. Indeed, at this period, Wittgenstein remained silent (or whistled), the Vienna Circle notoriously rejected metaphysical considerations and Bertrand Russell concentrated on mathematics. Meanwhile, Kotarbiński, as Roger Pouivet notes, was the first “analytical metaphysician”, who would “turn to logic in the treatment of fundamental metaphysical questions” (Pouivet 2009: 237). In a paper presenting his early metaphysical positions, written against the putative existence of ideal objects conceived as entia rationis, Kotarbiński devotes a few revealing lines to the question of whether Aristotle takes the notion of secondary substances at face value, in which case he would be a faithful follower of Plato, or whether he conceives them as simple metaphors, which would make him a “reist” (Kotarbiński 1920). The whole article is both a reflexion, sometimes amusing (a feature which is noteworthy because of its rarity), on classical notions of philosophy (ideal objects, essence of being, substance), and a critical review of notions that the author may have encountered during his exchanges with Kazimierz Twardowski (content of intention among others). Kotarbiński already knew that he was unhappy with purely philosophical entities and the seriousness with which this discipline presents itself, however he had not yet identified which of them were the most troublesome. Kotarbiński was more generally dissatisfied with philosophy as a discipline. While recalling the main lines of what he understands to be the empiricist position, represented by J.S. Mill and Bertrand Russell, according to which the apparent ontological debates are purely semantic, as well as the most well known reactions to it, coming from phenomenology and from Marxism, Kotarbiński asks a number of questions concerning the legitimacy (and the raison d’être) of these positions. He notes that the problem emerging from these debates is first of all methodological: what disciplines are legitimate in their inquiries concerning the fundamental features of being? What are the questions that we are to pursue in philosophy or in newly emerging disciplines, like psychology or sociology? These procedural questions remained central within the Lvov-Warsaw school and are still open today. It must be noted that they were the starting point for nearly all

Kotarbiński’s Strong Minimalist Ontology

�

41

of its members, whereas, at the same time the Vienna Circle at some point simply recognized the priority of natural sciences over other forms of discourse. The ontology itself is not free from methodological inquiries either. “Is there a sense in looking for the essence of an individual object?” Kotarbiński asked, noting that this question was barely mentioned by Aristotle. He thus opened a new field of reflexion for contemporary metaphysics, aware of the appeal of semantic reduction, yet ready to assume classical challenges. This interrogation is parallel to the research conducted at the same time by the British philosopher G.F. Stout, whose “ontological particularism”, certainly allowing the existence of universals yet in a very uncommon way,² is also marked by an urge to give all their place to concrete individuals. It is noteworthy that both philosophers refused to say that there is a hierarchy in order of beings. They did not think that, from a metaphysical point of view, there is anything more fundamental in an atom than in a person. It seems indeed that particularist ontologies like these set the agenda for future investigations, via e.g. the trope theory to come.³

�.� Elementy A comprehensive account of reism is to be found in Kotarbiński’s Elementy (1929), a book that was meant to be a textbook of the theory of knowledge, formal logic and methodology of science. After a well known and extensive criticism of Elementy proposed in 1930 by Ajdukiewicz, the question of categories became less important, and was partly replaced by discussions of names (empty, fictional, genuine, apparent etc.). This modification incited an interpretation according to which reism is not an ontological but (only) a semantic doctrine. This interpretation is wrong, and an analysis of Kotarbiński’s later papers will make that obvious. The semantic evolutions of reism will not be analyzed in detail in the present paper for, although they are significant given both the relation of the doctrine with Leśniewski’s logic and the importance of the philosophy of language for the LvovWarsaw school, they do not belong directly to the debate about ontology.⁴ The discussion concerning ontology is smoothly introduced in Elementy within a chapter on categories. Indeed, in Aristotle, categories are the ways to classify the predicates, yet these processes are easy to reify and, according to Kotarbiński, the classification is itself misleading, since it includes not only pred-

2 See (Nasim 2008). 3 See e.g. (Williams 1953). 4 For an extensive examination of these questions see (Smith 1990) and (Woleński 1989).

42 � Anna C. Zielinska icates proper but also subjects. These are not the main subject of his investigation though, since he is more interested in discussing the contemporary views represented by Kantian philosophers (Kotarbiński 1986 [1929]: 61/1966: 50), in particular by Wilhelm Wundt, an influential psychologist and philosopher at the turn of the century. Wundt distinguished four categories of objects: Ding, Eigenschaft, Tätigkeit, Relation (things, qualities/properties, activities and relations⁵), and Kotarbiński decided to scrutinize this proposal. It must be noted however that he deliberately avoided engaging in detailed analysis of Wundt’s interpretations of these terms, and wished to take them, so to speak, at face value. Kotarbiński’s project is then described as follows: We are interested in liberating ourselves from linguistic hypostases imposed by categories, from assumptions about the existence of certain objects simply because some words have certain meanings (Kotarbiński 1986 [1929]: 62/1966: 51).

His major step towards this liberation was to take the view opposed to Wundt’s and to try to defend it: one should adopt as hypothesis the idea that all categories can be reduced to a single one, the category of things (which includes persons). All expressions positing entities belonging to other categories have to be understood figuratively. The fact that our language allows us to create abbreviations like “Caesar’s death” does not imply that such an “event” genuinely exists, nor that our ontology needs to postulate anything other than Caesar himself, at a moment of time. The fact that we are able to talk about the relation of “similarity” does not imply that this “relation” genuinely exists: it is only a substitute for what is really there, i.e. two or more things that are somehow similar. The same is valid for qualities: neither “whiteness” nor “justice” exist, there are only concrete things (objects or persons) which are white or just. There is nothing more in our world than first substances in the Aristotelian sense, nothing other than things and they are the only real subjects of any discourse. If this discourse appears to speak of anything else (i.e. of events, relations or states), it only means that it uses apparent names, or “onomatoids”, potentially translatable into the language of concrete things or persons. Interestingly enough, this nominalist stance is not adopted towards either universals or abstract objects (although they were discussed elsewhere). Concerning universals, Kotarbiński thought that these did not exist, but he was also convinced that this question had already been dealt with by earlier thinkers and that nominalism provided proof of the inexistence of secondary substances in the Aris-

5 In 1958, Kotarbiński also had in mind the notion of Sachverhalt (state of affairs).

Kotarbiński’s Strong Minimalist Ontology

� 43

totelian sense (i.e. universals⁶). The notion of abstract objects, on the other hand, is not sufficiently clear, since it can include both universals and things. Kotarbiński’s choice of categories as the main target of his work now becomes obvious: these are properly philosophical entities and they are easy to grasp given their being either well defined or intuitively robust, unlike “abstract objects”. Elementy also raises the question of semantic (as opposed to ontological) categories. Two words belong to the same semantic category if, in a phrase, when one is replaced by the other, the phrase remains meaningful (dorzeczne). Take the following phrases: Paris is older than Madrid, Paris is newer than Madrid, Paris is either than Madrid. Older and newer belong to the same category, and even if the two first phrases are contradictory, they are both meaningful. The third one does not mean anything, since either belongs to a different category than comparative adjectives. The semantic question received a hasty treatment here, but this is precisely the aspect that was going to dominate the further discussions about reism.

�.� The Development of Reism After Ajdukiewicz’s 1930 criticism of Elementy, Kotarbiński understood that the very formulation of his position was not flawless, especially concerning the negative claims of reism: “No object is a relation” is not, according to the modified version of the doctrine, a negation of the phrase “Something is a relation”, but it simply marks a rejection of this latter phrase as nonsense. This modification does not substantially modify reism, which, as an ontological position, was fundamentally modified once, when Kotarbiński started to use the term “somatism” or “pansomatism” (he later introduced the notion of “concretism”, too). Somatist reism is a form of reism which only admits the existence of concrete bodies; it differs from dualist reism, which leaves a place for non-material concrete objects, e.g. souls (Brentano would be the most well-known example of such a dualist reist). He then noted (1935/1955) that every entity is “something corporeal or something sentient”, a body, “temporal, spatial and resistant” – either a “gravitating solid” (an animal, a tree, a lake), “a whole consisting of such solids”, “an object of the kind of which such solids consist” or “a whole consisting of elements each of which belongs to one of the preceding three types” (1935/1955: 489). There is nothing “atomistic” in this approach – all kinds of concrete bodily entities are legitimate, and no ontological priority is given to molecules (they appear in the

6 See (Kotarbiński 1986 [1929]: 66/ 1966: 55).

44 � Anna C. Zielinska third position, the first being occupied by the most intuitive complex ordinary objects). From the psychological point of view, Kotarbiński rejects the existence of “elements of the contents of ideas” or “immanent aspects”, such as colours, tones, feels, smells, which makes him a “radical realist” (1935/1955: 493⁷). He also admits that his position (i.e. somatic reism) is a form of materialism, although this latter name is often too ambiguous to be a useful tool⁸ and the materialists are often happy to say that “every psychical fact is identical with some physical fact”, which reintroduces the problematic notion of fact to the fundamental philosophical considerations. A reist would simply say: “A experiences thus: P” (Max is experiencing: I am hungry!); here, there is a person feeling or perceiving something, a person who is hungry, and we need not postulate any hunger as such whatsoever. This eradicates the need for the notion of mental state from the vocabulary of description of a person as an essentially dynamic being at different moments of time. The main virtue of this picture is that it precludes the possibility of thinking of the person as having a stable and pure identity upon which some mental states are superposed. On the contrary, the person is defined as constantly changing and thus the classical difficulties met in definitions of personal identity do not apply here (the persistence of the identity after memory loss as known from Locke or Hume’s bundle theory challenging the ordinary experience of stability). It must be noted that the inner character of experience does not make it somehow special; Kotarbiński denounced the myth of the inner, putting the description of the first-person and the third-person experience on the same level. The author himself was aware of the far-reaching consequences of such a project: “not only would all objects be reduced to bodies, but there would be only one source of experience – extrospection” (1935/1955: 500). This extrospectionism was not further developed.

�.� Defenceless Ontology In the years following the Second World War, Kotarbiński concentrated on more practical issues. He worked mainly on his praxiology, and started to write on ethics. Metaphysical questions were not crucial for his work, and he assumed that his ontology belonged to the past. He did not defend reism anymore, he only

7 The idea is already present in Elementy. 8 The materialism of his time seems to have been committed both to a mechanistic conception of the world and to some concrete scientific ideas about the emergence of sentient beings out of non sentient ones (Kotarbiński 1935/1955: 496) which are not required by reism.

Kotarbiński’s Strong Minimalist Ontology

� 45

postulated the “reistic or concretist attitude”, which consisted in inviting teachers to be extremely attentive to the ambiguities of several expressions which then provoked several useless debates. We should always be able to make disappear every apparent misleading name, since “it is only then that we will be fair towards the reality, consisting exclusively in things and people” (1949: 392). Ontological reism is there, nevertheless it is not really defended but only explained. A few years later, Kotarbiński would go so far as to admit that, indeed, he thinks that the basic ontological intuition of reism is true, but is unable to prove it. He notes then that “fully mature concretism is only able to have this kind of definitive program: it will try with all its forces to get free from apparent names, whenever it is possible” (Kotarbiński 1958: 402). In this late paper, Kotarbiński makes appeal to Leśniewski’s mereology, saying that if this doctrine is true, and if indeed classes in a mathematical sense were concrete entities, then concretism or reism would also be true as an interpretation of this theory (Kotarbiński even finds an interpretation of mereology inclusive of empty classes, which is rather unorthodox). Indeed, there are numerous problems concerning the interpretation of several scientific problems within the framework of reism; even if the algebra of classes can be developed here, more important difficulties emerge with, among others, the theory of decisions.⁹ In 1959 (in the preface to the second edition of Elementy), Kotarbiński admitted that the book was no longer meant to play its initial role as a textbook, yet it conserved another of its virtues, the aim of being an introduction to the early stages of Leśniewski’s ontology. The fact that Kotarbiński does not simply say that the book includes the first version of his own ontological position, reism, can be viewed as a sign of modesty, but it may also manifest the author’s difficulty with consistently defending his positions. Indeed, what was extremely straightforward in the 1929 book becomes much less so in later writings. Yet its ideas are preserved in his thinking about action, a possible onomatoid itself. The language of Kotarbiński’s theory of action or praxeology was meant to be understood by anyone, and the rigour of reistic translations was not something the Polish philosopher would insist on. Yet a few times, and most notably in discussions concerning agents, he made an effort to clarify some terms that might lead to ambiguities he wished to avoid. And while considering thought as a form of action, he carefully notes that “thought” is only an apparent name, and that only thinking agents should be taken into account. His paper on the notions of agent and culture includes the idea according to which:

9 See (Woleński 1989: 236-243).

46 � Anna C. Zielinska all description of culture as set results of actions [...] may be subsequently reduced to a description of culture as a set of what was created by actions. An example of a fundamental, ultimate reduction might be the following: to say that one’s ability to play an instrument is a result of one’s rehearsals means that the well-playing person is a creation of the person who rehearsed (in this case, a creation of oneself as compared to the preceding phase). (Kotarbiński 1964: 376)

This seems crucial: in a perfectly sterile ontological framework where only physical bodies are included, there is a place for thinking about culture, education, the dynamics of people and of institutions (collective structures can easily be accepted within reistic framework via Leśniewski’s collective conception of classes). Yet in order to defend the discourse from abusive reifications, which might then lead to even more abusive actions (if one takes too seriously the onomatoids like “nation”, “race” or “religion”), the idea of reminding about what there is remains relevant. It is interesting to ask how such a position would have evolved today. It seems that the major problems of reism come from formal sciences, particularly harsh with nominalistic tendencies in general. Today, when biology and neurosciences occupy the major place in the debate between philosophy and sciences, it might be easier to take reism seriously. The most immediate candidate to a reistic analysis is certainly the notion of qualia, i.e. phenomenal experiences, in the context of the philosophy of consciousness. Here, a simple linguistic postulate of “substituting real names for onomatoids” would not be satisfying, and a genuine ontological investigation would have to be launched.

3 Truth The problem of truth is not central to Kotarbiński’s philosophy, yet the way he apprehended it clarifies the very possibility of reism. The crucial element of Kotarbiński’s understanding of truth as absolute is the temporal dimension of particular sentences. This idea is already present in Twardowski (1999 [1900]), who thought that the idea of relativity of truth was based on a misunderstanding. Judgments were not true from time to time or in certain circumstances, but they are absolutely true. However, this approach necessitates a richer conception of what a judgment is. Judgment is not equivalent to the phrase “it rains”, but is a concrete mental product that takes into account the context of enunciation. The apparent paradox stating that the same judgment “it rains” is sometimes true and sometimes false disappears when we realise that this phrase cannot be understood as

Kotarbiński’s Strong Minimalist Ontology

� 47

being “the same” at different moments of time and space. The apparent simplicity of language is misleading. In 1913, Kotarbiński thought however (as stated in his Szkice praktyczne) that the absolute truth, as it was defended in Twardowski’s paper, posed a number of problems in our thinking about the future. Influenced by Jan Łukasiewicz, who postulated a third, undetermined truth value, Kotarbiński tried to propose a practical conception of truth, leaving place for action and for change. He abandoned this idea after Leśniewski’s harsh criticism (1992 [1913]). Leśniewski understood Kotarbiński’s worries concerning the putative ontology hidden behind sentences about the future, yet he also recalled that it is only sentences, and not states of affairs, that can be true or false, and these sentences are rooted in time and space, as they are asserted (and not in general).¹⁰ It was not a straightforward return to Twardowski’s ideas, but a certain idea of time-dependent absoluteness was preserved in this framework. We see here that there is a fundamental concern for the practical and actional dimension of theoretical investigation; the representational and static paradigm of idealist philosophers is definitely left behind. Indeed, “for a reist, reality is a bundle of things in constant motion and change” (Kotarbiński 1949: 394), and he certainly did not think that the world is “[w]hat is the case – a fact – is the existence of states of affairs” (Wittgenstein 1963 [1922]: §2)): this latter conception gives priority to a static Bild of the world, described by a set of atomic propositions.

4 Translational Projects: Brentano and Kotarbiński From a historical perspective it must be noted that although Kotarbiński’s reism shows strong similarities with respect to Brentano’s late ontology (some authors thought that he simply developed Brentano’s ideas¹¹), these two positions emerged quite independently. Kazimierz Twardowski, the founder of the LvovWarsaw School and Kotarbiński’s teacher, was deeply influenced by Brentano and one might simply think that Brentano’s parsimonious ontology is at the origins of Kotarbiński’s reism, yet this conception is false. Twardowski was close to Brentano’s early ideas, and did not have much regard for his late rejection of abstracta. In consequence, these ideas were completely absent from Twardowski’s

10 For a more detailed treatment of this question, see (Rojszczak 2005: 134-142). 11 See (Smith & Mulligan 1982: 95).

48 � Anna C. Zielinska teachings, and virtually unknown in Poland. It was only in 1929 that Twardowski, having read Kotarbiński’s Elementy, wrote him a letter stating that although he himself was not an enthusiast of this kind of approach, Brentano seemed to be close to Kotarbiński’s reism.¹² For Brentano, reism emerged from his reading of Aristotle on one hand, and, as it seems, from his interest in empirical issues on the other.¹³ “Should we not say instead that there is nothing whatever that corresponds to [...] abstracta?” Brentano asked already in 1901, and a few years later he assumed entirely his rejection of any entia rationis (Brentano 1966 [1904] and 1966 [1914a]). The two philosophers thought that there is something misleading both in the ordinary and the philosophical discourses, and that analysis is the correct means to find the real issues hidden behind ambiguous expressions. Brentano proposed a language criticism (Sprachkritik) showing that hypostases or abstractions are not necessary and that an Occam’s razor could be applied to them. A philosophical translation is apt to formulate phrases containing exclusively Reales, things or persons. It must be noted that these two were essentially different in Brentano’s view, whereas for Kotarbiński there was no fundamental ontological difference between things and people, both being material particular substances.¹⁴ Nevertheless, both philosophers rejected certain classes of putative entities not because they thought that metaphysics (where these entities belonged) was unsinnig,¹⁵ but precisely because they considered it crucial to any philosophical development; in this sense, their positions differed from the stances of Wittgenstein or the neo-positivists. This ambition to create a new language was already present in Leibniz’s work (Leibniz 1677: 156), and Kotarbiński did not conceal his admiration for the German thinker. Brentano also appreciated the way Leibniz kept in mind the idea of a possible translation: “it is good to remind ourselves of Leibniz’s pertinent observation: whenever we put anything into abstract terms, we should be prepared to translate it back into concrete terms, in order to be sure that we have not altered the sense” (Brentano 1966 [1914b]: 95). Again, this ambition to think of a new clear language (slightly clumsy in Kotarbiński’s case, and criticised by Ajdukiewicz) is not to be identified with the neo-positivist idea of a universal language of science.

12 See (Kotarbiński 1966). 13 See (Chrudzimski & Smith 2004). 14 Dariusz Łukasiewicz notes that the distinctions between bodies and souls introduced by Brentano were inspired by his descriptive psychology and the theory of representation more than by ontology. In such a framework, souls had zero dimensions while bodies had three (Łukasiewicz 2009: 23). 15 I use here Wittgenstein’s well known notion of “nonsense” (“Most of the propositions and questions to be found in philosophical works are not false but nonsensical” [1963 (1922): 4.003]).

Kotarbiński’s Strong Minimalist Ontology

� 49

Some of Kotarbiński’s remarks in 1929 show that he wanted to propose a reform not only of philosophical language, but also of ordinary language. This idea is not perfectly coherent with his insistence on both philosophical categories proposed by Wundt and the idea that what should count at the end of the day is not the surface grammar, but a possibility to translate our ordinary discourse into the language of particular entities, the language of reism.

5 Conclusion Ontological reism remains a defensible metaphysical position, and Kotarbiński’s unwillingness to propose a more robust defence of his views has some identifiable historical causes, i.e. his post-war engagement in practical philosophy (both praxeology and ethics), more relevant in the context of a war-ravaged country. There is however one more reason why it remains difficult to justify the ontological part of the doctrine. Kotarbiński assumes indeed that “the fundamental justification of concretism is both naively intuitive and ordinarily inductive” (Kotarbiński 1958: 402/1966: 434). Does this mean that no decent philosophical discussion can be proposed in defence of intuitively grasped ideas? Certainly not, as, once again, is shown by the contemporary works on qualia. Ontological reism should be pursued as a negative theory though, not as a positive one. This question of the burden of the proof was raised by Kotarbiński in 1966, surprised by the fact that it is the reist, the one who rejects imaginary entities, who is supposed to bring new evidence in favour of his positions, whereas those who assume their existence do not feel compelled to go beyond intuitive claims. It cannot rely on analysis of language alone, but has to engage in the investigation of the world. This idea is coherent with the naturalistic tendencies of Kotarbiński’s approach, even if they did not yield any substantial results on their own.

Robert Brisart

Objects as Posits from a Phenomenological Point of View Robert Brisart: Université Saint Louis-Bruxelles, Belgium

Let us start with the following claim repeated in many places of Husserl’s last work: The lifeworld, for us who wakingly live in it, is always already there, existing in advance for us [...]. The world is pregiven to us. (Husserl 1976: 145/1970: 142)

A few years before, in his Cartesian Meditations, Husserl wrote along the same lines that: The objective world is constantly there before me as already finished [fertig schon da], a datum of my livingly continuous objective experience. (Husserl 1950: 136/1960: 106)

These lines would be misleading if they suggested that, reaching the end of his philosophical journey, Husserl had partly given up the task he once attributed to transcendental phenomenology. In Ideas, for example, this task was very clearly to “parenthesize” or put out of action the objectivity of the world such as it imposes itself upon natural attitude, in order to shed light on the semantic or noematic activity of consciousness by which the world of objects is made up. Some commentators have so argued that, in his last work, Husserl instead emphasized the passive feature of our experience, showing that any constitution by consciousness presupposes a lower level of mental process where the world offers itself to us as already finished and that it is only from this basis that all constructions of consciousness are fleshed out. As a result, the same commentators have also defended the thesis that the later Husserl was very close to the Heideggerian idea of a more originary transcendence than the mere sense-bestowing intentionality itself. In other words, this more primitive transcendence would be similar to what Heidegger called the “being-in-the-world”. Thus, it is in the life-world – the Lebenswelt – that Husserl would have finally sought the roots of the semantic performances of mind, and at the same time, would have shown that all intentional activity finds its real conditions of possibility in a primary relationship to the worldly frame that thought does not produce but presupposes as a “pregiven” in order to exercise its own powers as thought.

52 � Robert Brisart Hubert Dreyfus is well known for having given such an interpretation of Husserl’s last work. He wrote for example: So when Husserl seeks to accommodate philosophically Heidegger’s point that things are encountered first in their everyday capacities as usabilia and only later as objective, scientifically accessible things, he proposes a special sort of “science” of the essence of the lifeworld: He finds that a passively given independent “thing” is presupposed by any encounter with equipment, and starting from that, all the higher level significances of the life-world (not to mention the theories of the objective sciences) can be, so to speak, added on. (Dreyfus & Haugeland 1978: 235)

In my view, this thesis is doubly wrong. First, it is wrong in respect of Heidegger, because the “world” in the meaning of the Heideggerian notion of “being in the world” has nothing to do with the Husserlian notion of life-world. The two concepts are even poles apart. But this is an aside. The point is rather that the same thesis is also wrong with regard to Husserl’s view about the world. Interpreting Husserl as if he led to the claim that the intentional activity of consciousness is dependent on an objective already finished world is tantamount to arguing that his phenomenology would finally have fallen into the “myth of the given”. By this term, I mean a certain idea of the given which is stronger than the idea criticized by Sellars. What I mean is more akin to what Putnam has called the idea of a “ready-made world” which he had himself denounced as a myth which is constitutive of the “metaphysical realism” (Putnam 1982).¹ So, my first question will be whether the life-world in the Husserlian sense of the term can be reduced to something like a ready-made world. It is undoubtedly true that, as Husserl says, the world is there before us as an already finished objectivity. But it is precisely a truth that belongs to the natural attitude of our ordinary experience in the world of things. However, in the Crisis, Husserl does not seem to have given up the project of a phenomenology of intentionality which has to turn towards the mind in order to grasp how the objectivity is produced, rather than to content itself with this objective world: The development of an actual method for grasping the fundamental essence of the spirit in its intentionalities, and for constructing from there an analysis of the spirit that is consistent in infinitum, led to transcendental phenomenology. It overcomes naturalistic objectivism and every sort of objectivism in the only possible way, namely, through the fact that he who philosophizes proceeds from his own ego, and this purely as the performer of all his validities [als dem Vollzieher aller seiner Geltungen], of which he becomes the purely theoretical spectator. In this attitude it is possible to construct an absolutely self-sufficient science of

1 See also (Putnam 1981: 146).

Objects as Posits from a Phenomenological Point of View � 53

the spirit in the form of consistently coming to terms with oneself and with the world as spiritual accomplishment [Verständigung der Welt als geistiger Leistung] [...]. Also, the ego is then no longer an isolated thing alongside other such things in a pregiven world. (Husserl 1976: 346/1970: 298)

Therefore, as we can see, the situation of a phenomenology of intentionality is certainly not the same as that of ordinary experience. This latter is necessarily endowed with a kind of spontaneous physics according to which the world still looks like a ready-made and, inside of it, any object appears as already constituted independently of mind and self-identifying as the sun, the moon, a rabbit or a duck. Of course, we will need further to add a little more to this, but, for the moment, let us just say that natural attitude considers things in such a way because it is vital for it not to consider them otherwise. On the contrary, phenomenology is this luxury that thought allows itself asking how were constituted all these objects that we commonly reckon to be already established. If you judge that superfluous, I will reply that there is at least one good philosophical reason for allowing oneself this luxury: we know that this same world within which we reify in speaking about sun, moon, ducks and rabbits, could be comprised of a quite different way if we reify it with other meanings like mass, molecules and electrons. I use here the verb “reify” exactly in the sense that Quine uses it in order to express how objects are posed by us for furnishing the world in various ways. In his short philosophical dictionary, Quine writes: Terms, some of them, refer to objects of one sort or another – stones, people, or abstract ones such as numbers. To declare there to be objects of a given sort, for purposes of one’s theory, is to reify them. Zoologists reify wombats and not unicorns, thereby reckoning the term “wombat” as referring and “unicorn” as not. (Quine 1987: 180)

The issue must be now to explain how we reify the world in such a way that we experience it usually and necessarily as a already finished given.

1 Reifications In order to carry out this examination, let us first look a little closer at what exactly Husserl means when he states that the world is given as already there. To follow his analysis, the world is pregiven to us not just as a mere reality, but rather in the form of what represents a set of values in our eyes. Objectivity is thus not primarily a matter of facts or of states of affairs, but, relatively to our most common behavior, it gives itself in the guise of what has value or validity for us:

54 � Robert Brisart Things, objects (always understood purely in the sense of the world of life) are “given” as being valid for us in each case [jeweils geltende]. (Husserl 1976: 146/1970: 143) In their human commissions and omissions, human beings are related to realities which are valid for them [geltende Realitäten]. (Husserl 1976: 266-267/1970: 2263) In this sense what is in question is not the world as it actually is but the particular world which is valid [jeweils geltenden Welt] for the persons, the world appearing to them with the particular properties it has in appearing to them; the question is how they, as persons, comport themselves in action and passion [...]. Persons are motivated only by what they are conscious of and in virtue of the way in which this [object of consciousness] exists for them in their consciousness of it, in virtue of its sense – how it is valid or not valid for them. (Husserl 1976: 296/1970: 317)

How are we to understand that validity or value as the main feature of things in our usual experiences of them? In reading the many passages devoted to this issue in the Crisis, our relationship to things is primarily practical and concerned with the needs of our own life within the world. Of course, in order to happen normally and not to be doomed to a systematic failure, this practical experience implies that faced with things we never found ourselves completely clueless, without knowing of what kind they are and so what we are allowed to expect from them. Therefore, as Husserl often repeats, the world, as much as all the things contained therein, always presents itself to our activity as “a set of fixed types” (Husserl 1976: 176/1970: 173) that it maintains into any of its own variations and changes. Therefore, this world is usually given to us as a set of evident elements which provide our activity into it with all the necessary confidence. So it is what essentially means that things always appear marked by value: the reason is not so much that they are understood as relating to the needs and interests of our life, but much more because they seem trustworthy enough for ensuring the smooth running of our activities. This is clearly expressed in Experience and Judgment, which in many respects, brings in the main theses of the Crisis: In the continuous validation of its being, the world present to consciousness as horizon, has the subjective general character of trustworthiness as horizon of existents known in general [...]. This indeterminate general trustworthiness is allotted to all things which attain separate validity as existent. (Husserl 1973 [1939]: 37)

But although one usually tends to think so, it is not in reality itself that confidence has its origins. If it so easy to find our way through reality, the reason is rather that we have reified it efficiently enough for coping into it without too many problems. Thanks to our reifications, indeed, we know what things are, what their properties

Objects as Posits from a Phenomenological Point of View � 55

are, the kinds of relationship that may exist between them, and thus what we can expect, or not, from all that. How are we to understand these reifications otherwise than products of the semantic power by which we can recognize and identify at first glance almost anything as being such or other sort of thing? What we call having experience precisely means being constantly able to immediately put a name or a meaning to anything that we encounter, even in the form of a mere sensation that, on the spot, we grasp as the outline of a particular object. This explains why, except in the very few cases where reality can surprise us, our usual experience is never confronted with an opacity which would leave us without words to say what we perceive and what we are dealing with. The usual ontology underpinning our experience makes it very unlikely that such a situation could arise. On the contrary, it makes everything seem transparent: we are able to give a meaning to everything that appears to us in one way or another – otherwise that does not appear –, and, on this base, there is no more sense to discuss the fact that a duck is a duck or that the moon is the moon. All reifications that meanings make possible and thus all objects constituted in this way are transparent simply because we do not call them into question. Note that, in a way, this transparency still implies a form of impenetrability and therefore of opacity: our objects are transparent because they are obvious, and because they are obvious we do not have to scrutinize them and generally do not attempt to do that. The obviousness of all objects fits a set of beliefs that provides assurance to our attendance in the world. Of course, this assurance is entirely doxic since it is only based on belief, acceptance, or presumption that the world is so constituted and we do not have to go back over that. As Husserl says, there is a “hypothesis of the world” or “the world is the constant presupposition” (Husserl 1976: 264265/1970: 261). But we have to add that this hypothesis is not made any way. It is underlined by our semantic system and it is the efficiency of this system – its ability to rule at best our experience among things – which makes the reliability of the general hypothesis by confirming it day after day. Again, this is not reality in itself that provides assurance to our experience but rather all the meanings through which we have reified the real in order to make it a set of determined objects. Thanks to meanings, reality looks like an objective world which is the most trustworthy because, from experience to experience, all our identifications are constantly confirmed so that we are fully assured that the surrounding world is exactly as we say or mean it. Meanings are therefore what is most essential to our experience simply because, through them, the world appears as a base of constant validities. They generate values doing so that our living experience in the world never loses its way.

56 � Robert Brisart

2 Posits In a well-known appendix to the Crisis whose text was the subject of two lectures given in Vienna on May 1935, Husserl evokes the natural attitude of man living within the world in terms of “mythical natural attitude” or “mythical-practical attitude” (Husserl 1976: 330-331/1970: 283 and 285). These expressions seem to me very fortunate even if it might be questionable whether Husserl uses them in the sense that I would give. In the lecture of 1935, the point is to distinguish the natural attitude and the philosophical attitude. While the former is marked by the interests of practical life, the latter has rather a theoretical vocation which is to understand how the validity of the objective world occurs, and this task first commands us to set aside the natural attitude of practical life for which this objective world always seems there, like an unquestionable given². In simpler terms, we would say that the aim is to produce a theory of experience, which is precisely the subject of transcendental phenomenology. However, in the lecture of 1935, Husserl remarks that the natural attitude itself already raises the first signs of the specifically human vocation to look away from its usual activity among things in order to find a more unified sense and value of the life-world. According to Husserl, this moment precisely coincides with the mythical or religious attitude with which the natural attitude is coupled, when it invents for itself higher powers supposed to rule the world and thereby also provides a firmer and more global assurance to our life within it. Nevertheless, the mythical attitude still remains part of the natural attitude, since all its inventiveness is only devoted to the practical needs of a humanity seeking still more assurance in the temporal flow of its experience. Here not only men and animals and other subhuman and subanimal beings but also superhuman beings belong to the mythical natural attitude. The gaze which encompasses it as a totality is practical. (Husserl 1976: 330/1970: 364)

Such an approach to the mythical attitude could seem trivial, but it is at least interesting for the light that it sheds on the process of reification that underlies the natural attitude whose mythical attitude is just an extension. Gods do not have another function than physical objects, whether ordinary objects of the usual experience or much more sophisticated objects of science. But, as objects of a special kind, Gods are probably of an utmost epistemological significance, because the mythologizing of the world from which they proceed reveals the nature of all the

2 See (Husserl 1976: §§38-39/1970: 143ff.)

Objects as Posits from a Phenomenological Point of View � 57

objective validities that give confidence to our usual experience. All these validities are only mythical and therefore all of them fall into what, generally speaking, we can call the myth of objects. In an excerpt from summer of 1936, Husserl seems well going in this direction: In contrast to philosophy having understood its meaning and direction, from its point of view, any concrete truth of life, in the universality of human historical life, and particularly in its highest completed forms (truth of religion [...] but also truth of daily and traditional life in the historical sense) is something like a mythical truth, not at all in the sense of a misrepresentation or an analogy, but in the sense of an expression of absolute reason which shows itself through the finiteness in the guise of these truths. (Husserl 1993: 225-226)

Let us leave aside the issue of the “absolute reason”. I will return to this below: it is another myth that Husserl obviously never managed to get rid of. For the moment, the important thing is to see that this passage reflects an idea that could certainly stand comparison with Quine’s words about what Homer’s Gods have in common with electrons, as with ducks and the moon. Let me quote these wellknown lines of “Two Dogmas of Empiricism” where Quine introduces the notion of posited entities or posits: For my part I do, qua lay physicist, believe in physical objects and not in Homer’s gods; and I consider it a scientific error to believe otherwise. But in point of epistemological footing the physical objects and the gods differ only in degree and not in kind. Both sorts of entities enter our conception only as cultural posits. The myth of physical objects is epistemologically superior to most in that it has proved more efficacious instrument than other myths as a device for working a manageable structure into the flux of experience. Positing does not stop with macroscopic physical objects. Objects at the atomic level are posited to make the laws of macroscopic objects, and ultimately the laws of experience, simpler and more manageable. (Quine 1951: 44)³

No doubt that, for Husserl too, all the products of science are ultimately related to the practical life-world where they are used only to strengthen the effectiveness of our experience regarding to which all these theoretical productions have finally value. In the Crisis, the task of transcendental phenomenology is precisely to show it against the naturalistic conception of science. Husserl writes about their products:

3 About the common sense bodies as posits, see also (Quine 1976: 250): “The positing of molecules differs from the positing of the bodies of common sense mainly in degree of sophistication”.

58 � Robert Brisart We see further that all these theoretical results have the character of validities for the lifeworld, adding themselves as such to its own composition [...]. The concrete life-world, then, is the grounding soil of the “scientifically true” world and at the same time encompasses it in its own universal concreteness. (Husserl 1976: 134/1970: 131)

Including all the objects of science, transcendental phenomenology dismisses everywhere the myth of the given preferring to it the myth of the objects. Its tasks are therefore only those implied by a theory of experience: to unravel how our reifications are taking place, how our referential systems are constituted and especially for what purpose this happens. So, starting from the very beginning, namely sensations, Husserl writes: One must no go straight back to the supposedly immediately given “sense data”, as if they were immediately characteristic of the purely intuitive data of the life-world. What is actually first, is the “merely subjective-relative” intuition of prescientific world-life. For us, to be sure, this “merely” has, as an old inheritance, the disdainful color of the doxa. In the prescientific life, of course, it has nothing of this; there it is a realm of good verification and, based on this, of well-verified predicative cognitions and of truths which are just as secure as is necessary for the practical projects of life that determine their sense. (Husserl 1976: 127-128/1970: 125)

Therefore, should we not admit that the ontological economy of transcendental phenomenology is entirely prescribed by a very pragmatist conception of what are the objects and what is the world at large? I think so. The Husserlian program is obviously not only to reactivate the process of semantic constitution that underlies the world such as it is pregiven in our experience, but also to point out that, for any kind of objects, this constitution is always determined by the practical needs of life, and therefore bearing the mark of the “subjective-relative”. Setting out elsewhere all the tasks required by “the systematic unravelling of the constitutional problems”, Husserl notes: The purpose of these unravellings, however, is none other, and can be none other than actually to disclose [...] the actualities and potentialities (or habitualities) of life in which the sense, world, has been built, and is continually being, built up. (Husserl 1974: §96b, 248249/1969: 242)

What remains concealed behind the myth of the given is unveiled by the myth of the objects if the latter simply means that all our productions of sense are just as well constitution of objects and more generally constitution of a world whose value and justification can only be measured in terms of the needs for human experience to have a better and better hold on reality. But if one calls to mind something like a semantic pragmatism speaking about transcendental phenomenology, there then raises a problem to which this phe-

Objects as Posits from a Phenomenological Point of View

� 59

nomenology is exposed. Indeed, in the philosophical tradition, pragmatism usually goes hand in hand with pluralism. So, why, then, does Husserl remain reluctant to admit the idea that there is more than one way to make the world, even if the validity requirements of these different versions are still prescribed by the sole vital needs of experience to have an effective structure in order to cope with reality in the best way?

3 Good and Bad Myths Of course, in several passages, Husserl recognizes that “all people do not share with us all the objects that constitute our life-world” (Husserl 1968: 496). On the ground of such claims, some generous interpretations have suggested reading the Crisis as concluding in favour of the plurality of worlds. For example, David Bell did not hesitate to assert that phenomenology has transformed itself, becoming “historically and culturally parochial” (Bell 1990: 231).⁴ But this kind of interpretations does not stand up to a more meticulous analysis of the texts, since the fact remains that more often than not Husserl sets against the plurality of worlds the uniqueness of a world to which all of us would belong, despite our different views of it. Therefore, if there are not objects given in advance, there would yet be well a pregivenness of the world which, according to Husserl, justifies that we must claim the existence of an “universal a priori which is in itself prior, precisely that of the pure life-world” (Husserl 1976: 143/1970: 141). If one wonders now about the reasons of such an a priori, one is merely stating the fact that they are all linked to the idea of a unique humanity founded on the intersubjective communication, the mutual understanding, and thus on the pure possibility for everyone to rise to “a truth about the objects which is unconditionally valid for all subjects” (Husserl 1976: 142/1970: 139). Among the numerous passages of the Crisis where such an idea is quite identically repeated, I quote this one: Let us consider the fact that everyone, in his commerce with others within his worldconsciousness, at the same time has consciousness of others in the form of particular others; that in an amazing fashion his intentionality reaches into that of the other and vice versa; that thereby one’s own and others’ ontic validities combine in modes of agreement and discrepancy; and that always and necessarily, through reciprocal correction, agreeing

4 See also (Carr 1987: 36), who emphasizes “relativistic implications of the concept of intentional object”.

60 � Robert Brisart consciousness of the same common world with the same things finally achieves validity [...]. The world-consciousness of any individual is always in advance – and indeed in the mode of ontic certainty – the consciousness of one and the same world for all. (Husserl 1976: 257/1970: 254)

For Husserl, thus something akin to a pregiven must persist. This pregiven is certainly not the objective world, but this form of consciousness through which we have the certainty to be member of a same world where the interpenetration of everyone’s experiences makes each life occur in a community of continual exchanges. Nevertheless, if we try to take a look at that which could make possible this consciousness to be part of a life-world determined by the intersubjective communication, then we are inevitably faced with something like a unique and common semantics. Indeed, how could we explain the possibility of an universal and mutual understanding – if such a thing exists – otherwise than in assuming that the semantics of any language has a universal objective feature? As Husserl himself says: I do not see what another sees, but, by the sense, I have in share the unity of our life, a life of communication, a life of exchange through expression, in the broadest sense through language. (Husserl 1993: 199)

Quine has showed that there are two kinds of myths: the good ones and the bad ones. The former are good because they are useful, while the latter are bad for the opposite reason. Probably one could say that Husserl’s phenomenology combines both. As we have seen, the myth of objects features high among what is the most useful to us. On the other hand, the idea of a unique and common semantics is another myth which is all the more pointless since it takes us mind off our own linguistic practices. Quine himself has denounced it as “the myth of the meaning” showing that its main argument is wrong: communication or, if one prefers, translation is always spoilt by an insuperable indeterminacy which speaks volumes about the diversity of referential systems or ontologies that humanity can have at its disposal simply for using their assumptions to help the organization of its own experience. As we can see, one thing is giving up the idea of a readymade world and giving preference to that of its semantic constitution, and another thing is being quite clear about the nature of semantics itself. In this regard, Dagfinn Føllesdal was right to claim that noema – or meaning in Husserl’s sense – is equivalent to Fregean Sinn. But this is not enough: it would still have to be demonstrated that this similar conception of meaning as something objective, universal and communicable, amounts to separate semantics from the ordinary languages, and to confine it in thought. If thought can be called noesis, then we see why meaning must be called noema.

Objects as Posits from a Phenomenological Point of View �

61

But the issue is therefore that this extralinguistic hypostasis by which meanings become mental entities cannot happen without losing sight of the plurality which is the main feature of referential systems of human experience. In other words, there are several possibilities of reification through semantics and so “a diversity of worlds in the making” (Goodman 1978: x). To grant this idea begins by admitting that “meanings are, first and foremost, meanings of language” (Quine 1969: 26). Husserl’s phenomenology clearly has not been able to take this plunge.

Bruno Leclercq

The Concept and its Object are (not) One and the Same The Functional View of Higher Order Objects in Carnap’s Work Bruno Leclercq: University of Liège, Belgium

Since we always use the word “object” in its widest sense, it follows that to every concept there belongs one and only one object: “its object”. [...] Thus, we will say that even general concepts have their “objects”. It makes no logical difference whether a given sign denotes the concept or the object, or whether a sentence holds for objects or concepts. [...] We can actually go even further and state boldly that the object and its concept are one and the same. (Carnap 1928/1967: §5, 10)

These claims, which Rudolf Carnap states in section 5 of The logical structure of the world (Der logische Aufbau der Welt), really seem surprising coming from someone whose work comes in the straight line of Gottlob Frege and Bertrand Russell’s logical analyses. Indeed, there is no logical distinction on which the fathers of both symbolic logic and analytic philosophy have put more stress than the clear-cut distinction between objects and concepts. Most of their logical – but also ontological and epistemological – theses rely on it. Furthermore, this distinction is one of the main issues at stake in the separation of analytic philosophy from other philosophical schools, and most notably from the views held in the Brentanian school about (intentional) objects (Leclercq 2011b: 53-96). Yet an important part of Carnap’s own work consisted of questioning such a distinction. First, he did so from Frege and Russell’s own extensionalist stance; then, reconsidering this stance itself, he did so by developing intentionalist positions. We will here describe this philosophical path which makes Carnap a rather atypical figure of early analytic philosophy, but also confirms the importance and accuracy of the object/concept distinction.

