REFERENCE TO THE NON-EXISTENT

John N. Deely

St. Mary's College

Notre Dame, Indiana

"DISCOURSE" IS THE English word which, perhaps, better than any other, designates the center of gravity or principal focus, as it were, of contemporary philosophical thought. For, from the standpoint of the sociology of knowledge, linguistic analysis and phenomenology are the two dominant movements or schools of contemporary philosophy, and " discourse," signifying, as it does, thought and language equally, is a rubric that covers--as well as any rubric can--the central concern of both of these characteristically different philosophical movements. As its name plainly tells, linguistic analysis is a philosophy concerned with the analysis of language, whether simply to get clear about common usage, and so obviate needless philosophical perplexities consequent upon careless speech (" ordinary language analysis "), or to supply for the deficiencies of common speech by substituting for it, at least in scientific and philosophical contexts, a technically exact, formalized mode of discourse (" logical analysis ").

Phenomenology, in contrast to both forms of analysis, is more " mentalistic " or " thought " oriented, inasmuch as it seeks not so much to clarify the patterns of speech as to explicate-- by a careful attention to and description of immediate experience in all its variety--the forms and laws according to which thought constitutes the objects given in and by experience.

In general terms, then, still speaking from the external standpoint of sociology of knowledge, philosophy today is centered on discourse.

Shifting now to an internal standpoint, we can say that one of the problems for any philosophy so centered is the problem of non-being, insofar as the question of non-being arises, more

or less ineluctably, out of the fact that discourse appears to refer to things regardless of whether or not they exist in fact. In the philosophy of St. Thomas, for example, the celestial spheres are repeatedly referred to as explanatory factors decisively involved the phenomena of life and death, the specific constancy of biological forms, the genesis of cognition from sensation, and many other philosophically significant occurrences.¹ Yet few today believe in the reality of the celestial spheres. Nor is such a profound confusion of non-being with being by any means limited to medieval times. The history of science and ethnology is jam packed with references to what does not exist--or at least is not regarded as existing by contemporary lights; and no doubt our own culture harbors its fair share of non-beings parading in the guise of beings.

Indeed, what would become of literature generally if human discourse did not have, or at least appear to have, the capacity to refer to what does not exist as if it did exist? It might almost be said that non-being, which plays no positive role in the physical world, finds a comfortable home indeed in the world of human discourse. It hardly .seems too much to say that the relativity of discourse to objects, and its indifference to the being and non-being of those objects, are the two properties that define discourse and reveal its essential character.

I. The Impasse over Non-Being

If we look at the ways in which the analytic and phenomenological traditions have construed the apparent capacity of discourse to refer to what does not exist, we find that they have come to terms with this phenomenon in ways that not only are characteristically different but also lead to a kind of fundamental impasse in the area of methodological assumptions. For, whereas the phenomenologists descended from Husserl regard this apparent indifference of discourse to the physical world as real and a fundamental given, the analysts

¹ See Thomas Litt, Les corps célestes dans l'univers de saint Thomas d'Aquin (Paris: Nauwelaerts, 1963),

descended from Russell regard it as a mere appearance, to be explained away with the help of the techniques of mathematical logic.

But while the programmatic statement and detailed working out of these two opposed programs is owing, respectively, to Husserl's theory of intentionality and Russell's theory of descriptions, the basic inspiration for both programs came from somewhat earlier background figures--Franz Brentano (1838-1917) and Gottlob Frege (1848-1925).

The terms of the later phenomenologists' attitude toward the possible non-being of the objects of discourse were set, as it turned out, in the original text wherein Brentano introduced the notion of intentionality into contemporary discussion:

Every mental phenomenon is characterized by what the scholastics of the Middle Ages called the intentional (and also mental) inexistence (Inexistenz) of an object (Gegenstand), and what we could call, although in not entirely unambiguous terms, the reference to a content, a direction upon an object (by which we are not to understand a reality in this case) or an immanent objectivity.²

Thoughts or mental acts, thus, are allegedly distinguished by virtue of placing the one who has or makes them into a relation with objects, regardless of the factual status--the " reality "-- of those objects. This alleged characteristic of the mental, subsequently known as the property of intentionality, became for Edmund Husserl--as Spiegelberg puts it--" the central insight in his phenomenological analysis of consciousness." " From now on," i. e., after Husserl, " the expressions ' intentional ' and ' intentionality ' stood for the relational property of having an intention, or being aimed at by it." ³

There we have the guiding view from which phenomenology comes to terms with the apparent capacity of discourse to refer

² Franz Brentano, " The Distinction between Mental and Physical Phenomena," trans. by D. B. Terrell, in Realism and the Background of Phenomenology, ed. by Roderick Chisholm (Glencoe, Ill.: The Free Press, 1960), p. 50.

³ Herbert Spiegelberg, The Phenomenological Movement (2nd ed., rev.; The Hague: Martinus Nijhoff, 1965), Vol. I, p. 107.

to what does not exist in fact. Thought itself, by virtue of its essential intentionality, constitutes the relation to its objects in the same way whether or not those objects have a further existence in fact. Since language is but the expression of thought, it is not to be wondered at that we can refer in speaking to what does not exist: the indifference of thought to being and non-being, which language merely records outwardly, and which constitutes the capacity of discourse to refer to what does not exist, is an immediately evident feature of common experience, a primary datum, a fundamentum inconcussum that is to phenomenology what the Cogito was to Cartesianism. That we can refer in discourse to what does not exist in fact, in short, is not something problematic for phenomenology--something that, over and above being noticed and named as of the essence of thought, calls for an explanation of its possibility. It is rather what determines the problematic of phenomenology and fixes the horizon within which phenomenological philosophy moves. The description and analysis of consciousness and experience generally begins from this very fact: discourse is not bounded by the world of physical realities.

Brentano himself did not accept the consistent and purified development of his doctrine of intentionality in the hands of Husserl and the phenomenologists. Indeed, in first introducing the notion, he had had his reservations. From the outset he looked with suspicion on the proposition that objects of discourse need not be real, and consequently, even on the proposition that discourse is genuinely relative to objects. In the case of discourse, seeing that " the term of the so-called relation need not be given at all in reality," he wrote, " one might doubt that we are here really dealing with something relative, and not rather with something only apparently relative, which one might accordingly call a relation-like thing." ⁴

It was the doubt voiced here that finally won out in Brentano's mind over the more straightforward doctrine of intentionality.

⁴ Franz Brentano, Psychologie vom empirischen Standpunkt (Hamburg: 1955 and 1959), Vol II, p. 134.

"All mental references refer to things,"⁵ he finally concluded, and the things referred to are really existing things, concrete individuals. The apparent capacity of discourse to refer to what does not exist, he held finally, is merely apparent--fictitious, not real. Pieces of discourse ostensibly referring to non-existent objects are but shorthand or abbreviations that would, on a fuller and more careful statement, be found to consist of a complicated discourse " whose terms refer," as Chisholm puts it, " only to ' genuine objects '--to individual concrete things."⁶ All mental references, appearances to the contrary notwithstanding, refer to things.

This is not to deny that in many cases the fiction that we have, as an object, something other than a real entity--for example, that that which lacks being as such, as well as that which has it, may be an object--proves itself harmless in logical operations; indeed, by means of this fiction these operations can be facilitated, because they are simplified in expression and even in thought itself. It is similar to the way mathematicians are accustomed to use with advantage the fictions of numbers less than zero, and many others. By this method a presentation and judgment, complicated in various ways, permit themselves to be handled as if they were simple, and one is spared the trouble (which is useless in some cases) of clarifying more exactly a confusedly grasped mental event. . . .