1 Concepts and Objects In the early 1890’s, Frege published three very famous papers (Frege 1980 [1891]; 1980a [1892]; and 1980b [1892]) which explained and clarified the principles of

64 � Bruno Leclercq the logical analysis that were at the ground of his own Ideography (Begriffsschrift) (Frege 1879/1972), and from then on have formed the foundations of contemporary predicate logic. Instead of Aristotle’s views on predication, which distinguished singular, particular and universal predicative judgements according to the generality of their subject-term, Frege suggests that only individual entities¹ should be considered as objects and genuine logical subjects of judgements, while generality would always be the mark of concepts and predicates. This means that, despite their linguistic similarity, statements such as “Socrates is mortal” and “All human beings are mortal” have very different logical forms. The first one states that to some individual entity – which is here designated by the proper name “Socrates” – belongs some property – mortality – and therefore that this object satisfies (or verifies) the concept of “being mortal”. In the second statement, two concepts – “being a human being” and “being mortal” – are at stake, and what is stated is that all individual entities which satisfy the first one also satisfy the second one; in contemporary logical notation: ∀x(Hx ⊃ Mx). If we now consider the particular judgement “Some human beings are mortal”, it states that not all objects which satisfy the concept of “being a human being” falsify the concept of “being mortal” – ¬∀x(Hx ⊃ ¬Mx) – and therefore that there are objects which satisfy both concepts – ∃x(Hx ∧ Mx). Now, this leads to Frege’s account of concepts as “functions”. By itself, “being a human being” is neither true nor false, but it is true of some objects and false of other ones; in Frege’s terms, the function “x is a human being” needs to be saturated by some object in order to become a proposition – which is why Russell will name it a “propositional function” –, which will be true or false depending on which object the variable “x” takes as a value. Similarly, some relational concepts such as “being the father of” – Fxy – need to be saturated by some ordered pair of objects – e.g. �George Herbert Bush, George Walker Bush�, �Bill Clinton, Barack Obama� – in order to become a proposition and be true (as in the first example) or false (as in the second one). Another way for propositional functions to become propositions and acquire a truth value is for their variables to be bound by quantifiers: “∀xMx” for “Everything is mortal” (a kind of abbreviation for the possibly infinite conjunction Ma ∧ Mb ∧ Mc ∧ ...), “∃xHx” for “There are human beings”

1 Frege did not provide any other characterization of individual entities apart from “self identity” (Frege 1979 [1883]: 62). But it appears from the examples he gave that he did not consider real entities (which are known by empirical acquaintance) as the only objects; some ideal objects such as numbers, or even truth values, also fall into the general category of (individual) objects (Frege 1980 [1891]: 23-30).

The Concept and its Object are (not) One and the Same � 65

(a kind of abbreviation for the possibly infinite disjunction Ha ∨ Hb ∨ Hc ∨ ...), “∀y∃xFxy” for “Everyone has a father”. Now, this logical analysis has a major philosophical significance. First of all, it leads us to think that the linguistic form of a proposition can be misleading and that the “surface” linguistic subject is not necessarily its genuine logical subject, i.e. the object it really bears on (Frege 1879/1972: §3, 112-113). In “All human beings are mortal”, the logical subject is not “all human beings”; indeed, “human being” itself is a concept and a predicate, which, just like “mortal”, is satisfied by some proper objects and falsified by other ones. Both “Hx” and “Mx” are propositional functions in ∀x(Hx ⊃ Mx), while the proper logical objects of this statement are the individual entities that can be the value of “x” and satisfy or not these concepts. Besides, as Frege’s Grundlagen der Arithmetik have shown in 1884, “Unicorns do not exist” should not be analyzed as denying the property of existence to (nonexistent) objects; “unicorn” is a concept, of which this statement says that it is satisfied by no object: ¬∃xUx. Existence, Frege says, is a second-level property, i.e. a property of concepts rather than of the objects satisfying them.² “Being eight” is not a property of the planets of the solar systems themselves, i.e. of Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus and Neptune, but is a property of the concept “planet of the solar system” – i.e. the property of being satisfied by eight objects or of having eight as the cardinal number of its extension. In the same way, “existence” is not a property which the unicorns themselves lack and the human beings themselves possess, but a property which the concepts “unicorns” and “human beings” will lack or have according to whether their extension is empty or not, i.e. whether they are not or are satisfied by some objects. And this leads to another important consequence of Frege’s logical analysis. As his paper “Sinn und Bedeuntung” shows, conceptual terms such as “unicorns” can make sense or have a “meaning” (Sinn) – i.e. clear defining features (Merkmale) – even though they lack any “reference” (Bedeutung), i.e. even though they are not satisfied by any object. The sentence “Unicorns do not exist” is perfectly meaningful even though its linguistic subject is an empty concept. And this contrasts with a sentence like “Fkjlmp is mortal”, which Frege would consider as devoid of meaning – and of any truth value – if the proper name “Fkjlmp” fails to designate any object. Unlike proper names, which are loaded by a presupposition

2 See in particular (Frege 1961 [1884]/1953: §§46 and 53). See also (Frege 1979 [1883]), and (Frege 1984 [1894]).

66 � Bruno Leclercq of reference (Frege 1980b [1892]: 68-69),³ conceptual terms are not weighted down with any ontological commitment; being general, they can happen to be satisfied by several objects as well as by a single one or by no object at all. This means that concepts can be conceived at will through their defining features without any new object appearing in the world; concepts are just meant to sort the objects of the world into those which, when taking the place of their variable, make them a true sentence and those which make them a false sentence. As Quine will later say, concepts do not enrich the ontology – i.e. the “furniture of the world” – but only the ideology – i.e. the set of our classificatory principles. Even though such a conclusion is not drawn by Frege himself, there is a strong nominalistic significance in his conception of concepts as functions: conceptual terms could just be seen as linguistic expressions allowing us to talk about objects; they would not designate objects on their own. Some parts of Frege’s theory, however, seem to resist the clear-cut distinction between concepts and objects. A first thing to notice is that, while Frege sees the Sinn of a conceptual term in its defining features (Merkmale), he seems to hesitate as far as its Bedeutung is concerned. On one hand, the Bedeutung of a conceptual term is its range of value (Wertverlauf ), i.e. the set of objects which make it a true proposition. We will thus say that two conceptual terms such as “daughter of Barack Obama” and “child of Michelle Robinson” have the same Bedeutung even though they do not have the same Sinn. On the other hand, Frege wants to say that a conceptual term does not refer to its Wertverlauf directly, but only through the designation of the concept itself, which is a kind of abstract object, that can indeed work like an object – i.e. take the place of the logical subject of another logical argument – in some special contexts such as the contexts of propositional attitudes: since I can admire the daughters of Barack Obama without admiring the children of Michelle Robinson (by lack of knowledge that they are the same individuals), Frege thinks that here the concept itself rather than the individual objects is the object of my admiration and the logical argument of the statement.⁴ Another tension arises in Frege’s theory due to the fact that, despite his clearcut distinction between (individual) objects and (general) concepts, Frege provides a theory of meaning for singular terms which is similar to the one he de-

3 Free – i.e. “pressupositionless” – logic precisely aims at replacing these referential presuppositions by explicit existential assertions. Since this requires that existence be afresh predicated to basic objects (and not only to concepts), free logic can be used as a tool for Brentanian-oriented logical and ontological analyses. 4 See (Frege 1980b [1892]: 58-59, 62-63, 66-67 and 77-78) and also the letter sent to Edmund Husserl on the 24th of May 1891 (Frege 1976: 96).

The Concept and its Object are (not) One and the Same �

67

velops for conceptual terms. He indeed thinks that two singular terms such as “Phosphorus” (the morning star) and “Hesperus” (the evening star) can happen to have the same “Bedeutung” – refer to the same object – even though they do not have the same “Sinn” (Frege 1980b [1892]: 56-62). Now this means that, like general terms, singular terms (including proper names) can also have a Sinn, i.e. defining features, which fix their Bedeutung. Against this view, Bertrand Russell, as is well known, will make a distinction between two kinds of singular terms: proper names such as “Barack Obama”, which do not have any Sinn and directly refer to an object, and definite descriptions such as “the actual President of the USA”, which isolate a single object through a conceptual characterization that is satisfied by this object only (Russell 1905a). Unlike proper names, definite descriptions behave like conceptual terms: their Bedeutung is fixed by their Sinn. The only specificity of definite descriptions is that to the conceptual term is added a specific operator – “the actual President of the USA” – which implicitly states that there is one and only one object satisfying this concept. In case there were in fact no object satisfying the concept (e.g. “the actual King of France”) or in case there were several objects satisfying it (e.g. “the daughter of Barack Obama”), this implicit statement would be false, in such a way that the sentences whose linguistic subject is “the actual King of France” or “the daughter of Barack Obama” would also be false, but not meaningless (since conceptual terms keep their Sinn even when they are devoid of Bedeutung). This case is therefore quite different from sentences like “Fkjlmp is mortal”, which are devoid of meaning and neither true nor false. Proper names involve a presupposition of reference, which is not the same as involving implicit existential statements as definite descriptions do. Now, what about Pegasus? Should we consider a statement like “Pegasus has big wings” as devoid of meaning – because of lack of reference of “Pegasus” – or just as false – because there is no object in the actual world which satisfies the definite description “the flying horse which was captured by Bellerophon”? Russell suggests that we should rather treat “Pegasus” in the second way, i.e. as a pseudoproper name, which is in fact an abbreviation for a definite description – thus for a conceptual expression –, and that we should therefore say that sentences about Pegasus are false rather than devoid of meaning. By discarding the notion of “singular term” and sharpening the distinction between proper names – directly referring to an individual object – and conceptual terms – which have a Sinn (of a general nature) and a Bedeutung (that can happen to be made of several objects, a single object or no object at all), Russell clearly reinforces the opposition between objects and concepts. While concepts have a definition, i.e. characteristic features (Merkmale), and can be satisfied or not by actual entities, objects are those individual entities which have some prop-

68 � Bruno Leclercq erties (according to which they satisfy some concepts) but do not have a definition. The proper name “Barack Obama” directly refers to an individual object without describing it; this object, which can be known by acquaintance, happens to have some properties but is not defined by any description. On the contrary, the definite description “the actual President of the USA” states what is required in order to be its reference; it is a conceptual term, which either points (under its de re interpretation) to the individual object that satisfies this description in the actual world, i.e. Barack Obama, or (under its de dicto interpretation) to whoever satisfies or would satisfy this description. Now, since the latter (the de dicto “object”) has defining features and can only be known by description, it is obviously of a conceptual nature.⁵ Frege had shown that “All human beings are mortal” and “Some human beings are mortal” should be respectively read as “What is human is mortal” – ∀x(Hx ⊃ Mx) – and “There are individuals that are both human and mortal” – ∃x(Hx ∧ Mx) –, where “human being” appears as a concept. Similarly, Russell shows that, in many singular judgments, the linguistic subject is of a conceptual nature, so that the genuine logical subject is the individual which satisfies this description: “The actual President of the USA was born in Hawaï” should be read as ∃x[Px ∧ ∀y(Py ≡ x = y) ∧ ∀x(Px ⊃ Hx)]. Here again, what Russell’s analysis suggests is that using the linguistic expression “the actual President of the USA” and making it the linguistic subject of statements does not commit us to consider this expression as the name of a new object “the actual President of the USA”, which should be admitted as part of the “furniture of the world” next to other objects such as Barack Obama, you and me. And this is still more obvious about the linguistic expression “the actual King of France”: we do not necessarily have to admit in the ontology an (inexistent) object whose properties would make false the statements about it. Just like “human being” and “unicorn”, “actual President of the USA” and “actual King of France” are conceptual expressions which do not name objects but only work as classificatory principles which eventually isolate the individual objects – were there several, one or none of them – that have some properties. Once again, the suggested logical analysis has a clear nominalistic significance. Now, this sharp distinction between the furniture of the world – which is made of individual objects that can be known by acquaintance and designated by proper names – and the conceptual apparatus – which is devoid of any own ontological commitment and only works as a set of classificatory principles allowing

5 See in particular (Russell 1959 [1912]: Chap. V). For Frege (1979 [1883]), it is superfluous to say of an object that it can be experienced.

The Concept and its Object are (not) One and the Same � 69

to sort the former – is exactly what Quine will respectively name “ontology” and “ideology”. For Quine, ontological commitments of some theory do not lie in their propositional functions, but rather in the individual objects which can be the arguments of – take the place of the variables in – these propositional functions, and satisfy them or not, i.e. in the individual objects that can be the values of the variables which are bounded by the (universal and existential) quantifiers of this theory (Quine 1948). Rephrasing sentences like “The golden mountain is made of gold” or “The round square is round” helps to grasp that they are not about inexistent or impossible objects but about the common individual objects of the world, none of them satisfying the concept of “being a golden mountain” or of “being a round square”. As Russell had already done, Quine here explicitly takes as his targets philosophers of the Brentanian school⁶. The latter tend to consider linguistic expressions such as “unicorns”, “Pegasus”, “the actual King of France” or “the round square” as referring to proper – though inexistent or even impossible – objects. Such expressions can be the logical subjects of statements (which, according to their properties, they will make true or false), but also the genuine objects of intentional attitudes, such as beliefs, desires, hopes, fears, feelings of admiration or hate, and so on: I can admire Pegasus, fear ghosts, hope to be soon the actual King of France or believe that I have found a round square (or the squaring of the circle). Meinong’s theory of objects is at the exact opposite end of the nominalistic trend that prevails in analytic philosophy from Russell to Quine on the grounds of Frege’s logical analyses. Meinong’s theory of objects – and the Meinongian logics that have been developed upon it⁷ – tend to weaken or even to blur the distinction between objects and concepts. Meinongian objects such as “the square in general” and “the round square”, “the smallest mountain” and “the golden mountain”, “the actual President of the USA” and “the actual King of France”, “Pegasus” and “Sherlock Holmes’ best friend”, but also “the square I am know thinking about” and “the square no one has ever thought about” typically have defining features (Merkmale) – a “Sosein” – even if these features are not satisfied by any actual object – even if they have no Sein (Meinong 1904/1960: §4 84-85). Typically such “objects” are thus known by description rather than by acquaintance.

6 See in particular (Brentano 1874/1973: 194-200), (Höfler 1890: 7), (Twardowski 1894/1977: §3, 8-10; §5, 21-26 and §6, 28) and (Meinong 1904/1960: §§1-4, 77-86 and §§8-9, 94-99). 7 See in particular (Findlay 1995 [1963]), (Chisholm 1972), (Chisholm 1973), (Routley 1980), (Parsons 1978), (Parsons 1980), (Rapaport 1978), (Zalta 1988), (Jacquette 1996), (Paśniczek 1998) and (Priest 2005).

70 � Bruno Leclercq Now, this, analytic philosophers would reply, makes them concepts – which have a Sinn, i.e. defining features (Merkmale), even when they have no Bedeutung, i.e. an empty range of values (Wertverlauf ) – rather than proper objects. Meinongians, Fregeans would say, seem to be trapped in language; every time language can build a new description, they think “there is” (in a neutral sense) an object corresponding to it, while analytic philosophers do not think that a new object has been added to the world but merely that a new classificatory principle has been conceived. Meinong’s logical analysis leads to Meinong’s “ontological jungle”, i.e. to a world of objects which is not only luxurious but also quite messy, since objects such as the actual King of France, the actual fat King of France, the actual bald King of France, the actual fat and bald King of France, and so on, either have to stand (crushed) next to each other in the jungle or to be somehow interlocked into each other.⁸ We will not list here the benefits and drawbacks of both logical analyses. As we have shown in other papers,⁹ Meinongian logics struggle with problems – such as synthetic predicative judgments, extensional identity between definite descriptions or de re modalities – which are more easily treated in extensionalist logic. On the other hand, Frege and Russell’s extensionalist logic surely fails to give a proper account of fictional statements, as well as of statements involving intentional attitudes or (other) modal operators. And, of course, this last failure is the reason why, after having tried to keep to extensionalism by treating such statements through the “formal mode of speech”¹⁰ – a strategy to which Quine himself will stay attached –, Carnap and other heirs of Frege will extend his logical analysis by developing modal logics. What we will here be concerned with is Rudolf Carnap’s own positions in this debate. For, surely, Carnap stands on the straight line that runs from Frege and Russell, whose disciple he explicitly claims to be, to Quine, who clearly claims to be Carnap’s disciple. How come such a philosopher, who is so strongly confirmed as an analytic philosopher and indebted to Frege’s logical analysis, can claim that “the object and its concept are one and the same” ?

8 See (Leclercq 2012). 9 See (Leclercq 2011a; 2010; 2012 and 2014). 10 Such a metalinguistic strategy – which Quine will name “semantic ascent” and to which he will himself stick – is a way to make explicit that Meinongians are trapped by language and that they believe to talk about the structure of the world while they just talk about the structure of our language. On this see (Carnap 1934). See also (Bouquiaux, Dubuisson & Leclercq 2014).

The Concept and its Object are (not) One and the Same �

71

2 Pseudo-Objects of Science The claim is made in one of Carnap’s earliest works, Der logische Aufbau der Welt, which was published in 1928. In this book, Carnap tries to extend Frege and Russell’s logicism by showing that not only arithmetic – as Frege claimed – or even the whole body of mathematics – as Russell claimed – but the whole rational structure of science (including empirical sciences) is – or can be – entirely grounded on logic. Contrary to what Kant’s Critique of Pure Reason maintained, there is no need of synthetic a priori principles in mathematics and no need of such principles in the rational structure of empirical sciences either. On their own, logical – i.e. analytical – principles can do the whole job of the rational organization of knowledge (Carnap 1928/1967: §106, 175-176). For the empirical sciences, this would mean that all the information comes from the empirical data while their rational treatment confines itself to reformulating these data in different ways and to drawing conclusions from them without adding any new content, unlike Kant’s synthetic a priori principles were supposed to do. This, as we know, is the view of logical empiricism. Now, in the Aufbau, Carnap tries to establish the plausibility of this view by showing that the whole of science can be rebuilt in such way that all its objects and concepts – all the ones that appear in its statements – be characterized on the ground of elementary empirical objects and relations¹¹ through purely logical means. Subjective (“autopsychical”) entities such as sensations or colors, objective entities such as atoms, molecules, cells, organs or organisms, heteropsychical entities such as beliefs, desires, fears or hopes, or even cultural entities such as social groups or values, ... all can be defined as logical complexes of similarity relations between elementary experiences. Such complexes – e.g. an atom of Hydrogen – will thus exist if these similarity relations actually occur, and they will in turn satisfy some new concepts and relations – e.g. be linked to an atom of Oxygen – if some further relations actually occur between the corresponding complexes of elementary relations. Even though it does not develop such a project in all its details – which would of course require much scientific information as well as a logical work that would still be much bigger than what the Principia Mathematica were for the mathematics –, Carnap’s Aufbau sketches the procedure for such a logical (re)construction

11 In Carnap’s system (Carnap 1928/1967: §67, 107-109 and §78, 127-128), the basic objects are elementary experiences and the basic relation between them is “recollection of similarity” (Ähnlichkeitserrinerung).

72 � Bruno Leclercq of the objects and concepts of the first few levels of science while it only suggests the broad outlines of the extension of this undertaking to all other levels. The main idea is that all scientific objects will, according to Russell’s theory of types, be built as higher order objects which are characterized as predicative or relational classes of lower level objects – e.g. molecules as sets of relations between atoms. Eventually, all scientific objects would thus be some logical objects based on the similarity relations satisfied by the elementary experiences. As the Aufbau explicitly claims, such a project would amount to a constitution theory like the ones Mach, Natorp, Husserl or Meinong had considered and attempted (Carnap 1928/1967: §3, 7-9),¹² with yet the very interesting specificity that the constitution of higher order objects on the grounds of lower objects would here be purely logical and not require any synthetic a priori principle. As for the basic objects of the system, Carnap does not want to start from “simple sensations” – such as a sensation of redness – as Mach does. Taking into account the works on experience of Stumpf, James, Husserl or the Gestalt theorists, he believes that “simple sensations” themselves are not immediate parts of the flow of experience but only result from some kind of abstraction on it (Carnap 1928/1967: §67, 107-109). Though they seem to be elements of these basic experiences, sensations should rather be considered as logical complexes of experiences. And this is why Carnap names “quasi-analysis” the process of building simple sensations on elementary experiences by a logical construction. For example, in case my immediate experience would include blue circular spots (not yet identified as such), they would satisfy similarity relations with blue triangular spots, blue square spots, blue hexagonal spots, ... as well as with red circular spots, yellow circular spots, green circular spots, ... And of course blue square spots will also endorse similarity relations with blue triangular spots and blue hexagonal

12 See also the system of Nicolai Hartmann, as explained in Roberto Poli’s paper in this volume. Like most theories of objects, the Aufbau asserts both that there are objects of different categories (physical objects, mental objects, spiritual objects, values, ...) and that these categories are grounded on each other : “There is only one domain of objects and therefore only one science. We can of course still differentiate various types of objects if they belong to different levels of the constructional system” (Carnap 1928/1967: §4, 9). The logical tool Carnap uses allows him to conceive such a system as a mere classificatory reorganization of one single domain of basic objects, i.e. of elementary experiences : “All objects of empirical sciences (except for the elementary experiences themselves, which correspond to the stars) are constellations of stars, together with their relations and connections, which are formed from propertyless, but orderable stars. The differences between the so-called object types, especially the difference between the physical and the psychological, merely indicate different types of constellation (or their connections) which are due to different modes of organization” (Carnap 1928/1967: §162, 259).

The Concept and its Object are (not) One and the Same � 73

spots, but they will not with yellow circular spots or green circular spots. So much so that, from the flow of experience, will emerge some equivalence classes on the ground of the relation of similarity:

Fig. 1

The “simple sensation” of redness can thus be characterized as one of these equivalence classes. a is blue iff it is a member of the equivalence class { a, b, c, d, e}, i.e. if it is such that aSb, aSc, aSd, aSe, bSc, bSd, bSe, cSd, cSe, dSe. Being equivalence classes, sensations are objects of a logical type which is higher than the elementary experiences they are built upon (or abstracted from). Therefore talking about sensations is just an economical way of talking about elementary experiences. According to Russell’s axiom of reducibility¹³ and Carnap’s thesis of extensionality, all that is said about the former – all statements which have the former as arguments satisfying higher order concepts and relations – can be reformulated so as to bear on the latter, i.e. in (rather complex) statements which have the latter as arguments satisfying similarity relations. And what is true of sensations is also true of all objects of science. They all belong to (increasingly) higher logical types, and all that is said of them can be said (in a less economical way) of lower type objects. All proper scientific statements are mere abbreviation phrases for complex sets of similarity relations between elementary experiences: The constructional system shows that all objects can be constructed from “my elementary experiences” as basic elements. In other words (and this is what is meant by the expression

13 See (Russell & Whitehead 1927: vol. I, 55-60).

74 � Bruno Leclercq “to construct”), all (scientific) statements can be transformed into statements about my experiences (more precisely, into statements about relations between my experiences) where the logical value is retained. Thus, each object which is not itself one of my experiences, is a quasi object; I use its name as a convenient abbreviation in order to speak about my experiences. In fact, within construction theory, and thus within rational science, its name is nothing but an abbreviation. (Carnap 1928/1967: §160, 255)

Even though it has a new epistemic value, the sentence about the higher order object is equivalent to – has the same logical value as – the sentence about the lower order objects. [...] the relativity of the concept “quasi object”, which holds for any object on any constructional level relative to the object on the preceding level, is especially obvious. It is now clear how the seemingly contradictory theses of the unity of the object domain [...] and the multiplicity of independent object types [...] are to be reconciled. In a constructional system, all objects are constructed from certain basic objects, but in step-by-step formulation. [...] statements about all objects are transformable into statements about the basic objects so that, as far as the logical meaning of its statements is concerned, science is concerned with only one domain. [...] in its practical procedures, science does not always make use of this transformability by actually transforming all its statements. As far as the logical form of its statements is concerned, science is concerned with many autonomous object types. (Carnap 1928/1967: §41, 70)

Now, this also provides Carnap with a clear-cut distinction between scientific and metaphysical – meaningless – statements: as soon as you can find the rigorous expression of a sentence in the Aufbau’s system, i.e. show how it can be translated into a statement that bears on the similarity relations between elementary experiences, you have done two things. First, you have shown that it is a well-formed sentence – it does not violate the rules of logical grammar (the rule of the logical types distinction, for example). Second, you have shown that it has well defined empirical truth conditions – it is true if and only if some similarity relations between elementary experiences occur (Carnap 1928/1967: §161, 256-257). We can now understand what Carnap means by “object” in the Aufbau and why he says that an object and its concept are one and the same. Except for the lower level objects – elementary experiences –, all objects of science are built as (predicative or relational) classes, which means that they are characterized by concepts which are satisfied by lower-level objects. According to Russell’s theory of types, classes themselves are objects which can satisfy higher order concepts, and, in this way, also form further higher order objects. On each level, the distinction between concepts and objects – propositional functions and their arguments – stays sharp, but the conceptual nature of all higher order objects is shown by

The Concept and its Object are (not) One and the Same � 75

the fact that they are “known by description”, i.e. characterized by predicates or relations satisfied by lower order objects. [...] to every concept there belongs one and only one object: “its object” (not to be confused with the objects that fall under the concept). [...] Thus, we will say that even general concepts have their “objects”. It makes no logical difference whether a given sign denotes the concept or the object, or whether a sentence holds for objects or concepts. [...] We can actually go even further and state boldly that the object and its concept are one and the same. This identification does not amount to a reification of the concept, but, on the contrary, is a “functionalization” of the object. (Carnap 1928/1967: §5, 10)¹⁴

This last statement clearly makes Carnap one of Frege and Russell’s heirs rather than one of Meinong’s.

3 Intensions Based on Russell’s theory of logical types with the axiom of reducibility, Carnap’s Aufbau explicitly relies on a very strong thesis of extensionality: “in every statement about a concept, this concept may be taken extensionally, i.e., it may be represented by its extension (class or relation extension)” (Carnap 1928/1967: §43, 72). Surely Carnap does not deny that some linguistic contexts – such as those which involve modal operators or verbs of intentional attitudes – are not extensional. But, as Frege and Russell had done (Frege 1980b [1892]: 58-59, 65-67 and 76-78),¹⁵ Carnap’s Aufbau deals with these intensional contexts by saying that they do not concern scientific propositional functions as such but their meaning or “sense” (Frege’s Sinn), which is a different object, that plays no role in science except as the content of some thoughts (i.e. of heteropsychological objects) (Carnap 1928/1967: §45, 76-77). In his Logical Syntax of Language, however, Carnap will give an account of such contexts by treating them like metalinguistic contexts or, as he says, by switching to the “formal mode of speech”: just like “Paris is part of It’s raining

14 “The position which treats general objects as quasi objects is closely related to nominalism. It must be emphasized, however, that this position concerns only the problem of the logical function of symbols (words) which designate general objects. The question whether these designata have reality (in the metaphysical sense) is not thereby answered in the negative, but is not even posed” (Carnap 1928/1967: §27, 50). 15 Like Frege, Russell admitted that some linguistic contexts were not extensional, but, like Frege, he claimed that science, or at least mathematics, could rely on extensional contexts only. See for example (Russell & Whitehead 1927: vol. I, 72-74).

76 � Bruno Leclercq on Paris” should be read in the formal mode as “ ‘Paris’ is part of ‘It’s raining on Paris’ ”, “Charles says that it is raining on Paris”, “Charles believes that it is raining on Paris” or “It is necessary that � + � = �” should respectively be read as “Charles says ‘It’s raining on Paris’ ”, “Charles believes ‘It’s raining on Paris’ ” and “ ‘� + � = �’ is analytic”, i.e. as formal mode statements that bear on other statements (Carnap 1934/1937: §§66-69, 243-255). Such a metalinguistic strategy – which Quine will name “semantic ascent” and to which he will himself stick (Quine 1960: §§44, 47 and 56) – gives the advantage of restoring extensionality: in intensional contexts, this strategy says, the objects of logical analysis are linguistic expressions which comply to the extensionalist principle of salva veritate interchangeability. This means that, contrary to what Brentanians thought, intensional contexts do not require intensional objects (objects characterized by some defining features or mode of apprehension). Besides individual objects and the classes made of them, all that is required are these other objects which are linguistic expressions. In the 1940’s, however, and to Quine’s great regret, Carnap departed from such a view. Taking an interest in semantics¹⁶, he progressively showed more and more concern with giving a proper account of intensional contexts. In Meaning and necessity (Carnap 1956 [1947]), he distinguished two kinds of equivalencerelations: factual (F-) and logical (L-)equivalence. While individual objects, classes and truth-values are the matter of the former, individual concepts, properties (attributes) and propositions are the matter of the latter. “Phosphorus” (the morning star) and “Hesperus” (the evening star) are F-equivalent but not L-equivalent, i.e. their designata happen to be one and the same in our world but they could have been different. Similarly, since it happened that Plymouth was founded by the passengers of the Mayflower, “founder of Plymouth” and “passenger of the Mayflower” are F-equivalent; but they are not L-equivalent as “bachelor” and “unmarried” are. And again, since “Washington is USA’s capital” and “Paris is France’s capital” are both true in the actual world, they are F-equivalent, but they are not L-equivalent (Carnap 1956 [1947]: §§2-6 7-32). Names (singular terms), predicators (general terms) and sentences, Carnap says, all have both an extension (respectively an individual object, a class of objects and a truth-value) and an intension (respectively an individual concept, a property and a proposition). In order to fix the meaning of a given linguistic expression, two successive operators are thus required: first, a semantical analysis of the expression with the aim of understanding it, i.e. of fixing its intension (i.e. its defining characters and satisfaction conditions as far as singular and general

16 This started with (Carnap 1946).

The Concept and its Object are (not) One and the Same �

77

terms are concerned, its truth conditions as far as sentences are concerned); and, secondly, in order to fix the expression’s actual extension, an empirical investigation concerning the factual situation which is referred to. This, of course, seems very close to the theory of meaning Frege developed in “Sinn und Bedeutung”, except that Frege believed the usual Sinn of an expression becomes its Bedeutung in intensional contexts, while Carnap claims that linguistic expressions keep the same extension and intension in all contexts but that the former or the latter matters for the sake of the truth value (Carnap 1956 [1947]: §§28-30, 118-133). If extension is what matters, the phrase “the morning star” can be replaced by the phrase “the evening star” salva veritate, as is the case in the following first example. If it cannot, intension matters, as is the case in the following last two examples: The morning star has a retrograde rotation. The morning star necessarily shines in the morning. The morning star is often mistaken for a (proper) star. On the other hand, this theory of meaning goes against Russell’s theory on denoting, since, for Carnap, all singular terms, including proper names, behave like general terms, i.e. they have both an intension and an extension (Carnap 1956 [1947]: §§7-9, 32-42).¹⁷ And this means that, besides the actual individual who happens to be its extension in our world, the name “Socrates” also refers to an individual concept (its intension) which is F-equivalent but not L-equivalent to the individual concepts “the master of Plato” or “the husband of Xanthippe”.¹⁸ In Carnap’s quantified modal logic, predicate variables thus have both value extensions – made of individual objects – and value intensions – made of individual concepts (Carnap 1956 [1947]: §10, 45). As Quine will rightly point out, this en-

17 The idea of individual concepts was already present in (Carnap 1928/1967: §158, 247). 18 On the contrary, Barcan Marcus and Kripke’s quantified modal logic will be built on the idea that, unlike most definite descriptions, proper names behave like “tags” or “rigid designators”, i.e. refer to the same individual object regardless of the properties it has (and thus of the description that can be given of it) in each world. See (Barcan Marcus 1961) and (Kripke 1980). This will tend to make another modal logic, as can, for example, be seen from the fact that the principle of necessity of identity – x = y ⊃ (x = y) – which, for Carnap, only concerned L-equivalence (rather than F-equivalence) and intensions (rather than extensions) – here concerns individual objects, which are claimed to be identical with themselves in all possible worlds. The statement of a contingent identity between a proper name and a definite description – “Barack Obama is the actual President of the USA” – or between two different definite descriptions – “the husband of Michelle Robinson is the actual President of the USA” – should therefore not be considered as proper identity statements, but, as Russell had suggested, should rather be formulated as predicative statements. On this, see (Leclercq 2011b).

78 � Bruno Leclercq tails an ontological commitment to semantic entities;¹⁹ in intensional contexts, the actual object Venus disappears at the benefit of de dicto entities such as the second planet of the solar system, the morning star or the evening star. Carnap’s quantified modal logic, Quine says, “leads us to hold that there is no such ball of matter as the so-called planet Venus, but rather three distinct entities: Venus, Evening-Star and Morning-Star” (Quine 1947: 47). Of course, this seems to make Carnap’s analysis close to the one made in Brentano’s school, especially in Meinong’s work. In intensional contexts, the phrase “the morning star” works as de dicto rather than de re, which means that an “entity” such as the morning star is an object (of intentional attitudes or of modal statements) while, at the same time, it is characterized by some defining features. Here again, the object and its concepts seem to be one and the same. Yet an important difference separates Carnap’s semantics from Meinong’s. Indeed, for Carnap, the same expression does not refer to the same “object” when occurring in an extensional or in an intensional judgment: in the first case – “the husband of Xanthippe taught Plato” – the judgment is about an individual object which actually existed; in the second case – “the husband of Xanthippe, whoever he was, was a married man” – the judgment is about an individual concept. And this, on its own, strongly resists Meinong’s claim that “there is” (in a neutral sense) an object such as the husband of Xanthippe, which has both the property of having taught to Plato and of having necessarily been a married man. Carnap’s clear-cut distinction between two meanings (extension and intension) of a linguistic expression such as a definite description backs up Frege’s semantics as well as its links with Frege’s partition between objects and concepts. Indeed, like Frege (and Russell) and unlike Meinong, Carnap rather talks about empty concepts than about inexistent or impossible objects. Existence, Frege used to say, is a second level property, as it is a property of concepts – of which it says that they are satisfied by at least one object – rather than of the objects themselves. Ontological statements, Carnap insists, concern intensions rather than extensions; in the case of singular terms, they concern individual concepts rather than individual objects. Therefore, Carnap’ analysis is in fact very different from Meinong’s (Carnap 1956 [1947]: §16, 66-67):

19 “Entity” is the general term Carnap uses for all values of variables, while “individuals” are those entities which are taken as the elements of the universe of discourse in the system (Carnap 1956 [1947]: §4, 22-23).

The Concept and its Object are (not) One and the Same � 79

MEINONG

CARNAP

Alexander’s white horse is an object which actually exists.

The individual concept “Alexander’s white horse” is not F-empty.²⁰

Alexander’s unicorn is a possible but not actual object.

The individual concept “Alexander’s unicorn” is F-empty but not L-empty.

Alexander’s round square is an impossible object.

The individual concept “Alexander’s round square” is L-empty.

Based on the contrast between the description state that is actually the case and all other possible (i.e. consistent, non-contradictory) description states, Carnap’s semantics for modal logics somehow implements Leibniz’ idea of possible (i.e. consistent) worlds. Such a semantics, however, can only give an account of logical necessity – analyticity – but is not really fit for other modal or intentional operators. For example, Carnap’s semantics cannot give an account of the statement that “the Man in the iron mask is believed by Voltaire to have been Louis XIV’s half brother”. Nor can it give an account of the statement that “Human beings not only happen to be mortal, but are so necessarily, i.e. by nature”. Indeed, unless we put it into the defining features of the concept of “human being”, there is no logical necessity that human beings be mortal, which means that there are consistent description states where human beings are not mortal. Therefore, even though Carnap can state that being a human being F-implies being mortal, he cannot state that there is any necessity to this. By enriching Leibniz’ idea of possible worlds with the idea of an accessibility relation between worlds – in such a way that not all “possible” (i.e. consistent) worlds are “possible for” (i.e. accessible to) the actual world –, Kripke’s possible world semantics will provide a solution to such a problem. Amongst all consistent worlds, we could select as accessible only those worlds which respect some laws of biology. Now, in all these worlds, human beings will be mortal (like all multicellular organisms), so that it will be biologically necessary that human beings be mortal even though this is not logically necessary. Similarly, if we select as doxas-

20 Such a strategy clearly is a semantic ascent, but, of course, it differs from the metalinguistic strategy that had been developed in the Logical Syntax of Language, since, to Quine’s regret, intensional statements are no more formulated so as to talk about linguistic expressions but here clearly talk about their sense (intension).

80 � Bruno Leclercq tically accessible only those worlds where all of Voltaire’s beliefs are true (worlds that would still differ from each other on lots of statements about which Voltaire had no beliefs), we can give an account of an equivalence of the Man in the iron Mask and Louis XIV’s half brother which would neither be F-equivalence (if that’s not really the case in our world) nor L-equivalence (the expressions are not synonymous), and by this way give an account of the statement that “The Man in the iron Mask is believed by Voltaire to have been Louis XIV’s half brother”. And this is how modal logics today treat most intensional linguistic contexts which, for Brentanians, seemed to require intentional objects.²¹ As the referents of de dicto definite descriptions, such intentional objects are here replaced by functions that relate each possible world to the individual object which satisfies this description in this world. Now, Hintikka says, since, unlike actual objects, these functions are conceptually characterized and are not themselves members of any possible world, they should be considered as part of our ideology rather than of our ontology (Hintikka 1969: 94-95 and 101-109; and 1975: 30 and 89-90).

21 Now, another problem occurs in Carnap’s semantics as well as in standard Kripke’s semantics. As Carnap himself acknowledges, L-equivalence between two linguistic expressions amounts to their strict identity or synonymy, but does not still guarantee salva veritate interchangeability in all contexts. For example, “John knows Harry is married” could be true while “John knows Harry is a bachelor” would be false, because of John’s not knowing the meaning of the word “bachelor”. Yet John could hope that his favorite football team takes the second place championship if last year’s winner wins again this year, but not hope that if his favorite football team does not take the second place, then last year’s winner does not win again this year’s championship, by lack of grasping that these two wishes are L-equivalent (since p ⊃ q is L-equivalent to ¬q ⊃ ¬p). Besides, all analytic statements will not be interchangeable salva veritate in all contexts even though they are L-equivalent (e.g. “two is an even prime number” and “two is between one and three”). To these problems, Carnap tries to find a solution by making a distinction between the intension – which is here common to p ⊃ q and ¬q ⊃ ¬p – and the intensional structure, by which these formulas differ (Carnap 1956 [1947]: §§14-15, 56-64). For a reflection on how the depth of logical analysis can intervene in epistemic contexts, see (Gillet & Gochet 1993), as well as Éric Gillet’s thesis (1992). Another and complementary way of treating such problems will also be to distinguish Lequivalent “objects” by appealing to the consideration of impossible (i.e. inconsistent) possible (i.e. accessible) worlds. Because of John not knowing the meaning of “bachelor”, this propositional function is not satisfied by any object in the worlds where all what he knows is true, and therefore, in those worlds, the objects which satisfy the propositional function “unmarried” do not satisfy the propositional function “bachelor”. This makes them inconsistent worlds (according to the meaning of both words), but they still are the only ones compatible with John’s knowledge. For this possible worlds semantics for relevant logic, see (Routley et al 1982) and (Brady 2003).

The Concept and its Object are (not) One and the Same � 81

4 Conclusion Carnap’s dual semantics is good at giving an account of both F-equivalence – such as “the husband of Xanthippe was the teacher of Plato” – and L-equivalence – such as “the husband of Xanthippe was the man who had Xanthippe as his wife” –, as well as of both synthetic judgments – such as “Socrates was a married man” – and analytic judgments – such as “the husband of Xanthippe (necessarily) was a married man”. Furthermore, by giving an account of modalities that covers purely logical necessity and possibility, Kripkean modal logics can also give an account of both de re – such as “John knows that the husband of Xanthippe was forced to drink poison hemlock” – and de dicto modalities – such as “John knows that the husband of Xanthippe was a man”. On the other hand, as we have shown in other papers (Leclercq 2012), Meinongian logics can only give an account of all these judgments and of the differences between them by distinguishing two kinds of predication – “encoding” and “exemplification” –, a distinction which then leads to a distinction between two kinds of objects: those which, like the referent of “the husband of Xanthippe” taken as de re, i.e. Socrates, both encode and exemplify their properties and those which, like the referent of “the husband of Xanthippe” taken as de dicto, only encode them. Such a distinction of course strongly mitigates the originality of Meinong’s logical analysis and somewhat draws it back to Carnap’s dual semantics.

| Part II: Objecthood Prodigality

Maria Gyemant

Objects or Intentional Objects? Twardowski and Husserl on Non-Existent Entities Maria Gyemant: University of Liège, Belgium

At the very origins of the Husserlian concept of intentionality we can find a discussion that took place in the context of the School of Brentano between several of his students, namely the discussion of the status of non-existent or intentional objects of representations. The part Husserl took in this discussion lead him to his own original position on intentionality, which will receive a full-blown description in the Logical Investigations. The main issue in Husserl’s discussion of Twardowski’s thesis concerned the ontological status of what Twardowski was the first one to call “intentional objects”, that is objects such as, for instance, round squares, that lack a physical space-time form of existence. What exactly are round squares was one of the questions that bothered many philosophers at the beginning of the 20th century, whether logicians or psychologists. The immediate answer would be “Nothing”: a round square doesn’t exist at all. Some of the said philosophers were glad to settle the issue this way. The others, however, had reasons for asking further questions: if a round square is nothing how could it be different from other inexistent entities, for instance, four angled triangles? What is it exactly that supports the contradictory properties of being round and square? And how can we even talk about round squares if they are nothing? These philosophers would then conclude that, in order for us to be able to state their inexistence, certain inexistent entities must be presented to our judgment, and thus they must possess a form of being, even if it is a different, a minimal form. This is what the Frege-Russell tradition has called Meinong’s paradox: if the proposition “Round squares do not exist” is true, than its object, the round square, must somehow possess some kind of being, in order to be judged as inexistent. So inexistent objects, in order to be objects, must also have (a form of) existence. Hence the paradox: inexistent objects must in a certain way exist. So now, those who said that round squares do not exist at all must find a suitable way to dismiss this paradox. They have to explain how we can decide that an object doesn’t exist without supposing a certain relation to it. On the other hand, those who say that inexistent objects such as round squares do exist must sooner

86 � Maria Gyemant or later start tinkering with the ontological status of these inexistent objects in order to make it different from that of existent objects. We can thus identify two ways of solving this dilemma. One way is to say that round squares are not really objects but simply something we can put into words.¹ This position typically introduces a new entity between our thoughts, or the words we use, and the objects they are about: the meaning. If there are cases where the reference of our words does not exist, we need something else to guide us, to determine what we are talking about. We need an intermediary entity to take upon itself the determinations of an object that does not exist. We have then a threelevel model where the third level, that of the reference, of the object we refer to, can be empty. But the second level, that of the meaning, can never be empty, even when there is no reference. Thus, when we think of round squares, we have the meaning “round square” but no reference. So inexistent objects don’t have to possess being anymore in order for us to be able to think about them, all we need is the meaning of our words. This is basically a Fregean position, but we can also attribute it to other logicians such as Bolzano. If the three-level model is less obvious in Bolzano, the theory of objective representations (Vorstellungen an sich) is one of the first forms of a theory of meaning: these objective representations can never be absent from any subjective act of representing, even though it is possible that the object presented does not exist.² The other way to solve the problem is a typically intentionalist two-level solution of a Brentanian inspiration. Brentano and the Brentanians would define representations as intentional acts related to objects. If a thought is a representation or based on a representation, defined itself by its relation to its object, then all our thoughts must have a reference, that is, all representations must represent something. In the case where the object does not or cannot really exist, the ontological status of the reference cannot be the same as when it exists. The strategy would be then to think of the inexistence of for instance round squares as a lesser form of existence, yet still a form of being. This is what Meinong (Meinong 1971: §3, 489-490 and Meinong 1973: 205-365) tried to do by introducing the distinction between being and being-so. A round square, even if there is no such thing, has a form of being, the being-so. It is something: for instance it is square. While both Twardowski and Husserl take an intentionalist position on this issue, they are far from reaching the same conclusions. On the contrary, Husserl criticizes rather ruthlessly Twardowski’s thesis and proposes instead a solution that includes the best in the two concurring positions. But in a certain way Twar-

1 This would be typically a Fregean position. See (Frege 1980b [1892]). 2 See (Bolzano 1987: §69).

Objects or Intentional Objects?

� 87

dowski already takes a step in this direction in his habilitation thesis of 1894, On the Content and Object of Presentations (Twardowski 1894/1977), which means that maybe Husserl’s critique might seem more radical than it actually is.³ Basically, Twardowski seems to introduce an intermediary entity, the intentional object, supposed to bear the determinations of the represented object even when it does not exist. In this respect, the intentional object in Twardowski bears some resemblance to Frege’s Sinn and, as we will see, there’s even a Fregean ring to the arguments Twardowski brings up on this issue. On the other hand though, Twardowski is a Brentanian intentionalist and thus he cannot accept that certain representations simply do not represent anything. This intermediary entity, the intentional object, has to play then the role of the object, the reference of the act, and not that of a psychological content or a Fregean meaning. And as an object, it also has to have its own ontological status: that of intentional inexistence. This is however the point on which Twardowski’s thesis gives way to Husserl’s attack. Even though Twardowski took a step in the right direction, in Husserl’s view he failed to make the right distinctions and ended up with an uneconomical theory and a new and, according to Husserl, unjustified type of ontological status. I will argue that, because Twardowski tries to knot together two different solutions he fails to bring forward a thoroughly coherent theory of the object of representations. The intentional objects, that should constitute the solution to the paradox, cause in return quite a few new difficulties concerning their ontological status and their psychological dependence. They do not function as real objects do, because they are immanent to the mind and dependent upon it, yet they are not on the side of the psychological content either. Their status is ambivalent, neither content, nor object: they function as pseudo-objects, as a second intermediary entity, beside the content, and this is also true in return, as I will show, for representations where the object does exist. So, in the first section of this paper I will give an account of Twardowski’s view on the issue of representations that present inexistent objects. My purpose

3 In the first book of his Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy Husserl goes as far as acknowledging Twardowski’s merit in having introduced the distinction between act, content and object. “The three words in this juxtaposition have become nothing short of slogans, particularly since Twardowski’s fine treatise. Yet, however great and doubtless the service of this author in having acutely discussed certain generally ordinary confusions and made their error evident, it must still be said that in the clarification of the relevant conceptual essence he did not get considerably beyond what was well-known to the philosophers of earlier generations (despite their incautious confusions). This is not, perchance, to be charged to him as a fault. A radical advance was just not at all possible before a systematic phenomenology of consciousness” (see Husserl 1977: 267/1982: 309).

88 � Maria Gyemant will be to show that the status of intentional objects is far from clear: they rather slide between the content side and the object side and constitute a third type of entity. Committing to intentional objects in the case of representations with nonexistent references forces us then, as I will show in a second section of this paper, to commit to them in all cases, multiplying thus the intermediary strata of entities that separate us from the world. In a third section, I will briefly recall Husserl’s critique of Twardowski’s theory of intentional objects. I will show that Husserl’s own theory – presented already in a text from 1894 called Intentional objects (Husserl 1979 [1894]: 303-348/1994: 345-387), and a form of which we also find in the Fifth Logical Investigation (Husserl 1913: V, 421-425/2001: vol. 2, 125-127) – even though not as different as it might seem from that of Twardowski, has the great advantage of putting an order among this collection of entities and avoid the ontological consequences of Twardowski’s intentional objects.

1 Twardowski: The Object and the Content of Representations Cannot be Confused The main purpose of Twardowski’s paper is to distinguish the act of representing from its content and to further distinguish this content, immanent to the act, from the transcendent object represented by the representation. Obviously, it is the second distinction that will demand a thorough analysis to which Twardowski dedicates most of his study. The difficulty comes from the use of the term “representation”. This use is equivocal: there are three different things that we usually call “representation”. We call “representation” the psychological act of representing but we can also call “representation” its object and the way the object is represented to the mind (its content). We can distinguish linguistically the first sense of the term from the other two by using the verb “representing” instead of the noun “representation”. This strategy, however, cannot be applied to the distinction between the object that exists in itself (an sich Bestehende), which is what the representation is about, and the content, conceived as a mental image or a sign of this object that exists, so to say, in our head (“an” uns Bestehende) (Twardowski 1894: 4/1977: 2). We say of both that they are that which is “presented” in the act of presenting. So if the language doesn’t serve us directly for this distinction, Twardowski will

Objects or Intentional Objects? �

89

try to use it indirectly, in a rather cunning way, by comparing representations and names.⁴ The first strategy in order to make clear the difference between the content and the object of the representation will pass then through an analogy with the way a name works. A name has a triple function. First, it suggests that the speaker is really thinking about something; it is a sign of the representations that go on in his mind.⁵ The other two functions of a name though, are very similar to the Fregean Sinn and Bedeutung. The second function of a name is to bring up a certain content, which is common to the speaker and his listener and could be compared with the Fregean Sinn, which is public and objective. And finally, the third function of a name is to name an object, which constitutes its reference. Similarly, a representation has a different relation to its content and to its object: it has a content but refers to an object. The second strategy used by Twardowski to contrast the content and the object of a representation is to use a much more complicated analogy in which we will also find the first signs of an accumulation of intermediary entities. The second analogy Twardowski provides is between a representation and a painting. A painting of a landscape represents a landscape that is really out there by means of an image made entirely of canvas and paint. The landscape and the image of the landscape, even though both existing in the same world, cannot be mistaken for one another. They stand each in a completely different relation to the act of painting. Now, if we compare the act of presenting with the act of painting, we clearly see that the painted landscape corresponds to the object of the representation while the image of this landscape is the analogue of the content. Object and content are as different as the landscape that unfurls in front of our eyes is different from the painting made of strokes of colour. But Twardowski pushes this analogy further. He makes use of another distinction, introduced by Brentano in his Psychology from an Empirical Standpoint (1874/1973: II, 7, §7), between determining and modifying properties. The classical example that illustrates this distinction is to compare expressions like “white man” with “dead man”. “White” is the kind of property that contributes to a better determination of the object in question, by opposing it, for instance, to “black man”. On the other hand, “dead” is the kind of property that does not simply underline certain characters of the object, but changes the object altogether. A dead

4 We could read this analogy as a first attempt to put together Brentanian and a Fregean models of the reference. Another tentative would be Husserl’s. 5 This function is left aside in the Fregean model because it is a purely psychological function that has no implications in explaining reference.

90 � Maria Gyemant man is not a man anymore. The same goes for a “false friend” or a “false diamond”. “False” is a property that turns any object into its own opposite. Now, in the case of the painting it would seem that the attribute “painted” when applied to the painted image should work differently from when it is applied to the real landscape represented in the painting. And the question is to know when “painted” is used as a determining property and when it is used as a modifying property. Of course, the point is for Twardowski simply to show that the same attribute is used differently for the image and for the landscape, and similarly the attribute “represented” is differently applied to the content and to the object of the representation. It is not so important to know which use is determining and which one is modifying. The important point is this: if for the content and the object of the same representation the same property is used differently, then the two cannot be confused with one another. Yet, if we take a closer look to this analysis, we discover that things are not as plain as they seem. Which use of the attribute “painted” is determining and which one modifying? Twardowski says that using “painted” in order to describe the painted image is using it as a determining attribute. The image is painted, as opposed to chiselled or engraved. Being painted is one of the determining qualities of the image: by acknowledging that it is painted we know more than before about the image in question. So, if “painted” is used as a determining attribute in respect to the image, than it should be used modifyingly about the painted landscape. A painted landscape is not a landscape at all: it is a painting. And yet, Twardowski hesitates on this point. Of course, a painted landscape is not a landscape, but a painting, the same as a false friend is not a friend, but a foe. Yet, there is a determining use of the attribute “painted” which can be applied directly to the landscape without modifying it. Between the determining properties of a landscape, its shapes and colours, there is one we could describe as the fact that it is this same landscape that has been chosen by a painter in order to be painted. It is a mountainous landscape, with trees, a river etc., an explored or unexplored landscape, a familiar or strange one and furthermore it is a landscape that has also been painted, fact that changes nothing to its being a landscape (Twardowski 1894: 13-14/1977: 12-13). Now this double use of a property for the same object can be applied to our inquiry about objects of representations. We can indeed use the property “represented” in relation to the object of a representation not in one but in two ways, one determining and the other modifying. This point, which is not clearly stated in Twardowski’s analysis, is however reinforced, as we will see, by his solution to the paradox of inexistent object. Twardowski’s point is to distinguish between content and object by using the same property differently, once in a determining sense and the other time as a

Objects or Intentional Objects? � 91

modifying property. With the analysis of the painting, he arrives nonetheless at a different model, where we apply the same property (“painted”, respectively “represented”) to the same object, yet once in a determining way and the other time in a modifying way. So the result is not a distinction between content and object, but between two kinds of objects. We have then three rather than two uses of the term “represented”. We say about the content of a representation that it is represented in a determining way. And we say about the object of a representation that it is represented both in a determining and in a modifying way. In a modifying way, a represented object is precisely the object become content of a representation, and thus not an object anymore, but part of the representation. The represented object in a modifying sense is the object insofar as it is the correlate of the representation.⁶ Yet, the object is also represented in an unmodifying, determining way, when we mean by it the reference towards which the representation is directed, the real object that we want to acknowledge by means of our representation.⁷ We could also explain this by saying that using the term “represented” for an object in a modifying sense is taking the object to be something like a Fregean Sinn, the thing we mean when we represent, that part of the representation that indicates what object we are referring to. And correlatively, an object is represented in a determining way when taken as the reference, the Fregean Bedeutung.⁸ It would be easy to settle this issue, as Twardowski seems to do at some point, by saying that a represented object in a modifying sense is nothing but the very content of the representation. Thus, modifying and determining properties serve Twardowski to distinguish not directly between content and object, but between two ways of understanding the object of a representation: in a modified sense the represented object is actually not an object at all, but the content of the representation, while in a determining sense it is the reference to which the representation is

6 See also the analysis in (Twardowski 1894: 43/1977: 39-40): “The presented apple and the representation of the apple seemed to be the same in every case, while this sameness only obtains between the meanings of the two expressions if one takes ‘presented’ in the modifying sense, that is, if one means by the presented apple the content of the representation of the apple; for the latter, too, is presented. Of course, the presented apple in this sense is not an apple, but the content of a representation.” 7 This could be a first form of the noesis-noema distinction and also the reason why Husserl retracts his critique of Twardowski in his Ideas 1 8 Of course this translation in Fregean terms only goes so far as we do not consider the fact that the whole analysis Twardowski develops here is a psychological analysis of mental acts of representation. Yet as far as the connection between content and object is concerned, and actual acts are left aside, the analogy with Frege doesn’t seem that far fetched.

92 � Maria Gyemant related. This would suffice for the purpose Twardowski has pursued in this paragraph: we now have a better comprehension of the distinction between content and object and are no longer in danger of confusing the two. Yet, as we will see, Twardowski will maintain the distinction between content and object even when the object is represented only in a modifying sense, thus becoming immanent, dependent upon the mental act. Even this modified object is still, as Twardowski’s analysis of inexistent objects will show, to be distinguished from the content. So far Twardowski has successfully managed to contrast content and object. The problems appear though as soon as we try to apply this model to the nonexistent objects. If the object does not exist, we face, of course, the impossibility of distinguishing it not only from the content but also from anything else, other non-existent objects included. So what Twardowski needs is to show that even when the object does not exist, and indeed could not exist (like, for instance, a round square), we still have an object, different from the content: the reference of a representation is not something a representation (in the Brentanian sense of the word, employed by Twardowski) could lack.

2 But What if the Object Does not Exist? Bolzano and Frege would have said that when we’re talking of non-existent objects such as round squares, there is no need to say that they exist in any way. There simply is no reference (Bedeutung) there, the Sinn is all there is. But Twardowski could not translate this solution into his Brentanian vocabulary.⁹ This translation would force him to say that when the object does not or cannot exist, the psychological content of the representation is all there is. There would be then no distinction between the content and the object of the representation of a round square because there simply is no object. To say that there is no object would sound however very strange in a Brentanian framework. For Brentano, one of the fundamental points is that there is no representation without an object, that all representations present something. Of course, in Brentano, the thing each representation presents is an object such that you can also call it the content of the representation.¹⁰ But the very point Twar-

9 “It is tempting to believe that in this case there is no real difference between content and object, but only a logical one; that in this case content and object are really one; and that this one entity appears sometimes as content, sometimes as object, because of the two points of view from which one can look at it.” (Twardowski 1894: 29/1977: 27) 10 See (Brentano 1874: Book II, Chap. I, §5).

Objects or Intentional Objects?

� 93

dowski wants to defend in his thesis is that content and object are not interchangeable, and this is true even when the object does not exist.¹¹ So if he gives up on this point, he also has to give up his whole thesis. Now this, of course, is not yet an argument for maintaining the distinction between content and object even when there is no object. In the 6th paragraph of his thesis Twardowski brings four arguments in order to defend this idea. The first argument consists in considering the existential judgements corresponding to the case of non-existent objects. The existential judgement of a round square is “The round square doesn’t exist”. Now this judgement could not deny the existence of the psychological content of the representation of a round square. Denying the content means denying the very existence of the act of presenting a round square itself. But if there is no representation there could be no existential judgement either, because, as Brentano tells us, every judgement is founded in the representation of its object. So the existential judgement cannot state that there is no content of the representation “round square”. It can only deny the existence of the object. Thus, even when the object does not exist, it cannot be confused with the content of the representation. The second argument concerns the properties of the content and the object. If they were the same they would have the same properties. Yet we can find properties in the object that cannot be properties of the psychological content. For instance, a golden mountain has certain spatial qualities that the content of the representation of a golden mountain cannot have. The most obvious is that the content cannot be itself golden. The tricky part, though, is that the golden mountain is not golden either, because it does not exist at all. If it existed, it would be golden. But, because it is supposed to be golden, it does not exist. Yet, even considering that it doesn’t exist, the golden mountain must be, in a certain way, golden. If it weren’t golden it could very well exist: many regular, non-golden mountains exist. So, it is its being golden that turns the mountain into a non-existent object. And this object is golden in virtue of the content of the representation: it is so that we present it. Yet, this of course doesn’t make the content itself golden. So, once again, there is a difference between content and object even when the object doesn’t exist. The third way of reinforcing this distinction is to use the Bolzanian eidetic variation in a series of examples very similar to Frege’s Morning Star and Evening Star example. We can change the content of a representation without changing

11 “To the contrary, a brief consideration shows that the difference between content and object of a representation which can be ascertained when the object exists also are present when the object does not exist.” (Twardowski 1894: 30/1977: 27)

94 � Maria Gyemant the object and vice-versa. Thus, thinking of the roman city of Juvavum or about Motzart’s birthplace is using two different contents to connect to the same object, the city of Salzburg. Even though, while thinking of Juvavum, we have in mind Romans and antique cities, while thinking of Motzart’s birthplace brings up fragments of Eine Kleine Nachtmusik, the reference we connect to through this very different contents is one and the same object. This argument is rather convincing when the object does exist. Yet, when it doesn’t we have a problem in identifying something that stays the same while we vary the content (Twardowski 1894: 32-33/1977: 30). The example used by Twardowski is that of the representation of a square with oblique angles and a square with unequal diagonals. The two contents of these representations have something in common: they represent both an impossible square. Yet, since they lack any reference, we have no means to say if it is the same non-existent square they represent. The solution Twardowski proposes is somewhat artificial: he supposes there is an object that has both the unequal diagonals and the oblique angles. This complex impossible object stays the same when we consider each one of its determinations separately. It is the same impossible square that we represent on the one hand trough the content “a square with oblique angles” and on the other trough the content “a square with unequal diagonals”. Thus, the contents can vary while the object, the square with both oblique angles and unequal diagonals, stays the same. So once again, even if the object does not exist, we can think about it and we can do it by using variable contents. However, the inexistent object exists as the unique reference of both representations only because we think it thus. The fourth argument is correlative to the third. If we can vary the content while the reference stays the same, we can also vary the objects while the content stays the same. This is the case when the content is a concept under which fall more than one object. For instance the content of my representation can be “a Greek philosopher” while the object we refer to is sometimes Socrates, sometimes Plato. If it were so, then, because we can vary the objects without varying the content the two would be different from one another. This, however, could not work for non-existent objects, because the concept that becomes the content of the representation has no extension whatsoever. And where there is no object, we obviously cannot vary the objects. In any case, Twardowski already rejects the idea that a concept has more than one object as its reference, even when the objects do exist. He dedicates the whole 15th paragraph to the demonstration of this point. We could resume it by saying that for Twardowski a general representation having a concept for its content doesn’t refer to the singular objects that form the extension of the concept but to a general object (Twardowski 1894: 105/1977: 100). This general object is an inexis-

Objects or Intentional Objects?

� 95

tent entity (Twardowski 1894: 106/1977: 101) formed by what the representations of each singular object that falls under the concept have in common. The variation argument, whether it concerns the content or the object of representations, seems thus problematic for the point Twardowski tries to make: that there is a difference between content and object even when the object does not exist. But even if we eliminate these last two arguments, the first two still stand and illustrate sufficiently that content and object are different in all representations, including those that represent non-existent objects. Now, if we have to maintain a difference between the content and the object of a representation even when this object’s existence is denied, we have to find a way to affirm the object at least as long as we haven’t yet denied its existence. A new type of objects, waiting for the existential judgement that affirms or denies their existence, is the price to pay for having objects and contents for each and every possible representation. These are the intentional objects, which are objects only in virtue of their being presented via a content. They are, as we have seen, nonetheless irreducible to this content even though their only determination is that they are being represented as being so and so. Now, the property of being represented functions here in a modifying sense. What it modifies is the ontological status of the objects considered. A represented object is, as Twardowski himself concedes in his 5th paragraph, not a real object (Twardowski 1894: 24-25). Its existence is modified to what Twardowski calls, following Brentano, phenomenal or intentional existence. However, unlike the conclusion of the discussion of determining and modifying properties, Twardowski will now maintain that this modified existence doesn’t reduce these objects to mere contents of representation. Modifying an object doesn’t automatically turn it into a content, but into another kind of object, an intentional object yet to be distinguished from the content. This is the point I wanted to stress in my analysis of Twardowski’s position. If a first version of the theory is a three-level intentional theory that only brings up two entities besides the act of presenting itself, the content and the object, the discussion of the case of inexistent objects introduces a new entity, the intentional object, which is the object as represented in a modifying sense, turning the theory into a four-level model. This modified object waits for a confirmation or negation that only comes with the existential judgement. But this has consequences also on the case where there is a real object. Even if the object really exists, we only consider its existence when we make a judgement about it. As far as we stay at the level of representations, all objects are simply intentional objects. So now we have four instead of three entities: the act, the content, the intentional represented object and the real objet. It is the intentional object that is presented, while it is the real object that is confirmed by the affirmative existen-

96 � Maria Gyemant tial judgement. Which is of course problematic, because it means that it is not the same object that is presented and judged.¹² To resume Twardowski’s position, we can see now that in order to solve the problem of representations of inexistent objects while maintaining the distinction between content and object, he has to introduce a new category of objects, the intentional objects. These objects are neither contents, nor real references but they come between the Twardowskian versions of Sinn and Bedeutung. If the contents always exist, and the objects can sometimes lack, intentional objects are also present in all representations, yet they cannot be confused with the content. They appear however to slide between the two, sometimes being assimilated to the content, other times to the object. In the rest of this work I would like to briefly recall the points that Husserl criticizes in Twardowski’s thesis and show that despite the very severe critique his position is actually closer to that of Twardowski than it would appear at first sight. However, Husserl’s way of combining a theory of meaning with an intentionalist point of view is much more coherent than Twardowski’s attempt and avoids the multiplication of entities.

3 Husserl’s Version: What is Wrong with Twardowski’s Intentional Objects Twardowski’s mistake is, according Husserl, that, although he has identified the objective character of representations, he has ranged it on the side of reference and not of content. In other words, it is true that each representation has an intentional object. It is not true that this intentional object is a type of object¹³: it is

12 This is one of the main critiques Husserl makes of Twardowski’s position in his text on the Intentional objects of 1894: “Each representation has an object. No doubt this expresses a truth. We call ‘intentional’ existence what this ‘having’ presupposes. And again: There is not an object corresponding to each representation; each one does not found an affirmative existential judgement – since many found a negative one. This too is undeniable; and we call that which the judgement of existence here presupposes ‘true’ existence [...] But some still believe this provides a resolution. They believe it possible to give this distinction a substance such that to each representation there is assigned an immanent object, but not to each a true object. Therefore I cannot help but see here once again that false duplication which also doomed the image theory: The immanent object can (univocally again presupposed) be none other than the true object in the cases where truth corresponds to the representation.” (Husserl 1979 [1894]: 308/1994: 350) 13 “Since such talk is practically unavoidable, even in cases where objects in truth do not exist, or where one must leave undecided their being or non-being, then the best one can do is to differ-

Objects or Intentional Objects? � 97

rather a type of content, moreover the only type of content that really matters in the description of representations: the objective content.¹⁴ For Husserl there are no such things as different types of objects. We cannot distinguish between existent and non-existent objects as if the non-existence of an object were some kind of property referred to an entity that has its own ontological status.¹⁵ Ontologically speaking, a non-existent object is no object at all. What there is in the case of representations of such objects is the representation itself and its own referring character. All representations refer, that is they are all directed towards objects. Yet some of them do not meet their reference in the real world. A representation has its referring character in virtue of its content. Yet once again, Husserl conceives the content differently from Twardowski. In Twardowski the content, even when assimilated to a meaning, was conceived as a psychological occurrence. For Husserl, these psychological occurrences are nothing more than a subjective content that can vary even between representations of the same thing for the same person. There is however another type of content, that lets us identify representations even if they occur at different times, in different persons, with totally different subjective contents: it is what Husserl calls the objective content, the ideal meaning.¹⁶ Each representation has a meaning that is identical in

entiate the mode of expression in such a way that the talk of ‘corresponding’ objects is established as authentic, and that of ‘represented’ objects as inauthentic. The proposition ‘To each representation an object corresponds’, is then false; the proposition ‘Each representation represents an object,’ true. The proper contradictory of the former states: The relevant existential judgement ‘V exists’ is not valid for each representation ‘V’. By contrast, the latter proposition states: Each representation can function under an assumption as if it were an unconditioned representation with respect to its object.” (Husserl 1979 [1894]: 335-336/1994: 375) 14 “However, no matter which are our psychological reasons that determine our way of speaking of represented objects, sometimes authentic, sometimes inauthentic, it is clear that only the signification is an indissoluble and primary determination of the subjective representation.” (Husserl in Schuhmann 1990/91: 165 [our translation]) 15 “It is a serious error to draw a real (reell) distinction between ‘merely immanent’ or ‘intentional’ objects, on the one hand, and ‘transcendent’, ‘actual’ objects, which may correspond to them on the other. [...] It need only be said to be acknowledged that the intentional object of a representation is the same as its actual object, and on occasion as its external object, and that it is absurd to distinguish between them. The transcendent object would not be the object of this representation, if it was not its intentional object.” (Husserl 1913: V, 424-425/2001: 240) 16 “It is regrettable that the author did not, right at the outset of the treatise, investigate the distinction between ‘auxiliary representations’ and significations with the conscientiousness otherwise characteristic of him. Certainly then his concept of content would have been different, and he never would have come to confuse, in such an inadmissible fashion, constituents of the objective concept with subjective ‘auxiliary representations’.” (Husserl 1979 [1896]: 356/1994: 395)

98 � Maria Gyemant all representations of the same object. In other words, there is something objective in the content itself: we do not have to go look for objectivity outside the representation. It is this objective content that makes us say that two people have the same representation. They both mean the same thing no matter how different are the images that come up in their minds. So even where there is no reference, there always is an objective content. We always mean something, but what we mean sometimes doesn’t exist. This objective content comes in Husserl’s theory to play the same part intentional objects played in Twardowski. The advantage in Husserl is that because it is not ranged on the side of reference, but on the side of the content, we do not need a special ontological status for them. We don’t need to say that inexistent objects exist in any way. For Husserl then all objects are intentional if we look at them in a certain way. But intentional objects are actually objective contents of the representation. On the other hand, as he insists in his 5th Logical Investigation, no object, whether real, imaginary or impossible, is a content of the act of representation directed towards it.¹⁷ As a reference, the object can never be immanent to the act, the way intentional objects were for Twardowski. Nevertheless, the fact that each representation refers to an object is precisely what is objective about its content. The referring to the object is part of the act and constitutes its objective content, and this is true even when the referent, the object we refer to, does not exist at all. This theory will however only come to a final form with the noesis-noema distinction discovered by Husserl around 1907.

4 Conclusion Twardowski and Husserl share a similar project. They both consider reference as correlative to mental acts of representation rather than to linguistic terms, yet they try to solve the problem of occasional lack of reference by introducing intermediary entities. The entities Twardowski introduces are however problematic because they are on the side of reference and need a particular ontological status. Husserl solves this problem by considering the fact of referring as different from the fact of actually having a reference. Thus, the objective content of representations only

17 “If I represent God to myself, or an angel, or an intelligible thing-in-itself, or a physical thing or a round square etc., I mean the transcendent object named in each case, in other words my intentional object: it makes no difference whether this object exists or is imaginary or absurd.” (Husserl 1913: V, 425/2001: 240)

Objects or Intentional Objects? �

99

tells us what the object would be if it existed, but has no ontological implications concerning this object. It could be argued that the objective content at this early stage of the theory is not really an intermediary entity either, and even when it turns into the noema of Husserl’s later phenomenology it can still be said that it is not really an entity at all, and that the model stays a two-level Brentanian, or even a one-level model. But this is a subject for further discussion.

Dale Jacquette

Domain Comprehension in Meinongian Object Theory Dale Jacquette: University of Bern, Switzerland

1 Elements of Object Theory Alexius Meinong developed his object theory as a continuation of his teacher Franz Brentano’s thesis that the intentionality of thought uniquely distinguishes psychological from purely physical phenomena.¹ Brentano in his 1874 Psychologie vom empirischen Standpunkt taught that psychological occurrences are essentially marked by their aboutness or direction upon intended objects.² Belief, doubt, hope, fear, love, hate, expectation, dread, and the like, are all about something, which is their intended object. We believe, doubt, hope, fear something. Nonpsychological entities and events in contrast are never in and of themselves intrinsically about anything at all. The moon is not about anything; it intends no object. Neither does a waterfall, a rainbow, a neurophysiological electro-chemical brain event, or any other purely physical phenomenon, considered only in and of itself.³ Meinong in his Gegenstandstheorie boldly goes in a direction that Brentano did not approve. He reasons that if all psychological occurrences are intentional, intending some object or other, then some thoughts must be about objects that do not happen to exist, or that even in some sense cannot possibly exist. If I entertain the proposition that Sherlock Holmes is a detective, then, true or false, my thought seems ostensibly to be about Sherlock Holmes, despite the fact that

1 A discussion with bibliographical references of Meinong’s object theory relation to Brentano is found in (Jacquette 1990/91). 2 “Every psychic phenomenon is characterized by what the Scholastics of the Middle Ages called intentional (also indeed mental) in-existence of an object, and which we, although not with an entirely unambiguous expression, will call the relation to a content, the direction toward an object (by which here a reality is not understood), or an immanent objectivity. Every [psychic phenomenon] contains something as an object within itself, though not every one in the same way. In presentation something is presented, in judgment something acknowledged or rejected, in love loved, in hate hated, in desire desired, and so on.” (Brentano 1874: 115 [translation by Jacquette]) 3 See (Jacquette, 2004), (Jacquette 2001b), (Jacquette 2002a) and (Jacquette 2006).

102 � Dale Jacquette Sherlock Holmes is a nonexistent fictional character (Jacquette 1989).⁴ Intention alone is certainly not powerful enough to create or imply existence. The question is only what we should say concerning the logic, semantics, philosophy of language, and philosophy of mind surrounding the intending especially of nonexistent objects. If we are thorough-going extensionalists, as the dominant tradition in analytic philosophy has powerfully and often unthinkingly encouraged, then we may staunchly deny that the meaning of thought, language, art and other artifacts, can ever require a reference domain consisting of anything other than actually existing objects. These existents can be intended, and none others. Brentano himself, early and late, seems to have adopted two different versions of this view, possibly as a reflection of his deep-reaching Aristotelianism and commitment to primary substances as the only and ultimately existent individuals, also reflected in his later metaphysical doctrine of reism. Reism is Brentano’s later Aristotelianism, that only individual existent physical entities can be the subject of true or false predications (Jacquette 1990/91).⁵ Meinong took his philosophy in a different direction toward object theory. Brentano’s path requires ingenious and not always natural paraphrase of discourse that still appears to be as much if not more so about nonexistent as it is about existent objects. Meinong proceeds ontically indiscriminately, democratically, where the ontic status of the ostensible intended objects of thought are concerned. He does not need to know whether an object of thought exists or does not, in order to know that it is a particular object of a particular thought, and potentially of many thoughts. Existent or nonexistent, they are the intended objects of thought, and they must have adequate identity conditions. Meinong generalizes Leibniz’s intensional, constitutive property-based, law of the identity of indiscernibles and indiscernibility of identicals to all intended objects, with the result that all objects, regardless of their ontic status, are individualized by virtue of their possession of and correspondence with a particular choice of properties. Property-based intensional identity conditions are available for all ostensible intended objects, without consideration of their ontic status. After all, how do we really know what’s out there? The advantage of a Meinongian object theory is seen in the logical possibility that our clearest and most distinct perceptions do not actually correspond to anything existent in an external world outside the contents of our thoughts. Even so, in that extreme case we would continue to understand

4 See also (Jacquette 1996: 256-264). 5 On Brentano’s Aristotelianism in ontology, see (Jacquette 2011a). Concerning Brentano’s later reism, see especially his posthumously edited (Mayer-Hillebrand 1966); and including a Letter from Brentano to Anton Marty, April 20, 1910 (Brentano 1966: 225-228).

Domain Comprehension in Meinongian Object Theory �

103

the meaning of our thoughts ostensibly about nonexistent objects. You will not think I am gabbling gibberish, if I try to start up a conversation about Sherlock Holmes. The advantage of Meinong’s ontically agnostic approach to the meaning of thought and its expression is that, by agreeing with Brentano that thoughts generally intend objects, then thoughts that appear to be about nonexistent objects, possible and even constitutively impossible, have a presumption also of intending nonexistent objects, including, for starts, the notorious golden mountain and round square. Meinong held as a consequence of these assumptions that a reference domain must include not only existent objects, spatiotemporal and abstract, if such there be, but also “homeless” objects (heimatlose Gegenstände) that do not and cannot exist, but that in their logical role as purely intended objects, are ontically neutral, beyond being and nonbeing (jenseits von Sein und Nichtsein).⁶

2 Meinongian Intensional versus Fregean Extensional Reference Domains The advantages of a Meinongian object theory are evident in comparison with the poverty of a purely extensionalist semantics, and complementary physically reductivist or eliminativist philosophy of mind. Meinong opens the door to a wide range of intensionalist possibilities in logic and language that were suggested but never systematically pursued by Brentano’s intentionalism. With a Meinongian object theory at the foundation of logic in place of a Fregean Begriffsschrift reference domain, we can do parallel semantic justice to scientific theories that falsely posit the existence of actually nonexistent objects, whether as idealizations or posited in false explanatory hypotheses. Examples of the falsely posited are usually conceded to include vortices, phlogiston, the planet Vulcan, the Philosopher’s Stone, the æther, teleologies.⁷ There are also an extraordinary number of obligatory nonexistent ideal objects. Idealizations of many kinds are encountered even in the most rigorous contemporary physical science. They are found already in kinematics, in the physics of moving projectiles unimpeded by impressed forces. No such entities actually exist, because all bodies in motion are in fact impeded by impressed forces, beginning with universal gravity. Applied mathematical laws would scarcely be dis-

6 Meinong offers a frequently consulted overview of his mature object theory in his (1904/1960). See (Jacquette 2001a). 7 See (Jacquette 1992).

104 � Dale Jacquette cernible in observed actual phenomena if scientists did not smooth the edges off their acquired data of actual phenomena on grounds of practical measurement discrepancies or compromising background factors. A simplifying continuous curve represented by an elegant function is superimposed on a varia of actual measurements seen as revealing an essential underlying lawlike commonality. The practice itself is not criticized, underscoring the fact that nonexistent entities including false hypotheses and idealizations in the formulation of applied mathematical natural laws are rife in scientific discourse. We cannot understand the language of science without a correct semantics. A purely extensionalist semantics with a Fregean existence-presuppositional reference domain does not seem adequate, whereas an intensionalist semantics, defining objects as any logically possible combination of properties, comprehends and accommodates phlogiston as one nonexistent thing distinct from the Philosopher’s Stone, or the planet Vulcan, among other ostensibly distinct nonexistent things, as well as from every existent thing. We cannot otherwise hope to understand the history of science as a progression of distinct true and false ideas and idealizations (Jacquette 1996: 238-255). Meinongian logic and semantics, in comparison again with a Fregean extensionalist reference domain, additionally holds out the prospect of offering a more natural and satisfying explanation of the distinct meanings of imaginative works of fiction. David Hume famously writes that there need be no discernible textual differences between some works of history and novels: “[Imagination expressed in works of fiction] can feign a train of events with all the appearance of reality, ascribe to them a particular time and place, conceive them as existent, and paint them out to itself with every circumstance, that belongs to any historical fact” (Hume 1975 [1748]: Section V, Part II, §39, 47). The difference is only that histories in principle are supposed to contain only truths concerning existent objects, whereas novels collectively represent at least some events and objects that never existed or occurred. A reader, without proper cues or background information, relying entirely on the internal content of a book, might never know which. Fictions, unlike histories, are the products of a freely imaginative combination of properties in creating the characters, settings, events and plots in which they participate. The writer of a history as much as a novel has a narrative voice, but a history means to speak only of existent things, persons and occurrences, whereas, unless seriously deluded, the author of a fiction may purport to represent reality, but understands that, despite possible mention of some external existent things, the references of the story imparted are to distinct nonexistent objects, that are not part of the actual world. A Fregean reference domain in a semantics for interpreting the sentences of a writing in fiction, like that invoked in the semantics of false and ideal scientific ex-

Domain Comprehension in Meinongian Object Theory �

105

pressions, cannot extensionally distinguish one nonexistent intended object from another. If there is to be any discriminating determination of the meaning, then it is an intensional Meinongian logic rather than extensional Fregean logic that is needed for the proper interpretation of the false sentences of a work of fiction.⁸ Nor should we lose sight of the fact that histories, like scientific discourse, not only often incorporate but are themselves nonexistent idealizations. It is a naive idealization in the first place to suppose that histories can be interpreted by means of a Fregean extensionalist semantics because unlike fictions they contain only true propositions concerning existent objects. There are frequent disputes in historical research as to whether a certain person or event existed, whether Troy or King Arthur of Great Britain actually existed, and if so whether there was truly a court at Camelot with the participants known to legend, whether Moses of the Old Testament was a real person or a later composite of historical leaders, and countless other things about which there can be disagreement as to the facts of their ontic status. There are ostensible references aplenty to nonexistent objects on both sides of such disputes, regardless of who turns out to be right, or whether the matter is ever finally settled. We expect both parties to such disagreement to be making meaningful pronouncements, which must be the case in order for one of them finally to be defending a true proposition, and for opponents to advance a false but still equally meaningful contrary proposition. The same considerations apply potentially to religious discourse that happens ostensibly to refer to nonexistent objects. Meinong rather than Frege offers the more general, and naturally applicable logic and semantics. The Meinongian reference domain of existent, actual and abstract, and nonexistent, alike metaphysically possible and impossible objects, incorporates everything that can be said in an extensionalist existence-presuppositional reference domain, and so much more besides. The advantage of Meinong’s intensionalist semantics is that it incorporates the more restricted Fregean extensionalist semantics as a proper part. We can intend existent as well as nonexistent objects in Meinongian logic. Any semantics charged with explaining the meaning conditions of large parts of putatively meaningful discourse outside of the logical foundations of mathematics or the true or false scientific description of logically contingent existent states of affairs must go beyond the limitations of the purely extensional semantics available in a Fregean reference domain.⁹

8 See (Cartwright 1954), (Chomsky & Scheffler 1958) and (Jubien 1972). 9 For criticisms of Fregean extensionalist existence-presuppositional reference domains see (Jacquette 2010: especially 22-140) and (Jacquette 2011c).

106 � Dale Jacquette This battle remains hard fought. Philosophers continue to propose ingenious but ultimately implausible paraphrases of discourse putatively about nonexistent objects, in order to avoid reference to nonexistents. These extensional efforts have not met with impressive success, and usually end up offering informal words of consolation about the need to limit meaning and live without the literal meaningfulness of considerable parts of discourse that are otherwise considered not only nominally meaningful, but true or false. If Hume’s insight is right, then it is more reasonable to suppose that there should be an exactly parallel semantics for expressions that purport to make true or false assertions both about existent and nonexistent objects. Such a mirror image account of the meaning of ostensible reference and true predication of properties to both existent and nonexistent objects in science, history and fiction, makes it obvious that only a version of Meinong’s object theory can possibly succeed, whereas Frege’s purely extensionalist existence presuppositional-reference domain will not distinguish and hence not adequately comprehend the nonexistent objects assumed by false science, false history, fiction, myth and religion. If we do not want to be driven into admitting that it is false or meaningless to say that Sherlock Holmes is a detective, or that a flying horse is a horse, that nonexistent objects cannot truly have any constitutive properties, then we must graduate from an existence-restricted Fregean reference domain to an intensional Meinongian reference domain that includes an existent or nonexistent object for every logically possible combination of constitutive properties. Some combinations constitute all the existent objects already included in a Fregean reference domain, and others all the nonexistent objects excluded from a Fregean reference domain. There is nothing we can do in philosophical logic and semantics with a Fregean reference domain that we cannot also and equally do by means of a more comprehensive Meinongian reference domain. Whereas there are many valuable logical and semantic services provided by a Meinongian reference domain that cannot be satisfied by a Fregean reference domain (Jacquette 1998; and 2011b).

3 Comprehension Principle for Meinongian Object Theory The development of a Meinongian logic and semantics of existence and nonexistence requires a comprehension principle by which all and only the existent and nonexistent objects are included in a reference domain. This is more challenging than it may at first appear. There are hidden difficulties and hazards in trying to

Domain Comprehension in Meinongian Object Theory

� 107

specify precisely what objects are to be included in the Meinongian domain, and on what basis other candidate objects may need to be excluded. In the counterpart Fregean reference domain we have described as a foil to the Meinongian, the division is more straightforward. We include all and only the existent objects, actual and abstract, excluding all others, and the work is done. The Meinongian domain on the contrary is in one sense easily described at a high level of generality and informality, but poses difficulties when specific applications are considered. We can say without fear of contradiction that the Meinongian domain contains every logically possible intended object, everything we can think about, existent or nonexistent. True as far as it goes, although the formula leaves open an assortment of recalcitrant cases. We must ask in the first place what are the objects we can think about, what things and kinds of things are to be included among the intended objects of logically possible intentions. A correct comprehension principle for Meinongian object theory must include all and only the existent and nonexistent objects that are distinguishable from one another by virtue of differences in their properties. This is to say as a consequence of their intensional identity conditions. In Meinong’s mature formulation of object theory, he seems to think it almost a tautology that the objects can only be distinguished by their specifically constitutive properties.

4 Russell’s Problem of the Existent Golden Mountain An intensional semantics is based on properties. It defines objects in terms of particular choices of combinations from among the aggregate of all logically possible properties. Whereas an extensional semantics begins with the existent objects and prescribes truth conditions for propositions concerning existent objects as included in or excluded from the extensions of all existent objects possessing a certain property, an intensional semantics comprehends all combinations of properties, some of which will then turn out to belong to existent objects, like Napoleon and the Taj Mahal, and others will turn out to correspond to metaphysically contingently or necessarily nonexistent objects, like Sherlock Holmes, the golden mountain and round square. The suggestion is that a Meinongian reference domain comprehends existent and nonexistent objects alike corresponding to any and every combination of properties. Unfortunately, things are not this simple. We cannot include an intended object in a Meinongian reference domain for any and every combination

108 � Dale Jacquette of properties, for then we could have such combinations of unfiltered properties as [existent, golden, mountain] and [metaphysically possible, round, square]. Bertrand Russell first called attention to this problem in Meinong’s object theory (Russell 1905a: 45; and 1905b). As we work toward increasingly discriminating refinements of the Meinongian comprehension principle to shore it up against counterexamples like Russell’s, as it is possible to do, we stray further and further from the theory’s intuitive intentionalist basis. We can think of any object we like by putting together a collection of properties, but not any and all kinds of properties, and not in combination with certain other kinds of properties. Meinong’s solution has often been criticized as an abject effort to patch up an account that is too general and comprehensive for its own good. The desired fit with naive phenomenological expectations and the meaning of false and idealized science, history and fiction, myth and religious discourse, is made increasingly elusive as we put pressure on what presumably intended objects that can and those that should not be included in a Meinongian reference domain.

5 Meinong’s Watering-Down of the Modal Moment Solution How are we to exclude counterexamples like the existent golden mountain? Meinong wants to say that the golden mountain is golden and a mountain, albeit a nonexistent mountain, whereas the existent golden mountain is not golden, a mountain, albeit a nonexistent mountain and existent. Meinong no more courts outright logical inconsistency than any other sane philosopher.¹⁰ Meinong’s so-called independence of Sosein from Sein thesis bestows propertyhood even on nonexistent objects (Meinong 1910/1983). If the thesis is interpreted as applying also to stipulative existence predications for manifestly nonexistent intended objects, then nothing precludes excessive intensionalist comprehension of objects like Russell’s existent golden mountain in the reference domain. The existent golden mountain would appear moreover to be truly existent, just as it is truly supposed to be golden and a mountain, despite the fact that no golden mountain exists. If we wanted to include such synthetic experiments as the existent golden mountain or metaphysically possible round square in a Meinongian reference domain, since we seem to be referring to them in some sort even

10 Meinong’s solution to Russell’s problem of the existent golden mountain is presented in (Meinong 1973 [1907]: 278-282).

Domain Comprehension in Meinongian Object Theory �

109

now, we might try to do so by establishing a special category of nonexistent objects including any corresponding to combinations of properties including or implying the property of existence. For these, the existent golden mountain is golden, a mountain, and predicatively but not actually existent. The idea of being predicatively existent could be developed as merely said-to-be-existent, in a kind of Quinean semantic ascent leaving it open in individual applications whether the otherwise designated intended object actually exists or fails to exist.¹¹ Meinong offers something like this solution in his distinction between a property possessing or failing to possess the modal moment (das Modalmoment), without which a property is said to be watered-down, depotenziert.¹² A predicatively existent existent golden mountain in this terminology implies only watered-down existence lacking what Meinong calls full-strength factuality, since manifestly no golden mountain actually exists. To have the modal moment, a property in an intensional said-to-be combination of properties must actually have the property, and not merely be said to have it, predicatively only or lacking in the modal moment of full-strength factuality.¹³ The existent golden mountain is then not numerically identical to the golden mountain, according to intensional identity conditions, based on explicit differences in objects’ associated property combinations, and we can be relieved that the existent golden mountain does not actually exist. It is the golden mountain that is (falsely) said to exist, as opposed to the golden mountain that is (truly) said not to exist, [nonexistent, golden, mountain], and as opposed to the golden mountain without further qualification, [golden, mountain], of which neither existence nor nonexistence is predicated. We can distinguish the existent golden mountain from all other Meinongian objects by its possession of watered-down existence, without suffering the inconvenience of supposing that the existent golden mountain, lacking the modal moment of full-strength factuality, confounds the facts of the world by actually existing. Russell historically lost patience with Meinong at this point. Relying on his distinction between properties possessing and lacking the modal moment, Meinong replies to Russell that the existent golden mountain is existent, even

11 See (Quine 1960: 271-276), on the concept of semantic ascent. 12 See (Routley 1980: 496), (Jacquette 1986) and (Jacquette 1996: 80-91). 13 Meinong’s concept of the modal moment and watering-down extranuclear properties to nuclear versions is presented in (Meinong 1972 [1915]: 266). See also (Findlay 1995 [1963]: 103-104 [all references to this edition]).