The fact that such fictions are useful in logic has led many to believe that logic has non-things as well as things as its object and, accordingly, that the concept of its object is more general than that of the real. This is, however, thoroughly incorrect; indeed, according to what has been said, it is downright impossible, for there cannot be anything at all other than real objects, and the same homogeneous concept of the real, as the most general concept of all, comprehends everything which is truly an object. Also, the terms of ordinary language are most often not psychologically, but only grammatically, names. They do not name things, but it remains none the less true that the discourse in which they are involved is concerned with nothing other than things.⁷

⁵ Brentano, " Genuine and Fictitious Objects," trans. by D. B. Terrell, in Realism and the Background of Phenomenology, p. 71.

⁶ Editor's Introduction to Realism and the Background of Phenomenology, p. 5.

⁷ Brentano, " Genuine and Fictitious Objects," loc. cit., p. 75,

Thus the Brentano who inspired Husserl is someone quite different from the Brentano who philosophized in his own right. The original statement of the doctrine of intentionality, purified of ambiguities and rendered consistent by Husserl and others, affirmed that discourse really does just what it appears to do--make reference to objects regardless of whether those objects exist in fact as concrete individual things. Brentano, however, finally and firmly denied the reality of discourse's apparent capacity to refer to the non-existent. Sentences appearing to refer to non-existent objects, he held, are translatable with the loss only of convenience and confusion into other sentences that refer exclusively to objects existing in fact.

This later view of Brentano, little known in comparison with his earlier doctrine of intentionality, is substantially the same as the view that early came to the fore in the analytic tradition's dealings with the apparent capacity of discourse to refer to the non-existent. The analysts, however, unlike the phenomenologists, do not owe to Brentano the psychologist their initial inspiration for how to come to terms with the nonexistent. The analysts found their clue rather in the writings of Frege the logician, who had maintained from the first that " it is a defect of languages that expressions are possible within them, which, in their grammatical form, seemingly determined to designate an object, nevertheless do not fulfill this condition in special cases; because this depends on the truth of the sentence." ⁸

There are two aspects to Frege's position on this matter. First, and most important for the subsequent developments in analysis, is Frege's firm identification of the apparent capacity of discourse to refer to what does not exist as a " defect," "fault," or "imperfection," together with his programmatic statement--which we will come to in a moment--of how this defect might be remedied, this imperfection eliminated. Second,

⁸ Gottlob Frege, " On Sense and Nominatum, " trans, by Herbert Feigl, in Contemporary Readings in Logical Theory, ed. by I. M. Copi and J. A. Gould (New York: Macmillan, 1967), p. 85. Cf. Max Black's trans. of this same proposition in The Philosophical Review, LVII (May, 1948), p. 222.

and of equal interest from the standpoint of logical theory, but of lesser interest from the standpoint of a philosophy of discourse, is the non-programmatic, ad hoc means chosen by Frege in his own work to compensate for this alleged " defect " or " imperfection of language of which, by the way, even the symbolic language of analysis is not entirely free·" ⁹

With regard to the first point, Frege not only leaves no doubt that, so far as he is concerned, reference to what does not exist is a " defect of languages," an " imperfection," and " a major source of fallacies," but he also implies that it could and should be eliminated from discourse--at least from discourse that has been logically perfected--for once and all:

It is to be demanded that in a logically perfect language (logical symbolism) every expression constructed as a proper name in a grammatically correct manner out of already introduced symbols, in fact designate an object; and that no symbol be introduced as a proper name without assurance that it have a nominatum. It is customary in logic texts to warn against the ambiguity of expressions as a source of fallacies. I deem it at least as appropriate to issue a warning against apparent proper names that have no nominata. The history of mathematics has many a tale to tell of errors which originated from this source. The demagogic misuse is as close (perhaps closer) at hand as is the case of ambiguous expressions. ' The will of the people ' may serve as an example in this regard; for it is easily established that there is no generally accepted nominatum of that expression. Thus it is obviously not without importance to obstruct once for all the source of these errors, at least as regards their occurrence in science. Then such objections as the one discussed above will become impossible, for then it will be seen that whether a proper name has a nominatum can never depend upon the truth of a proposition.¹⁰

But, having seen and said this much, Frege fails to take the further step of actually finding the means " to obstruct once and for all the source of these errors " that creep in to our theoretical and practical affairs through the door of non-being. Instead, Frege adopts an ad hoc solution to the problem based

⁹ Ibid.

¹⁰ Ibid., p. 86.

only on reasons of convenience. " Whenever something is asserted," he says, whether by means of ordinary discourse or of the symbolic language of analysis, "then the presupposition taken for granted is that the employed proper names, simple or compound "--i. e., the linguistic means employed, be it a word or a sentence--" have nominata," i. e., refer to objects existing in fact.¹¹ This being the case, even though " we may be in error as regards that assumption, and such errors have occurred on occasion," Frege considers that " it will suffice for the moment to refer to our intention in speaking and thinking in order to justify our reference to the nominatum of a sign, even if we have to make the proviso: if there is such a nominatum." ¹² Frege then proceeds to circumvent the demand that a logically perfected language designate only objects existing in fact by adopting a convention that allows him to postulate what an object referred to shall be in a given case.

It is the privileged position of Bertrand Russell in the analytic tradition to have faced the recognized demand squarely and to have refused to be satisfied with anything less than a logical device that would in principle allow for the elimination from discourse of any reference to what does not exist, and so would remove from discourse the deceptive appearance of being able to deal indifferently with being and non-being. The measure of any philosopher is to see clearly the requirements of an issue, to choose among the alternatives according to the end in view, and to work out in principle the requirements of the chosen alternative to the point where the proportion between means and end comes unmistakably into view. By this criterion, it is easy to see why Russell is in fact the most important single figure in the tradition of analytic philosophy, and why it is from him, rather than from Frege, that the analytic tradition derives its distinctive heritage, even as the phenomenological tradition stems from Husserl rather than Brentano. Brentano and Frege planted seed in stony ground.

¹¹ Ibid., p. 85. See also p. 77.

¹² Ibid., p. 79.

Husserl and Russell transplanted the seeds, germinated them, and made them blossom.

Given the apparent indifference of discourse to being and non-being, and given the assumption that this appearance is merely an appearance and not the reality of the matter, Russell saw clearly both the required alternative to Frege's treatment of the alleged logical imperfection of language and (what is partly the same) the radical inappropriateness of Frege's means to the desired end. Consider the cases in which the denotation of denoting phrases used in propositions appears to be absent.

Now it is plain that propositions do not become nonsense merely because their hypotheses are false. The King in " The Tempest " might say, " If Ferdinand is not drowned, Ferdinand is my only son." Now " my only son " is a denoting phrase, which, on the face of it, has a denotation when, and only when, I have exactly one son. But the above statement would nevertheless have remained true if Ferdinand had been in fact drowned. Thus we must either provide a denotation in cases in which it is at first sight absent, or we must abandon the view that the denotation is what is concerned in propositions which contain denoting phrases. The latter is the course that I advocate. The former course . . . . is adopted by Frege, who provides by definition some purely conventional denotation for the cases in which otherwise there would be none. Thus " the King of France," is to denote the null-class; " the only son of Mr. So-and-so " (who has a fine family of ten), is to denote the class of all his sons; and so on. But this procedure, though it may not lead to actual logical error, is plainly artificial, and does not give an exact analysis of the matter.¹³

"Taking the latter course," i.e., finding a way to analyze propositions without having to allow--whether by convention (Frege) or by confusion (the later Brentano)--for anything unreal among the objects referred to, is the genius of Russell's celebrated theory of descriptions and the reason why it is justly described, in F. P. Ramsey's words, as " a paradigm of philosophy."

¹³ Bertrand Russell, " On Denoting," in Logic and Knowledge, ed. by R. C. Marsh, pp. 46-47.