110 � Dale Jacquette though it does not exist.¹⁴ Meinong means that the existent golden mountain has a watered-down existence simulacrum property of existence in its Sosein. As such, the [existent, golden, mountain] is said to be existent or to have existence predicated of it, making it different intensionally from all other intended objects. Lacking the modal moment of full-strength factuality, the property of existence included as an intensional identity condition for the intended object, the existent golden mountain, does not imply that the existent golden mountain actually exists, but rather allows without logical inconsistency that nonexistent [existent, golden, mountain]. If we do not make a principled distinction between these two technical senses of the word “existent”, as Meinong does, then to answer Russell’s problem as Meinong does looks indeed to be errant nonsense. Nonsense it nevertheless most emphatically is not, although it may not be the best solution to Russell’s problem of the existent golden mountain (Routley 1980: 496).¹⁵

6 Constitutive versus Extra-Constitutive Properties Solution An alternative way to constrain the Meinongian reference domain to exclude nonexistent objects said to be existent, is to outlaw comprehension for certain categories of properties. The distinctions become increasingly complex, although they are arguably unavoidable for a fully adequate true-to-its-roots intentionalist phenomenological Meinongian object theory. The restriction required to avoid logical inconsistency allows an object into the Meinongian reference domain only provided that its properties belong to one distinguished category rather than

14 Rudolf Carnap discusses Russell’s objection to Meinong’s object theory comprehension in (Carnap 1956 [1947]: 65). 15 “[L]ogically important though the modal moment is, the [nuclear – extranuclear] property distinction alone, properly applied, is enough to meet all objections to theories of objects based on illegitimate appeals to the Characterisation Postulate [Routley’s version of Meinong’s thesis of the Independence of Sosein (so-being) from Sein (being)]. The Meinong whose theory includes an unrestricted Characterisation Postulate is accordingly, like Meinong the super-platonist, a mythological Meinong.” I once thought that the constitutive versus extra-constitutive property distinction was sufficient to forestall Russell’s problem of the existent golden mountain and its variants, but as the present argument makes clear, I no longer believe that object theory can be adequately defended without combining the constitutive and extra-constitutive property distinction with Meinong’s watering-down of extra-constitutive properties lacking the modal moment of full-strength factuality.

Domain Comprehension in Meinongian Object Theory �

111

another. This discrimination inevitably invites questions, counterexamples, and grey area cases that bring the logical integrity and intuitive phenomenological appeal of the project into further doubt. The essential distinction is already introduced by Meinong as that between constitutive and extra-constitutive properties. These are also known after J.N. Findlay’s classic study and in his translation respectively as nuclear and extra-nuclear properties, although we shall largely ignore the latter terminology here as less descriptive.¹⁶ Constitutive or nuclear properties are the ordinary properties like being red, round, rude and their complements; extra-constitutive or extra-nuclear properties are properties with ontic implications that are not assumable according to Meinong in establishing intensional identity conditions for particular objects comprehended in the Meinongian reference domain. The excluded properties, as suggested by Russell’s problem of the existent golden mountain, being existent, possible, complete, consistent, and the like, that have no more than what the previously considered solution designated as the predicative status of the property of having been said to exist, be possible, complete, consistent, and so on. Here we are disappointed to find that Meinong does not offer a clearcut basis for distinguishing constitutive from extra-constitutive properties, which means that the Meinongian reference domain determined by intensional identity conditions is not fixed. This is a serious problem, although Meinong may personally not have been aware of or bothered by it, or regarded it as a project for future object theory proponents to resolve (Meinong 1972 [1915]: 176-177). Several criteria have been proposed to distinguish constitutive from extraconstitutive properties. If this can be done, then we need only consider intensionally all possible combinations of constitutive properties in order to answer the question of exactly what intended objects are comprehended as belonging to the Meinongian reference domain. Then the domain includes the golden mountain [golden, mountain], but does not include a distinct existent golden mountain as an [existent, golden, mountain]. The latter combination is precluded, because the permissible combinations are taken only from among the set of all constitutive properties, whereas being existent is paradigmatically extra-constitutive. The trouble is that if we have no good way to sort out the constitutive from extra-constitutive properties, then we do not know exactly which property combinations generally comprehend objects and which do not. We can say in general that an extra-constitutive property is one the complement predication of which is

16 Findlay (1995 [1963]: 176) proposes the English equivalents “nuclear” and “extranuclear” for Meinong’s distinction between konstitutorische and ausserkonstitutorische Bestimmungen. See (Jacquette 2001a).

112 � Dale Jacquette logically equivalent to the predication’s negation, that being nonexistent is logically equivalent to the property of not being existent, whereas the constitutive property of being nonround is not logically equivalent to the property of not being round. The Meinongian round square is round and nonround, by virtue of being square, but the round square is not such that it is both the case that it is round and it is not the case that it is round. The trouble remains of deciding in which cases the decisive logical equivalence holds and in which cases it does not, and there are some interesting ambivalent cases. We run up against difficulties in applying the constitutive versus extraconstitutive distinction to exclude the existent golden mountain as an intended object in the Meinongian reference domain distinct from the further unqualified golden mountain or golden mountain simpliciter. What shall we say of the [unintendable, golden, mountain]? Is it a distinct intended object? Is the property of being intendable or its complement of being unintendable constitutive or extra-constitutive? Or, in the case of the [uncomprehended, golden, mountain]? Meinong’s more logically inclined student Ernst Mally seems to have thought that the property of being unintended or un-thought-of was assumable and in that sense nuclear or constitutive, in his 1914 essay, “Über die Unabhängigkeit der Gegenstände vom Denken”. It is on the basis of such intensional comprehension that Mally argues that the Meinongian reference domain is independent of the logical contingencies of the actual intendings of thinkers intending objects in real time psychological episodes, because it must include unintended objects for which by definition no mind can possibly be responsible. The Meinongian reference domain thereby acquires an objectivity that it would otherwise lack if its comprehension principle were limited to the objects that real thinkers have actually intended, that have stood in intending relations to a thinker’s thought and its expression (Mally 1914/1989).¹⁷

7 Converse Intentional Properties as Intensional Identity Conditions Suppose someone writes a novel in which a detective alerts a criminal by posting two messages, one of which was intended and the other of which was unintended. This looks to be a difference in a Meinongian object’s converse intentional proper-

17 Meinong refers to Mally’s constructions like the unintended golden mountain as “defekte Gegenstände”. See (Jacquette 1982).

Domain Comprehension in Meinongian Object Theory

� 113

ties.¹⁸ In understanding the events of the novel, we may want to make reference to the fictional detective’s unintended as opposed to his intended message. For this purpose, we have no better provision than to distinguish between the two nonexistent objects: [intended, message, posted by the detective] and [unintended, message, posted by the detective]. Such examples argue for including the property of being unintended in at least some property combinations applied in an intensional Meinongian reference domain comprehension principle.¹⁹ All that seems fair enough. However, there are other applications in which it appears more natural and even essential in understanding the meaning of discourse to classify the property of being unintended, un-thought-of, or even uncomprehended, as extra-constitutive or extra-nuclear rather than nuclear or constitutive. An example is to speak of an unintended round square. Is the unintended round square unintended? One might say that it cannot be unintended, for then paradoxically it is intended in the very consideration of the question. If we are not availing ourselves of Meinong’s other solution involving a distinction between egalitarian properties with or without the modal moment, then we need to exclude at least some unintended objects from the Meinongian reference domain. Under the present proposal, there is no way to accomplish this discrimination except by classifying the property of being unintended as extra-constitutive

18 I adopt Roderick M. Chisholm’s terminology, from his essay (1982b). If I love Paris, then I have the intentional property of loving Paris, and Paris has the converse intentional property of being loved by me. One question about Meinongian domain comprehension has been whether converse intentional properties under any name are constitutive or extra-constitutive, and if one or the other what we should think of their contribution to their exemplification by intended objects that seem to be individuated by reference to what may be their unexemplified extra-constitutive converse intentional properties. The [unintended, golden, mountain] of Mally’s argument combines in its intensional identity conditions the constitutive properties of being golden and a mountain, and the extra-constitutive property of being unintended. If we hope like Mally to establish the mind-independent objectivity of the Meinongian object theory reference domain by virtue of the unintendedness of the freely assumed unintended golden mountain, then we may agree with Mally that converse intentional properties, or perhaps only some of them, are constitutive or encoded rather than extra-constitutive or exemplified by Meinongian objects. See note 19 immediately below. 19 I previously proposed classifying converse intentional properties as constitutive or nuclear in order to save Mally’s argument for the mind-independent objectivity of the Meinongian object theory domain. See, for example, (Jacquette 1996: 73-78). I have since reversed my opinion and now regard converse intentional properties as extra-constitutive. My reason is an increased appreciation for the fact that extra-constitutive properties sometimes need to be included among an intended object’s intensional identity conditions, but that the actual exemplification of extraconstitutive properties does not follow logically from their identity-serving function. That Sosein is independent of Sein for Meinong is a two-way street.

114 � Dale Jacquette rather than constitutive. We have already seen that there are some applications like the detective posting two messages in which on the contrary it seems more correct to classify the property of being unintended as a constitutive rather than extra-constitutive property. Nor is the property of being unintended the only difficult case. What are we to say of the unexemplified color? Colors are sometimes invented that previously did not exist in nature, but result only from previously untried chemical processes or combinations of pre-existent colors.²⁰ Prior to their invention, these colors are unexemplified. So we should be able intelligibly to speak of an unexemplified color. While we are laboring to create the new color, we identify it for referential purposes intensionally as the property of having certain constitutive chromatic values C and of being unexemplified, [unexemplified, chromatic values C]. What’s good for the goose is good for the gander. So, if being unexemplified is constitutive, then so must be the complementary property of being exemplified. This means just as in the [existent, golden, mountain] we should be able to postulate the comprehension of an object in the Meinongian reference domain defined as [exemplified, property of choice], whereby we stipulatively conjure into existence any property we choose. If it works, then we can all exemplify the property of being millionaires. Since the attempt does not succeed, we must reconsider the proposal of treating the property of being unexemplified in the case of a specific color as constitutive rather than extra-constitutive. If the color is created at a certain time, then we may nevertheless want to distinguish constitutively between the previously unexemplified and later exemplified color with its specific set of chromatic values C. When we investigate the scope and limits of the distinction between constitutive and extra-constitutive properties, we meet with an unexpected and somewhat unsystematic mixture of properties that do not readily lend themselves to classification exclusively as either constitutive or extra-constitutive. Some applications seem constitutive, and others extra-constitutive. The applications themselves can be relativized readily and reasonably enough. The trouble is that the basis for relativizing some of these dual aspect properties is the effect of their combination with other properties. As such, they cannot freely and independently enter into combination with other exclusively constitutive properties in establishing inten-

20 There is a fascinating history of the invention of what were then new colors. It does not happen often, and it requires work to find something not already present in nature, but it is possible a fortiori because it has happened historically in the development of synthetic chemistry in the case of Prussian blue and other colors. See such classics as (Scheele 1966 [1778]) and (Coleby 1939).

Domain Comprehension in Meinongian Object Theory

� 115

sional identity conditions for objects in the Meinongian reference domain. Each logically possible combination would need to be considered individually in order to determine in that case whether the dual aspect property were behaving constitutively or extra-constitutively in that combination. The requirement is evidently incompatible not only in practical application, but with the very theoretical concept of a comprehension principle for a Meinongian object theory. These distinctions cannot be subjectively nuanced or negotiated in a theory like Meinong’s, but must stand on well-founded objective principles. Meinong’s is not a theory of how or what people think, but of what they can think. They can think descriptively or under nominalizations of any object associated with any combination of constitutive properties as their distinguishing intensional identity conditions. The theory at one level or another after a series of retreats from its initial intuitive appeal and apparent generality, leaves a trail of further qualifications and distinctions that finally do not belong either to the category of constitutive or extra-constitutive properties univocally. One seemingly ad hoc provision is added to shore up object theory against inconsistencies arising primarily from Meinong’s assumption that thought is free to intend any object distinctly constituted by any choice of constitutive properties. The unlimited freedom of assumption (unbeschränkte Annahmefreiheit) together with the object theory’s intensional identity presuppositions, pairing combinations of constitutive properties with objects in a semantic reference domain regardless of their ontic status, are supposed to comprehend all of the existent and nonexistent intended objects, actual, abstract and beingless, metaphysically possible and impossible. The Meinongian reference domain consequently cannot be properly comprehended without a prior correct division of all and only constitutive from all and only extraconstitutive properties (Meinong 1910: 346-347; and 1972 [1915]: 283).

8 Synthesis of Alternative Complementary Solutions to Russell’s Problem What about such null intensional property combinations as the object lacking any constitutive properties? It seems to be something we can think about, so it should be included by any adequate comprehension principle in the Meinongian object theory reference domain. As such, it must have its own distinguishing intensional identity conditions, which in the nature of the case cannot include any constitutive properties. It must be different as an intended object from any objects possessing at least one constitutive property, unless implausibly it is ranked not as

116 � Dale Jacquette an intendable object at all, and hence to be excluded altogether from the Meinongian object theory reference domain. The dilemma is that either we must conclude that we cannot think about such an object, or we must do so by means of unexemplified extra-constitutive properties by which the object is nevertheless intensionally identified. From any direction, the first option seems unjustifiably restrictive. Why, in thinking about the object lacking all constitutive properties, should intensional identity conditions break down? We cannot distinguish the object lacking all constitutive properties from other objects about which we might also think. Otherwise, we can distinguish such objects only by including extra-constitutive properties among their intensional identity conditions. In the latter case, we step away from the Meinongian principle that the objects are comprehended by identity conditions associating a distinct object with every unique combination exclusively of constitutive properties. Meinong’s own distinction between properties possessing or lacking the modal moment of full-strength factuality can be understood as making precisely this latter compromise. To challenge this Meinong-inspired picture, we might consider existence as a supposedly clearcut case of an extra-constitutive property. Since distinct Meinongian objects in Meinong’s historical development of object theory can only be individuated and differentiated by means of differences in their respective constitutive properties, existence is excluded. The awkward alternative seems to be that Russell’s existent golden mountain is existent, and hence that a golden mountain exists. We do not expect the paradigmatically extra-constitutive property of existence to enter into any Meinongian object’s Sosein of identifying and individuating so-being. Imagine that someone freely creates a fiction in which a detective describes two villains, one of whom within the story is said to exist, and the other of whom within the same story again is said not to exist. Such a distinction might even be crucial to understanding the meaning of the story as a whole, including key elements of plot. How are we to distinguish between the references made in the fiction to the existent and nonexistent villains, without including some sort of extra-constitutive existence and nonexistence properties, as among the intensional identity conditions for these fictional intended objects? If knowledge of the two villains is as limited as this, then it can happen that only [existent, villain]  = [nonexistent, villain]; otherwise, if existence and nonexistence are precluded from application of the intensional identity conditions for comprehension of objects in an object theory reference domain, then we are paradoxically driven to denying the reflexivity of identity, [villain]  = [villain]. Accordingly, there appears no better alternative than to adopt some version of Meinong’s original solution to Russell’s problem of the existent golden mountain.

Domain Comprehension in Meinongian Object Theory � 117

We should consider the distinction between extra-constitutive properties with or without the modal moment. We can mark the difference symbolically by making an exception for extra-constitutive e-properties imported into the bracketed intensional identity conditions for a special range of nonexistent Meinongian objects. Whereas ordinarily for constitutive c-properties we can write that the object identified by the so-being of constitutive properties Pc1, Pc2 has in truth both of those properties, indicated by the brackets and ⇒ double arrow. [Pc1, Pc2] ⇒ Pc1[Pc1, Pc2] says that if there is an object in the object theory domain with intensional identity conditions Pc1, Pc2, then that object truly possesses property Pc1. Similarly: [Pc1, Pc2] ⇒ Pc2[Pc1, Pc2]

AUTHORIZED

Alternatively, where extra-constitutive properties Pe1, Pe2, etc., are recruited as intensional identity conditions, no such inference is authorized. Thus, we have: [Pc1, Pe1] ⇒ Pc1[Pc1, Pe1]

AUTHORIZED

[Pc1, Pe1] ⇒ Pe1[Pc1, Pe1]

UNAUTHORIZED

But not:

An extra-constitutive property enclosed within the brackets lacks what Meinong speaks of as the modal moment of full-strength factuality, but is on the contrary a modally watered-down counterpart of an extra-constitutive property. With respect to Russell’s original problem of the existent golden mountain, we can then say that: [existent, golden, mountain] ⇒ golden[existent, golden, mountain] And: [existent, golden, mountain] ⇒ mountain[existent, golden, mountain] But not: [existent, golden, mountain] ⇒ existent[existent, golden, mountain] We cannot deduce that the existent golden mountain exists from the object’s uniquely individuating intensional identity conditions. We are free to maintain on independent object theory grounds that the existent golden mountain does not exist, if it is a fact that: nonexistent[existent, golden, mountain] Meinong’s answer to Russell’s problem of the existent golden mountain that it is existent even though it does not exist, therefore, need not be rejected out of

118 � Dale Jacquette hand. It can be provided with not only an intelligible but justifiable background interpretation. There is no inconsistency in the proposition that the existent golden mountain is non-modally-momentous existent, even though it does not modallymomentously exist. It is only a matter of keeping the intensional identity conditions rich enough to distinguish any phenomenologically distinct intended objects on the basis of differences in their individuating properties, even if these should turn out to include extra-constitutive properties like existence or possibility and their complements, from which their actual exemplification cannot be deductively inferred. Converse intentional properties like being intended or unintended, like being comprehended or uncomprehended, are also classified without exception as extra-constitutive. The unintended or uncomprehended golden mountain is a distinct object on intensional identity conditions from the golden mountain, and even than the existent golden mountain, but for that reason alone need not be unintended or uncomprehended. By application of the same distinction as that appealed to further above, it does not follow that: [unintended, golden, mountain] ⇒ unintended[unintended, golden, mountain] Although: [unintended, golden, mountain] � = [golden, mountain] And: [unintended, golden, mountain] � = [intended, golden, mountain] Moreover, for the same reason, there is no contradictio in adjecto in the predication: intended[unintended, golden, mountain] Leaving it equally open as a logical possibility, depending on the logically contingent state of the world and the existence and direction of thought: unintended[unintended, golden, mountain] Similarly for the [uncomprehended, golden, mountain]. There is no true predication of the uncomprehended[uncomprehended, golden, mountain] following merely from the free intention of the [uncomprehended, golden, mountain], in the way that the golden[uncomprehended, golden, mountain] and mountain[uncomprehended, golden, mountain] follow from intending the [uncomprehended, golden, mountain]. The important difference between the present

Domain Comprehension in Meinongian Object Theory

� 119

proposal and multiple modes of predication solutions to Russell’s problem of the existent golden mountain (and its variants), is that the distinction here depends essentially on applying Meinong’s original distinction between the categories of constitutive and extra-constitutive properties, rather than invoking the very different distinction between dual modes of predications. If converse intentional properties like being intended or being unintended are extra-constitutive, then an object comprehended in the Meinongian object theory reference domain need not possess those properties by virtue of including them among its intensional identity conditions. This application undermines Mally’s 1914 argument for the mind-independent objectivity of the Meinongian object theory domain. Mally maintains that there are at least some unintended mind-independent objects implied by the freely assumed [unintended, golden, mountain] ⇒ unintended[unintended, golden, mountain]. If being unintended is an extra-constitutive property instead, then inferring from the properties as intensional identity conditions of an intended object to the object’s actual exemplification of any of its extra-constitutive, as opposed to any of its constitutive, properties, is logically unauthorized. The unintended golden mountain may or may not actually be unintended, if to be unintended is extra-constitutive. Although in another sense it must also certainly be unintended, golden, and a mountain, in order to be exactly that and no other actually intended object. There is an equivocation in two senses of “unintended” in the proposed solution, one that is brought into intensional identity conditions and the other that may or may not be actually exemplified by the object so identified.

9 Mind-Independent Objectivity of the Meinongian Domain The important lesson of successful efforts to establish a comprehension principle for a Meinongian object theory reference domain is perhaps this. That the best response to Russell’s problem of the existent golden mountain is not that object theorists must choose alternatively to enforce a distinction between constitutive and extra-constitutive properties, on the one hand, versus some form of the modal moment solution, on the other. The distinction between properties possessing or lacking the modal moment of full-strength factuality may be indispensable to avoiding the full range of Russell-inspired counterexamples. It functions properly, however, only with respect to a previously distinguished division of extraconstitutive properties that can thereafter enter into an object’s intensional iden-

120 � Dale Jacquette tity conditions without necessarily being exemplified by the objects they serve to identify. What is required is a sympathetic synthesis of both solutions. It is specifically extra-constitutive properties, based on that prior distinction, that can appear within intensional identity conditions for a Meinongian object, independently of its ontic status as actually existent or nonexistent. We cannot forestall a limitless stream of Russell-encouraged counterexamples to any comprehension principle for Meinong’s object theory without judiciously observing both the distinction between constitutive and extra-constitutive properties, and, among the extraconstitutive properties, between those possessing and those lacking the modal moment of full-strength factuality. These, a fortified revisionary Meinongian object theory must allow, are assumable as intensional identity conditions without their implying their actual exemplification. The two approaches are therefore not rival, but complementary solutions, neither of which is sufficient in and of itself to solve the family of counterexamples suggested by Russell’s commonsense problem of the existent golden mountain. Including converse intentional properties among the extra-constitutive properties in a revisionary Meinongian object theory invalidates Mally’s 1914 argument for the objectivity of an object theory reference domain (Mally 1914/1989).²¹ It is not axiomatic in the first place that a phenomenological Brentano-motivated intentionalist object theory needs an objective mind-independent comprehension principle for the objects in its reference domain. Objectivity in the sense of mindindependence of the sort Mally sought to attain in any case is achieved in another way, without benefit of Mally’s argument, incapacitated by the classification of converse intentional properties like being unintended as extra-constitutive.²² A Meinongian reference domain of all existent and nonexistent potentially intended

21 I discuss these alternative approaches to semantics and metaphysics in depth in (Jacquette 2002b: especially 158-181). 22 I omit discussion of Edward N. Zalta’s dual modes of predication in this context, although it is sometimes described as Meinongian, because Zalta’s abstract objects are Platonic and Fregean existents or subsistents, in the old-fashioned terminology, rather than Meinongian beingless objects, neither physical nor abstract. See (Zalta 1983) and (Zalta 1988). The burden of my argument is that the constitutive versus extra-constitutive property distinction like dual modes of predication cannot solve the full range of Russell-inspired counterexamples to a principled comprehension of the Meinongian object theory domain in lieu of judicial application of Meinong’s concept of the watering-down of extra-constitutive properties and modal moment of full-strength factuality. The solution is unavailable to any dual modes of predication approach that rejects the constitutive versus extra-constitutive property distinction. Plural modes of predication in a Meinongian context were originally suggested by Mally in his (1909a) and (1909b). See (Jacquette 2008).

Domain Comprehension in Meinongian Object Theory

� 121

objects is objectively comprehended by the abstract operations through which a distinct object is added to the domain. The principle of inclusion is the mindindependent mathematical combinatorics governing all logically possible assemblages of constitutive and extra-constitutive properties. Appropriate restrictions are then required to block inference to an object’s actual exemplification of merely predicated extra-constitutive properties, even when they are included among the object’s intensional identity conditions, when they are not independently known to be actually exemplified. The original impetus for a Meinongian object theory may have depended on an interpretation of Brentano’s 1874 thesis of the intentionality of all and only the psychological. Its reference domain is nevertheless objectively determinable without phenomenology, descriptive psychology or exercise of inner perception, as the complete range of all the logically possible intendable objects whose intensional identity conditions are determined by every logically possible combination of properties, constitutive and extra-constitutive. Actual exemplification of constitutive properties is logically guaranteed. Actual exemplification of extraconstitutive properties, in contrast, is not logically guaranteed, but is often a matter of correspondence with relevant truth-making states of the world when it is not logically implied or denied. Converse intentional properties are unequivocally classified as extra-constitutive, which, like the extra-constitutive property of existence, does not prevent them also from entering into a domain object’s intensional identity conditions, even when they do not actually hold true of and are not actually exemplified by the object.²³

23 I have benefited from discussions with critics and enthusiasts of Meinong’s object theory most recently at the International Colloquium, Objects and Pseudo-Objects: Ontological Deserts and Jungles from Meinong to Carnap, Université de Liège, Liège, Belgium, May 15-16, 2012.

Sébastien Richard

Meinong and Early Husserl on Objects and States of Affairs Sébastien Richard: F.N.R.S., Free University of Brussels, Belgium

Both Alexius Meinong and Edmund Husserl were pupils of Franz Brentano at the end of the 19th century, but they were not in Vienna at the same time, and probably never met. They discussed their respective ideas mainly by exchanging letters from 1891. These letters and their respective writings are full of accusations and reproaches, with the result that Husserl and Meinong put an end to their correspondence in 1904. Their views concerning the notion of object were among the points of contention. They both agreed that the notion should be extended beyond mere real individual objects in order to include ideal ones. Among these new objects we find the objectual correlates of judgements. But while Husserl called them “states of affairs”, Meinong called them “objective”. More fundamentally the ontological and semantic functions that these two notions have to fulfil are not the same. Meinong and Husserl have also divergent positions on the extent which has to be given to the notion of object: should it include impossible objects? Meinong answered affirmatively, but Husserl did not.¹

1 Brentano and the Brentanians on the Object of Presentations In a famous passage from his Psychology from an Empirical Standpoint of 1874, Brentano set forth a distinction between psychological phenomena and mental ones: Every mental phenomenon is characterized by what the scholastics of the Middle Ages called the intentional (or mental) in-existence [intentionale Inexistenz] of an object and what we might call, though not wholly unambiguously, the reference to a content [Beziehung auf

1 This paper owes much to the work of Peter Simons on Meinong and the Brentanian School (for an up to date summary see especially [Simons 2012]). We would like to thank him here for the several discussions we have had with him on the topic of the present paper.

124 � Sébastien Richard einen Inhalt], a direction towards an object [Richtung auf ein Objekt] (which is not to be understood here as meaning a thing), or an immanent objectivity. Each mental phenomenon includes something as object within itself, although they do not all do so in the same way. In presentation something is presented, in judgment something is affirmed or denied, in love loved, in hate hated, in desire desired, and so on. (Brentano 1874: 124-125/1973: 88)

The mental phenomena are then, in opposition to the physical ones, “those phenomena which contain an object [Gegenstand] intentionally within themselves”. The “intentional inexistence” of the object of a mental phenomenon must be understood in two ways. First as claiming that this object actually exists but not in the same way as reality does. It exists only immanently; it possesses an “objective being” in the mind. The scholastics used to say that the object contained in a mental phenomenon does not have an esse reale but an esse objective. This is the first meaning of the prefix “in” in Inexistenz. Secondly the intentional inexistence of the object of a mental phenomenon stresses that an existing object in the extra-mental reality must not correspond to it. For example, I can desire something which does not exist, but this does not mean that I do not desire an object. Brentano’s thesis has been judged to be unclear. One of the main difficulties concerns the ambiguity which affects the status of the object of the mental phenomena. On the one hand, we are naturally led to consider this object as being an extra-mental thing, whereas it is immanent to the mental sphere, and on the other hand Brentano rejects the interpretation of the object of consciousness as being the object as it is “presented” (vorgestellt). The first attempt to solve this ambiguity came from one of Brentano’s students: Alois Höfler. In 1890 – at the beginning of what Peter Simons has called the “decisive decade” for the concept of intentionality (Simons 1992) – he published a logic textbook intended for the students of Austrian secondary schools. In §6 of this book he distinguishes the “content” (Inhalt) from the “object” (Gegenstand) of an act of presentation: 1. What we called above the “content of the presentation and of the judgment” lies entirely within the subject, like the act of presentation and the act of judgment itself. 2. The words “Gegenstand” and “Objekt” are used in two senses: on the one hand, for that thing existing in itself [an sich Bestehende] [...] on what our presenting and our judging is directed so to speak, on the other hand, for the psychical “image” [Bild] existing “in” us, more or less approximating this reality, which quasi-image (more accurately: sign) is identical with what has been called “content” in 1. To distinguish it from the object assumed to be independent of the thinking one also calls the content of a presenting and jugding (the same for: feeling and willing) the “immanent or intentional object” of those psychical phenomena. (Höfler 1890: §6, 7)

Meinong and Early Husserl on Objects and States of Affairs

�

125

So, according to Höfler, when we have a presentation of something we have an “act of presentation” and what is presented in this act is a “content”, which can also be called an “immanent or intentional object”. We must then distinguish the content from the object properly called, i.e. the object “independent of the thinking” toward which the presentation is directed. Höfler, and with him Meinong, differentiates here what was only confusedly conceived by Brentano: the difference between what is immanent to the mental sphere – the content of a presentation – and what is extra-mental – the object of this presentation. This first clarification was then carried further by yet another of Brentano’s students, Kazimierz Twardowski, in his habilitation thesis, On the Content and Object of Presentations (Twardowski 1894/1977). If the object of a presentation is not something immanent to the mental sphere, but something extra-mental, the question arises as to whether there is always an object corresponding to a presentation. In other words, are there presentations which, like “round square” and “golden mountain”, are “objectless” (gegenstandlos), as Bolzano called them (1837: §67)? Twardowski’s answer to this question is unambiguous: no, there is no presentation lacking any object corresponding to it. As Twardowski’s arguments on this matter have already been considered elsewhere in the present volume,² we won’t discuss them again. What interests us more is the metaphysical conclusion drawn by Twardowski from his assumption according to which every presentation has an object corresponding to it: The object is something different from the existent; some objects have existence in addition to their objecthood [Gegenständlichkeit], that is, in addition to their property of being presented (which is the real sense of the word “essentia”); others do not. What exists is an object (ens habens actualem existentiam), as is also what merely could exist (ens possibile); even what never can exist but can only be conceived of (ens rationis) is an object; in short, everything which is not nothing, but which in some sense is “something” is an object. (Twardowski 1894: 37-38/1977: 35)

Everything, except the pure nothing – what the metaphysical tradition called the nihil negativum –, is an object, be it existent or not. For Twardowski the object (Gegenstand) is dissociated from the existence.

2 See the paper by Gyemant in this volume, and also (Richard 2012).

126 � Sébastien Richard

� Meinong: Objects and Objectives Meinong, another of Brentano’s students, accepted Twardowski’s distinction between content and object,³ and also his claim that there are objects about which we think and which do not exist. So he thought that he could legitimately say that “there are [es gibt] objects of which it is true that there are no [nicht gibt] such objects” (Meinong 1904: 490/1960: 83). This last claim only sounds like a paradox: the two “there are” must of course not be understood in the same way. Only the second one expresses an “ontological commitment”, while the first one ranges over all objects indiscriminately. But if there are other objects than the existing ones, if there even are objects “beyond being and non-being”, is it possible to consider a theory of properties and laws pertaining to all objects, existent and non-existent alike – what Meinong calls a “theory of objects” (Gegenstandstheory)?

�.� The Theory of Objects: Beyond Being and Non-Being At the beginning of his 1904 “Theory of Objects” Meinong explains clearly the question guiding his text: [...] we wish to know whether, among the sciences that are accredited by scientific tradition, there is one within which we could attempt a theoretical consideration of the Object as such [des Gegenstandes als solchen], or from which we could at least demand this. (Meinong 1904: 485/1960: 78)

Traditional metaphysics cannot be the science sought, because metaphysics is not sufficiently universal, being restricted to what actually exists. For Meinong metaphysics is the “science of the actual” (Wirklichkeitswissenschaft) (Meinong 1978 [1921]: 48), the science of “real objects” (real Gegenstände) – “actuality” (Wirklichkeit) being here synonymous with “reality” (Realität). But this does not mean that actuality is restricted to what exists now. Meinong defines real objects as objects which “exist or at least, by nature, can exist” (Meinong 1971 [1899]: 394/1978: 150). It seems then that he defines reality in the same sense as the modern ontologists did, namely as what can exist, and consequently that with which existence

3 See (Meinong 1971 [1899]: 381/1978: 141). After all Höfler’s Logik was written with the help of Meinong, the former being the student of the latter. See (Marek 2001: 262-264).

Meinong and Early Husserl on Objects and States of Affairs

�

127

does not conflict.⁴ From this point of view, the golden mountain is an object as real as the Erzberg.⁵ Being restricted to real objects, metaphysics is subject to what Meinong calls “the prejudice in favour of the actual” (das Vorurteil zugunsten des Wirklichen). However, besides real object we also have to make room for “ideal objects” (ideale Gegenstände). These objects cannot exist but have another form of being: they “subsist” (bestehen). Contrary to existence, subsistence (Bestand) is a nontemporal mode of being,⁶ so that ideal objects are timeless, in contrast to real objects. The objects which subsist but do not exist are not just mental objects: like every object, they possess a nature independent of our ability to know, even if they can be known. The relations of identity, of difference, and the mathematical objects are all ideal objects of this kind. With real and ideal objects we are still in the domain of all that is, i.e. the domain of beings (Seienden), be they actual or non-actual. Yet there is also all that is not (Nichtseiendes), objects which neither exist nor subsist. In the first edition of On Assumptions (1902), Meinong drew a distinction between the “thetic function” and the “synthetic function” of thought.⁷ This distinction is required because we need to separate the simple existence position from the predicative attribution, which are two different modalities of judgments.⁸ In 1904, Meinong moved this distinction from the theory of judgments to the theory of objects. This mainly means that in the two functions what is grasped is an object: in the thetic function the thought grasps a “being” (Sein) and in the synthetic function it grasps a “so-being” (Sosein) (Meinong 1904: 489/1960: 81). Yet the grasping of the latter does not presuppose the former:

4 See (Wolff 1962 [1730]: §134). 5 See (Meinong 1971 [1899]: 457/1978: 192). 6 See (Meinong 1902: 467). 7 See (Meinong 1902: 420-427). 8 Let us remember that if first Brentano had claimed that every categorical judgment was reducible to an existential one, his position on this point evolved from the end of the 1870s. He finally acknowledged categorical judgments to have a specificity. This is the famous “double judgment theory” in which a judgment of the form “a S is P” is equivalent to, on the one hand, the judgment “a S is”, which only posits the subject, and, on the other hand, once the predicate has been identified to the subject in the presentation, the judgment “there is a SP” (es gibt ein SP) or “there is a S which is P” (es gibt ein P seiendes S), which adds the predicate to the subject already admitted in the first judgment. Therefore the existence or inexistence claim in a categorical judgment is about both S and SP. There is in every categorical judgment a double thetic modality and the synthesis of two existential judgments is needed to produce a categorical judgment. Of course in an existential judgment such as “a S is” this synthesis is not required since there is only one thetic modality.

128 � Sébastien Richard If someone judges, e.g., “A perpetual motion machine does not exist”, it is doubtless clear that the object whose existence is denied must have properties, namely those in terms of which it can be characterized, and that without these properties, the conviction of nonexistence could have neither sense nor justification. To have properties, of course, is tantamount to “being thus and so” [sosein]. But this so-being does not have existence as its necessary condition, for existence is precisely what is being denied – and correctly denied, moreover. (Meinong 1910: §12, 79/1983: 61)

Consequently Meinong maintains a “principle of the independence of so-being from being”,⁹ i.e. the claim that the so-being of an object, the fact that it has such and such property, cannot be affected by the fact that this object is or is not. A property can then be attributed to an object even if this object is impossible or contradictory. For instance, it is legitimate to say that the golden mountain is made of gold and that the round-square is round as much as it is square.¹⁰ In fact, for Meinong, when we judge truly that an object does not exist, we must not only be able to grasp its so-being before its being or non-being, but also the object itself: If I say, “Blue does not exist,” I am thinking just of blue, and not at all of a presentation and the capacities it may have. It is as if the blue must have being in the first place, before we can raise the question of its being or non-being. But in order not to fall into new paradoxes or actual absurdities, perhaps the following turn of expression may be appropriate: Blue, or any other Object whatsoever, is somewhat given prior to our determination of its being or non-being, in a way that does not carry any prejudice to its non-being. (Meinong 1904: 491/1960: 83-84)

What is then the status of this object given before any determination of its being or non-being? Obviously it cannot be an ontological status, a third mode of being next to existence and subsistence, since a judgment about an impossible object will reveal that such an object has no being at all. Meinong says that such an object is “outside being” (außersein), indifferent by nature to being and non-being. In other words the domain of such an object is ontologically neutral. Being “beyond being and non-being” (jenseits von Sein und Nichtsein), this object is a “pure object” (reiner Gegenstand) (Meinong 1904: 494/1960: 86) whose “proper essence” is constituted by its so-being. The object grasped in a judgment is not created by the subject prior to the judgment. It has a consistency, an essence, which can be grasped by the subject but which is independent from him:

9 See (Meinong 1904: 490/1960: 82). This principle was first formulated by Ernst Mally. See (Mally 1904: 126). 10 See (Meinong 1904: 490/1960: 82).

Meinong and Early Husserl on Objects and States of Affairs

� 129

In front of the grasping [Dem Erfassen], its object is each time what is logically prior, even when this object follows the grasping chronologically. This is why the grasping can never create its object or modify it in any way but only select it in a way, extracting it from the diversity of what is previously given (at least as being outside being). (Meinong 1978 [1921]: 45)

In the end, the Meinongian object seems to be understood in the etymological sense of Objectum¹¹ or Gegenstand as what stands in front of us. It refers then to the experiences which grasp the object, while stressing at the same time its independence from the mind.

�.� On Objectives When we introduced the principle of the independence of so-being from being in the last section we also noted that being and so-being were objects. They are not objects in the usual sense, i.e. they are note “objecta” (Objekte), which are the objects of presentations. They rather are what Meinong calls “objectives” (Objektive).¹² This notion first appeared in the Meinongian theory of judgments. It can be seen as a generalization of the relation between a presentation and its object to judgments. In the Brentanian tradition the objectual correlate of a presentation is usually taken to be an object to which we get access with the help of the content of this presentation. The difference between a presentation and a judgment does not lie in their respective objectual correlate but in the relation to it: in the judgment a recognition is added to the simple consideration of the object. Here Meinong splits off from his fellow Brentanians by identifying two factors in the judgment, instead of just one: the moment of “conviction” (Überzeugung) and the moment of assertion or denial, “the position within the antithesis of yes and no” (1910: 3/1983: 10). It is in the gap between these two moments of the judgment that we find one of Meinong’s most important innovations: the “assumption” (Annahme). Its distinctive characteristic is that it only possesses the second moment of judgments. An assumption has an affirmative or negative position – which differentiates it from a presentation – but it completely lacks the moment of conviction. Meinong gives us the following example. Let us suppose that the Boers won the 1899-1902 war

11 See (Dewan 1981). 12 Meinong says that he preferred the word Objektiv to the word Sachverhalt, already used by Stumpf and Husserl, because he finds it strange to speak of a non-subsisting Sachverhalt (Meinong 1910: §14, 101/1983: 76).

130 � Sébastien Richard against the British Empire. In doing so we assume that the Boers won the 18991902 war, we make a hypothesis which we know is false, i.e. a hypothesis which is deprived of any conviction. The opposition of yes and no, a common feature of assumptions, must not be confused with the positive or negative stance proper to judgments in the Brentanian tradition, because the latter is equivalent to an acceptance or a denial of existence. For Meinong the assumptive phenomenon is independent of the question of the existence or non-existence of what it is about. We can now ask what is intended by a judgment or an assumption, i.e. what is their objectual correlate? We could think that the objects corresponding to judgments and assumptions are identical to the objects of presentations – the objecta – on which they are based. In this case Socrates would be the object of the judgment “Socrates exists”. If Meinong questions the simple identification of the object of a presentation with the object of a judgment, he does not completely reject this solution. It is just that Socrates is not the object of the judgment “Socrates exists” and of the presentation “Socrates” in the same sense. For Meinong, Socrates is the object of the former only secondarily, by virtue of the first object of the judgment. This first object is what is judged strictly speaking. When I judge that Socrates exists, my judgment is about Socrates but Socrates is not what I judge. What I judge is that Socrates exists, i.e. what Meinong calls an “objective” (Objektive). The positive stance is taken toward the objective that Socrates exists, and not toward the objectum Socrates which is judged about. While the latter can be mentally presented, the former can only be judged or assumed (Sierszulska 2005: 40). The notion of assumption has the followings features¹³: 1.

2.

The judgment and the assumption have two “objects” (Gegenstände): on the one hand the object “judged” (geurteilt) or “assumed” (angenommen), which is an objective, and on the other hand the object of which something is judged or assumed, that is the object which something is judged about (beurteilt) or assumed about (beannahmt), which is an objectum (Meinong 1910: 44/1983: 38). Truth and falsity do not pertain to objectives. An objective which is recognized as true is said to be a “fact” (Tatsache). This is a direct consequence of the Meinongian distinction between judgments and assumptions. Since an objective is the objectual correlate of an assumption, it must be possible to consider it independently of the belief in its content. This belief only belongs to judgments.

13 See (De Calan 2006: 109).

Meinong and Early Husserl on Objects and States of Affairs

3.

� 131

Although the objectives are independent of the opposition between the true and the false, we can nevertheless apply to them the antithesis of the positive and the negative, in accordance with their status as objectual correlates of assumptions.

What is the ontological status of objectives, their mode of being? Being the objectual correlates of both judgments and assumptions it is clear that their mode of being cannot be existence, which pertains only to judgments. When I judge truly that Barack Obama exists, Barack Obama exists indeed but the fact (the objective which is the case) that Barack Obama exists does not itself exist. “The existence itself (as any other objective) does not exist” (Meinong 1978 [1921]: 18). To say otherwise would lead to an infinite multiplication of existences (Findlay 1995 [1963]: 73-74). For example, if the existence of Barack Obama exists, then the existence of the existence of Barack Obama also exists, and so on. However if objectives do not exist, they still possess a mode of being: the “subsistence” (Bestand). They are subsisting ideal objects: [...] “that A exists” or yet “that it does not exist”, that “subsists”, in case the given judgment has been uttered rightly, but that does not exist so to speak yet again. (Meinong 1902: 187)

The condition here added is important for it stresses that all the objectives do not subsist; some of them are non-subsisting objects. An objective which subsists is said to be the case – it is a fact – while an objective which does not subsist is not the case – it is not a fact. Consequently the truth or falsity of a judgment results from the “factuality” (Tatsächlichkeit) or “counterfactuality” (Untatsächlichkeit), the subsistence or the non-subsistence of the corresponding objective.¹⁴ For instance, that Barack Obama does not exist is a non-subsisting objective and that Pegasus does not exist is a subsisting objective; while the second one is a fact, the first one is not. Another important feature of the Meinongian objectives is that they are higher order ideal objects (Meinong 1978 [1921]: 17/1974: 227). Higher order objects are objects which are “built” upon other objects in the sense that the first ones are in a one-sided dependence relation with the second ones. For example, we can say that the superius “dissimilarity” is a higher order object “founded” upon the inferiora “red” and “green”, which are lower order objects. The same applies to objectives which must always have an object “under” themselves. They are necessarily dependent on the objects which constitute their “material” (Meinong

14 See (Meinong 1910: 69).

132 � Sébastien Richard 1910: 63/1983: 51). For instance, the objective that the snow is white is a superius founded on the objectum snow, the latter being the material of the former. Obviously the material of an objective can itself consist of other objectives. For instance, the objective that the snow is white is the material of the objective that it is a fact that the snow is white. We can thus have series of objectives organized into a hierarchy (Meinong 1978 [1921]: 18/1974: 227).¹⁵ We must also distinguish the objective from what Meinong calls a “complex” (Komplexion). Whereas the former can only subsist, the latter is able to exist. For instance, a chord is an existent complex made up of different existent tones.¹⁶ When played simultaneously these tones produce also an objective, that is the objective that these tones are being played simultaneously, but this objective never exists, it only subsists. Hence, “though an objective is in some sense complex, it cannot be identified with a complex” (Findlay 1995 [1963]: 97). Although objectives and complexes are different, they are not totally independent of each other: we can apprehend a complex, only if we also apprehend the objective of which the complex is a moment (Meinong 1910: §46, 280/1983: 202). So a chord can only be apprehended if we also apprehend the objective that some tones produce this chord when played simultaneously. After this brief presentation of Meinong’s conception of objects and objectives, let us now compare it to Husserl’s own conception of objects and states of affairs.