Russell's theory of descriptions first appeared in Mind of 1905 in the form of an article titled "On Denoting." The most technical and logically simple exposition of the theory is found, as one might suspect, in the Principia Mathematica, specifically, in the third introductory chapter, " Incomplete Symbols." The most philosophically important exposition of the theory, however, at least insofar as philosophy is more concerned with underlying assumptions and principles than with the resolution of problems formed into definite conclusions, is to be found in chapters 15-17 of his Introduction to Mathematical Philosophy. For, from the point of view of the assumption that discourse is only apparently and owing to confusion on the part of its users indifferent to the being and non-being of its objects (rather than really and essentially thus indifferent) , we may grant to Russell not only that " language is misleading, as well as. . . . diffuse and inexact," but also that " logical symbolism is absolutely necessary to any exact or thorough treatment " of the objects of discourse, " in particular as regards existence and descriptions."¹⁴ But if we wish to understand the motives for adopting this particular assumption, we must do so apart from the particular conception of the so-called " symbolic " or " mathematical logic " that depends for its philosophical applicability on one's having already assented to the view that a reference to what does not exist is only apparent and not real or " true." The exposition of the theory of descriptions in the Principia presupposes the truth of this assumption; but the less technical exposition in the Introduction to Mathematical Philosophy, precisely because it is not " only available to those who have mastered logical symbolism," is required to make explicit just what is involved in making the crucial assumption upon which is based, as Russell puts it, " the method by which mathematical logic can be made helpful in investigating the traditional problems of philosophy." ¹⁵

¹⁴ Bertrand Russell, Introduction to Mathematical Philosophy (New York: Humanities, 1919), p. 205.

page 263

What one finds is that an uncompromising passion for reality and truth lies behind Russell's formulation of his theory of descriptions:

In obedience to the feeling of reality, we shall insist that, in the analysis of propositions, nothing " unreal " is to be admitted. But, after all, if there is nothing unreal, how, it may be asked, could we admit anything unreal? The reply is that, in dealing with propositions, we are dealing in the first instance with symbols, and if we attribute significance to groups of symbols which have no significance, we shall fall into the error of admitting unrealities, in the only sense in which this is possible, namely, as objects described.¹⁶

Thus Russell clearly recognizes that non-being enters into history only through discourse and does so in the form of objects. Wars have been fought, lives have been lost, and endless miseries have been inflicted upon men because of non-being parading itself--thanks to discourse--in the guise of being. Might it not be, then, that a correct analysis of discourse would put an end to the tyranny of non-being over human affaire and " obstruct once for all," as Frege proposed, " the source of these errors, at least as regards their occurrence in science " ? Such a goal goes beyond the narrow concerns of logic and is surely worthy of the philosopher. It is the way to this goal that Russell set himself to find. " My theory of descriptions was never intended as an analysis of the state of mind of those who utter sentences containing descriptions," Russell testifies; " I was concerned to find a more accurate and analysed thought to replace the somewhat confused thoughts which most people at most times have in their heads."¹⁷

Consider again the apparent indifference of discourse to being and non-being in its relativity to objects:

The question of " unreality," which confronts us at this point, is a very important one. Misled by grammar, the great majority of those logicians who have dealt with this question have dealt with it on mistaken lines. They have regarded grammatical form

¹⁶ Ibid., p. 170.

¹⁷ Bertrand Russell, "Mr. Strawson on Referring," Mind, LXVI (July, 1957), p. 388.

as a surer guide in analysis than, in fact, it is. And they have not known what differences in grammatical form are important. " I met Jones " and " I met a man " would count traditionally as propositions of the same form, but in actual fact they are of quite different forms: the first names an actual person, Jones; while the second involves a propositional function, and becomes, when made explicit: " The function ' I met and is human ' is sometimes true." . . . This proposition is obviously not of the form " I met x," which accounts for the existence of the proposition " I met a unicorn " in spite of the fact that there is no such thing as " a unicorn." ¹⁸