3 Husserl: Objects and States of Affairs In his book on the mathematical infinite Louis Couturat said that [...] the most general concept, which includes every conceivable object, is the concept “object” itself, understood in the widest sense and denoting everything that can be intended. If all the objects are thus countable, it is because they are subsumed by the concept “object”. Whatever their differences, whatever the kinds to which they belong, we can never say that they have nothing in common, nothing identical between them: since they will always have the common feature of being intended, and they will always be identical as objects of thought. (Couturat 1896: 515)

15 A higher order object cannot consist of an infinity of objectives. The analysis of an objective must always end up – after a finite number of steps – in an objective built on objecta and not on other objectives. 16 See (Meinong 1971 [1899]: 395/1978: 150).

Meinong and Early Husserl on Objects and States of Affairs

� 133

He then adds in a footnote that this use of the term “object” is identical to Husserl’s use of the term “something” (Etwas) – an “object in general” as he will also call it – in the Philosophy of Arithmetic. In 1891 the father of phenomenology thus seems to have had a position similar to Meinong’s one: both of them understood the notion of object in a very broad way, including in it everything, be it real or only thought, possible or impossible, non-contradictory or contradictory.¹⁷ Nevertheless Husserl modified his understanding of the concept of object when he read Twardowski’s 1894 dissertation¹⁸ and tried to develop his own solution to the problem of objectless presentations. According to him this problem results from two contradictory propositions¹⁹: 1. 2.

Every presentation has an object which corresponds to it. There are objectless presentations.

Husserl first criticizes the “popular solution” to this problem which regards presentation as a mental “depiction” (Abbild) of the object. According to this conception, just as we can have pictures of something which does not exist, we can also have presentations of non-existing objects outside the mind. For instance, the presentation of a round-square is a mental depiction of a round-square but no object corresponds to this depiction outside the mind. If every presentation has an object, then the depiction must be the presented object and it seems that we have two objects presented: the depiction which exists and the outside object which can either exist or not. Husserl argues that there is always only one presented object. It is the same object which is aimed at by a presentation and which does not exist: The same Berlin that I present also exists, and the same Berlin would no longer exist if judgment were brought down as in the case of Sodom and Gomorrah. The same centaur Chiron, of which I now speak and which I consequently present, does not exist. And it is similar in every case where the presentation is univalent. (Husserl 1979 [1894]: 305-306/1999: 253)

Husserl therefore objects to the depiction theory that when we present or judge an object, it is the object outside the mind that we present or judge, and not its subjective depiction. It is to this object outside the mind that our joy or our pain relates; it is this object which is red or green, and not its depiction.

17 For Husserl see especially (1891: 86/2003: 84). 18 In 1896 Husserl wrote a review of Twardowski’s dissertation and a rather long essay entitled “Intentional Objects”, which was only published in 1979 in the volume XXII of the Husserliana. 19 See (Husserl 1979 [1894]: 303/1999: 251).

134 � Sébastien Richard However, although Twardowski adopted a kind of depictive conception of presentation,²⁰ the theory set out here is not his own solution to the problem of objectless presentations. For the Polish philosopher there is only one object in the strict sense of the word. This is the reason why he says, for instance, that in a negative existential judgment the object about which we claim that it does not exist is not the object presented in the presentation, i.e. the “content”, but the object presented by the presentation, i.e. an object outside the mind. At first sight, the actual problem of Twardowski’s theory of presentation is not that it distinguishes two objects, but that it assigns two kinds of existence to the object of a presentation: the “true existence” and the “intentional existence”, which is not an existence properly speaking, but only an existence “as being presented”.²¹ Such a solution to the problem of objectless presentations is absurd according to Husserl because, all things considered, it has the same flaw as the depiction theory: it must ultimately accept two objects, one which is immanent and the other one which is true.²² But Husserl would not be more satisfied if we could show that there is only one object and two modes of existence because, for him, the notion of existence is unambiguous. The division relevant to the solution to the problem of objectless presentations does not pertain to objects or to existences, but to presentations themselves. More precisely we must draw a distinction between two kinds of “logical functions” (Husserl 1979 [1894]: 311/1999: 256) of presentations. First we must say that for Husserl all the objects are intentional,²³ i.e. they are intended, “presented in the presentation”. Therefore if we accept the thesis that every presentation presents an object, it means that every presentation presents an intentional object. But we must also say that a “true” object – an existing one – is an object which corresponds to a “valid” presentation, a presentation which is the base for an affirmative existential proposition. For instance the presentation “the current president of the United States” is a valid presentation in this sense since it is the base for the truth of “Barack Obama exists” and there is a true object which corresponds to it. On the contrary, an objectless presentation is not valid because it is not the base for a true existential proposition. For instance, the presentation “a number which yields the result −1 when it is squared” has a con-

20 He does not conceive the mental depiction as a copy of the object. He rejects the idea that there is “a kind of photographic resemblance between content and object” (Twardowski 1894: 67-68/1977: 64). 21 See (Husserl 1979 [1894]: 307/1999: 254). 22 See (Husserl 1979 [1894]: 308/1999: 255). However it must be stressed that for Twardowski existence as being presented is not anymore – in the strict sense of the word – existence. 23 See (Benoist 2001: 204).

Meinong and Early Husserl on Objects and States of Affairs

� 135

tent whose characteristic marks conflict each other, so that this presentation is the base for the truth of “a number which yields the result −� when it is squared does not exist”. In this case the object is said to be “merely” (bloss) intentional. Nevertheless this does not mean that we have then an object: a merely intentional object is an object which exists in no sense of the term. It is an object which is only intended. The intention is real, but its corresponding object is, in fact, nothing. In the end, for Husserl, there is only one notion of object – the existing object: [...] the expressions “an object” and “an existent, true, real, proper object” are fully equivalent. (Husserl 1979 [1894]: 315/1999: 259)

The demarcation line does not lie between objects, but between presentations: there are presentations which are valid and others which are not, presentations whose contents have the property of intending an object and others whose contents do not have this property. If there is only one notion of object, identical to the intentional object, there is however an “improper” sense according to which we can talk about an object in the case of an objectless presentation. But this way of speaking of an object in the case of an objectless presentation is always a deviant way, true in an improper sense. It seems that for Husserl when we are able to speak about an object we are able to identify this object, to express identity statements about this object. In other words: no object without identity. The difference between the proper and the improper way of speaking of objects lies then in the kind of identity tied up with these objects: in the proper way of speaking of objects the identity statements are unconditional whereas they are conditional in the improper way of speaking.²⁴ When we talk about the same fictional object with the help of two different presentations P and P′ , what we mean in fact is that these two presentations intend the same object under a certain assumption. For instance, when we say that “Zeus” and the “the greatest of the Olympian gods” are two presentations which intend the same object, what we implicitly mean is that they intend the same object “under the assumption” of Greek mythology. Most of the time the addendum “under the assumption of the Greek mythology” is omitted, because it is evident. If Greek mythology was discovered to be true in the end, then the identity between “Zeus” and “the greatest of the Olympian gods” would be an unconditional one.

24 See (Husserl 1979 [1894]: 316/1999: 260).

136 � Sébastien Richard

�.� On Categorial Objects The critique of the Twardowskian solution to the problem of objectless presentations demonstrates that Husserl advocates a conception in which only the true object, the transcendent object, the object intended by a valid presentation exists: Strictly and truly speaking, the presented object of invalid presentations is therefore nothing; that is to say, it is not something. Something and something-which-is are equivalent concepts; “not something” is “something which is not”. (Husserl 1979 [1894]: 331/1999: 270)

In the Logical Investigations of 1900-1901, Husserl still maintains that the notion of object is unambiguous: every object of presentation is an intentional object. Nevertheless it seems that he has now abandoned his theory of assumptions as a solution to the problem of objectless presentations in aid of the distinction between two intentional modalities: the intuitive and signitive intentions. When I perceive some written or uttered marks an intention can come into play which bestows a meaning on these marks. In this way I relate myself to an object. But it would be a mistake to believe that because the written marks have a meaning, there is necessarily an object corresponding to that meaning. What is necessary in the meaning phenomenon is the fact of being related, not of being related to something. It is only if the object is intuitively given in the way it is intended that we can truly say that there is an object. The signitive intention is then said to be “realized” or “fulfilled”. For instance, with the words “round square” I am unable to intend any object, i.e. the signitive intention which bestows a meaning to these words cannot be fulfilled, it is “empty”, and this for an a priori reason: because the meaning through which I intend an object with these words is contradictory. Therefore, in the Logical Investigations, Husserl still rejects the possibility of objects which do not exist; what exists is only the intending of an object. Nevertheless, if in 1900-1901, Husserl keeps a conception of objects narrower than the one of Twardowski or Meinong, his introduction of the notion of intuitive fulfilment makes it possible to conceive an extension of the domain of objects beyond the individual and sensible ones. Indeed, if Husserl considers intuition to be the only true provider of objects, he also extends this intuition beyond the simple sensible one: there is also a “categorial intuition”.²⁵ With this new intuition it becomes possible to consider, beyond the sphere of real objects, i.e. individual objects which are spatio-temporal and perceptible in a sensible way, a new domain

25 On the Husserlian categorial intuition, see (Lohmar 2002).

Meinong and Early Husserl on Objects and States of Affairs

� 137

of categorical or ideal objects which are non-temporal²⁶ and are not “objects of possible sense-perception” (Husserl 1913: VI, §43, 139/2001: vol. 2, 278). Let us consider the statement “I see that this paper is white”, uttered after some perception. Do perceptions correspond to all the parts of this statement? Husserl firmly rejects such a naïve parallelism between grammar and perception. He distinguishes rather the “mater” (Stoff ), which is fulfilled directly by simple intuitions (perception, imagination, remembering etc.), from the “form” (Form), which is not immediately fulfilled.²⁷ What Husserl calls the “categorial forms” (kategoriale Formen), the functors “and”, “not”, “or” etc. are forms in this sense. They give the statement its syntactical structure, its grammatical articulation. The fact that the categorial forms are not immediately fulfilled does not mean they are not fulfilled in any sense. It is just that their fulfilment is not sensible. Categorial acts, such as the categorial intention and the categorial intuition, are “founded” (fundiert) acts. This means that these acts are one-sidedly dependent on other acts; they can only appear if other acts, upon which they are founded, appear first. The corresponding categorial or ideal objects are “objects of higher levels” (Husserl 1913: VI, §46, 145/2001: vol. 2, 282), i.e. they can only be given to us on the basis of lower levels objects. This being said, let us consider again the sentence “this paper is white”. The nominal expression “this paper” can receive its intuitive fulfilment from perception or imagination. However the meaning of such an expression cannot be reduced to perception. According to Husserl, even with the names there is always a “surplus” (Überschuß) of the meaning intention in relation to the simple intuition which corresponds to it. This surplus is its (nominal) form. The same applies to an adjective such as “white”²⁸: its intention is only partly fulfilled by the colour “moment” of the paper – this particular white.²⁹ The reason for this is that in the sentence “this paper is white” the expression “white” does not intend a particular white, but the white, i.e. the white in specie, which is an ideal object which can only be fulfilled by a special kind of categorial intuition – what Husserl calls an “eidetic intuition”.³⁰

26 For the distinction of the real being and the ideal being see (Husserl 1913: II, §8). 27 See (Husserl 1913: VI, §42, 135-136/2001: vol. 2, 276). 28 See (Husserl 1913: VI, §40). 29 A moment in the Husserlian sense is a concrete instance of an ideal species, a dependant part of the whole in which it is instantiated. A moment is more or less identical to what is nowadays called a “trope” in analytic metaphysics. 30 On this special kind of categorial intuition, see (Lohmar 2002: 140 ff.)

138 � Sébastien Richard What about the copula “is”? Obviously it is nothing perceptible – in a sensible way – in the objects;³¹ it is not a real predicate, as Kant would have said. It is a categorial form through which we intend an ideal objectual correlate on the basis of other intentional acts. This categorial object is what Husserl calls a “state of affairs” (Sachverhalt), and the objectual correlate proper to the copula is only a moment – the “relational being” – of this state of affairs.

�.� On States of Affairs The technical term “state of affairs” was introduced in the Brentano School by Stumpf in his logic lectures of 1888.³² It denoted then the “specific content of a judgment” and was expressed linguistically by a that-clause. If this notion already appeared in Husserl’s work as early as 1894,³³ it is only in the Logical Investigations of 1900-1901 that it was really studied in itself. Husserl introduced it in §12 of the first Investigation. Every expression says something about something. The latter is the object of the expression and must not be confused with its meaning. Being an expression, a judgment must also intend a kind of object, what Husserl calls an “objectual correlate” (Gegenständlichkeit), which is a notion larger than the object in the narrow sense.³⁴ This objectual correlate proper to judgments is not the object named by the subject of the judgment, i.e. the one about which something is said. Let us consider the two following propositions: (1)

a is bigger than b.

(2)

b is smaller than a.

These propositions have a different meaning. We could say, following Frege, that they do not express the same thought. Moreover the object about which they say something is not the same – just as what they say about it. In (1), this object is the one named by “a” and in (2), it is the one named by “b”. But even so both propositions seem to intend the same thing: the state of affairs that a is bigger than b. This one is the analogon of the object denoted by a name at the level of the sentence;³⁵ it is the objectual correlate peculiar to sentences.

31 See (Husserl 1913: VI, §43, 137-138/2001: vol. 2, 277). 32 See (Stumpf 1999 [1888]: 312-313). The term already appeared in Lotze and Bergmann’s work: see (Smith 1989a). 33 See (Husserl 1979 [1894]: 304/1999: 251). 34 This Husserlian notion of objectual correlate can mean a state of affairs, a real or categorial form etc. See the footnote in (Husserl 1913; I, §9). 35 See (Husserl 1913: I, §12, 48).

Meinong and Early Husserl on Objects and States of Affairs

�

139

This means that a judgment intends a state of affairs in the same way as a presentation intends its object. The judgment is not a presentation which has for intentional correlate a complex object. In a sensible presentation an object may appear in a sensible way, it may be given to us “as having full-bodied existence” (Husserl 1913: V, §28, 445/2001: vol. 2, 139). However in the judgment what appears is not the existent sensible objet, “but the fact that this is” (die Tatsache, daß er ist). That is not to say that a state of affairs can in no way be the objectual correlate of a presentation. In fact, this happens in the “nominalization” (Nominalisierung) of a state of affairs, as for instance when we say “the fact that S is P is delightful”.³⁶ In this case, the state of affairs itself functions as the subject about which we judge: what is delightful is obviously not the judgment “S is P” but the state of affairs that S is P. With the help of nominalization we can thus “name” states of affairs and have a presentation of them. Like the numbers or the red in general, but unlike the individual objects that they may contain, states of affairs are “ideal objects”. They are found neither in space nor in time, and they are “the same for everyone and always” (Tran Duc Thao 1986 [1951]: 52). For instance, a table may cease to exist but that this same table has four legs is identical for everyone and always is. Being ideal entities, states of affairs do not exist: they subsist (bestehen) or do not subsist, they obtain or do not obtain.³⁷ In this sense, they have a concept of being different from the spatiotemporal existence given in sensible perception, which applies to real objects. What is the relation between states of affairs and their foundational objects? According to Husserl, the former include the latter “as real parts in themselves” (Husserl 1913: VI, §47, 147/2001: vol. 2, 283). For instance, the table which may not exist anymore is included in the state of affairs that the table has four legs – it is one of its parts. Thus a non-temporal object can contain a temporal object. How this is possible is not perfectly clear. In the example of the table it seems that even when it ceases to exist the table is still a part of the state of affairs that the table has four legs, because it is still given to us as part of it in an intuitive way. When the table exists it is given to us in a sensible perception and when it ceases to exist it is given to us in imagination, but in the two situations the table is given to us as the same real part of the state of affairs.

36 See (Husserl 1913: V, §33, 460/2001: vol. 2, 148). 37 See (Husserl 1913: VI, §39, 126/2001: vol. 2, 266).

140 � Sébastien Richard This rather sketchy account of Husserl’s theory of states of affairs³⁸ demonstrates how similar it is to Meinong’s theory of objectives. Let us now conclude with the main difference.

4 Concluding Remarks In arguing that all mental phenomena intentionally contain an object, Brentano first passed a confusion and then a puzzle on to his pupils. The confusion occurred between the content and the object of a mental phenomenon. But once it was clarified by Höfler and Twardowski, a puzzle appeared: the Brentanian claim that every presentation has an object which corresponds to it seems to be in contradiction with the Bolzanian claim that there are objectless presentations. In their answer to this intentional puzzle both Meinong (through Twardowski) and Husserl enlarged the traditional notion of object. They made a broad use of the term “object” by applying it not only to real existing objects but also to the objectual correlates of judgments: states of affairs or objectives. Both of them are ideal higher order objects. What essentially separates Husserl and Meinong with regard to these kinds of objects – despite many similarities – is the role they play in their respective theories of meaning. According to Meinong, objectives are not only the objects primarily intended by judgments and assumptions, they are also the meanings of sentences³⁹ – that is, the linguistic expressions of judgments and assumptions. Thus, as Reinach noted, Meinong’s concept of objective “runs together the two completely different concepts of proposition (in the logical sense) and state of affairs” (Reinach 1911b/1982: 374). On the other hand, for Husserl, the meaning of an expression must never be confused with the object intended through this expression, and consequently the meaning of a sentence cannot be identified with the corresponding state of affairs.⁴⁰ It is in fact a species, an ideal content called “proposition” (Satz) instantiated in the real content or “matter” of the judging act. These propositions are according to Husserl the actual truth-bearers, whereas Meinong ascribes truth and falsity to objectives. We must say that Husserl’s theory seems here to be in a better position. Indeed, since in Meinong’s theory there is no room for a notion of proposition distinct from the notion of objective, it is dif-

38 For a more detailed account see (Mulligan 1989). 39 See (Meinong 1910: §10, 58/1983: 48). 40 Contrary to Tugendhat’s claim in (1976/1982: 117).

Meinong and Early Husserl on Objects and States of Affairs

� 141

ficult to understand how the founder of the Graz School could account for cases where we have different meanings but the same objective, as for instance in the following two sentences “the victor of Jena was the Emperor of the French” and “the loser at Waterloo was the Emperor of the French”. Although Meinong and Husserl both include specific objectual correlates for judgments in the domain of objects, Husserl’s use of the term “object” is not as broad as Meinong’s use. Indeed the founder of the Graz school includes impossibilia such as the golden mountain or the round square among objects, something which the founder of the phenomenological movement refuses to do. This is because they do not accept the same access to the domain of objects. For Meinong an object is whatever is the intention of a mental act, so that there are real and ideal objects but also objects which neither exist nor subsist. Husserl has a more restricted understanding of the access conditions to objects: in order to have an object it does not suffice to have an intention, what is intended must be given in an intuitive way. Of course Husserl does not restrict intuition to perception but also includes in it founded acts. These give us access to the domain of categorial or ideal objects, such as essences, meanings, numbers, relations, laws, states of affairs, etc. These two broadened understandings of the notion of object have given rise to two different ontologies at the turn of the 19th and 20th centuries: the “theory of objects” for Meinong and the “formal ontology” for Husserl, which differ in a number of ways that we cannot investigate here.⁴¹

41 We pursue such an investigation in (Richard 2014).

Guillaume Fréchette

Essential Laws On Ideal Objects and their Properties in Early Phenomenology Guillaume Fréchette: University of Salzburg, Austria

1 A Bit of History It is still not widely known that shortly after their publication in 1900-01, and thanks to their early discovery by Johannes Daubert,¹ Husserl’s Logical Investigations (LI) received a particularly enthusiastic reception among the students of Theodor Lipps in Munich. Through their discussion of Husserl’s work in the Akademischer Verein für Psychologie, an academic circle for psychology founded by Lipps, the Munich students were soon led to form their own phenomenological circle, trying at the same time to find a position liberated from what they recognized, thanks to Husserl, as Lipps’ psychologism, but also to contrast their own position with Husserl’s conception of phenomenology. This position is interesting and important at least for two reasons: first of all, it is the first natural and direct descendant of the phenomenology developed in the Logical Investigations. Indeed, the Munich phenomenologists expanded Husserl’s analyses to vast domains of philosophy in general and ontology in particular: emotion theory, social ontology, action theory, aesthetics, the philosophy of perception, self-consciousness and intentionality. This expansion was made possible by the central role attributed by them to essences in phenomenological analysis. Correlatively, the position defended by the Munich phenomenologists also shows that the transcendental reduction is not a real part of phenomenological analysis, a fact that, if not forgotten, still remains highly debated today.² Who were the Munich and Göttingen phenomenologists? Since the first discussions of Husserl’s LI in the Akademischer Verein, many different orientations had come to be represented among the Munich phenomenologists. At the time of the “Munich Invasion of Göttingen”³ in 1905, at least two different groups must

1 See, for instance, (Schuhmann 1977: 72). 2 I discuss this point in (Fréchette 2013). 3 See (Spiegelberg 1959: 157).

144 � Guillaume Fréchette be distinguished: on the one hand, those who to a large extent remained relatively faithful to Lipps, such as August Gallinger, Aloys Fischer, Fritz Weinmann and Max Ettlinger.⁴ On the other hand, another group of philosophers from the Akademischer Verein was showing more than a mere interest in phenomenology already in 1905 and progressively abandoned most of the Lippsean conceptions after 1906. Among the members of this group were Theodor Conrad, Johannes Daubert, Adolf Reinach and Moritz Geiger.⁵ Again, these two groups shouldn’t be confused with a third group, namely Husserl’s own students in Göttingen, who found their domain invaded by the Munich phenomenologists in the summer of 1905: among them, we find Wilhelm Schapp, Karl Neuhaus, Alfred von Sybel, Alexander Rosenblum, Dietrich Mahnke, Heinrich Hofmann, David Katz and Erich Heinrich.⁶ Remembering the encounter between the members of these three groups in the summer semester of 1905, Wilhelm Schapp sketches an interesting picture of the “Munich Invasion”: One day, it must have been in 1907, the Munich people were there, the Munich invasion of Göttingen. I think it was a summer semester. They were Reinach, Conrad and the young Hildebrand. Geiger appeared occasionally. We used every opportunity, day and night, to engage in philosophical discussions with the Munichers. In our opinion, they were much ahead of us in every aspect. They did not have the devoutness that we had. Reinach blamed Husserl for his turn to the Marburg School, a turn that was already noticed in Munich. [...] We formed at that time a phenomenological association, which met every week and which was led for a while by Conrad. I remember that he tried to get more clarity about things by investigating the “meaning” of words, certainly in connection with the Munich investigations. Again and again, we were investigating word complexes, such as red wine, a wine being red,

4 On August Gallinger (1871-1959), see (Schorcht 1990: 134ff.); on Aloys Fischer (1880-1937), see (Kreitmair 1950); on Fritz Weinmann (1878-1905), see (Schuhmann 1973: 130); on Max Ettlinger (1877-1929), see (Smid 1982: 115). 5 Theodor Conrad (1881-1969) was one of the first of the Munich phenomenologists to go to Göttingen. He published very few articles. Among them, see (Conrad 1911), which was well received in the Munich circle. On Conrad, see (Scaramuzza 1998). Johannes Daubert (1877-1947) was definitely considered as the Husserl-man in Munich (see the letter of Otto Schultze to Aloys Fischer from 17 July 1903, quoted in [Leijenhorst & Steenbakker 2004: 291]). On Adolf Reinach, see (Mulligan 1987). On Geiger, see, among others, (Zeltner 1960). I leave aside here the case of Max Scheler, which would need a treatment of its own. 6 For recent works on Wilhelm Schapp (1884-1965), see (Joisten 2010). Karl Neuhaus was Husserl’s first doctoral student. He completed his degree in 1908. According to Theodor Conrad, he was the Leiter of the Philosophische Gesellschaft in Göttingen from 1910 to 1912, but very little is known about him. See (Avé-Lallemant & Schuhmann 1992). On David Katz, see (Spiegelberg 1972: 42-52). Dietrich Mahnke was an early follower of Husserl, but got his PhD only later in the twenties. On Mahnke’s later works, see (Biller 1987: 691-692). On von Sybel, Rosenblum, Hofmann and Heinrich, see (Schuhmann 1977).

Essential Laws

� 145

the wine is red. We looked for the relationship between word and meaning, concept and object. Sometimes, a word was said about the Munich standpoint, about the way they focused on the Platonic doctrine of metexein, the doctrine of participation in concepts, about the way they boldly advanced the doctrine, in continuation of Husserl, that there is not only the “two” as ideal object, as Husserl taught at that time, but that there must be many, infinitely many twos. The Munich people did not believe anymore in the sensation as constituent of perception and declared all such statements as constructions; they still believed in acts and psychology, those weren’t called into question. (Schapp 1959: 21)⁷

Schapp underlines here three important aspects of Munich phenomenology: its specific manner of philosophical investigation, wherein the analysis of meaning, of what we mean (meinen) by an expression, is put at the forefront;⁸ its particu-

7 “Eines Tages, es muß wohl 1907 gewesen sein, waren die Münchener da, die Invasion aus München. Ich meine, es wäre ein Sommersemester gewesen. Es waren Reinach, Conrad und der junge Hildebrand. Geiger tauchte gelegentlich auf. Wir benutzten jede Gelegenheit, um mit den Münchnern Tag und Nacht philosophische Gespräche zu führen. Sie waren uns nach unserer Meinung in jeder Beziehung weit voraus. Sie hatten nicht die Gläubigkeit, die wir hatten. Reinach warf Husserl seine Wendung zur Marburger Schule vor, die damals in München schon bemerkt war. [...] Wir hatten damals einen phänomenologischen Verein gegründet, der wohl wöchentlich zusammenkam und in dem Conrad zeitweise die Leitung der Diskussion übernahm. Ich entsinne mich, daß er wohl im Anschluß an Münchener Untersuchungen versuchte, über die ‘Bedeutung’ eines Wortes zu größerer Klarheit zu kommen. Wir prüften immer von neuem Wortgefüge, wie roter Wein, rotseiender Wein, der Wein ist rot. Wir suchten nach dem Zusammenhang von Wort und Bedeutung, Begriff und Gegenstand. Zuweilen fiel dann auch ein Wort über den Standpunkt der Münchner, wie diese die Platonische Lehre vom metexein, die Lehre von der Teilhabe an den Begriffen in den Mittelpunkt stellten, wie sie ferner im Anschluß und in Fortführung von Husserl kühn die Lehre aufstellten, es gäbe nicht nur als idealen Gegenstand die ‘zwei’, wie Husserl damals wohl lehrte, sondern es müß te viele, unendlich viele Zweien geben. Die Münchner glaubten nicht mehr an die Empfindung als constituens der Wahrnehmung und erklärten alle entsprechenden Aussagen als Konstruktion; sie glaubten aber noch an Akte und an Psychologie, diese waren noch nicht in Zweifel gezogen.” 8 Compare Schapp’s report with Daubert’s notes on phenomenological and critical investigation (Phänomenologische und kritische Fragestellung) from December 1905 in MS A I 1/34: ‘‘in der phänomenologischen Fragestellung kehrt immer wieder die Frage: ‘Was meinen wir damit’ oder ‘Was meinen wir, wenn wir sagen’. ...” The importance of MS A I 1/34 was already shown by Smid (1982, 140). Besides his reflections on the topic in his published words, Reinach’s focus on the Meinen is also apparent in a letter to Conrad on 14 April 1904, quoted here in Schuhmann and Smith’s translation: “[t]he question: how does the child know that grown-up people ‘mean’ something by their words, is answered by Lipps thus: it sees how they point to something and simultaneously hears a complex of sound. [But] the problem was: how does the child come to understand an expression, and more specifically the expression of words? To this one surely cannot give an answer which involves appeal to another form of expression, to ‘pointing’. For then of course the question still remains: How does the child know that by moving the arms etc. something is meant?” See (Schuhmann & Smith 1987: 7). Another good example can be found in August

146 � Guillaume Fréchette lar conception of ideal objects and, finally, its conception of perception, in which sensations are considered irrelevant to phenomenological analysis. Kevin Mulligan has discussed the first aspect in numerous papers.⁹ In (Fréchette 2013), I deal with the third aspect of Munich phenomenology according to Schapp. In the present paper, I will try to shed some light on the second aspect pointed out by Schapp, namely the Munich-Göttingen conception of essences, laws of essence, and ideal objects. I will first start with a preliminary account of their conception of the synthetic a priori at the basis of their conception of essence (§2); I will then offer a first characterization of this conception, which I will label as metaphysical realism (§3), highlighting its key concept: foundation (§4). In the last four sections (§§5-8), I discuss different outcomes of this conception of essences: the nature of laws of essences (§5), different categories of essences (§6) and anumericity (§7). Since the accounts dealt with in the present paper are barely known, even to phenomenological circles, the aim of this paper is merely descriptive. The point here is to grasp the ‘essence’ of the Munich and Göttingen account of essences in phenomenology.

2 Synthetic A priori Hume’s distinction between relations of ideas and matters of fact is often understood as an ancestor of the distinction between the necessary connections between meanings or concepts and the absence of such connection between the elements of the world. However, this is not exactly what Hume had in mind, when we look for instance at his account of the interrelations among our ideas of color: It is evident, that even different simple ideas may have similarity or resemblance to each other; nor is it necessary that the point or circumstance of resemblance should be distinct or separable from that in which they differ. Blue and green are different simple ideas, but are more resembling than blue and scarlet; though their perfect simplicity excludes all possibility of separation or distinction. (Hume 1978: 675)

In other words, the truth of a proposition about the relations between our ideas of colors cannot be established on the basis of an analysis of ideas, since our ideas of

Gallinger’s study on objective possibility, which begins with the following question: “What does it mean, something is possible, what do we mean, what is the meaning of this, when we describe an object as possible? [Was heißt es, etwas ist möglich, was meinen wir, was bedeutet es, wenn wir einen Gegenstand als möglich bezeichnen?]” See (Gallinger (1912, 18). 9 Some of them are available in French in (Mulligan 2012).

Essential Laws

� 147

colors are simple ideas and hence cannot be analyzed. Following Reinach’s reading of Hume, “no point of view is conceivable from which one could say that two colors and their dissimilarity contradict each other in the logical sense.” (Reinach 1911a/1976: 176) Such propositions, as Smith puts it, are seen by Hume as “reflecting objectively existing interrelations among the phenomena themselves.” (Smith 1986: 7) Hume does leave room for necessary truths that are not analytic, but contrary to Kant, these necessary truths which are not analytic are not exclusively propositions about our knowledge. Rather, they stand somewhere in between epistemological and ontological propositions. This understanding of Hume’s synthetic a priori particularly championed by Adolf Reinach, who criticizes Kant for having erroneously taken Hume’s standpoint on the analyticity of mathematical judgments to be concerning Kant’s own concept of analyticity, is a central element in the Munich-Göttingen conception of the a priori. The root of this error, according to Reinach, was Kant’s assimilation of Hume’s “ideas” to his “concepts”, thereby missing the real sense of the a priori developed by Hume. tis from the idea of a triangle, that we discover the relation of equality, which is three angles bear to two right ones; and this relation is invariable, as long as our idea remains the same. (Hume 1875 [first volume]: 372)

In other words, we know about relations of ideas without matter of facts. How does Reinach interpret this claim by Hume? In our vocabulary, this means that he knows essential structures. He thereby knows what we found to be the basis of the a priori. (Reinach 1911a: 176)

What is the specific sense of the Humean synthetic a priori which Reinach tries to disclose? The main distinction which is made here by Reinach is between what he calls material and modal necessity. According to Kant, Hume inquired exclusively into modal necessity, while Reinach believes that Hume was interested solely in material necessity. Following his reading of Hume, Reinach means that when I feel heat and conclude that there must be fire, the inference I am making is not grounded on modal necessity, but on material necessity, that is on the necessary connection “of such a sort that heat always requires fire.” (1911a: 184). The material necessity characterizes the succession from fire to heat as causal and is a grounding relation. In other terms, the state of affairs that the oven is hot holds in virtue of the state of affairs that there is some fire in it. Following Hume and Reinach, the necessary connection involved here is determined exclusively by the essence of the terms of the connection exactly in the same sense as similarity between colors is determined by the essence of colors. But Kant missed that point:

148 � Guillaume Fréchette According to Kant, Hume saw only two possibilities. Either the foundation of the causal judgment in pure reason, or the explanation of it from experience, i.e. from the mechanism of association and the “subjective necessity arising from it”, which is falsely taken to be objective. That for Hume there is a third possibility – the immediate grounding of necessity through experience – is overlooked by Kant and, from Kant’s standpoint, must be overlooked. (Reinach 1911a: 186)

The distinction between the so-called modal and material necessity and the concomitant focus made on the latter kind of necessity are inherited by Reinach from Husserl’s Logical Investigations, where the concept of necessity relevant for Husserl is an “ideal or a priori necessity grounded in the essences of things” (Husserl 1913: III, §7, 240/2001: vol. 2, 12¹⁰). What is the kind of necessity involved here and why doesn’t it have the epistemological implications of the Kantian synthetic a priori? As for the second question, the answer is relatively simple: Kant postulates an undeterminable X which is the ground on which pure reason identifies a judgment as synthetic a priori. A law such as “Every event has a cause” is a true synthetic a priori judgment in virtue of our ability to construct the world according to this law. Reflection on the conditions of possibility of our experiences furthermore gives us a way of identifying these true synthetic a priori judgments: for instance, it would be impossible for us to experience a world in which some events would fail to have a cause. Following that line, the question of the grounds for our true synthetic a priori judgments is transformed into a question on the conditions of possibility of our experiences. The focus made by Reinach on material necessity is precisely going against Kant’s idea of grounding true synthetic a priori judgments in modal necessity. The distinction is central: while Kant thinks of necessity as a feature conditioning the structure of our experiences, Reinach and the Munich phenomenologists see necessity as grounded directly in the essences of things. The concept of necessity central to Husserl, Reinach and the Munich phenomenologists is sometimes called “metaphysical necessity” nowadays, and the strong distinction advocated today by Kit Fine between essence and modality clearly belongs to the early phenomenological tradition. What does it mean for a conception of material (or metaphysical) necessity to be “grounded in the essence of things”? First, it means to identify the most basic relations concerning objects. In this respect, the Munich phenomenologists are continuing the tradition inaugurated by Brentano’s metaphysics: they identify these relations as the ones between the whole and its parts. Every object is a (real

10 English translation modified.

Essential Laws �

149

or possible) part, i.e. there are (real or possible) wholes, which include it.¹¹ Such relations hold for any kind of object whatsoever: take for instance the blue color of that book on my desk: it doesn’t exist without some extension or shape of which it is the color. In that sense, not only that specific shade of blue wouldn’t exist without being the color of that specific book, but any color wouldn’t exist without being extended. The relation here expressed is a relation of ontological necessitation between dependent species, but these relations between dependent species can hold as one-sided or mutual dependence relations. According to the Munich phenomenologists, we find many such relations not only in the field of perception (for example between color hue and brightness, or between a tone and its height, which are relations of mutual dependence), but also in the field of social institutions (every promise brings an obligation with it, which is a relation of one-sided dependence: you might have an obligation without promising anything, but you can’t promise anything without being in the obligation of fulfilling your promise) and in many other fields. As a matter of fact, most, if not all, Munich phenomenologists and Husserl’s Göttingen students acquainted with them conceived their contribution to phenomenology as investigations into the sphere of material necessity: this is the case for instance in the field of emotion theory with Scheler, Kolnai, Voigtländer, Geiger, in the field of aesthetics with Waldemar Conrad, Geiger, Schapp and Ingarden, in the field of social philosophy with Stein, Walther and Conrad-Martius; in psychology with Pfänder, Conrad and Beck, or in logic and ontology proper with Reinach, Pfänder, Héring, Ingarden and Spiegelberg. Another important characteristic of the conception of the synthetic a priori defended by Munich phenomenologists is the specific status they give to the synthetic a priori. Between 1901 and 1908, Husserl still seemed hesitant as to whether the proposition judged or the state of affairs was to be regarded as the bearer of the truth and necessity. In 1906, he considered this question as a matter of perspective: from the perspective of judgment, equivalent propositions denote different states of affairs, but from an ontological perspective, taking states of affairs to be objective complexes which are independent from the acts, equivalent proposition do express the same state of affairs.¹² This ambivalence was strongly criticized by Reinach.¹³ For him, and for most of the Munich and Göttingen phenomenolo-

11 See (Husserl 1913: 226). 12 See (Fréchette 2003) on this question. In A I 10/69, Daubert gives an account of a discussion with Husserl from August 15th 1906 where he defends the perspective account of states of affairs. The account developed two years later in 1908 (Husserl 1987: 28f) seems to be based on that earlier account. 13 See (Reinach 1989: 116 and 526).

150 � Guillaume Fréchette gists, the bearers of necessity, possibility, subsistence, etc. are always – and only – states of affairs. Therefore, synthetic a priori judgments like ‘every promise entails an obligation’ are true in virtue of the laws of essence concerning promises and obligations. These laws being nothing but “general principles expressing relations between states of affairs” (Reinach 1911b/1982: 339), the ultimate bearers of the material necessity expressed in such judgments are states of affairs.

3 Metaphysical Realism An obvious objection against the kind of account of logic championed by Reinach and other Munich and Göttingen phenomenologists would be to say that the position advocated here is merely a nominal variation on the position defended by logical realists like Bolzano for instance. After all, both Bolzano and Munich phenomenologists agree that propositional attitudes like “Anne thinks that the door is closed” have an objectual correlate as object, namely the “proposition” or “state of affairs” expressed in the that-clause.¹⁴ Propositions (or states of affairs) are also said to be standing in relation of ground and consequence¹⁵, they are conceived of as bearer of modalities¹⁶, they (and not judgments) stand in relation of contradiction. While truth and falsity are typically properties of propositions, metaphysical realists like Reinach like to think that a judgment like “the door is closed” is not correct because it expresses a true proposition, but because the state of affairs corresponding to it subsists: A judgment is correct if the state of affairs corresponding to it subsists; and two contradictory judgments cannot both be correct because two contradictory states of affairs cannot both subsist. The law relating to judgments thus obtains its foundation from the corresponding law relating to states of affairs. (Reinach 1911b/1982: 376)

Does that mean that Reinach would agree to say, with Meinong, that truth and falsity are nothing but subsistence and non-subsistence, i.e. properties of states of affairs? Reinach doesn’t agree with Meinong in that respect:

14 On the side of logical realists, see for example (Bolzano 1837: I, 154) for the conception of objective propositions as content (Stoff ) of judgments. On the side of metaphysical realists, see (Reinach 1911b/1982: 336). 15 On the side of logical realists, see (Bolzano 1837: I, §168, 177). On the side of metaphysical realists, see (Reinach 1911b/1982: 338). 16 On the side of logical realists, see for example (Bolzano 1837: 563). On the side of metaphysical realists, see (Reinach 1911b/1982: 339).

Essential Laws � 151

We acknowledge freely the difference between judgment and “proposition in itself”; but just as the proposition must be separated from the judgment, so also must it be separated from the state of affairs. A proposition is true when the state of affairs which is correlated with it subsists. And two contradictory propositions cannot both be true because two contradictory states of affairs cannot both subsist. (Reinach 1911b/1982: 376)

Still, the reason for this disagreement is unclear: which conceptual role is left to propositions? Because of his untimely death, Reinach didn’t have the opportunity to offer a detailed account of the conceptual role of propositions. In his Logic from 1921, Pfänder offers a more detailed account of the relation between propositions – which he calls judgments¹⁷ – and states of affairs: To each particular judgment there corresponds a state of affairs. To the judgment, “Sulfur is yellow”, there corresponds a state of affairs that consists of the material species, sulfur, and its being-yellow. The judgment projects this state of affairs out of itself. [...] It is also true that no judgment can be formulated without projecting a state of affairs. But the projected state of affairs is not for that reason identical with the formulated judgment. Rather, the state of affairs is the counterpart, the “intentional correlate” of the judgment that projects it. (Pfänder 1921/2009: 35)

In his lectures on logic and theory of knowledge, Pfänder presents the distinction in the following way: The proposition is made up of words, and words are made up of letters [...] the judgment is not made up of words or letters [...] the judgment may be true or false [while] the proposition can only be true or false in a metaphorical sense, or in another sense... One can make a judgment without constructing a proposition. [...] In the proposition, the judgment comes to expression [...] the expression relation is not a mere association. The judgment is the sense [Sinn] of the proposition, the thought construction [Gedankengebilde] inserted in it. (Pfänder 1912/1913: 18 November 1912)¹⁸

In other words, propositions express judgments which project states of affairs, which are the intentional correlates of judgments. But judgments are not ontolog-

17 Pfänder distinguishes between proposition (Satz) and judgment (Urteil) in a different way than Husserl, Bolzano, and Reinach himself. For Pfänder, propositions (Sätze) are purely linguistic entities, they are composed of words and they are not the bearer of truth or falsity. The real bearers of truth and falsity are judgments. 18 “Der Satz besteht aus Wörter, die Wörter aus Buchstaben [...] das Urteil besteht nicht aus Wörter oder Buchstaben [...] Das Urteil kann wahr oder falsch sein, der Satz kann nur in übertragenem Sinne, oder in einem anderen Sinne wahr oder falsch sein. [...] Man kann Urteile fällen, ohne Sätze zu bilden. [...] Im Satz kommt das Urteil zum Ausdruck. [...] Die Ausdrucksbeziehung ist keine bloße Association. Das Urteil ist der Sinn des Satzes, das ihm eingelegte Gedankengebilde.”

152 � Guillaume Fréchette ically independent of their bearers: the judgers. They are “thought constructions” (Gedankengebilde). On the other hand, if the state of affairs is the intentional correlate of the judgment, you can’t have a state of affairs without having a judgment. The projection metaphor also supports this mutual dependence. We find in Pfänder a sensibly different setting than the one found in the Bolzanian model, according to which propositions in themselves are independent of actual thoughts, sentences etc. about them. Contrary to Bolzano, Pfänder and Reinach see the truth-bearers as dependent on actual judgers. Bolzano doesn’t use the term “truth-making” to describe the relation between the world (objects and their properties) and the propositions.¹⁹ A proposition is true, according to him, if the object designated by the subject-idea (Subjektvorstellung) has the property that the proposition ascribes to it. Propositions, concepts (Vorstellungen), objects and properties are necessary for this account, but states of affairs are not. In other words, Bolzano, Pfänder and Reinach attribute very different ontological properties to truth-bearers. Furthermore, according to Pfänder, the analogy between propositions and states of affairs would not do since the state of affairs is the intentional correlate of the judgment. An analogous statement concerning propositions would be unacceptable for Bolzano. The features highlighted here might certainly help putting some flesh on the label of “metaphysical realism” as a characterization of the orientation defended by Munich and Göttingen phenomenologists, but there is more to say in this respect. Mulligan (2006) stressed another particular aspect that plays a central role in the Bavarian metaphysical realism: the concept of foundation.²⁰ This aspect is discussed extensively in Pfänder’s Logik of 1921: It lies in the essence of every judgement to make a claim to truth. Truth, as we have seen, is, according to its very essence, something which cannot attach to a judgement all by itself but only in a certain relation to something else, namely in the relation of agreement with the objects dealt with by the judgement. Only if this relation obtains can the judgement be true. But this relation requires necessarily in order to obtain two foundations, namely the judgement on the one hand and the behaviour of the objects the judgement deals with on the other hand. [...] Thus if a judgement is not only to lay claim to truth but also to have truth then the corresponding behaviour of the objects is absolutely necessary as a ground. The truth of a judgement, according to its essence, only obtains...if this reason is a sufficient

19 Interestingly, Bolzano holds that the variation of ideas (Vorstellungen) contained in propositions make given propositions true relative to specified variables. But here, the truth making relation holds between ideas and propositions. This is the only use of “wahrmachen” found in Bolzano. See (Bolzano 1837:II; §155, 114, 122 and §156, 133). 20 See also (Mulligan 2008).

Essential Laws

� 153

reason. It follows that every judgement, in order to be true, stands necessarily in need of a sufficient reason. (Pfänder 1921/2009: 231-232)

Not only is truth not to be confused with subsistence, but the different relations of foundation show that in order for the judgment to be true, there must be a bearer of the judgment on the one hand – as we already said, no judgment without judgers – and, on the other hand, the subsistence of a state of affairs. This difference might be illustrated and contrasted with a Bolzanian example: according to Bolzano, there is a true proposition stating the number of grapes that grew on the Italian soil in 1837, although there is no record of any sort giving the right number. The truth of that proposition is not dependent on any knowledge or piece of evidence concerning that number. Pfänder and Reinach would interpret this example in a different way. Although, for some contingent reasons, there is no true judgment about the number of grapes which grew on the Italian soil in 1837, there are (there were) grapes and a relation between them, in short: there could have been a state of affairs that could have grounded a true judgment about these grapes. A judgment is true, according to Pfänder, if the “relation of agreement with the object dealt with by the judgment obtains”. Since there is no relation of agreement in that context, there is no true judgment. In fact, there couldn’t be any true judgment about the number of such grapes since these grapes were never counted and don’t exist anymore. The important concept here at play is the concept of grounding or foundation. That a truth “attaches” to a judgment thanks to a “relation of agreement with the objects dealt with by the judgment” is what Pfänder and many other Munich and Göttingen phenomenologists call a law of essence. What this law expresses is a grounding relation: the judger (and his judgment) and the state of affairs judged both ground the true judgment.

154 � Guillaume Fréchette

4 Grounding and Essences: Logic and the Theory of Object Bolzano offers a semantic account of grounding, understood as a relation between true propositions.²¹ The Munich phenomenologists have a different take on this issue. We have seen in the last section that the state of affairs that p and the judging that p are both grounding the true judgment p. If that true judgment exists, “there must necessarily be a sufficient reason for its existence” (Pfänder 1921/2009: 260). What is then the ground for the principle of sufficient reason? This is where essences and grounding are brought into close connection: Its own sufficient reason lies, accordingly, in the nature of the judgment and the nature of truth. The truth of the principle of sufficient reason follows not from concepts (therefore, not from the concept of the judgment and the concept of truth), but from the characteristic essence of the judgment and of truth itself. (Pfänder 1921/2009: 260)

In this respect, Pfänder departs significantly from the Bolzanian account already discussed. According to Bolzano, and as it is stressed by Mulligan (2004: 413): As far as our knowledge is concerned, an object is no more than what we represent in our minds, whenever we believe we represent it. Thus in logic its idea constitutes its essence. (Bolzano 1837/I: §111)

In other words, what Bolzano calls “essences” are nothing but the concepts of objects, a position that Pfänder rejects. The reason for his rejection of foundation through concepts is similar to the one behind his conception of judgments as bound to real judgers. The same kind of relation holds between concepts and objects as between judgments and states of affairs: the latter is the intentional correlate of the former (Pfänder 1921/2009: 144). Analogously to states of affairs that are projected by judgments, objects are projected by concepts. Such objects are called by Pfänder formal objects, while the objects in themselves are called material objects. Interestingly, Pfänder traces the line between logic and ontology precisely to that distinction: the objects dealt with in logic are purely formal objects, which

21 But as Mulligan pointed out, Bolzano uses the terminology of “folget, herleiten, ableiten, vermöge des bloßen Begriffes, etc.” not exclusively for relations between true propositions. In fact, in §502, he appeals to the relation between ground and consequence to account for the concept of essence. See (Mulligan 2004: 414f.)

Essential Laws

� 155

one also could call intentional correlates, while ontology deals with material objects. Also, formal (or intentional) objects are defective by nature: If, for example, the concept “quadrilateral” means nothing but a plane figure described by four intersecting lines, then it does not belong to the intentional correlate of the concept “quadrilateral” to have four interior angles. (Pfänder 1921/2009: 144)

This point is interesting because it shows that the nature of concepts is to mean something in a particular way. Not only are concepts necessarily linked to thinkers, but they also are linked with defective objects: in a similar way, Pfänder should be bound to say that the concept expressed by “creature with a heart” has neither human nor any animal whatsoever as its formal object (and doesn’t have the same extension as the concept expressed by “creature with a kidney”) but only some of the following defective objects having only these two properties: having a heart and being a creature. This is of course an important departure from the more standard theory of concepts found in Bolzano for instance, according to which the objects of concepts are their extension. The difference is so important that one wonders if Pfänder is not thinking of the formal object of concepts simply as another term for their intension. This is not the case: one shouldn’t confuse the content of a concept with the sum of the characteristics of the object (Pfänder 1921/2009: 147). This distinction is basically an ontological distinction, in the sense that one shouldn’t confuse what belongs to the content or intension of a concept with what is the formal object. According to his view, the intension of the concept expressed by “triangle” is “object with three angles” and it is a semantic category, while the formal object is an object with three angles: it is an ontological category. Of course, there are many objects with three angles that are not triangles, but this doesn’t represent a problem for Pfänder’s account, since the formal object, as we said, is thoroughly determined by the concept. Another point should be added here: Pfänder distinguishes between implicitly and explicitly compound concepts. For instance, the concept expressed by “gold” has implicitly the parts “shiny, yellow metal”. These parts have correlates in the formal object. But what tells us that they belong to the formal object and not simply to the material object? Pfänder answers: Nevertheless, we must bear in mind that only those partial concepts can be implicitly contained in an object-concept that are really found there; only those, therefore, that posit their objective correlate in the corresponding formal object – and not those concepts that posit something that is found only in the material object, but is not at all co-intended by the objectconcept. (Pfänder 1921/2009: 150)

156 � Guillaume Fréchette But then, how to draw the line between partial concepts that are implicitly contained in an object concept and simply different concepts that don’t belong to it? The line is difficult to trace since Pfänder is advocating for a conception of extension qua formal object. In the case of the concepts expressed by “equiangular triangle” and “equilateral triangle”, for example, these can’t be said to be equivalent since they have different formal objects. Here, it seems that relations of identity of extension between concepts depend only on their material object, and not on their formal objects. Even the absence of an extension (as in the case of the concept expressed by “round square”), it seems that the formal object just depicts the structure of the content of the concept, while the being or non-being of a material object in this case is simply left aside. The distinction between the formal and the material object is a central distinction. It is on the basis of this distinction that Pfänder dissociates logic (as a theory of the formal object) from phenomenology understood as a theory of the material object: If it is certain [...] that the being S of an object, according to its essence, cannot exist without involving the being P of the object, then all objects that are really S are necessarily also P, and the universal judgment is an assertoric-categorial one. The epistemological question of how it is possible to recognize whether or not the being S of an object necessarily, or according to its essence, involves the being P of the same object, lies outside the circle of logical problems. (Pfänder 1921/2009: 399) Logic investigates thought-structures not only in themselves, but also purely for their own sake; while phenomenology considers thoughts only as the ideal content of certain acts of thinking, directing itself to the essential relationship of the act of thinking to other acts of thinking, to object-consciousness and to intentional objects. (Pfänder 1921/2009: 28)

Logic is therefore to be distinguished from a theory of essences. As a matter of fact, Pfänder holds that even the law of identity is itself not a real logical law: [T]his principle is not at all a genuine logical principle. It tells us nothing directly about any logical object – except, of course, that, insofar as logical entities are objects, they are as such (according to this principle) identical with themselves. This is because the principle of identity as given above refers to objects in general, and no logical investigation is needed to recognize it as true. (Pfänder 1921/2009: 207)²²

22 “Er besagt nämlich direkt nichts über irgendeinen logischen Gegenstand, außer insofern auch die logischen Gebilde überhaupt Gegenstände sind und als solche natürlich, wie es der Satz behauptet, mit sich selbst identisch sind. Denn der obige Satz der Identität bezieht sich auf Gegenstände überhaupt, und er bedarf, um als wahr erkannt zu werden, keinerlei logischer Betrachtung und Untersuchung” (Pfänder 1921: 182).

Essential Laws

� 157

In that sense, there is no co-extensivity between logic and object theory. Moreover, the law of identity is a law of object theory that is deprived of logical content: “[a]ll that is required for its validity is the universal and formal nature of objects in general, not the specific nature of logical objects.” (Pfänder 1921/2009: 207). The law of identity is a law of formal ontology and not a law of logic. The principle of identity is, therefore, neither “immediately evident,” nor shown to be true through either psychological insights or on the basis of an inductive generalization from the study of individual examples. Its truth must rather be made apparent in another way. (Pfänder 1921/2009: 215)²³ That every object is identical with itself is, of course, immediately evident; for its self-identity is grounded immediately and finally in the nature of the object in general. The principle that declares this, however, is not a logical principle but, as already noted, a principle from the general theory of objects or formal ontology. Nevertheless, the state of affairs it posits forms the ultimate basis for the truth of the logical principle of identity. For only if this state of affairs obtains is the logical principle also correct. To this state of affairs one must necessarily return if the truth of the logical principle is to become evident. Because of this relationship of grounding between the formal-ontological state of affairs and the logical principle, it is understandable that traditional logic usually only refers to the ontological principle and forgets to formulate the genuine logical principle expressly. (Pfänder 1921/2009: 216)²⁴

Spiegelberg, a student of Pfänder, also disapproves of Husserl conception of the relation between ontology and logic. Ontology, he holds, is not a “pure logic”: “logic is based on ontology and is basically impossible without it” (Spiegelberg 1930: 6).

23 “Der Satz von der Identität ist also weder ‘unmittelbar evident’ noch durch psychologischen Erkenntnisse noch durch inductive Verallgemeinerung aus untersuchten einzelnen Beispielsurteilen als wahr zu erweisen. Seine Wahrheit muß vielmehr in anderer Weise ersichtlich gemacht werden” (Pfänder 1921: 190). 24 German original in (Pfänder 1921: 191): “Der Satz (der Identität) ist freilich kein logischer Satz, sondern wie schon bemerkt, ein Satz der allgemeinen Gegenstandstheorie oder der formalen Ontologie. Aber der Sachverhalt, den er setzt, bildet doch die letzte Grundlage für die Wahrheit des logischen Satzes von der Identität. Denn nur, wenn dieser Sachverhalt besteht, ist auch der logische Satz berechtigt. Zu diesem Sachverhalt muß man zurückgehen, wenn die Wahrheit des logischen Satzes ersichtlich werden soll. Aus diesem Begründungsverhältnis zwischen dem formalontologischen Sachverhalt und dem logischen Satz wird verständlich, daß die überlieferte Logik letzten Endes auf einer formal-ontologischen Tatsache basiert, so ist sie doch nicht mit der allgemeinen Gegenstandstheorie oder formalen Ontologie identisch.”

158 � Guillaume Fréchette

5 Laws of essence What do realist phenomenologists understand as laws of essences? Essences are at the core of what Reinach calls an “axiom of phenomenology”: To every objectual domain is assigned a sphere of aprioric content, an a priori regularity of essence. This sphere must be investigated prior to any empirical observations. (Reinach, 1989: 440)

At least in Reinach’s works, the expression “regularities of essence” (Wesensgesetzlichkeit) is often used and is more often used than the expression “law of essence” (Wesensgesetz). There is one obvious reason for this: laws are propositions. If the way essences relate to one another is regulated by propositions, then propositions are more fundamental than essences, and this would of course go against the point made by the realist phenomenologists. The term “regularity of essence” designates here not a proposition but a state of affairs that has essences as parts and which can be expressed by a law of essence. States of affairs being, according to the realist phenomenologists, more fundamental than propositions, the “laws of essence” formulate relations between essences. The expression of these “laws” gives us a way to grasp the relations between essences. Regularities of essence are based on a grounding relation between the predicate and the subject expressed in the proposition designating this regularity (Reinach 1989: 363). Red and blue are different. According to Reinach, the sentence “red and blue are different” expresses an essential law, which has the form “being-b grounded in A and C is ascribed to it” (where b is the predicate “different”, A is the essence and C is “red and blue”). In other terms, if properties are the ontological pendant of predicates, these are not the essence of the object. The essence is what grounds (gründet) the property of being different, or being a color, etc. Interestingly, the importance of laws of essence for Munich and Göttingen phenomenology comes from Husserl’s Logical Investigations, especially from his explanation of the relation of dependence: Hue, saturation and brightness are for instance properties that every color necessary has. There is no specific blue color which hasn’t a specific hue, a degree of saturation and of brightness. As Husserl writes in the third Logical Investigation: The inability-to-exist-by-itself of a non-independent part points therefore to a law of essence, according to which the existence of a content belonging to the parts’s pure species (e.g. the species of color, form etc.) presupposes the existence of contents of certain pertinent pure species.... Non independent objects are objects belonging to such pure species as are

Essential Laws �

159

governed by a law of essence to the effect that they only exist (if at all) as parts of more inclusive wholes of a certain appropriate species. (Husserl 1913: III, §7, 240/2001: vol. 2, 12)

The relation of dependence illustrated here by Husserl is what Brentano calls the metaphysical parts of a whole. Metaphysical parts like the hue, brightness and saturation of a color (here the whole) were called by Brentano the essences (Essenzen) already at the end of the 1860s.²⁵ Do synthetic a priori laws necessitate instantiation? According to Reinach, these laws don’t need a single instantiation. A single example suffices to illustrate a law of essence: It is intuitively graspable, from one example (of conviction), that every judgment can only have one state of affairs, one being-such, as correlate, according to its essence. Similarly, it belongs to the essence of moods that they don’t need an intentional correlate; it belongs to the essence of genuine questions that they don’t have their source in certitude, but that they are rather grounded in incertitude. (Reinach 1989: 439)²⁶

6 Essences, Ideas, Eide, Morphes Jean Héring (1890-1966), an Alsatian student of Husserl in Göttingen, developed an account of essences that has been very influential on the later developments of the Munich and Göttingen phenomenology, particularly in (Ingarden 1925; 1928; 1964-1965), (Stein 1950), (Conrad-Martius 1957) and (Spiegelberg 1930). Héring (1921) distinguishes between five central categories: the individual object, its essence (Wesen), the essentiality (Wesenheit, eidos), the quiddity (Washaftigkeit, Morphe) and the idea (Idee). According to him, every individual object has only one essence, which it doesn’t share with any other individual object. The set of two white coffee cups on my table is composed of two objects

25 See (Brentano 1867a: 137). Husserl had a transcription of these lectures. The importance of the influence of Stumpf’s concept of partial contents (Teilinhalte) in (Stumpf 1873: 109) is certainly undeniable, but since Stumpf himself was influenced by Brentano’s 1867 lecture on metaphysics, it just seems more to the point to link Husserl directly to Brentano in this respect. In his Descriptive Psychology, Brentano will change the name of these parts for sich durchwohnenden Teile but the concept remains quite the same. See (Brentano 1982 [1891]/1995). 26 German original: “An einem Beispiel (von Überzeugung) ist intuitiv zu erfassen, daß jedes Urteil wesensmäßig nur ein Sosein, einen Sachverhalt, zum Korrelat haben kann und muß. Ebenso liegt es im Wesen von Stimmungen, daß sie intentionaler Korrelate nicht bedürfen; im Wesen von echten Fragen, daß sie nicht aus Gewißheit entspringen, sondern in Ungewißheit fundiert sind.”

160 � Guillaume Fréchette and, belonging to them, two different essences. Héring says that the essence of the individual object is its Sosein. It belongs to the essence of a feather to be able to write finely, but it doesn’t belong to this essence to be a feather that lies on my table. Such essences are also individual in themselves. Héring distinguishes further two relations: x belongs to the essence of y (rel1) x follows from the essence of y (rel2) Rel1 is involved in the case of the capability for a feather to write. In that sense, it seems that even dispositions would be categorized as essences. What belongs to the essence of something is its essence kernel (Wesenskern). Rel2 is involved in cases like the one of a sphere of one meter of circumference. It follows from the essence of that sphere that it is smaller than the blue coffee cup on my table. In the case of rel2, X seems to be a state of affairs, while in the case of rel1, it might be a state of affairs, but it also may be a disposition.²⁷ Therefore, necessary properties of an object are not parts of its essence: a 50m2 apartment is necessarily smaller than a 100m2 apartment, but it only follows from the essence of the 50m2 apartment, it doesn’t belong to its essence to be smaller. According to Héring, not only general objects (like kinds or sorts) have essence, but individual objects as well. And both general and individual objects can be ideated, or put into ideas. Another feature of Héring’s concept of essence (Wesen, which he sometimes calls the Sosein of objects) is that it belongs to its object in such a way that it ceases to exist when the object ceases to exist. This is of course not the case with the idea. The Wesen has to be distinguished from the Washaftigkeit, or its quiddity. If this wine stain has an essence (to which for instance it belongs to be a stain on some surface that absorbs liquid to a minimal extent), it also has its morphe, something like “wine-stainness”, which makes the wine stain what it is. The morphe is the morphe of a specific object, the object’s morphe. In that sense, Héring’s morphes are quite similar to what is usually called today a trope. But it also makes sense to speak, according to Héring, of the “wine-stainness” not of a specific stain, but “an und für sich genommen”, or of “the winestainness as such”. In such cases, we then mean something which is in itself completely free of any relation to objects, something which “is what it is”, independently of the existence or not of real or ideal worlds of objects. We can think them [= the essentialities] without the world [...] they are autonomous and rest in themselves. (Héring 1921: 510)

27 See here also (Ingarden 2007: 50f.)

Essential Laws �

161

The winestainness as such is what Héring calls the essentiality (Wesenheit, eidos). Finally, the fifth category in Héring’s ontology is called the idea. I may have bought twice a specific lamp sold by Ikea, once for the bedroom, the other for the living room: but there is only one idea of the lamp, its model so to speak. This is what Héring calls an idea. Héring’s ontology of essences could be schematized in the following way:

Idea (of a red stain) if

ex em pl

if

es

pl em ex es

Red stain 2

participates in

Red stain 1 Essence being a stain of the most expensive Austrian Wine adhering to a surface Essence kernel

(Wesenskern)

Morphe (Redness ot the red of the stain) Eidos: Redness as such

instantiates

Fig. 2

Eide (Wesenheiten) are realized in morphes thanks to the object (here the red stain 1), which is also called its realizator (Héring 1921: 510). It is of course problematic to have a distinction between ideas and eide, at least from a strictly terminological point of view. But there are fundamental distinctions to be made here: ideas have what Ingarden (2007: 56) calls variables (Veränderliche) in their content. For instance, the idea of a red stain may be subject to different variations, such as “red stain of wine”, “red stain on the carpet”, etc. This is not the case with the eidos: eide are in contrast with ideas completely determined. Furthermore, morphes are realizations of eide on the basis of the object which is then said to be the realizator (Héring 1921: 510). Also, ideas are conceivable from two different points of view: from the point of view of their ideal mode of being, but also from the point of view of the objects they exemplify.²⁸ It is also important to stress here that ideas are not concepts (Héring 1921: 533). Concepts are unintuitively intented meaning units, while ideas can at least

28 See (Héring 1921: 530).

162 � Guillaume Fréchette in principle be intuited. Another difference between idea and concepts is for Ingarden (2007: 63) that there are “contradictory concepts”, or as Bolzano would call them, “objectless notions” (gegenstandlose Vorstellungen). But there are no “objectless ideas” in Ingarden’s and Héring’s understanding of the term. The distinction between object, essence (and essence kernel), morphes, eide, and idea is a relatively complex one. The relation between the morphe and the eidos corresponds, in Husserl’s terminology, to the relation between the moment and the species. So what is the role played by Héring’s essences (Wesen), if this role is not precisely played by morphes? Morphes are moments (or instanciations, or tropes) of eide, or in other terms, one could say that they are particularized properties, in the sense that they are realizations (as Héring puts it) of properties qua universals (redness as such, for example). As such, one single object can participate in different eide, which are then instantiated by different morphes. In the case of essences, it is different. Essences are individuals of a slightly different kind: they attach directly to objects, they are not instantiations of universals and are in no relations with universals. As a matter of fact, they are Unikate, insofar as there is for each individual object one single individual essence. Interestingly, Héring calls essences also the Sosein of objects, in a sense that might remind us of Meinong. Indeed, Meinong proposed to distinguish between the being (Sein) and the being-so (Sosein) of objects, the first being correlative to acts of judgments, while the second being correlative to assumptions. In 1904, Meinong formulated the principle of independence of being-so from being, according to which the being-so of an object is not affected by its non-being.²⁹ Héring would obviously reject that principle since essences are correlates of objects and cease to exist when their object cease to exist. Furthermore, Meinong’s talk of the Sosein in the sense of “an object’s having properties” goes in another direction than Héring’s account of essence. Essences are not properties (even not particularized properties) of objects but rather seem to be some kind of set of traits (Züge) belonging to the object. As the result of a glass of wine accidentally falling on the carpet, the red stain has an essential trait (Wesenszug) of being caused by the falling of the glass. This trait might or might not be an essence kernel of the red stain.³⁰ But the existence of the set of these traits (as of each of them individually), though dependent upon the existence of the object, is not to be understood exactly as a trope, in the sense of morphes. In that sense, essences are not tropes of the red of the stain, but a trope of the object tout court. Essences seem more to

29 See (Meinong 1904: 8). 30 Hering’s point on essence kernels is not clear, since his examples are using only ideal objects, and not concrete objects. See (Hering 1921: 499).

Essential Laws

� 163

be like some kind of perspectival or subjective objects, which adapt to our knowledge of the objects they are the essence of, while these objects themselves are as they are independently of our perspective on them. Phenomenological analysis understood as a Wesenschau, or intuition of essences, would simply mean here that what we describe is the essence of the object, which is in itself variable. In that sense, the empirical discovery of Metis and Thebe, the two moons of Jupiter discovered by Voyager 1 in the 1970s, changes nothing about the system of Jupiter as an object, but it completes, or add more traits to its set of essential traits, or its essence.

7 Anumericity Héring’s article influenced many early phenomenologists to go against Husserl’s use of “Wesen’‘ (essence) and “Idee” (idea) as synonyms.³¹ One of the outputs of these criticisms is that Husserl comes to the knowledge of ideas by making abstraction of the individuality of the object’s essence.³² But Spiegelberg is not satisfied with this conclusion. He asks the question of the foundation of the character of idea. In other words: what is then the principle of ideality? Abstracting from the essence of individual objects will only give us individual parts. And to say simply that an idea has no reality is not enough, since not only ideas don’t have reality – illusions, fictions, etc. do lack reality as well. Even numbers, which have no individuality, cannot be considered as ideas. According to Héring, as we have seen, ideas are general in the sense that they are undetermined. That ideas allow variation in their content is precisely a sign of this generality. In his 1930 dissertation, Spiegelberg also focuses on this distinction, insisting not only that ideas are undetermined, but also that they in fact are anumerical: The difference between idea and essence in the sense of an essence kernel doesn’t really need a further specification of its own. The essence kernel is an exceptional group of inner elements which are distinguishable similarly in the idea and in the exemplars. The essence kernel of an individual is individual, the essence kernel of an idea is materially anumerical. The essence kernel of an individual can’t be an anumerical idea. This is completely excluded by both essences. (Spiegelberg 1930: 222)³³

31 See for example (Spiegelberg 1930: 2199) and (Pöll 1936: 31). 32 See (Spiegelberg 1930: 214f.) 33 “Der Unterschied von Idee und Wesen im Sinne des Wesenskerns braucht kaum noch eigens herausgestellt zu werden. Der Wesenskern ist eine ausgezeichnete Gruppe von inneren Ele-

164 � Guillaume Fréchette Comparing ideas with essences, or even with objects, Spiegelberg notes Regarding such a question [whether they are numerically one or many], ideas behave with a complete indifference, because they are numerically without any quantity, anumerical, without any number. [...] Two-in-general is [not] numerically determined, in contrast with the individual twos. (Spiegelberg 1930: 99)³⁴

We find here the point addressed by Schapp in his remarks about the “infinitely many twos”. There are infinitely many twos: as a matter of fact, the essence of being two comes to every set of two objects which is considered as a composed object. Since essences are individual, the object composed by the reunion of the table and the coffee cup has the essence of being two. When this object will cease to exist, the essence of being two (as the essence of this specific object) will cease to exist as well. On the other hand, the idea of the two (die Zwei-überhaupt) has no quantitative property. As Héring and Ingarden underlined first, ideas are undetermined and have variation places in their content, what Spiegelberg, Pöll and Beck call anumericity.³⁵ According to Spiegelberg, it is precisely because ideas are undetermined that they are multiply realizable. Therefore, the standard Platonic conception of ideas is not optimal: There is nothing in principle to prevent replacing a “model” (paradeigma) with a whole group that is formed by it in the same way. Only in this way can multipliability be ruled out if the foundation of all multiplication, the numerical one-hood, is absent [...] [The idea] is the one object in whose case it remains undetermined whether it is internally structured like one or however many exemplars. Only the qualitative aspect is fully developed in it as in the case of other numerical objects, whereas the numerical aspect is altogether absent in it. (Spiegelberg 1930: 100)³⁶

menten, die sich gleichmäßig in der Idee wie in den Exemplaren unterscheiden lassen. Der Wesenskern eines Individuums ist individuell, der der Idee stofflich anumerisch. Niemals kann der Wesenskern eines Individuums eine anumerische Idee sein. Das ist durch beiden Wesen völlig ausgeschlossen.” 34 “Die Ideen verhalten sich einer solchen Frage [ob sie numerisch eins oder viele seien] gegenüber gänzlich indifferent, weil sie überhaupt numerisch quantitätslos, anumerisch, anzahllos sind. [...] [D]ie Zwei-überhaupt [ist ebensowenig] numerisch bestimmt, im Gegensatz zu den einzelnen Zweien.” 35 See (Beck 1929). Beck thinks also that ideas are behind all numbers. The idea is not multipliable, but can exist in different synthesis. 36 German original: “es besteht kein prinzipielles Hindernis, an Stelle des einen ‘Musterbildes’ (paradeigma) eine ganze Reihe von ihm gleichgeformten anzusetzen. Nur dann kann Vervielfältigbarkeit ausgeschlossen sein, wenn die Grundlage aller Vervielfältigung, die numerische Einsheit, fehlt [...] [die Idee] ist der eine Gegenstand, bei dem es wesensmäßig in der Schwebe bleibt, ob er innerlich wie ein oder wie viele Exemplare gebaut ist. Nur die qualitative

Essential Laws

� 165

An idea is what Spiegelberg calls “unidividuelles und anumerisches Quale”. In this sense, Schapp is right when he says that there are “infinitely many twos” according to the Munich phenomenologists. But this is possible only on the basis that ideas are purely qualitative objects and by definition (or rather by essence) deprived of numericality.³⁷

8 Final remarks The early phenomenologists attributed a central importance to the notion of essence and its related family members: laws and regularities of essences and ideal objects. I showed first that this importance was directly dependent upon their conception of the synthetic a priori. According to them, material necessity is at the basis of the synthetic a priori, and not the modal necessity, as Kant understood it. Causal connections are in this sense not grounded in pure reason or in subjective necessity, but in the essence of things. Therefore, synthetic a priori judgments like “every promise entails an obligation” are true in virtue of the laws of essence concerning promises and obligations, and not in virtue of modal necessity. These laws are nothing but “general principles expressing relations between states of affairs.” (Reinach 1911b/1982: 339) I also distinguished this position, which I labeled a “metaphysical realism”, from the logical realism championed by philosophers like Bolzano, according to which propositions are the bearer of truth, modalities, and are standing in relation of ground and consequence. This distinction presupposes that propositions and states of affairs are to be distinguished, as proposed by Munich and Göttingen phenomenologists. For them, while truth and falsity are genuinely properties of propositions (they agree with logical realists on that respect), the ground for the correctness of propositions is to be found in the subsistence of states of affairs. Therefore, subsisting states of affairs are the foundation of true propositions. This position, according to which ontology is the foundation of logics, was particularly defended by Pfänder, as we have shown. In the last four sections of the paper, I discussed different elements of the metaphysical realism exposited in the first part. The first element is the relation of

Seite ist bei ihm voll entwickelt wie bei den numerischen Gegenständen, die numerische fehlt ihm vollständig.” 37 Pöll, a student of Geyser and Pfänder in Munich, followed Spiegelberg’s insight according to which the lack of numericality should count as the principle of ideity (Ideitätsprinzip) (see Pöll 1936: 94). The idea is, according to its essence, external to the order of numbers.

166 � Guillaume Fréchette foundation advocated by the early phenomenologists between ontology and logics: Ontology provides the ground for logics. The second element is the distinction between essence (Wesen) and idea (Idee). On this point, early phenomenologists are in disagreement with Husserl. Finally, the last aspect was the property of anumericity, a property possessed exclusively by ideas. This is the reason why there are “infinitely many twos” according to the Munich phenomenologists, as Schapp pointed out: the two white cups participate in the essence of being two white cups, but they exemplify the two ‘in general’ which, in its turn, is said to be anumerical.

Acknowledgement: This paper was supported by the Austrian Science Fund (FWF), project M1403-G15.

Denis Seron

Adolf Reinach’s Philosophy of Logic Denis Seron: F.N.R.S., University of Liège, Belgium

Reinach was one of the most eminent representatives of the so-called “realist phenomenology” – a term which usually encompasses Husserl’s first pupils in Göttingen. In spite of his untimely death in 1917 at the age of 33, Reinach left a profound and powerful work, the significance of which went unrecognized until rediscovered by Barry Smith, Kevin Mulligan, James DuBois, and Karl Schuhmann about twenty years ago.¹ His numerous writings are devoted to a wide range of philosophical topics, especially language, right, action, and realist ontology. The present study is about one of the most innovative aspects of Reinach’s philosophy, namely his theory of states of affairs. Its purpose is to show that this theory – like, to a lesser extent, Meinong’s theory of object, with which it has striking similarities – may be regarded as an insightful alternative to Bolzanian semantic objectivism, including its phenomenological variant in Husserl’s Logical Investigations (Smith 1987b: 192; and 1989a: 52). For one of Reinach’s most original views on logic is that it should, at least to some extent, be redefined as a theory of states of affairs. This view of logic was very different from Husserl’s, and explicitly rejected by other realist phenomenologists, such as Alexander Pfänder.² Reinach’s theory of states of affairs provides at least a healthy questioning of the Fregean view, at best a fruitful alternative to the “Myth of Meaning”. In any case, it raises some fundamental questions which are too often neglected, especially the question of what logic is about.³ My aim is to sketch the contours of Reinach’s philosophy of logic, and to discuss its most important philosophical implications and difficulties. I shall first briefly recall what a state of affairs is in Reinach’s view, and outline his attempt to reduce logic to a theory of states affairs. Next, I shall try to show that Reinach’s phenomenological realism is a direct consequence of his bias in favor of states affairs against propositions. I shall conclude

1 For a bio-bibliographical survey see (Schuhmann & Smith 1987), (Smith 1987a) and (Schuhmann 1984). 2 See (Pfänder 1921: 165), who transfers states of affairs to the theory of knowledge and limits logic to “judgments considered purely as such”. 3 See (DuBois 1995: 123).

168 � Denis Seron by pointing out some difficulties of Reinach’s approach, which are due less to its realist than to its Meinongian background.

1 States of affairs The notion of state of affairs (Sachverhalt) plays in Reinach’s philosophy a central role which has no equivalent, except perhaps in Wittgenstein’s Tractatus (Smith 1982: 293; and 1987b: 223).⁴ On this subject Reinach mainly refers to Husserl and Meinong. However, his states of affairs are, in key respects, more akin to Meinong’s “objectives” than to Husserl’s states of affairs. There are at least three important differences between his and Husserl’s states of affairs. First, in a way that recalls the Tractatus, Reinach draws a sharp distinction between states of affairs and objects (Gegenstände). States of affairs certainly are “objectual correlates” of judgments, but they are not “objects in the strict sense” (Reinach 1911b: 111 and 114/1982: 334 and 338). The red flower I see is an object; the fact that the flower is red, “its being-red”, is a state of affairs. On the other hand, object and state of affairs, the flower and its being-red, are both “objectities” (Gegenständlichkeiten) or “objective formations” (gegenständliche Gebilde), these terms being generic titles for any entity of any kind. Secondly, Reinach opposes the existence (Existenz) of objects to the “occurrence” or “obtaining” (Bestand) of states of affairs. This distinction is more than merely of a terminological kind. A noteworthy difference is that, given a state of affairs /S is P/, this state of affairs’s non-obtaining involves there being an obtaining negative state of affairs /S is not P/, while the inexistence of an object A does not imply the existence of a negative object non-A. For example, the fact that the flower’s being red does not obtain implies that the flower’s not being red obtains, but the inexistence of Santa Claus does not imply the existence of a non-Santa Claus (Reinach 1911b: 117/1982: 341). Husserl, too, used the verb bestehen to qualify states of affairs, but it does not seem that he ever treated occurrence and existence as mutually exclusive. Reinach’s view seems much closer, in this respect, to Meinong’s theory of “objectives”, with this difference, however, that for Meinong not only states of affairs

4 As Barry Smith (1989a) has pointed out, the philosophical notion of state of affairs can be traced back to Julius Bergmann (1879: 2ff., 252ff.), who defines the correctness of judgment as an “adequation with the state of affairs” (Übereinstimmung mit dem Sachverhalt). Another, betterknown source is the 1888 lecture by Carl Stumpf to which Reinach refers in his essay on negative judgment. See (Reinach 1911b: 114/1982: 373).

Adolf Reinach’s Philosophy of Logic

� 169

“obtain”, but also ideal objects such as numbers and geometrical figures (Reinach 1911b: 116/1982: 374). Thirdly, Reinach and Husserl also disagree on non-obtaining states of affairs. The former clearly affirms, in his essay on negative judgment, that “occurrence is by no means included as an essential moment within the concept of a state of affairs” (Reinach 1911b: 116/1982: 374). This means that a state of affairs does not need to obtain in order to be a state of affairs, or that states of affairs may obtain or not obtain. Hence we may suppose that a complete theory of states of affairs should deal with non-obtaining states of affairs such as the being-gold of a mountain, or even with necessarily non-obtaining states of affairs, like the being-round of a square. Here again, Meinong’s influence is obvious and explicit (Reinach 1911b: 116-117/1982: 340-341). From the point of view of Husserl’s Logical Investigations, by contrast, the being-gold of the mountain is not an object, nor, a fortiori, a state of affairs. It is nothing at all, even though “the mountain is golden” is an existing (false) proposition. In his essay on negative judgment, Reinach proposes five necessary and sufficient conditions for states of affairs (Reinach 1911b: 114/1982: 338ff.; see 1908: 343-344; and 1913: 427): (1)

A state of affairs may be “believed or asserted” (geglaubt oder behauptet).

In other words, states of affairs, and only them, are subject to “assertions” (Behauptungen) and to beliefs or “convictions” (Überzeugungen). It is impossible to assert or believe an object. As only beliefs and assertions are usually called judgments (Urteile), states of affairs may be defined as objectual correlates of judgments. (2)

States of affairs, and only them, may stand in relations of implication.

If we are in presence of antecedent-consequent or “foundation” (Begründung) relations, then we can be sure that we are in presence of states of affairs.⁵ (3)

States of affairs, and only them, present modalities.

There are states of affairs not only of the form /S is P/, but also of the form /S is presumably P/, /S is possibly P/, etc. This very strong thesis is in marked contrast to the views of Husserl and Pfänder, who both admit ontological as well as logical

5 See (Reinach 1913: 460): relations of Grund and Folge are expressed by “hypothetical propositions”.

170 � Denis Seron modalities.⁶ Likewise, Husserl does not seem to confine ontological modalities to propositional acts. (4)

States of affairs, and only them, may be positive or negative.

Thus, it is only in a “derivative” sense that propositions and judgments are said to be positive or negative. This condition corresponds to the view, usually associated with Aristotle and also defended by Brentano, that there can be no negative presentations. There are states of affairs of the negative form /S is not P/, but it makes no sense to talk of negative objects, for example of non-Socrates or a non-pen. (5)

States of affairs, and only them, may be known (erkannt).

This condition should not be confused with condition (1). For knowledge, according to Reinach, is different from both belief and assertion. “Knowledge”, which Reinach also calls “evidence” (Evidenz), is the mode of givenness peculiar to states of affairs (Reinach 1911b: 123/1982: 375). Objects, when perceived or imagined, are “seen” (gesehen), or geschaut; states of affairs are not seen, but “known”, or erschaut. Once again, Reinach’s analysis sharply contrasts with the Logical Investigations, in which states of affairs were held to be given in some special type of categorial intuition, in “acts of synthesis”. Reinach explicitly distinguishes his evidences from Husserl’s categorial intuitions, which he claims always give us objects, not states of affairs (Husserl 1900-01: VI, §§48-49; see Reinach 1911b: 119/1982: 343). According to Reinach, beliefs and assertions presuppose the knowledge of the state of affairs. Suppose, for instance, that you discuss with a colleague what color this or that object in the room is.⁷ You get closer and find that the object is red. This means that a state of affairs, the being-red of the object, is given to you, and it is only on the basis of this evidence that you can believe that the object is red. Next, you come back to your colleague and say: “This is red”. What is expressed by this statement, Reinach notes, is not merely a belief. There is also an act of asserting by which I “intend” or “mean” (meine) a state of affairs that is no longer given or “known”. So, belief and assertion are independent of the knowledge of the state of affairs, but at the same time they can possibly be based on such knowledge. Interestingly enough, these distinctions are supposed to apply to negative states of affairs as well. The idea is that the belief that the object is not yellow does not merely presuppose that the subject is given a positive state of affairs /the object is red/, but also that she “grasps” a necessary implication relation between

6 See (DuBois 1995: 118ff.) 7 See (Reinach 1911b: 97/1982: 317).

Adolf Reinach’s Philosophy of Logic

� 171

this state of affairs and the negative state of affairs /the object is not yellow/. Now, in Reinach’s view, both constitute a genuine knowledge or positive evidence of the negative state of affairs /the object is not yellow/ (Reinach 1911b: 124/1982: 353).⁸ The five conditions above are necessary and sufficient conditions. They can be fulfilled only by states of affairs: “These determinations are to this extent sufficient, that every entity to which they apply is of necessity a state of affairs.” (Reinach 1911b: 117/1982: 341) Note that Reinach’s lessons of 1913 published in the Sämtliche Werke as “Introduction to Philosophy” provide a slightly different list (Reinach 1913: 427). Reinach substitutes for condition (5) – the possibility of being “known” – the following criterion: unlike objets, which exist or do not exist, states of affairs and only them “obtain [bestehen] or do not obtain”. This criterion is closely connected to condition (4). The main difference, as we have seen, is that the non-obtaining of a state of affairs entails the obtaining of the contradictory state of affairs. For example: the fact that the flower’s being-red does not obtain entails that the flower’s not-being-red obtains. Nothing similar is true of objects: the inexistence of Santa Claus does not imply the existence of the negative object “non-Santa Claus”.⁹

2 Propositions In Husserl’s Logical Investigations, logic was defined as a theory of science or a theory of theories. Since a theory, in Husserl’s view, is a system of propositions, logic was equally defined to be a theory of propositions. According to Husserl, just as for Bolzano and Frege, logic is a theory whose objects are theories and parts of theories, that is, propositions as well as parts and systems of propositions, in short: meanings (Bedeutungen). Logic is thus distinct from formal ontology, which is the formal theory of objects, including states of affairs. Furthermore, logic in Husserl’s sense is also different from the psychology of judgment. According to the Investigations, the logician studies propositions in specie, “in themselves”, regardless of their instantiation in judgments, thoughts, questions, etc. This allows her to enunciate a priori or “ideal” laws, whereas descriptive psychology is confined to inductive or “real” laws.

8 “[W]ird auch dieser negative Sachverhalt erkannt.” 9 The realist phenomenologist Jean Héring (1921: 497), made a similar remark about properties: while for any state of affairs /S is P/ there is a contradictorily opposed negative state of affairs /S is not P/, there is no negative property contradictorily opposed to the property P.

172 � Denis Seron Reinach never defended such a Bolzanian-style view of logic, and this certainly makes his approach both unique and innovative when compared to Husserl’s Logical Investigations. For him, as we shall see, logic must be, first and foremost, a theory of states of affairs. Thus, the first question to be asked is what becomes of propositions in Reinach’s philosophy of logic. Reinach’s preserved work gives only few hints about this crucial question, which is left largely unexplored by commentators. A first point to be taken into account is that Reinach, far from rejecting the notion of proposition altogether, explicitly appropriates the distinction between proposition and state of affairs (Reinach 1911b: 138/1982: 376; see Künne 1987: 184). In a passage from the Nachlass quoted by Barry Smith, Reinach regrets that “all Austrians always confuse propositions and states of affairs.” (B II 5, 375)¹⁰ Likewise, in two footnotes of his essay on negative judgment, he objects to Meinong that his notion of “objective” rests upon a confusion between proposition and state of affairs, which leads him to erroneously hold them as true or false (Reinach 1911b: 114 and 116/1982: 373-374 and 374). In fact, he argues, “states of affairs obtain or do not obtain; propositions are true or false” (Reinach 1911b: 116/1982: 374). Just as Husserl in the Logical Investigations, Reinach thus distinguishes between judgments which are correct or incorrect (richtig oder unrichtig), propositions which are true or false, and states of affairs which obtain or do not obtain. In a lecture dated December 1910, Reinach defines proposition as follows: A proposition [Satz] in our sense is not a grammatical proposition, but the meaning of a group of words, the thought that serves as its foundation. The proposition is independent of the judging experiences and factual cases. It is ideal and extra-temporal. (Reinach 1910b: 351)

It is plausible to say that the term “thought” (Gedanke) is borrowed from Frege, whose works, as we know from the lecture On Phenomenology of 1914, were known to Reinach. However, Reinach uses the word in a broader sense, according to which not only propositional (Satzgedanken), but also nominal meanings are Gedanken (Reinach 1913: 419-420).¹¹ In any case, this definition of proposition is hardly distinguishable from Husserl’s definition in the Logical Investigations: the proposition is some variety of meaning (Bedeutung), it is in some sense “objective”, its truth or falsity is independent of the mental act in which it is thought or asserted, etc.

10 Quoted by Smith (1982: 310, footnote 16). By maintaining the distinction between truthbearer and truthmaker, Reinach clearly departs from Meinong’s approach. See (Salice 2009: 262f.) 11 See (Salice 2009: 250). The same use can be found in (Pfänder 1921: 141f., 158-159).