Thus, by the simple expedient of distinguishing between propositions and propositional functions, Russell has, in principle, a means for eliminating from discourse all reference to what does not exist. Whenever a statement appears to refer to a non-existent object and to assert something about it, that statement must not be taken, as traditional or "Aristotelian " logic would take it, as expressing a simple proposition; it must be taken, rather, as the symbolic or " mathematical " logic would take it, as expressing a complex combining a proposition and a propositional function together into what, grammatically considered, appears to be a simple proposition but really is not. Thus all parts of discourse, seemingly constituted by propositions about unreal objects, are really compounds of propositional functions and false propositions about the real world, and this is what the analysis of such propositions by the techniques of symbolic logic reveals, though it is what analysis by the techniques of traditional logic conceals and glosses over. " For want of the apparatus of propositional functions," Russell concludes, " many logicians have been driven to the conclusion that there are unreal objects."¹⁹ This conclusion--or confusion--can henceforward, in principle at least, be avoided, now that the propositional function has been discovered (or perhaps we should say: invented). " For clear thinking, in many very diverse directions, the habit of keeping propositional

¹⁸ Russell, Introduction to Mathematical Philosophy, pp. 168-169.

¹⁹ Ibid., p. 169.

functions sharply separated from propositions is of the utmost importance, and the failure to do so in the past has been a disgrace to philosophy." ²⁰ It is with an eye to the absence of this distinction in traditional logic that Russell claims of the modern truth-value logic that " the description of the subject as symbolical logic is an inadequate one. I should like to describe it simply as logic, on the ground that nothing else really is logic." ²¹

Let us grant the ingenuity of Russell's device whereby, as Harman remarks of Quine's developments of it, " talk about ideal objects is reduced to loose talk about real objects." ²² Let us grant, too (at least for purposes of the present discussion), Russell's claim that it is through the habit of distinguishing propositions from prepositional functions that the modern, symbolic, or mathematical logic achieves its decisive superiority over the traditional or Aristotelian logic.

It remains that the theory of descriptions does not exclude non-beings from among the objects of discourse but merely enables us to eliminate in a technical manner all references to objects which we believe, for reasons of our own, to be unreal. Of course, our belief in point of any given object or class of objects might be mistaken (heaven forbid). But there is nothing within the system of modern logic--or of traditional logic, for that matter--that enables us to know that the beliefs about reality, upon which our regimentation or formalization of some theory is based, are correct. If they are not, though our formalization will eliminate from discourse all apparent reference to objects other than those we assert to be real, the discourse will yet contain reference to unreal objects. Even if our beliefs happened to be completely correct, apart from the fact that we would have no way of knowing this, it would remain the case that our formalized discourse would still appear and be taken to refer to unreal objects by anyone not sharing

²⁰ Ibid., p. 166.

²¹ Bertrand Russell, " The Philosophy of Logical Atomism," in Logic and Knowledge, p. 271.

²² Gilbert Harman, " Quine on Meaning and Existence, II," The Review of Metaphysics, XXI (December, 1967), p. 361.

our beliefs. How one will formalize or " regiment " (express in the quantificational terms of modern logic) any given philosophical theory, therefore, will depend: (1) on what one understands that theory to assert to be real; (2) on what one believes concerning those assertions.

Suppose, for example, two modern logicians, one who believes exactly like Paul VI and one who believes like Jean-Paul Sartre, set out to formalize the Summa of St. Thomas. First, let us suppose that they would each formalize it in terms of the various entities they took Thomas to assert to be real. And then they would each formalize it a second time, but now according to what they themselves take to be real. In the first case, the quality and worth of the formalizations would exactly match the quality and worth of their respective understanding of the doctrine of St. Thomas and would differ accordingly. In the second case, the manner of binding the variables--referring to objects--would differ wildly between the two men, not according to their different understanding of St. Thomas now but according to their radically opposed conceptions of what there is to reality. In both cases the techniques of formal logic serve no more than to express differences in understanding; they provide no means whatever for adjudicating or eliminating those differences. The extent to which mathematical logic " appears," as Russell thought, " to invalidate much traditional philosophy, and even a good deal of what is current in the present day," ²³ is a deceptive appearance indeed. For it has its whole being from the extent to which a given user of symbolic logic disagrees--for whatever reasons, and rightly or wrongly--with the traditional or current views, and in terms of which disagreement he gives symbolic expression to these views in order to " invalidate " them.