Adolf Reinach’s Philosophy of Logic

� 173

In fact, the disagreement between Reinach and Husserl is not about the distinction between judgment, proposition, and state of affairs. The substantive difference in their positions turns on how they conceive of the mutual relationships between those three terms. On a closer look, it seems as if there were no more room for propositions in Reinach’s philosophy. One might even be tempted to see Reinach’s concept of proposition as an unnecessary import from the Logical Investigations. The import seems unnecessary, mainly because Reinach, unlike Husserl, does not need to appeal to propositions to account for the fact that an incorrect judgment has sense. Again, this results from his broadly Meinongian starting point. While the possibility for a proposition to be false allows Husserl to reject non-obtaining states of affairs, the acceptance of non-obtaining states of affairs allows Reinach to do without propositions in themselves. The question, then, is which parsimony is to be preferred. Reinach’s depreciation of the propositional or semantic dimension creates a number of difficulties which are characteristic of his theory of states of affairs. On the one hand, the question arises whether it is possible in all cases to replace propositions with states of affairs. Thus, Reinach sometimes tends to ascribe to states of affairs certain functions which seem distinctive of propositions. Despite his – Husserlian – definition of proposition as the meaning of a statement, Reinach sometimes suggests that the meaning of the statement is provided by the corresponding state of affairs. Hence one can ask whether he did not make the same mistake he criticized in Meinong and “the Austrians”, namely the mistake of confusing propositions and states of affairs. For example, in the essay on negative judgment (Reinach 1911b: 112-113/1982: 336), Reinach raises the question of whether states of affairs are necessarily relational. Consider the rose’s being red, which can be expressed by a statement such as “the rose is red”. This state of affairs, he notes, can be equated with obviously relational states of affairs such as /the rose forms the substrate of the red/ or /the red inheres in the rose/. But does this entail that the state of affairs /the rose is red/ is relational? Surely not, Reinach suggests. Although equivalent, that is, built up upon one and the same thing “the red rose”, the three states of affairs are different. Now, they are different, Reinach tells us, in that the three statements are “different in sense” (bedeutungsverschieden) (Reinach 1913: 112, footnote). The statements “the rose is red”, “the rose forms the substrate of the red”, and “the red inheres in the rose” are not synonymous, and the difference in meaning indicates that they refer to different states of affairs. However, this approach presents difficulties in the case of incorrect affirmative judgments. In Reinach’s view, two incorrect affirmative judgments, if they are different in sense, must refer to two different states of affairs. But if so, then both of these must be non-obtaining states of affairs. Why, then,

174 � Denis Seron talk about states of affairs rather than propositions? It is now the notion of state of affairs that seems unnecessary. More on this below.

3 Logic as a theory of states of affairs What is “logic”? And what do we take “logical” to mean when we talk about “logical principles”? Husserl and Reinach sharply differ in how they answer this question, and it is in that respect that Reinach’s philosophy of logic seems the most novel and forceful if compared with the Logical Investigations.¹² The answer to this question is to be found in a footnote at the end of the essay on negative judgment, which I shall take as a starting point. The problem at stake is about the principle of non-contradiction, which Reinach construes in terms of states of affairs. Basically, he claims, the principle of non-contradiction means that two states of affairs /S is P/ and /S is not P/ cannot both obtain at the same time. The footnote generalizes this result to all logical laws, or at least to a large part of them: It will be seen that these principles relate to states and their obtaining (Bestand); the same holds for the other fundamental principles of traditional logic. These have normally been related to judgments, e.g.: two contradictory judgments cannot both be correct. This principle is certainly incontestable, but it is a derived and not a primitive principle. A judgment is correct if the state of affairs corresponding to it obtains; and two contradictory judgments cannot both be correct because two contradictory states of affairs cannot both obtain. The law pertaining to judgments thus obtains its foundation from the corresponding law which relates to states of affairs. Attempts have been made from other quarters to relate this law not to judgments but to propositions. Two contradictory propositions, it is now said, cannot both be true. We acknowledge freely the difference between judgment and “proposition in itself”; but just as the proposition must be separated from the judgment, so also must it be separated from the state of affairs. A proposition is true when the state of affairs which is correlated with it obtains. And two contradictory propositions cannot both be true because two contradictory states of affairs cannot both subsist. Thus here too the propositional law is reducible to a law which relates to states of affairs. At the same time this provides an example which may indicate the sense of our claim above, that the major part of traditional logic will prove to have its foundations in a general theory of states of affairs (daß große

12 The contemporary philosopher Reinach most resembles in this respect may be R. Chisholm, who defined propositions as states of affairs “such that the laws of propositional logic may be interpreted as being applicable to them” (Chisholm 1970: 19). See also (Chisholm 1971: 180): “I also believe that the theorems of logic are most plausibly interpreted as pertaining to what I have called states of affairs.”

Adolf Reinach’s Philosophy of Logic

� 175

Teile der traditionellen Logik sich ihrem Fundamente nach als allgemeine Sachverhaltslehre herausstellen werden). (Reinach 1911b: 138/1982: 376 [transl. by Smith, slightly modified])

To some extent, these views still converge with the Logical Investigations. According to the first and fourth Investigations (Husserl 1900-01: I, §29; and IV, §14), certain logical laws can be formulated in an equivalent way as ontological laws. The logical principle of non-contradiction may thus be translated into an formalontological law stating that two contradictory states of affairs cannot both obtain. The difference between both authors is about what this equivalence relation really means and implies. In fact, Reinach did much more than call attention on this equivalence as Husserl had done before him (Seron 2011). He claimed that the laws of proposition “are reducible to [zurückführen auf ] laws which relate to states of affairs”, and that “the major part of traditional logic will therefore prove to have its foundations in a general theory of states of affairs.”¹³ One should be careful not to overestimate the extent of this claim. Reinach’s ambition was certainly not to boil down all logical into ontological distinctions between states of affairs. On the contrary, it is consistant with the view that logical categories are, to some extent, autonomous vis-a-vis ontological categories. Reinach’s argument is that one can make positive judgments about negative states of affairs (I assert that the flower is not red) and negative judgments about positive states of affairs (I deny that the flower is red). This shows, he claims, that the logical difference between positive and negative judgments cannot be merely reduced to the ontological difference between positive and negative states of affairs. In this sense, logical classifications lie in part beyond the sphere of states of affairs: “A logic which systematically carried through the distinction between judgment and judged state of affairs could scarcely decline to classify judgments according to the characteristics of their correlated states of affairs.” (Reinach 1911b: 122/1982: 351) Hence we may suppose that actually Reinach’s logic is not confined to states of affairs and applies to propositions and judgments as well (DuBois 1995: 117-118; see Smith 1987b: 192). This approach is clearly illustrated in an obituary of William James published in 1910 (Reinach 1910a: 50). In this text, Reinach starts by rejecting the pragmatist view of truth in the name of correspondentism. Then, in (more unexpected) terms which recall neo-Kantianism, he attributes to James the merit of having re-

13 On this point my interpretation differs from Salice’s (2009: 266f.). Salice is certainly right that Reinach maintains the distinction between propositions and states of affairs, but in my view, this does not imply that logic, for Reinach, must be a theory of propositions as opposed to an ontology of states of affairs.

176 � Denis Seron futed the copy theory of truth. Now, his line of argument against the Abbildtheorie fits well with the idea that logical structures do not merely reflect ontological structures. Most interestingly, the conception of truth Reinach adopts is certainly correspondentist, but not explicitly maximalist. “A judgment is true, he argues, if the state of affairs it refers to obtains.” (Reinach 1910a: 49)¹⁴ The relation between truth and the state of affairs’ obtaining is conceived here as a relation of mere implication, not of equivalence. Herein lies an important difference between Reinach and Wittgenstein’s Tractatus, which defends maximalism about elementary propositions (Wittgenstein 1963 [1922]: 4.25).¹⁵ The former does not rule out the possibility that a true proposition refers to a non-obtaining state of affairs.¹⁶ It thus seems that Reinach’s aim is not merely to reduce propositional logic to an ontology of states of affairs. Nevertheless, it remains true that, in another sense, Reinach’s logic is no more than a theory of states of affairs. But other explanations are required in order to understand this idea. The Introduction to Philosophy provides a precise definition of what logic is: Logic is (1) a general theory of science, as far as it provides the laws of foundation [Begründungsgesetze] in general and thus gives the form for the relations of foundation that apply to all science; (2) a special theory of science, as far as it refers to the laws of foundation relevant for special sciences, and to types of special science. (Reinach 1913: 453)

In short: logic deals with relations of inference. It is a theory which supplies rules for grounding a judgment in another judgment, hence a theory whose objects are relations of antecedent and consequent. Now, as we have seen, Reinach takes such relations to hold between states of affairs rather than between propositions or judgments. The powerful originality of his philosophy of logic lies in the fact that he undertook to ontologize the relations of implication and inference.¹⁷ Since these originally lie in the states of affairs, it is only in a derivative sense that implication is a relation between propo-

14 “Ein Urteil ist wahr, wenn der Sachverhalt, auf den es sich bezieht, besteht.” Compare (Reinach 1911b: 138/1982: 376): “Ein Urteil ist richtig, wenn der zugehörige Sachverhalt besteht.” 15 “Ist der Elementarsatz wahr, so besteht der Sachverhalt; ist der Elementarsatz falsch, so besteht der Sachverhalt nicht.” 16 The question of whether all true proposition corresponds to an (obtaining) state of affairs should not be confused with the question of whether all (true or false) proposition corresponds to an (obtaining or non-obtaining) state of affairs. While Reinach’s answer to the first question is clearly yes, his answer to the second one is far from clear. 17 See (Reinach 1913: 453): “It is not contingent that certain types of foundation relations are connected with certain objective assumptions. The sphere of logic is obviously larger than the sphere of the foundation relations.”

Adolf Reinach’s Philosophy of Logic

� 177

sitions or judgments. As a consequence, it is also in a derivative sense that logical principles, as rules of inference, apply to propositions or judgments. The principle of non-contradiction originally means that two contradictory states of affairs cannot both obtain; the principle of the excluded middle means that “there is no third between the obtaining and the non-obtaining of a state of affairs” (Reinach 1913: 477); etc. Reinach clearly equates implication with inference, even going so far as to talk of “axiomatic” (axiomatische) and “theorematic” (gefolgerte) states of affairs (Reinach 1913: 404). But such an approach is fraught with difficulties. It may seem not too difficult to locate implication relations in the “things themselves”, at least in the sense that true conditional propositions must correspond to obtaining states of affairs. But the same hardly applies to inference. For inferences, unlike conditionals, are not true or false, but valid or not valid.¹⁸

4 Reinach’s realism and its difficulties Of course, the fact that Reinach replaces talk of propositions by talk of states of affairs is not a detail. It is at least plausible to believe that the two approaches will have different ontological implications. Consider, for example, an incorrect judgment that S is P. The judgment is incorrect because the corresponding state of affairs, S’s being P, does not obtain. From Husserl’s point of view, the judgment has sense, that is, has a false proposition as its content, even though there is no corresponding state of affairs. In a sense, this makes it possible to achieve a notable ontological economy. For propositions, in the Logical Investigations, are defined to be “species of judgment”, that is, intentional contents which the psychologist considers as “intentional matters” and the logician as meanings “in specie”.¹⁹ Certainly, the logician, for Husserl, considers propositions independently of their realization in mental acts. But this is just one face of the coin. The other is that the psychologist or phenomenologist reveals to the logician that her propositions really are abstract parts of mental acts. The “phenomenological foundation of logic” precisely means that the logician achieves her “ideations” on the basis of empirical materials which are, so to speak, provided by the psychologist. The general thought seems to be that, ab-

18 See (Gochet & Gribomont 1990: 35-37). 19 The interpretation here outlined – which is much indebted to (Kusch 1995) – is developed in more detail in (Seron 2012b) and (Seron forthcoming).

178 � Denis Seron solutely speaking, propositions ontologically commit us to the existence of no objects except the mental acts that instantiate them.²⁰ Hence, as Barry Smith has rightly suggested, Husserl’s semantic objectivism in the Investigations is better seen as an Aristotelism of intentional contents than as a Fregean-style Platonism of meanings (Smith 1989b; Seron 2012; and Seron forthcoming). How are incorrect positive judgments to be described within the frame of Reinach’s logic of states of affairs? First, Reinach assumes that non-obtaining states of affairs cannot be “known”. Yet it remains that “we speak of states of affairs like the being golden of mountains or the being round of squares” (Reinach 1911b: 117/1982: 340-341). In other words, non-obtaining states of affairs can be “meant” (gemeint). States of affairs – as opposed to objects, which are presented (vorgestellt) – are “meant” or “in view” (abzielen auf ). What best characterizes the “meaning” of a state of affairs is that it always has a linguistic dimension (Reinach 1911b: 102/1982: 323) and is independent of any intuitive givenness. Intuition – that is, “knowledge” or “evidence” – is not necessary for the state of affairs to be meant or thought. Intuition is to be seen as an “accompanying” (begleitend) content, which is not immanent in the state of affairs or its meaning (Reinach 1911b: 106-108/1982: 328-330). Briefly: there still are states of affairs to be thought even where there are no states of affairs to be known, as it may be the case with incorrect positive judgment. These views have difficulties. Suppose that an incorrect positive judgment “means” a non-obtaining state of affairs. This may suggest that non-obtaining states of affairs have some kind of objectivity. But obvious difficulties arise at this point: What would this objectivity consist in, if it must be distinct from the state of affair’s obtaining? Where are non-obtaining states of affairs to be found, if they cannot reside in the mental act?²¹

20 See (Husserl 1900-01: V, 322/2001: vol. 2, 79): “As with all ideal unities, there are real possibilities, and perhaps actualities, which correspond to meanings: to meanings in specie correspond acts of meaning, the former being nothing but ideally apprehended act characters of the latter (jene sind nichts anderes als die ideal gefaßten Aktcharaktere dieser).” (Husserl 1913: 343): “[...] als ideal gefaßte Momente aus diesen.” 21 Benoist (2005: 27) rightly argues that these difficulties are caused by a confusion of intentionality with presentation – a confusion which Reinach precisely avoids by his notion of “meaning”. Unlike presented objects, states of affairs do not need to exist or be present. See (Reinach 1913: 419): “Ein gemeinter Gegenstand braucht nicht zu existieren”; and (Reinach 1908: 339): “Das Denken ist ein Abzielen auf Nichtgegenwärtiges, das Vorstellen ist das Vorsichhaben eines Gegenwärtigen.” However, this does not release us from the need to identify the ontological status of states of affairs and to explicate why they cannot be replaced with propositions.

Adolf Reinach’s Philosophy of Logic

� 179

It may be noted in passing that it is not always clear that non-obtaining states of affairs, for Reinach, must be “meant” rather than “known”. In the essay on negative judgment, Reinach thus considers the case of a negative conviction in a positive state of affairs (Reinach 1911b: 123-124/1982: 352-353). For example, I refuse to believe that this flower is yellow. The conviction, Reinach observes, rests on some positive evidence: I resist the belief that the flower is yellow because I see that it is red. But on the other hand, he also claims that the state of affairs /the flower is yellow/ provides some special form of evidence or knowledge, which he calls “negative evidence”. This seems to imply that non-obtaining states of affairs, too, are subject to evidence. Barry Smith has clearly pointed out the problem in the introduction to his translation of the essay on negative judgment (Smith 1982: 294-295). He first attributed to Bolzano, Frege, Meinong, and Chisholm a radical view according to which every correct or incorrect judgment requires a “statal entity”. Next, he argued that “a view of this kind is defensible only where it relates to entities belonging to the sphere of meaning (to Frege’s “realm of sense”) or, as in Meinong’s case, to some hybrid sphere of quasi-meanings.” To put it another way: for such a view to be tenable, the statal entities must be logical objects, namely Fregeanstyle propositions or (to some extent) Meinongian “objectives”. The opposite view, Smith continues, is to conceive the states of affairs incorrect positive judgments refer to merely as “object-entities”, hence as truth-makers rather than as truth-bearers. But this view involves a very counterintuitive version of Platonism: “What mind-independent external referent, what constituent part or contour of the world, could correspond, for example, to a false sentence, to a counterfactual conditional, or to a judgment concerning the indefinite future?” Now, there are two options to escape the difficulty. The first one – which is the option chosen by Russell, for example – is to deny that incorrect positive judgments have statal correlates. The second option is the one favored by Reinach. It consists in “distinguishing amongst the totality of autonomously existing states of affairs, subsistent states of affairs corresponding to true judgments, and nonsubsistent states of affairs corresponding to those that are false.” Smith attributes this position to Reinach, Meinong, and the Tractarian Wittgenstein. From this perspective, Husserl’s approach may seem more economical. Barry Smith thus defended the view that Husserl’s theory of states of affairs is naturalist rather than Platonist (Smith 1989a: 63). Husserl’s version of Platonism, if there is any, is about propositions, not states of affairs – which he considers to be mere truth-making components of the real world. Reinach, by contrast, is a Platonist about states of affairs, in that for him the timelessness of truth must be accounted

180 � Denis Seron for by appealing to a timeless sphere of (obtaining or non-obtaining) states of affairs.²² I have suggested that the key claim of Husserl’s attempt at a “phenomenological foundation of logic” is that propositions are “species of judgment”, and hence abstract parts of mental acts. This means that, although they can be studied “in themselves”, propositions, in last resort, exist only “in” the mental acts that instantiate them. As Husserl argues in the first Investigation and in his review of Palágyi, the case of propositions is analogous to that of colors (Husserl 1979: 157; and 1900-01: I, 100-101/2001: vol. 1, 230). It is no doubt possible to study colors in specie, with no concern whatever for colored things, so as to state a priori laws such as “the additive synthesis of green and red is yellow”. Yet, it remains that colors, absolutely speaking, exist only in spatially extended things. Likewise, Husserl considers the existence of “ideal” logical laws consistent with the fact that propositions are abstracta which exist only in mental particulars. Grounding logic, in his view, means revealing or “clarifying” the intuitive basis of logic, that is, descriptive materials for the logician’s ideations. But since logical objects are abstracta which exist only in mental acts, this intuitive basis must be psychological or phenomenological. To put it otherwise: the logician’s ideation is, as Husserl says in the sixth Logical Investigation, a “founded act”, namely an act which is based upon internal perception – this being, in the first edition of the Investigations just as for Brentano, the very experience that constitutes psychology as an empirical science. Accordingly, the foundation of logic requires one to assume no more than the existence of mental acts. This foundation is “metaphysically neutral” as to the objective world, for the simple reason that, in Brentano’s view, mental phenomena are the only things the descriptive psychologist needs to posit as really existent. Now, what happens if we accept Reinach’s claim that logical laws are not about propositions, but about states of affairs? The divergence lies in the fact that for Reinach, as he repeatedly insists, the state of affairs’s obtaining or nonobtaining is independent of the subject. A proposition is true in itself, independently of the subject, because the corresponding state of affairs obtains in itself, independently of the subject (Reinach 1911b: 137/1982: 369-370; and 1914: 544-545). In other words: the foundation of logic must be realist rather than psychological or phenomenological (in the Husserlian sense). Certainly, the task still consists, as in the Investigations, in clarifying the intuitive sources of logical laws. But the intuition concerned cannot be the psychologist’s experience. Rather, this intuition

22 See (Salice 2009: 261).

Adolf Reinach’s Philosophy of Logic

� 181

must be the “knowledge” of states of affairs which obtain mind-independently in the objective world. Obviously, such a foundation, if it is to be called “phenomenological”, can be called so only in a very different sense than in the Logical Investigations. Reinach’s phenomenology is not identical with descriptive psychology as in the first edition of the Investigations; nor has it anything to do with the later transcendental phenomenology. Nevertheless, the comparison with the Investigations helps to clarify exactly what a “realist phenomenology” is supposed to be. One of the key ideas in On Phenomenology is that phenomenology is not distinguished by the type of objects it deals with. Phenomenology is not, say, that science which concerns itself with purely immanent objects – with objects which should be marked off from the transcendent objects through phenomenological reduction. As James DuBois (1995: 148) has pointed out, the word “epoché” does not occur even once in Reinach’s published writings. Rather, Reinach’s phenomenology is a matter of “attitude”. It involves some new perspective on the same real world we deal with in our everyday life, namely a perspective grounded in the “intuition of essences”.²³ The theory of states of affairs has to play a central role in this respect. For intuiting essences, for Reinach, always means intuiting a priori states of affairs, namely essential connections in things. As Reinach puts in the Introduction to Philosophy of 1913, phenomenological analysis is intended to give us access to the “things themselves” by “disclosing objectities [...] in such a way that one can intuit [erschauen] their essences.”²⁴ In Reinach’s terminology, erschauen means intuiting a state of affairs, as opposed to schauen which means intuiting an object (see above). This approach is original when compared to the views of other realist phenomenologists, like Johannes Daubert and Jean Héring. These two philosophers, too, consider essence laws to be about foundation relations among essences.²⁵ But Daubert does not mention states of affairs in this connection, and Héring explicitly rejects the view that these relations could be states of affairs (Héring 1921: 524; see 496-497). These last remarks bring us into the heart of Reinach’s realism. We might venture to say that the difference between his philosophy of logic and that of

23 In this sense, the eidetic dimension aside, Reinach’s conception of phenomenology displays striking similarities with Drummond’s and Sokolowski’s externalist readings of Husserl. 24 “Much work and effort is needed for one to know the essential connections and to get to the things themselves [an die Sachen zu gelangen]. [...] It has been said that phenomenology starts with the ‘intuition of essences’ [Wesenserschauung]. This sounds mystical, but is clear. Analysis supersedes simple vision [tritt nämlich an die Stelle des blossen Dahinsehens] so as to disclose objectities [...] in such a way that one can intuit [erschauen] their essences.” (Reinach 1913: 448) 25 On Daubert, see (Fréchette 2001: 312ff.)

182 � Denis Seron Husserl in the Prolegomena somehow coincides with the difference between the logical objectivity of truth and the metaphysical objectivity of states of affairs. This brings us back to Barry Smith’s discussion of Reinach’s Platonism. In the Investigations, Husserl certainly considers a proposition’s being true as trivially synonymous with its being true in itself. But on the other hand, he also conceives of logic as a reflective science,²⁶ in the sense that its “foundation” necessitates psychological or phenomenological description of the corresponding mental acts.²⁷ In Reinach’s view, by contrast, the foundation of logic requires us to turn to the extra-mental world. Logical properties, relations, truths in general primarily lie in the states of affairs, which exist in themselves, independently of any mental acts. For this reason, the central idea of Reinach’s realism is not merely that the obtaining of states of affairs is logically prior to the truth of propositions and the correctness of judgments. Reinach also needs the further premise that there are mind-independent states of affairs – a premise which Reinach (controversially) views as alien to Husserl’s transcendental phenomenology.²⁸ The results of our discussions can be recapitulated as follows: First, Reinach’s aim is to replace Husserl’s idea of grounding propositional logic on psychological description by the idea of grounding a statal logic on the “knowledge” of a priori states of affairs. So, such a foundation can be called “phenomenological” only in quite a different sense. It is no longer the descriptive psychologist’s internal experience that defines the phenomenological attitude, but the intuition of states of affairs of a certain type. Now, Reinach’s characterization of states of affairs as essentially mind-independent involves a realist line of thought which surely represents a move away from the metaphysical neutrality of phenomenology in the Logical Investigations. This clearly shows the strategic role of the theory of states of affairs in Reinach’s philosophical project. For it is precisely the theory of states of affairs that makes phenomenology compatible with realism. If the intuition of a priori states of affairs is to define phenomenology, then the essential mind-independence of states of affairs entails that phenomenology must be intrinsically realist.

26 See (Husserl 1900-01: I, §34). 27 Matters are somewhat more complicated, however, since Husserl also holds in the Investigations that a proposition’s being true in itself implies a state of affairs’ obtaining in itself as its “necessary correlate” (Husserl 1900-01: Prolegomena, §62). But what is important here is that the objectivity of states of affairs is the business of the logician (or the “formal-ontologist”), not of the phenomenologist. 28 See (Reinach 1913: 484): “The essential [wesensgeseztliche] possibility of independent existence is here admitted, in fact assumed.” See also (Reinach 1989: 737) about Husserl: “streiche ich das Bewußtsein, so streiche ich die Welt.”

Claire Ortiz Hill

Husserl’s Way Out of Frege’s Jungle

1 Introduction¹ Influential twentieth-century philosophers and mathematicians turned philosophers leapt upon new theories of representation, meaning, judgment, arithmetic and sets elaborated during the late 19th century to remodel, if not all together eliminate, many traditional ideas about what Bertrand Russell once colorfully called the ultimate furniture of the universe. One of the principal strategies adopted by Russell, Rudolf Carnap, Willard Van Orman Quine and like-minded philosophers was to use logic to throw out the old furniture and replace it with new furniture of their liking. The full implications of those new theories about the ultimate structure of reality have yet to be drawn because, for many reasons, many avenues of research have yet to be pursued. Here, I wish to add new dimensions to standard discussions by looking at the ontological implications for analytic philosophy of Edmund Husserl’s theory that numbers and sets function in an entirely different way in the sphere of propositions and states of affairs than in arithmetic and in set theory (Husserl 2008 [1906/07]: §18c). To begin with we need to visit Husserl’s world of the purely logical.

2 Husserl’s World of the Purely Logical Exploration of what Husserl once called the strange world of the purely logical brought him to detect a natural order in formal logic and to broaden its domain to include two levels above the traditional Aristotelian logic of subject and predicates and states of affairs. He considered this new understanding of the structure of the world of pure logic to be of prime importance for the understanding of logic and philosophy. These levels are explored in Introduction to Logic and Theory of

1 Dedicated to the memory of Professor Paul Gochet of Liège and of Ruth Barcan Marcus who forsook the desert landscapes of extensional logic and crossed the line into the jungle of Aristotelian essentialism

184 � Claire Ortiz Hill Knowledge (Husserl 2008 [1906/07]: §§18-19). Part I of Formal and Transcendental Logic is devoted to describing them. In the introduction, Husserl says that he considered this stratification, which had not yet been fully detected in the Logical Investigations, to be of the greatest significance, not only for a real understanding of the genuine sense of logic, but also for all of philosophy. He believed it to be a matter of a radical clarification of the relationship between formal logic and formal mathematics and that with it emerged a definitive clarification of the sense of pure formal mathematics as a pure analytics of non-contradiction (Husserl 1974/1969: 11). On the first level of this edifice Husserl placed the traditional Aristotelian apophantic logic of subject and predicate propositions and states of affairs that deals with what is stated about objects in general from a possible perspective. He stressed that, although the concept of predicative judgment stood at the center of formal logic as it developed historically, it was but a small area of pure logic, the extended sphere of pure logic that includes the mathematical disciplines being immense in range and wealth of content in comparison (Husserl 1973 [1939]: §1; and 2008 [1906/07]: §18c). In Husserl’s theory of the forms of subject-predicate propositions of this first level, number only occurs as form, but not as an object about which something is predicated. He emphasized that only the forms of the plural numerical predication about objects as such belong in a simple theory of objects in general and the forms of their states of affairs and that in statements, propositions or states of affairs, forms are dependent. We can make such forms independent, but then new higher order objects, hypostasizations of forms emerge that are not objects in their own right. For Husserl, this means that numbers function entirely differently on this first level than they do in the arithmetic of the second level (Husserl 2008 [1906/07]: §18c). I shall illustrate Husserl’s point about the ontological status of numbers in subject-predicate propositions by borrowing from Frege, for the sake of unity of argumentation, the proposition, “Jupiter has four moons”, that the latter used to illustrate his own theory about numbers in The Foundations of Arithmetic (Frege 1961 [1884]: §57). According to Husserl’s analysis in Introduction to Logic and Theory of Knowledge, this would be a statement about Jupiter’s moons in which the number characteristic four occurs as form and is thereby dependent. If one says w and x and y and z are φ, Husserl reasoned, then one has combined the objects w... z by “and”. Here the “and” is form and grounds the unitary form of the plural predication. Corresponding to this is a cardinal number. However, Husserl stresses, this is a new thought configuration, for it is one thing to make statements about objects in which number properties occur as form and are thereby dependent and another thing to make statements about numbers as such in such a way

Husserl’s Way Out of Frege’s Jungle �

185

that the numbers are the objects. Statements about numbers in which numbers are the objects have their place on the second level of Husserl’s hierarchy (Husserl 2008 [1906/07]: §18c). Sets as objects do not occur on Husserl’s first level of the logic of subjects and predicates any more than numbers do. He taught that in set theory, we make judgments universally about sets that in a certain way are higher order objects. We do not make judgments directly about elements, but about whole totalities of elements and arbitrary elements, and the whole totalities, the sets to be precise, are the objects-about-which. Corresponding to every plural is a set, but in the theory of proposition forms, or forms of states of affairs, the set does not occur as object. In it, the objectsabout-which are thoroughly indeterminate A B.... Rather, only the plural occurs in it, which constitutes a form of predication about arbitrary objects. (Husserl 2008 [1906/07]: §18c)

On the two levels rising above the level of subject-predicate propositions, it is no longer a question of objects as such about which one might predicate something, but of investigating what is valid for higher order object formations like cardinal number and set, which are determined in purely formal terms grounded in the essence of logical forms and deal with objects in indeterminate, general ways (Husserl 2008 [1906/07]: §18d). Husserl conceived of the second level as an expanded, completely developed analytics in which one proceeds in a purely formal manner since every single concept used is analytic. One calculates, reasons deductively, with concepts and propositions. Signs and rules of calculation suffice because each procedure is purely logical. One manipulates signs, which acquire their meaning in the game through the rules of the game. One may proceed mechanically in this way and the result will prove accurate and justified (Husserl 1996: §58). Husserl gave these examples of numbers occurring as objects in arithmetical propositions of the second level: 1. 2. 3. 4. 5.

“Any number can be added to any number”. “If a is a number and b a number, then a + b is as well”. “Any number can be decreased or increased by one”. “The numbers form a series continuing from 0 in infinitum”. The different laws of addition, subtraction, multiplication, etc. (Husserl 2008 [1906/07]: §18c)

Sets also function entirely differently in the set theory of the second level where statements are not made directly about elements, but about whole totalities of arbitrary elements. The set theory of the second level asks what is valid for the

186 � Claire Ortiz Hill higher-order objects called sets. Husserl gave the following examples of the truths about sets as objects-about-which that make up set theory: 1. 2.

3.

“2 sets can each be joined into a new set”. “2 sets a b are each related to one another in such a way that either a is part of b or b is part of a, or that they intersect (a set having a part in common), or that it turns out that they are identical, coincide”. “The set formed of the elements A B C is part of the set formed of the elements A B C D containing ‘more elements”’ (Husserl 2008 [1906/07]: §18c).

On the third level of Husserl’s hierarchy is the theory of manifolds. He described manifolds as pure forms of possible theories which, like molds, remain totally undetermined as to their content, but to which thought must necessarily conform in order to be thought and known in a theoretical manner. On this level, formal logic deals with whole systems of propositions making up possible deductive theories. It is a matter of theorizing about possible fields of knowledge conceived of in a general, undetermined way and purely and simply determined by the fact that the objects stand in certain relations that are themselves subject to certain fundamental laws of such and such determined form. Husserl saw the general theory of manifolds, or science of theory forms, as a field of free, creative investigation made possible once it was discovered that deductions, series of deductions, continue to be meaningful and to remain valid when one assigns another meaning to the symbols. No longer restricted to operating in terms of a particular field of knowledge, one is free to reason completely on the level of pure forms. Operating within this sphere of pure forms, one can vary the systems in different ways. Nothing more need be presupposed than the fact that the objects figuring in them are such that, for them, a certain connective supplies new objects and does so in such a way that the form determined is assuredly valid for them. One finds ways of constructing an infinite number of forms of possible disciplines (Husserl 1913: Prolegomena, §§69-70; see 2008 [1906/07]: §§18-19; 1977: §§71-72; 1996: Chap. 11; and 1974: §33). In the methodology of manifolds, one speaks of numbers, but one does not mean cardinal numbers, but anything for which formal axioms of the arithmetical prototype hold. If we drop the cardinal number meaning of the letters in the ordinary theory of cardinal numbers and substitute the thought of objects in general for which axioms of the arithmetical form a+b = b+a is to hold, we no longer have arithmetic, but a purely logical class prototype of theory forms to which, besides innumerably many possible domains, the domain of cardinal numbers is also subject. One may then speak of numbers in the formal sense, but they are not cardinal numbers. They are objects indeterminately, universally defined by

Husserl’s Way Out of Frege’s Jungle

� 187

axiom forms as they are especially actually found for cardinal numbers. For cardinal numbers, ab = ba holds. In constructing a manifold, though, one may just as well stipulate that ab  = ba, or, for example, ab = −ba. And, likewise for the other basic principles (Husserl 2008 [1906/07]: §19b and §19d). Husserl concluded that formal constraints banning meaningless expressions, meaningless imaginary concepts, reference to non-existent and impossible objects unduly restrict us in our theoretical, deductive work, but that resorting to the infinity of pure forms and transformations of forms frees us from such conditions and explains why having used imaginaries, what is meaningless, must lead, not to meaningless, but to true results (Husserl 1996: §57). In Logical Investigations, he expressed his conviction that his theory of complete manifolds was the key to only possible solution as to how in the realm of numbers impossible, nonexistent, meaningless concepts might be dealt with as real ones (Husserl 1913: Prolegomena, §70).

3 Entering Frege’s Jungle Now I want to look at Frege’s theory of numbers and sets. Frege considered the prime problem of arithmetic to be that of how one apprehends logical objects, in particular numbers, and asked what justifies one in recognizing numbers as objects (Frege 1903: 224). It is vital to realize in addition that the logical objects that Frege needed to apprehend were independent, self-subsistent logical objects. Operating only on the lowest level of Husserl’s hierarchy, that of the traditional logic of subjects and predicates, Part IV (Frege 1961 [1884]: §§55-86) of The Foundations of Arithmetic is devoted to analyzing the concepts of arithmetic. Frege argued there that to obtain the concept of number, it was a matter of fixing the sense of an identity (Frege 1961 [1884]: §§62-70 and 106) and that only in the case of objects could there be any question of identity (Frege 1979: 182 and 120). The first portion (Frege 1961 [1884]: §§55-61) of Part IV is entitled “Every individual number is an independent object” and is devoted to affirming the independency of numbers. In contrast to Husserl, Frege maintained “that the number studied by arithmetic must be conceived not as a dependent attribute, but substantivally” (Frege 1961 [1884]: §106 and note). “Precisely because it forms only an element in what is asserted”, he reasoned, “the individual number shows itself for what it is, an independent object”. He considered that the presence of the definite article ‘the’ in expressions like ‘the number 1’ served “to class it as an object” and that in arithmetic “this independence comes out at every turn, as for example in the identity

188 � Claire Ortiz Hill � + � = �” (Frege 1961 [1884]: §§57 and 106). He argued that since it was a matter of arriving at a concept of number usable for scientific purposes, “we should not, therefore, be deterred by the fact that in the language of everyday life number appears also in attributive constructions” for that “can always be got around” and proposed that a subject-predicate statement proposition like “Jupiter has four moons” can be converted into “the number of Jupiter’s moons is four”. [...] we can say: “the number of Jupiter’s moons is the number four, or 4”. Here “is” has the sense of “is identical with” or “is the same as”. So that what we have is an identity, stating that the expression “the number of Jupiter’s moons” signifies the same object as the word “four”. (Frege 1961 [1884]: §57)

Frege concludes his argument with the important proviso that the independence that he is “claiming for number is not to be taken to mean that a number word signifies something when removed from the context of a proposition, but only to preclude the use of such words as predicates or attributes, which appreciably alters their meaning” (Frege 1961 [1884]: §60). After devoting several pages to ferreting out some very basic problems that he saw sticking to his theory, and unable to silence questions, doubts and suspicions about the undesirable consequences that he saw proceeding from it, he acknowledged that he could not by those methods obtain any satisfactory concept of number. He said that left unmodified his technique was liable to lead to false or nonsensical conclusions or be sterile and unproductive. So to forestall the problems that he foresaw would vitiate his theories, against his better judgment, he introduced extensions. He settled for the definition: “The Number which belongs to the concept F is the extension of the concept ‘concept equal to the concept F’ ” (Frege 1961 [1884]:, §§66-69, 107). He took logical law into his own hands and devised Basic Law V to allow logicians to pass from a concept to its extension. Frege knew well that what he wished to sanction through his law was “forbidden by the basic difference between first and second level relations”, but he temporarily convinced himself that, though a proof could “scarcely be furnished” and “an unprovable law” would have to be assumed, a transformation might “take place, in which concepts correspond to extensions of concepts” (Frege 1979: 182). Upon learning of Russell’s paradox, Frege tested the validity of the chain of inferences leading up to the contradiction and concluded that his law about extensions was false and his explanations did not suffice to secure a reference for his combinations of signs in all cases. He confessed that he had been reluctant to use classes, but that that was the only answer that he had found to the question as to how to apprehend logical objects (Frege 1980 [1952]: 132, 140-41). “I do not see how arithmetic can be scientifically established; how numbers can be appre-

Husserl’s Way Out of Frege’s Jungle � 189

hended as logical objects, and brought under review; unless we are permitted – at least conditionally – to pass from a concept to its extension”, Frege agonized in his study of the contradiction in his 1903 appendix to Basic Laws II (Frege 1903: 214). Frege was the first to admit that his logic led into a jungle. When asked late in life about the causes of the paradoxes of set theory, Frege replied that it was language’s propensity to undermine the reliability of thinking by forming apparent proper names to which no objects correspond that had allowed concept-words to be transformed into proper names and so come to be in places unsuited to them that had “dealt the death blow” to his set theory (Frege 1976/1980: 55). Frege warned that the difficulties that this idiosyncrasy of language entangles us in are incalculable and threaten to undermine the reliability of thinking (Frege 1979: 269-270; and 1976/1980: 54-55). He described what he called that the “essence of the procedure which leads us into a thicket of contradictions” as consisting in regarding the objects falling under F as a whole, as an object designated by the name “set of Fs”, “extension of F”, or “class of Fs” etc. This is inadmissible, he explained, because of the essential difference between concept and object, which is covered up in our word languages. Such a transformation of a concept into an object is inadmissible, because the set formed only seems to be an object, while in truth there is no such object at all. Experience, he said, had shown him “how easily this can get one into a morass”. Confusion, is bound to arise if, Frege warned at the end of his life, as a result of its transformation into a proper name, a concept word comes to be in a place for which it is unsuited (Frege 1976/1980: 54-55). Blurring distinctions between dependent and independent meanings by allowing a concept word to be transformed into a proper name and to come to figure on the wrong tier in the hierarchy of meaning broke the logical structure that Frege had professed to be so intent upon preserving and opens the way to confusion. Once logical structure is broken and meaning categories are violated, contradictions, paradoxes, antinomies, fallacies, nonsense, confusion, absurdity, pseudoobjects result. Frege chose a jungle, but he could have had a sterile, unproductive desert. Russell reported Henri Poincaré had rejoiced over the contradictions announcing triumphantly, “mathematical logic is no longer sterile, it begets antinomies” (Russell 1959: 59; and 1903: xii).

190 � Claire Ortiz Hill

4 Russell Strives to Prune Frege’s Jungle Intellectual sorrow descended upon Bertrand Russell in full measure when he came upon the contradiction about classes that are not members of themselves that put an end to an “intellectual honeymoon” unlike any he had ever enjoyed before or would ever again enjoy (Russell 1959: 56). Early in his search for a solution to the problem, Russell believed that “the key to the whole mystery” would be found in the distinguishing of logical types (Russell 1903: §104). When two words have two different types of meanings, Russell once warned, the relations of those words to what they stand for are also of different types and the failure to realize this is “a very potent source of error and confusion in philosophy” (Russell 1956: 133). So he established a hierarchy of classes according to which the first type of classes would be composed of classes made up entirely of particulars, the second type composed of classes whose members are classes of the first type, the third type composed of classes whose members are classes of the second type, and so on. The types obtained would be mutually exclusive, making the notion of a class being a member of itself meaningless (Russell 1973: 201). Russell believed that the theory of types he developed led to the “avoidance” of contradictions and to the detection of the fallacy that produced them (Russell & Whitehead 1927: 1). And he believed that no solution to the contradictions was technically possible without it. However, he saw that it was not “the key to the whole mystery”.² Deeper problems caused them to break out again and “further subtleties” were needed to solve them. Russell had originally believed that: When we say that a number of objects all have a certain property, we naturally suppose that the property is a definite object, which can be considered apart from any or all of the objects, which have, or may be supposed to have, the property in question. We also naturally suppose that the objects which have the property form a class, and that the class is in some sense a new single entity, distinct, in general, from each member of the class. (Russell 1973: 163-64)

However, the contradiction about the classes that are not members of themselves showed him that classes could not be independent entities. He said that it had taught him that if a word or a phrase that is devoid of meaning when separated from its context is wrongly assumed to have an independent meaning, false ab-

2 See (Russell 1919: 135) and (Russell 1956: 333).

Husserl’s Way Out of Frege’s Jungle

� 191

stractions, pseudo-objects, and paradoxes and contradictions are apt to result. He came to believe that if one assumes that the class is an entity, one cannot escape the contradiction about the class of classes that are not members of themselves. As he explained, “if you think for a moment that classes are things in the same sense in which things are things, you will then have to say that the class consisting of all the things in the world is itself a thing in the world, and that therefore this class is a member of itself” (Russell 1956: 260-265; and 1973: 163-165 and 171). The idea that classes were not entities shed some light on the ontological nature of classes by saying what they were not, but Russell had to do more than that. He had to find a way of making them disappear from the reasoning in which they were present without really completely letting go of them (Russell 1919: 184), because he believed that “without a single object to represent an extension Mathematics crumbles” (Russell 1903: §489). While struggling to get to the bottom of the problem of fake objects, Russell found parallels existing between the problems that arise when classes are treated as objects and problems that come up when descriptions are treated as names. These analogies plus the success he had with his 1905 theory of definite descriptions gave him an idea as to how classes might be analyzed away much as descriptions had been, and so gave him a concrete idea as to how he might sweep his problems away (Russell 1959: 49). Statements containing descriptions had proven amenable to further analysis. So once Russell was satisfied that classes and descriptions both fell into the same logical category of non-entities represented by incomplete symbols, he decided to extend his ideas about analyzing away descriptions to include class symbols. He reasoned that since: we cannot accept “class” as a primitive idea. We must seek a definition on the same lines as the definition of descriptions, i.e. a definition which will assign a meaning to propositions in whose verbal or symbolic expression words or symbols apparently representing classes occur, but which will assign a meaning that altogether eliminates all mention of classes from a right analysis of such propositions. We shall then be able to say that the symbols for classes are mere conveniences, not representing objects called “classes,” and that classes are in fact, like descriptions, logical fictions, or (as we say) “incomplete symbols”. (Russell 1919: 181-82)

The theory of definite descriptions was a way of making an object fit to go proxy for what was said about it (Russell & Whitehead 1927: 187). This means of drawing objects out of descriptions provided Russell with a practical model of how to make non-entities function as entities without incurring contradictory results. Russell finally felt obliged to introduce the axiom of reducibility to cleanse Principia Mathematica of unwanted entities. This specially designed axiom would

192 � Claire Ortiz Hill be “equivalent to the assumption that ‘any combination or disjunction of predicates is equivalent to a single predicate’ ” (Russell 1973: 250; and 1927: 58-59) and would provide a way of dealing with any function of a particular argument by means of some formally equivalent function of a particular type. It would thus yield most of the results that would otherwise require recourse to the problematical notions of all functions or all properties, and so legitimize a great mass of reasoning apparently dependent on such notions (Russell & Whitehead 1927: 56). He claimed that it embodied all that was really essential in his theory of classes (Russell 1919: 191; 1927: 58, and 166-167; and 1956: 82) and he leaned it at every crucial point in his definition of classes in Principia Mathematica (Russell 1973: 250; and 1927: 58-59 and 75-81). He considered that many of the proofs of Principia “become fallacious when the axiom of reducibility is not assumed, and in some cases new proofs can only be obtained with considerable labour” (Russell & Whitehead 1927: xliii). He called it “a dubious assumption” and a “defect” (Russell 1919: 192193). “This axiom,” he confessed, “has a purely pragmatic justification: it leads to the desired results, and to no others. But clearly it is not the sort of axiom with which we can rest content” (Russell & Whitehead 1927: xiv). Russell credited Occam’s razor with having given him “a more clean-shaven picture of reality”. It did not, he said, prove to him the non-reality of entities that it had showed him were unnecessary, but it abolished for him the arguments in favor of their reality. He said he did not think it possible to disprove the reality of integers or points or instants or the Gods of Olympus, but did not believe there was the faintest reason to think they existed (Russell 1959: 49). The man who once wrote that “Mathematics, rightly viewed, possesses not only truth, but supreme beauty – a beauty cold and austere [...] sublimely pure, and capable of a stern perfection” (Russell 1917: 63-64) concluded that, while the “aesthetic pleasure to be derived from an elegant piece of mathematical reasoning” remained, the “solution of the contradictions... seemed to be only possible by adopting theories which might be true but were not beautiful” and that the “splendid certainty” he had “always hoped to find in mathematics had become lost in a bewildering maze” (Russell 1959: 155-157). The logic that had appeared so convenient, simple and austerely beautiful had spawned error, contradiction, ugliness and messiness.

5 Frege’s and Russell’s Problems Live On Bewildering maze or not, analytic philosophers determined not to be driven out of the paradise Frege created for them went on to integrate the logic of Principia

Husserl’s Way Out of Frege’s Jungle �

193

Mathematica into mainstream philosophy. They called it “classical” logic and trying to solve the jungle of problems it produced became the stuff of logic and much of philosophy in our times (Hill 1997; and 2002). The emblematic figure in this was Quine, who fought hard to defend what he thought of as the “clear extensional ontology” (Quine 1969: 152) of his desert paradise safeguarded by strong extensional calculi. He made exposing and bewailing any hint of connivance with unwanted ontological notions one of the main planks of his philosophical program. He recognized that the logic that he loved harbored unsolved difficulties, but he and his followers found it desirable for achieving their ends, elegant, and aesthetically pleasing. He admonished philosophers to remain within the confines of the hard won metaphysically pure territory conquerable by faithful logicians enforcing a policy of extensional cleansing. He warned of the ontological crisis that would erupt were logicians to disobey his strictures and begin a retreat back into what he called “the metaphysical jungle of Aristotelian essentialism”. Appealing to philosophers’ weaker nature, he actually advised philosophers to run away from the problems, to flee creatures of darkness and curiously idealistic ontologies that repudiated material objects (Quine 1947: 43 and 47; 1976: 176 and 188; 1960: 191-232; and 1961: 158; Hill 1997: Chap. 11). Fortunately, a handful of philosophers braved the strictures and took a bolder attitude toward limning the true and ultimate structure of reality. They ventured beyond the narrow confines sterile environment created by strong extensional calculi and developed intensional languages capable of analyzing the many nonextensional statements which figure significantly in the empirical sciences, law, medicine, ethics, engineering, politics, and much of ordinary philosophy, for example, but which had been deemed unfit for study by the analytic philosophical establishment because they complicate matters by not conforming to the rigid standards for admission into the stark, sterile logical world Quine and so many others have found so beautiful (Barcan Marcus 1993: 5, 8 and 76; Hill 1997: 124). So, more and more reasons for not shoving reasoning into an extensional mold began gathering right in the “beautiful” desert world they were so intent upon preserving.

6 The Road Not Taken Husserl had initially experienced revulsion toward idealistic ontology akin to Quine’s and Russell’s, but found his early empirico-naturalistic Brentanian approach to arithmetic and sets incapable of providing the continuity and clarity needed in science and knowledge in general (Husserl 1913/2001: vol. 1, 42). After

194 � Claire Ortiz Hill wrestling with his doubts, he developed a way of finding clarity with respect to the central traits of reality that he believed positioned him “far from any mysticometaphysical exploitation of ‘Ideas’, ideal possibilities and such” (Husserl 1994: 39). He had initially interpreted Bolzano’s thoughts about ideas, propositions and truths ‘in themselves’ as metaphysical abstrusities”, as “mythical entities, suspended between being and non-being”. Then he realized that the first two volumes of Bolzano’s theories about ideas in themselves and propositions in themselves were to be seen as a first attempt at a unified presentation of the area of pure ideal doctrines, that a complete plan of a “pure” logic was already available there (Husserl 1913: 37; and 1994: 201-202). The ideal entities so unpleasant for empiricistic logic and so consistently disregarded by it, Husserl began teaching, were not artificially devised either by himself or by Bolzano. They were given beforehand by the meaning of the universal talk of propositions and truths that is indispensable in all the sciences. And that indubitable fact had to be the starting point of all logic, for science was a web of theories, and so of proofs, propositions, inferences, concepts, meanings (Husserl 2003 [1908/09]: 45; and 1994: 201-202). From the mid-1890s on, Husserl defended the view, which he attributed to Frege’s teacher Hermann Lotze, that pure arithmetic was basically no more than a branch of logic that had undergone independent development. Lotze, Husserl explained, had correctly recognized cardinal number as a specific differentiation of the concept multiplicity (Vielheit) and multiplicity as the most universal logical concept combining objects in general. This most universal concept of multiplicity splits into a series of different special forms and these are the cardinal numbers (Husserl 2001 [1896]: 102, 241-242, and 271-272; 2001 [1902/03]: 19 and 34; 2008 [1906/07]: §15; and 1996: §36b). In his logic courses, Husserl taught that all of arithmetic is grounded in the arithmetical axioms. The unending profusion of theories that arithmetic develops is already fixed, enfolded in the axioms, and theoretical-systematic deduction effects the unfolding of them following systematic, simple procedures. The concept of cardinal number is derived from given purely logical concepts and axioms and from perspicuous laws grounded in the essence of these purely logical categories. Each genuine axiom is a proposition that unfolds the idea of cardinal number from some side or unfolds some of the ideas inseparably connected with the idea of cardinal number. The meaning of cardinal number, he said, was the answer to the question: “How many?” Since each and every thing can be counted as one, he reasoned, to conceive the concept of number, or that of any arbitrarily defined number, we only need the concept of something in general. One is something in general. Anything can be counted as one and out of the units all cardinal

Husserl’s Way Out of Frege’s Jungle �

195

numbers are built (Husserl 2001 [1902/03]: 32-35, 39, 49, 231-232, and 239-249; 2008 [1906/07]: §§13c and 19d, and Appendix B.V, 434). Husserl never tried to transform subject-predicate propositions in which numbers figure as dependent properties into propositions about self-sufficient objects and never invented a law mandating that the way Frege did. For Husserl, concepts like cardinal number and set do not express essential forms of propositions and the laws that pertain to them are not laws for truths grounded in the essence of the proposition in general. They are grounded in the universal idea of objectivity that makes them applicable in every possible field of knowledge. They a priori express possible object prototypes and what is grounded in their formal essence (Husserl 2008 [1906/07]: §18a). Reflection upon the naturally broadest universality of the concepts number and set and also upon the concepts unit and element determining them showed him that the theory of sets and that of theory of cardinal numbers relate to any object whatsoever with a formal universality and are derivative formations of the concept of anything-whatsoever. Their fundamental concepts are syntactical formations of the empty something. The theory of cardinal numbers deals with numbers as differentiations of forms of sets and set theory with sets as made of any objects whatsoever that are taken together (Husserl 1974: §§24 and 27a). In contrast to Russell’s belief that “without a single object to represent an extension mathematics crumbles” (Russell 1903: §489), Husserl believed that the essential thing in mathematics was not the objects, but was its method that naturally flows into a purely symbolic technique (Husserl 2008 [1906/07]: §19a). He complained that extensions generate contradictions requiring every kind of artful device to make them safe for use in mathematical reasoning (Husserl 1974/1969: 74, 76, and 83).

7 Conclusion Russell wrote that “The characteristic excellence of mathematics is only to be found where the reasoning is rigidly logical: the rules of logic are to mathematics what those of structure are to architecture. In the most beautiful work, a chain of argument is present in which every link is important on its own account, in which there is an air of ease and lucidity throughout, and the premises achieve more than would have been thought possible, by means which appear natural and inevitable” (Russell 1959 [1917]: 61). Husserl claimed no less and it seems to me that his theory of the purely logical approximates those standards far more surely than the soi-disant clean, clear

196 � Claire Ortiz Hill beautiful extensional ontology that analytic philosophy wanted to have and to hold. Never bedazzled by the beauty of what George Boolos called Frege’s Eden, or what David Hilbert called Cantor’s paradise, Husserl developed a theory of formal logic that is a blueprint for limning the true and ultimate structure of reality every bit as much as Quine’s is. Husserl’s world of the purely logical is not particularly ugly and his theories about numbers, arithmetic and set theory are not particularly bad. They are just untried. Moreover, there is nothing especially beautiful about Russell’s bewildering maze, Frege’s thicket of contradictions or Quine’s fragmented world of rabbit parts, river stages and kinship, where the ontologies of physical and mathematical objects are but myths relative to an epistemological view (Quine 1960; and 1969). The analytic logical establishment said they wanted a beautiful logic, but what they really wanted was a logic that would permit them to undo many of the great issues of traditional ontology, issues which, ironically, the likes of Frege and Cantor did not find distressing. It was indeed a matter of trading ontology for a metaphysics, or really a lack of metaphysics, that they considered aesthetically pleasing. In fact, the logic whose virtues they chose to defend was ultimately neither true nor beautiful. In conclusion, I see this as yet another indication that, as Frege, Russell and Husserl all concluded, fundamental differences between dependent meanings and independent meanings lying concealed behind inconspicuous grammatical distinctions and ultimately prove inviolable because they are “founded deep in the nature of things” (Frege 1980 [1891]: 41) in such a way that contradictions, paradoxes, antinomies, fallacies, nonsense, confusion, absurdity inevitably result when they are not respected and that this is a topic of prime importance for the understanding of major issues in twentieth century western philosophy (Hill 2003; and 2010).

| Part III: Modes of Being

Arkadiusz Chrudzimski

Ingarden on Modes of Being Arkadiusz Chrudzimski: Szczecin University, Poland

Roman Ingarden’s theory of modes of being belongs to the most interesting and original parts of his ontology. According to Ingarden the word “being” is used with a number of different meanings. Not only have we to distinguish between existential “is” (“There is no Santa Claus”), copulative “is” (“This apple is red”) and “is” as transitive verb (“My dog is in your garden”). Also the philosophically most celebrated existential “is” can be – as another well known thinker used to say – predicated in various ways. According to Ingarden these various ways can be analyzed into several existential moments, which basically correspond to various ways of existential dependence. Possible modes of being are obtained by combining existential moments. Finally we get a highly articulated existential ontology. Modes of being aren’t regarded as primitive and unanalyzable. They have a kind of internal structure that helps us to understand ontological dependencies in which entities enjoying a given mode of being are involved. Moreover, the combinatorial idea makes it possible to construct many modes of being out of a relatively small number of existential moments, which makes the resulting dependence network ontologically transparent and cognitively accessible.¹

1 Quantification and Existence If one has to outline a map of the most fundamental metaphysical positions, one can begin with the distinction between the partisans of univocality of existential “is” and those who insist that being can be predicated in various ways. Among the first mentioned we find Quine, the late Brentano, and the majority of contemporary ontologists, while the most salient expositors of the second view were Aristotle and Alexius Meinong. Moreover it turns out that there is a fundamental difference between Aristotelian and Meinongian understanding of the equivocal-

1 The first two sections of this paper are based on (Chrudzimski 2003). My work was supported by the Polish National Centre of Science (NCN, 2012/07/B/HS1/01595) and the Foundation for Polish Science (“Master” programme, directed by T. Szubka).

200 � Arkadiusz Chrudzimski ity of being. However in order to explain these distinctions it will be convenient to begin with Quine. Philosophers of the analytic persuasion tend to understand being in Quinean terms (Quine 1948). After his construal “to be” means “to be a value of a bound variable”. This sounds simple, but in fact involves at least three independent claims none of which can be regarded as unproblematic: 1. 2. 3.

The quantified variables are the only grammatical locus where the ontological commitments of our discourse can manifest themselves. All quantification is ontologically committing. Quantification is univocal (which means that there is only one syntactical position that can be quantified over).

Let me begin with the first claim. Its justification involves certain substantial assumptions about the deep grammar (or logical form) of our language. The assumptions in question can be called “Russellian”, since they have been directly derived from the way Russell dealt with Meinong’s problem concerning the present king of France.² We all remember the story. Meinong observed that certain sentences of the subject-predicate form with non-referring terms in the subject position seem to be clearly true. His favorite examples were negative existential statements (like “Zeus doesn’t exist”) or analytic sentences in narrow Kantian sense (like “The round square is round”).³ Now it seems that for such a sentence to be true (and for that matter also to be false) there must be something of which the involved predicate is truly (or falsely) predicated. Ergo there must be an object to which the “empty term” in the subject-position refers. Since in many cases we know for sure that the object in question doesn’t exist (like in the case of the round square), it seems that we are forced to introduce nonexistent objects into our ontology. This is, in a nutshell, Meinong’s argument leading him to his (in)famous doctrine of objects “beyond being and non-being”.⁴ Russell’s answer is well known. He says that we don’t need all those Meinongian entities in the role of reference objects for empty names, because there are no empty names at all. Sentences that so impressed Meinong have a misleading surface grammar, but as soon as we identify their true logical form, the apparent names like “The present king of France”, “The round square”, or “The golden

2 See (Russell 1905a). 3 Contemporary Meinongians often add trivial identities like “Zeus = Zeus”. 4 See (Meinong 1904).

Ingarden on Modes of Being � 201

mountain” disappear. What we finally get is Russell’s theory of definite descriptions. The sentence: (1)

The present King of France is bald,

which surface grammar suggests the simple form: (2)

Ba

(with “a” symbolizing “the present king of France”, and “B” standing for “is bald”), becomes, in Russell’s hands: (3)

∃ x[Rx ∧ ∀y(Ry ⊃ y = x) ∧ Bx],

where “R” is a short form for something like: “has presently a royal power over France”. The interesting thing about the transition from (2) to (3) is the elimination of the empty name “the present king of France”. According to Russell all such (apparent) names are in fact definite descriptions, of the form: “the only x, such that...”, that can be eliminated according to schema (3) above. In a logically regimented language we have therefore no empty names at all. They have all been replaced by predicates, quantifiers and bound variables. Russell’s claim was thus that many expressions that we treat as names are in fact eliminable definite descriptions, but he still preserved the notion of a genuine proper name.⁵ Quine goes one step further. According to him for every proper name an appropriate predicate can be introduced that allows us to eliminate this name along the Russellian lines. Consider the sentence: (4)

Socrates is dancing.

Quine says that we have only to introduce the corresponding predicate “socratizes” (“S” for short) and we can write down a nice Russellian translation of (4) (“D” stands for “is dancing”): (5)

∃ x[Sx ∧ ∀y(Sy ⊃ y = x) ∧ Dx].

So the first important assumption of Quine’s doctrine of ontological commitments is the claim that a logically analyzed language doesn’t contain any singular terms. Not only definite descriptions but also all proper names are in principle eliminable. In a reasonably regimented language we are therefore left with nothing but predicates, quantifiers and bound variables.

5 However the Russellian class of genuine proper names is rather strange. It includes names of sense data, names of directly accessible universals and the expression “I” when used in Cartesian way.

202 � Arkadiusz Chrudzimski This means that one of the syntactic categories where the ontological commitments have been traditionally placed – proper names – just disappeared. Instead we have syntactically complicated structures involving quantified variables and this is no surprise that exactly this syntactical device has been chosen by Quine as the place at which the ontological commitments of our discourse can be checked. However this cannot be the end of the story, because beside proper names there is also another part of speech which traditionally inspired the development of heavy metaphysical machinery, and what I mean is precisely... predicates, with the help of which Quine tries to eliminate proper names. In fact the best known metaphysical controversies were focused on the ontological status of properties rather than on the Meinongian riddles of non-existent objects. We can therefore expect that the semantic values of predicates will be at least as metaphysically interesting as the putative referents of empty singular terms. So have we just replaced one troublemaker by another one? In this context Quine says something astonishing. He just denies that predicates carry any ontological commitment. We read that they express only “ideology” of our language, not its “ontology”. True enough, singular terms are supposed to “refer”, while predicates should “predicate”, and this is doubtless an interesting difference. Nonetheless our first intuition is that, in order for a predication to be true, there must be in the world something that corresponds not only to the referring expressions, but also to the predicate in question; and that this “something” deserves to be taken ontologically seriously. Many metaphysicians believe that there are certain entities corresponding to predicates, and that truth or falsity of a sentence of the form “Fa” consists precisely in obtaining or not obtaining of a certain relation between the entities corresponding to the predicate and to the subject of the sentence. A version of this view is the well known doctrine of substrates, properties, and relation of exemplification obtaining between them. Now Quine rejects such a metaphysics. As long as we don’t quantify over predicate variables we don’t need to worry about the metaphysics of properties (or other entities we choose as semantic values for predicates). The fact that certain predicates are true of certain objects must be regarded as a primitive fact about the universe that cannot be further analyzed.⁶ So we have here a reconstruction of the conceptual way leading to the first of Quine’s theses – that the quantified variables are the only grammatical locus where the ontological commitments of our discourse can manifest themselves. As

6 There is some justification of this policy having to do with a possible infinite regress. See (Devitt 1980).

Ingarden on Modes of Being � 203

we see it involves two claims: (i) that all singular terms are in principle reducible and (ii) that predicates are ontologically innocent. Now if we take these two suppositions – the supposition of eliminability of singular terms and the supposition of the ontological innocence of predicates – for granted, then there indeed seems to be no other places for deciphering ontological commitments than variables bound by quantifiers. But how should we assess these two claims? The first one is based on Russell’s authority but it must be said that in the light of the newer (but already classical) works by Kripke and Putnam (Kripke 1980; and Putnam 1975) it seems to be patently false. What the mentioned authors have shown is precisely that the majority of proper names don’t refer by description, which means inter alia that they are not eliminable in the Russellian way. And it is by far not the end of sad news, because the Kripke-Putnam style causal semantics casts also a deep shadow on the thesis of ontological innocence of predicates. What I mean is the causal semantics for the so called “natural kind terms” (like “gold”, “water” or “lion”). To grasp the real importance of this theory for Quine’s criterion, notice that according to the Fregean construal of the logical grammar (assumed both by Russell and Quine) there are no general terms at all. Every natural kind term (like “gold” or “horse”) becomes a part of a predicate (”is of gold”, “is a lion”). After this move, which was wholeheartedly accepted by Quine and the majority of contemporary logicians, it is very natural to assimilate the semantics of such composed predicates of the form “is a N” to the semantics of the grammatically straightforward predicates (like “run” or “sing”). We tend to assume that every predicate just predicates a certain property and a general name being only a grammatically non-independent part of a predicate is therefore bound to “refer by description”. Now this turned out to be a huge mistake. What all the thought experiments with twin earths have shown, is that the semantics of the natural kind terms resembles rather the semantics of our proper names. Both kinds of expressions refer directly, without any mediation by “bundles of properties”. After this the claim that predicates are ontologically innocent can be regarded only as a bad joke. It seems thus that already after analyzing the claim (i) we can safely conclude that Quine was fundamentally wrong and just forget about his theory of ontological commitment.

204 � Arkadiusz Chrudzimski

2 Equivocality of Being: Meinong, and Aristotle I think that this conclusion is fully justified, but for the sake of sketching a map of my further investigation let me review also the further two assumptions: that all quantification is ontologically committing; and that quantification is univocal (which means that there is only one syntactical position that can be quantified over). The first of these claims is anti-Meinongian, while the second antiAristotelian in spirit. Now in Quine’s original arguments these two issues are hopelessly mixed. Consider his much celebrated worry about “possible men in that doorway”. (Quine 1948: 4) Quine analyses a Meinongian theory introducing non-existent objects, and claims that we should refrain from such a theory, because it would be a source of “disorderly elements”. Consider the possible fat man in that doorway and the possible bald man in that doorway. Are they one and the same object, or rather two different ones, asks Quine, and suggests that we have here no clear decision procedure because of the lack of any clear criteria of identity. Quine is again wrong. He gets the impression that the criteria of identity are unclear only because he mixed together two kinds of entities. The possible fat man at that doorway can be considered as a “complete”, fully determined, concrete individual, but he can be also construed as an “incomplete” object involving only three properties: being a human, being fat and being in that doorway. Taken as incomplete objects the possible fat man in that doorway and the possible bald man in that doorway are clearly different, and when we take them as complete individuals they sometimes coincide and sometimes not. In a Meinongian universe there are many fat men in that doorway and many bald men in that doorway. Some of these fat men in that doorway are bald and some are not. Needless to say, there can be good reasons for avoiding this kind of ontology, but my point is that Quine’s reasons just weren’t good. Sadly enough they are based on a relatively simple confusion. An important lesson from this confusion is that one has to distinguish between two issues: (i) the question of introducing non-existent entities (actual and possible men in that doorway); and (ii) the question of introducing entities belonging to various ontological categories (complete and incomplete men in that doorway). Problems revolving around the first topic are broadly “Meinongian”, while the questions concerning the second move can be called “Aristotelian”. These two themes are mutually independent. We can adopt an ontology accepting only one category (say individuals or tropes) and at the same time claim that beside the existing individuals (or tropes) there are also non-existing ones. We can on the other hand accept a multi-categorial ontology (e.g. involving indi-

Ingarden on Modes of Being � 205

viduals, properties and states of affairs) and insist that there are no non-existent entities (i.e. there are only existent individuals, instantiated properties, and obtaining states of affairs). And finally we can also say that there are many categories of entities and for every category we have existing and non existing examples (i.e. there are existent and non-existent individuals, instantiated and uninstantiated properties, obtaining and non-obtaining states of affairs etc.). The last position can of course be further differentiated. There are for example many philosophers who accept uninstantiated properties, or non-obtaining states of affairs, but no non-existent individuals. A word of warning will be in order here. A philosopher who accepts nonexemplified properties or non-obtaining states of affairs is by no way forced to assign to them any strange way of being. Indeed the majority of contemporary Platonists do exactly the opposite. They introduce non-exemplified properties and non-obtaining states of affairs in good faith that they can be quantified over in the exactly as committing way as existing individuals. In a word: a partisan of nonexemplified entities can construe them in a Meinongian or in an actualist way, and the majority of contemporary philosophers follow the actualist way. From a Quinean point of view Meinongian ontology can be interpreted as introducing ontologically neutral quantification.⁷ If we symbolize the existence predicate as “E!”, then Quine’s claim can be expressed as: (Q)

¬∃x¬E!x,

while Meinong’s dictum, that there are nonexistent objects, reads: (M)

∃ x¬E!x.

And if we have such a neutral existence quantifier (or maybe better “being quantifier” – symbolized as “∃M ”) and the existence predicate (“E!”) at our disposal, we can easily define a Quinean, ontologically committing quantifier (“∃Q ”): (M-Q)

∃ Q xFx ≡ ∃M x(Fx ∧ E!x).⁸

On the other hand a multi-categorial, Aristotelian approach can be, when expressed in Quinean terms, construed as a position allowing for quantification of not only individual variables but also predicate or sentence variables. And such a multi-categorial quantification can be again interpreted either univocally or

7 This interpretation is, I believe, philosophically the most interesting and at the same time the most truthful to the historical Meinong, but there are of course other possibilities of interpretation. See (Chrudzimski 2005b) and (Chrudzimski 2007: Chap. 8). 8 See (Routley 1980: 187).

206 � Arkadiusz Chrudzimski equivocally. In other words: an Aristotelian can claim that there are individuals, properties, relations and all they exist in the same sense, but she can also claim that each of these categories enjoys its own unique mode of being. Needless to say, Aristotle himself defended this latter position.

3 Ingarden’s Place on our Map After this journey through the jungle of Quinean, Meinongian and Aristotelian themes we are in a position to sketch a rudimentary map of possible views concerning the controversy over the univocality and equivocality of being. To begin with, we remember that we have to distinguish between a Meinongian and an Aristotelian sense of equivocality of being: Meinong says that beside existent objects there are also nonexistent ones and Aristotle claims that beside concrete individuals (Aristotle’s substances) there are other “things” of which being can be predicated. These are of course two quite different claims. Second, the Aristotelian multi-categorical ontology can, but need not, be combined with the thesis of the equivocality of being, while Meinong’s distinction between ontologically committing existence and ontologically neutral being seems to directly involve this thesis.⁹ In the following classification I will therefore distinguish between the univocality/equivocality in Meinong’s sense (M-univocality/M-equivocity) and univocality/equivocality in Aristotle’s sense (A-univocality/A-equivocality). Consider the following table:

Non-Meinongian approaches Meinongian approaches

One-category Ontology D.C. Williams

Multi-category Ontology Armstrong Aristotle Lowe Ingarden Loux Meinong

A-univocal being

M-univocal being M-equivocal being

A-equivocal being

Fig. 3

9 This is true only under the interpretation assumed here. There are some “Meinongians” who claim that the so-called nonexistent objects exist in exactly the same sense as trees and stones among us, and the existence predicate in fact doesn’t have anything to do with modes of being, but means something like “being spatiotemporally connected with us” (and so expresses a “real property” after all). See again (Chrudzimski 2005b) and (Chrudzimski 2007: Chap. 8).

Ingarden on Modes of Being � 207

As we see, Ingarden has been placed together with Aristotle among non-Meinongians and partisans of multi-category approach who believe that being can be predicated in many ways. That means two things. First of all, Ingarden belongs to philosophers who don’t believe in Meinongian non-actual entities; and second, he claims that the word “being” is nonetheless used with a number of different meanings, when applied to different ontological categories. Contemporary Platonists and Aristotelians (like D.M. Armstrong, Jonathan Lowe or Michael Loux) are placed in the cell uniting multi-category ontology with A-univocal understanding of being. They accept various ontological categories but don’t claim that entities belonging to them also exist in different ways. As a proponent of one-category non-Meinongian ontology I have chosen a “tropist”, D.C. Williams. Meinong has been classified as a multi-categorial Meinongian who accepts an A-univocal construal of being. This is a bit controversial. Many scholars would interpret Meinong as a partisan of an A-equivocal reading and therefore shift him one cell to the right. This decision has some support in the way Meinong expresses his views. He indeed makes a verbal difference between existing of concrete individuals and obtaining of states-of-affairs (which Meinong calls objectives),¹⁰ but at the end of the day this distinction turns out to be merely verbal.¹¹ Finally, one-category Meinongians would claim that there is only one type of entities (say concrete individuals, states of affairs or tropes) but that there are among them both existent and non existent ones. Under the partisans of individual properties probably John Bacon could be placed in this rubric.¹² Now let us return to Ingarden. Concerning his anti-Meinongianism there are some passages that can suggest the opposite and now I will address them. First of all Ingarden sometimes speaks of pure possibilities, which sounds very Meinongian indeed. These passages are in fact very perplexing, but at the end of the day it always turns out that these possibilities are reducible to the relations obtaining between pure qualities that in Ingarden’s system play roughly the role

10 The German word used by Meinong is “bestehen”. For some mysterious reason it is often translated as “subsist”, which is a very bad choice, since it suggests some rather misleading medieval associations. The most straightforward and natural translation is “obtain”. 11 I have no room to argue for this thesis here. For details see my book (Chrudzimski 2007). In any event this issue isn’t important for Ingarden’s theory of modes of being that I will speak about in this paper. 12 See (Bacon 1995: 7).

208 � Arkadiusz Chrudzimski of universals.¹³ In fact Ingarden’s analysis of the modal operators corresponds to the standard policy adopted by metaphysical Platonists. The claim is that there are no non-actual (merely possible) entities; and the modal discourse is to be explained in terms of some relations (like exclusion, entailment etc.) obtaining between universals,¹⁴ according to the following (or similar) definitions: Possibly: ∃x(Fx ∧ Gx) = Df . There is no exclusion relation between the properties [F] and [G] Necessarily: ∀x(Fx ⊃ Gx) = Df . There is an entailment relation between the properties [F] and [G] The second point concerns the universals themselves. As said above, Ingarden has in his ontology entities resembling Platonic universals (he calls them pure qualities) that need not to be exemplified by individuals. True enough, Ingarden is in this sense Platonist and this point differentiates him clearly from Aristotle, but on the other hand the mode of existence he ascribes to his non-exemplified universals is not Meinongian at all. In this respect Ingarden is again like the majority of contemporary Platonists. Third, when speaking of ideal and purely intentional objects Ingarden often says that they are non-actual. At the first sight it can therefore seem that according to Ingarden such entities as non-exemplified universals or fictitious characters are to be construed as merely possible Meinongian objects. Now this would again be a mistaken interpretation. In those contexts Ingarden uses the word “non-actual” not in the meaning of “merely-possible”, but rather as opposite of “actual” in its temporal and so to say “substantial” meaning. Physical objects that exist now have both actual presence in time and actual causal powers. Both ideal and purely intentional objects lack these features and that’s why they are called “non-actual”. I will talk about it in the last section. So Ingarden’s way is generally not Meinongian but rather Aristotelian and there is another important point of coincidence between Aristotle’s and Ingarden’s thesis of equivocality of being. As said above the claim that we have to differentiate between existing of individuals, obtaining of states of affairs, and properties being exemplified can sometimes be understood as a purely grammatical distinction. It can mean that it would be weird or inconvenient to say that the state of affairs exists instead of that it obtains. In this reading it is a proposal of a certain

13 So in §5 of (Ingarden 1964-1965) we read that ontology investigates “pure possibilities” but as it turns out we learn that these possibilities are ultimately generated by relations obtaining between “pure qualities”. See (Haefliger 1994: 63). 14 For the details of this approach see e.g. (Chisholm 1989: 143f).

Ingarden on Modes of Being � 209

terminological convention. But when we hear such a thesis from a philosopher, then it typically means something different and metaphysically deeper. We are informed that between various modes of being there are certain “real”, not purely grammatical, differences; that entities belonging to different categories really are in different ways. To give a classical example: it is often said that individual objects exist “independently”, while properties exist only in individual objects. And very often it is claimed that this real difference culminates, as it were, in different “strengths“ of being. The mode of existence enjoyed by properties is said to be weaker then the mode of existence of their bearers. We all know that this was the way of Aristotle. He claims that various senses in which being can be predicated shouldn’t be construed as an arbitrary bundle of concepts, but instead make up a well ordered family. The principle of this ordering he calls “analogy” and its ultimate grounding lies in relations of ontological dependence obtaining between the entities belonging to given categories. According to this picture the only beings in the fullest sense of the word are concrete individuals (called by Aristotle “substances”), because they are the only entities that are ontologically independent (in the relevant sense). The other kinds of entities exist weaker, and the “strength” of their existence is proportional to their “nearness” to the substance. For example monadic absolute accidents are directly anchored “in” their substances, and their mode of existence is therefore relatively strong, relative accidents, on the other hand, need at least two substances, and enjoy a significantly inferior mode of being.¹⁵ Ingarden’s understanding of equivocality of being is very similar. The various modes of being are based on ontological dependencies and can be understood as relative strengths or weaknesses of existence. Moreover, also the overall structure of dependence in Ingarden’s universe is conceived in a broadly Aristotelian way. Concrete individuals (Aristotelian substances) stand in the centre, as the firmest starting points or ontological foundations, and other existents exist only insofar as they are connected with substances by more or less complicated chains of ontological dependencies. The big difference between Ingarden and Aristotle is of course that in Ingarden’s universe we have in addition a huge independent domain of ideal entities. In this respect Ingarden is not an Aristotelian but a Platonist. The fact that Ingarden’s analysis of modes of being ultimately amounts to the analysis of dependence relations, is the reason why some commentators say that in fact we have here not a doctrine of modes of being, but rather a theory of onto-

15 See (Metaphysics: 1088a30-31).

210 � Arkadiusz Chrudzimski logical dependence.¹⁶ The same can be of course stated against Aristotle. I think that this diagnosis is correct, but will not stress this issue in my paper. I believe that Ingarden’s existential ontology is interesting even if we take it as a theory of ontological dependence in disguise.

4 A Combinatorial Existential Ontology Ingarden’s original idea was that all modes of being (Seinsmodi) can be analyzed into several existential moments.¹⁷ In general by “moments” in metaphysics we mean non-independent “parts” or aspects of something, and existential moments are non-independent aspects of modes of being. The claim that entities such as physical objects, Platonic universals or intentional objects have different modes of being thus amounts to the claim that their modes of being consist in different combinations of existential moments. Ingarden’s existential moments concern various ways of ontological dependence or independence, and this approach has much to do with sociological context of Ingarden’s work. Ironically enough, Ingarden’s monumental ontological opus magnum was meant by him only as a preparatory study to a definitive refutation of Husserl’s transcendental idealism.¹⁸ But while this refutation has never been completed, the preparatory studies proved to be very interesting on their own.¹⁹ Now Husserl’s idealist claim was, that the external world is somehow existentially dependent on the intentional acts of transcendental consciousness, and the crucial question in interpreting Husserl’s late position was therefore how exactly this dependence is to be understood. In his investigation of this issue Ingar-

16 See (Simons 2005). Also Półtawski claims that Ingarden’s existential ontology investigating modes of being and existential moments concern in fact mainly dependency relations between objects, and therefore should be classified not as a distinct philosophical discipline, but rather as a part of formal ontology. See (Półtawski 2005: 195). 17 For an overview see (Simons 2005) and (Johansson 2013). 18 An early formulation is to be found in (Ingarden 1929). The opus classicus is The Controversy over the Existence of the World (Ingarden 1964-1965). 19 I fully agree with Haefliger and Küng who write: “Ingarden set himself the task of examining Husserl’s transcendental idealism. The systematic results of these investigations are contained in his monumental main work, Der Streit um die Existenz der Welt (The Controversy over the Existence of the World). A contemporary ontologist may find it more interesting to read this study, which in fact remained unfinished, not in connection with the realism/idealism controversy, but as a treasury of numerous ontological distinctions, problem formulations, and detailed analyses.” (Haefliger & Küng 2005: 9)

Ingarden on Modes of Being � 211

den soon realized that there are many ways in which one entity can be said to be existentially dependent on another one. To give a simple example: the way Sherlock Holmes is existentially dependent on Conan Doyle is quite different from the way George W. Bush existentially depends on George Bush Senior. In fact every entity is ontologically dependent or independent in various ways, and existential moments are precisely the devices designed to capture those various kinds of dependence. Ingarden’s existential moments often build binary oppositions. To grasp the idea assume that we have exactly three kinds of existential dependence (a, b, c), and for each kind we have two possibilities: either a given entity is ontologically dependent of something in the mentioned way, or it is not. Now for any object we can ask if it is a-dependent, b-dependent, and c-dependent of some other entity. Purely combinatorially we get 23 , i.e. eight, possibilities: 1. 2. 3. 4. 5. 6. 7. 8.

a-independent, b-independent, c-independent; a-independent, b-independent, c-dependent; a-independent, b-dependent, c-independent; a-independent, b-dependent, c-dependent; a-dependent, b-independent, c-independent; a-dependent, b-independent, c-dependent; a-dependent, b-dependent, c-independent; a-dependent, b-dependent, c-dependent.

Not all existential moments introduced by Ingarden behave in such nice ways, but the main idea is what has just been indicated. To sum up, Ingarden’s combinatorial theory of modes of being is in a sense just another variation on the old Aristotelian theme, but his project is without doubt very ambitious and philosophically interesting. According to Ingarden modes of being aren’t primitive and unanalysable. They have a kind of internal structure that can be articulated and that helps us to understand ontological dependencies in which entities enjoying this mode of being are involved. Moreover, the combinatorial idea makes it possible that many modes of being be constructed out of a relatively small number of existential moments; and this restricted arsenal of primitive moments helps us to make the resulting dependence network ontologically transparent and cognitively accessible. These are doubtless significant advantages of Ingarden’s approach.

212 � Arkadiusz Chrudzimski

� Four Main Dichotomies At the very beginning of his investigation Ingarden distinguishes four pairs of existential moments. From the point of view of existential ontology an entity can be (i) (ii) (iv) (v)

autonomous or heteronomous (Ingarden 1964-1965: §12); original or derivative (Ingarden 1964-1965: §13); self-sufficient or non-self-sufficient (Ingarden 1964-1965: §14); independent or dependent (Ingarden 1964-1965: §15).²⁰

�.� Autonomous and Heteronomous Let me begin with the first dichotomy: autonomous or heteronomous.²¹ This pair is very important, both for Ingarden’s original project of existential ontology and for proper understanding of many other parts of his philosophy. The paradigmatic example of a heteronomous entity is namely a purely intentional object – entity which, according to Ingarden, is literally created by every intentional act. Ingarden’s theory of intentionality belongs to the family of theories of intentional objects. In the philosophy of 19th and 20th centuries this approach has been initiated by Franz Brentano.²² According to this theory every intentional reference simply consists in such a “production” of a purely intentional object that stands before the subject’s mental eye to represent the external reference object (or to replace it, as in the case of error or conscious fiction).²³ Now according to Ingarden’s reading, Husserl’s transcendental idealism involves a strange claim, that what we take for an autonomously existing, external world is in fact a purely intentional correlate of intentional activity of the transcendental consciousness. The category of purely intentional objects was thus crucial for Ingarden’s planed refutation of Husserl’s idealism. As already mentioned, this refutation has never been completed, but as

20 The English terminology is taken from (Simons 2005). 21 German: seinsautonom/seinsheteronom, Polish: autonomiczny/heteronomiczny, or samoistny/niesamoistny. 22 Brentano first formulated this theory in his Psychology from an Empirical Standpoint (Brentano 1874). For an ontologically more articulated development see his Descriptive Psychology (Brentano 1982 [1891]/1995). There is a fascinating prehistory of this approach in medieval philosophy, of which Brentano was fully aware. What I primarily have in mind here is the doctrine of esse obiectivum championed by Duns Scott and his followers. See (Perler 2002). 23 On Ingarden’s theory of intentionality and his concept of purely intentional object see (Chrudzimski 2005a), (Chrudzimski 2002) and (Johansson 2010).

Ingarden on Modes of Being �

213

a by-product of this enterprise Ingarden developed a very interesting ontology of fiction that remains the best known part of his philosophy.²⁴ Heteronomous purely intentional objects have been contrasted with the category of autonomous entities. The paradigmatic example of this category is a threedimensional physical object enduring in time. What is the main difference? An autonomous object is, according to Ingarden, one which “has its existential fundament in itself”. In this sense it can be said to exist in itself. A heteronomous object, in contrast, “owes” its existence to something outside itself. In this sense it exists “in virtue of something else”.²⁵ We can capture the notion of heteronomy by the following principle: (A) (H)

∀ x[x is autonomous ≡ ¬∃y(y � = x ∧ y ⇒ x)] ∀ x[x is heteronomous ≡ ¬(x is autonomous)]

where the two-place predicate “⇒” is to be read as “is ontological fundament of”. Ingarden concedes that a heteronomous object can have its fundament in another heteronomous object, but at the same time claims that such a chain of existential grounding can neither go ad infinitum, nor be circular. Finally there must be a certain autonomous object which constitutes the ultimate existential fundament of the whole chain.²⁶

�.� Original and Derivative The second opposition: originality versus derivativeness²⁷ has much to do with classical topics in philosophical theology. Roughly derivativeness means “being created” and originality is the opposite of it. If it were all, then the notions could be captured by the following principles:

24 Ingarden developed these ideas mainly in (Ingarden 1972 [1931]) and (Ingarden 1968). For an overview of Ingarden’s theory of fiction see e.g. (Mitscherling 2005). 25 According to Ingarden, beside purely intentional objects also future empirical possibilities enjoy this heteronomous mode of being. They have their ontological fundament in present entities from which they can causally originate. I will talk about them in the next section. 26 Simons (2005: 42f.) remarks that in Ingarden’s theory “it appears possible to envisage an object that is partly autonomous and partly heteronomous, in that its fundament lies partly within itself and partly without. Andrzej Półtawski has suggested to me that Ingarden would certainly not have accepted such a possibility, since for him even a jot of heteronomy means an object is not autonomous.” Now it seems that in Ingarden’s ontology we can indeed find such hybrid entities. I mean here the negative states of affairs. A detailed discussion of this lies outside the scope of this paper, but see (Chrudzimski 2010). 27 German: seinsursprünglich/seinsabgeleitet, Polish: pierwotny/pochodny.

214 � Arkadiusz Chrudzimski (O) (D)

∀x[x is original ≡ ¬∃ y(y � = x ∧ y C x)] ∀ x[x is derivative ≡ ¬(x is original)] i.e. ∀x[x is derivative ≡ ∃y(y � = x ∧ y C x)]

where predicate “C” means of course “creates” or “is the creator of”.²⁸ Now the story gets more complicated, because Ingarden’s original formulation uses modal terms. Ingarden writes that an object is original if it cannot be created by any other object, and it is derivative if it can be so created. Right at the beginning of §13 of (Ingarden 1964-1965) we read: An object is existentially original if, by its essence, it cannot be created by any other object.

It seems thus that principle (O) should be at least modalised. (O♦)

∀x[x is original ≡ ¬♦∃ y(y � = x ∧ y C x)]

So far so good, but look what happens, when we try to apply our (D): (D)

∀ x[x is derivative ≡ ¬(x is original)]

to the modalised principle (O♦). What we get is the following formulation: (D♦*)

∀x[x is derivative ≡ ♦∃ y(y � = x ∧ y C x)]

which doesn’t look bad, but is definitely not what Ingarden wants. He writes: An object is existentially derivative if, by its essence, it can exist only by being created by another object.

A corresponding formal principle would look like this: (D♦)

∀ x[x is derivative ≡ �∃ y(y � = x ∧ y C x)]

In fact (D♦) doesn’t just contradict (O♦). There are some cases between (O♦) and (D♦), namely precisely those allowed by (D♦*) and excluded by (D♦). It seems that Ingarden doesn’t consider those middle-cases a real ontological possibility. According to him every entity has to be either necessarily created or necessarily not-created and it seems that this requirement is just a part of his definition of relation C.

28 Ingarden hastens to add that the relation of creating involved in this distinction shouldn’t be misconstrued in causal terms. To mention only one important point: causation can take place only between events or processes, while the original/derivative distinction concerns all categories of being.

Ingarden on Modes of Being �

215

We see that Ingarden’s concept of the original/derivative dichotomy is definitely less clear than the previous opposition of autonomy and heteronomy. The matter gets even less transparent by the fact that just after the quoted sentence: [1] An object is existentially original if, by its essence, it cannot be created by any other object.

we read: [2] If it exists at all, then it is only because, by its very essence, it cannot fail to exist.

Formally: (O�)

∀x[x is original ≡ ¬♦¬∃ y(y = x)]

The fact that Ingarden doesn’t make any comments about the transition from [1] to [2] suggests strongly, that he regards the second sentence quoted here simply as a consequence of the first. But we know that [2] doesn’t follow from [1]. It is imaginable that there is an object that cannot be created by any other object, and nonetheless can fail to exist. Would it exist, its existence would be, as it were, a kind of strange accident. After a while we read: [3] It follows from this, that, if it exists, then it cannot be destroyed by any other object [...].

Well, this follows from sentence [2] but not from [1]. It seems that we can easily imagine an object which cannot be created, but cannot be destroyed. Simons (2005: 45) remarks correctly that in his explanations of the original/derivative distinction Ingarden systematically mixes the notion of having no creator with the notion of existing necessarily. The dichotomies autonomy/heteronomy and originality/derivativeness are not mutually independent. Ingarden adopts a principle to the effect that originality excludes heteronomy: (HD)

∀ x(x is heteronomous ⊃ x is derivative)

The simplest justification for (HD) would be a theory to the effect that the relation of having its ontological fundament in is to be understood as a special case of the relation of being created by: ∀ x∀y[(x ⇒ y) ⊃ (x C y)],

and probably that’s precisely what Ingarden means, but in principle (HD) states only that if an object x is heteronomous (i.e. has its ontological fundament in an-

216 � Arkadiusz Chrudzimski other object y), then it is also created by a certain object z, but it isn’t explicitly stated that y and z must be one and the same entity.

�.� Self-Sufficient and Non-Self-Sufficient The third opposition is between self-sufficient and non-self-sufficient entities²⁹ (Ingarden 1964-1965: §14). In explaining these notions Ingarden refers to Husserl’s theory of parts and wholes, developed in his III. Logical Investigation (Husserl 1913). An object is self-sufficient if its essence doesn’t require that it can exist only as a part of some larger whole. An object is non-self-sufficient if its essence involves such a requirement. We see that we have here again modal notions. Formally they can be defined as follows: (¬SS) (SS)

∀ x[x is non-self-sufficient ≡ ∃y∃ z(y � = x ∧ x < z ∧ y < z)] ∀ x[x is self-sufficient ≡ ¬(x is non-self-sufficient)]

where the predicate “