Subjectivity, in the form of the beliefs of the user, is built in to the very foundations of the logistic method.²⁴ All the

²³ Introduction to Mathematical Philosophy, p. xii.

²⁴ This point is well remarked by Guido Küng, Ontology and the Logistic Analysis of Language (rev. ed.; New York: Humanities, 1967), pp. 8-9: " Because they are artificially made to conform to a logical ideal and because of the precision of their rules, logistic languages present a completely different case than do the natural languages. We still find a wide range of possible syntactical systems or ' grammars ". . . but these consciously contrived logistic languages are all alike in the fact that they are determined explicitly by the definite logical and ontological views of their inventors.

" The translation from a natural into a logistic language, i. e., the logistic analysis of language, involves an ontological commitment for every sentence translated."

properly philosophical differences over what there is precede any use that can be made of the method. Far from providing a tool for the adjudication of philosophical disputes, philosophical discussion ceases to be possible as soon as and as long as one has recourse to the logistic method, for that method cannot be employed at all except to the extent that one has decided--rightly or wrongly--what a given theory-to-be-formalized asserts to be real.

Carnap's distinction between questions internal and external to a framework or language-system is useful here. Prior to or independent of the selection of a language-system L with a domain D of objects as its fundamental domain, the quantifiers are without meaning. They are given meaning only with the formulation of L, i. e., by the explicit listing of the syntactical and semantical rules determinative of L.²⁵

For purposes of expressing philosophical differences over the fundamental units of being, modern logic achieves, at least initially, unparalleled clarity. For purposes of coming to understand the reasons behind those differences, however (which is, after all, the most proper concern of the philosopher), modern logic serves no essential use at all, for the reason finely stated by Alston: " Just as no sentence is necessarily misleading, so none is guaranteed, by its form, to be used without confusion. The supposition to the contrary is one of the unfortunate effects of philosophic preoccupation with artificial languages."²⁶

²⁵ R. M. Martin, " Existential Quantification and the ' Regimentation ' of Ordinary Language," Mind, LXXI (October, 1962), pp. 528-529.

²⁶ William P. Alston, " Ontological Commitments," in Philosophy of Mathematics, ed. by P. Benacerraf and H. Putnam (Englewood Cliffs, N. J.: Prentice-Hall, 1964), p. 257. Reprinted from Philosophical Studies, 9 (1958), pp. 8-17.

The apparatus of symbolic logic, as incorporating Russell's theory of descriptions, therefore, does not solve the problem of non-being in discourse any more than does the phenomenological method based on Husserl's theory of intentionality. The phenomenologists' doctrine offers no explanation of the appearance of non-being in discourse. But the analysts' doctrine, while it does show the possibility in principle of eliminating non-beings from among the objects of discourse, also signally fails to show how in fact this elimination could be successfully carried through, for, to borrow again Alston's words:

In any context where questions of existence arise the problem is whether or not we shall assert that so-and-so exists, not whether we shall choose some particular way of making this assertion. This means that assertion of existence, commitment to existence, etc., does not consist in the inflexible preference for one verbal formulation over any other, however gratifying such preferences may be to logicians.²⁷

Yet this is short of saying that the whole enterprise of mathematical logic has been, after all, without substantive philosophical point. When the delusions of grandeur generated by its practitioners have been seen for what they are, it can also be seen that " the point of the translation " of any given piece of discourse into symbolic form, is, in Alston's phrase,²⁸ " essentially a strategic one," specifically, it enables us to neutralize, " wherever they arise," confusions as to what we are affirming to be real, insofar as such confusions might arise from the ambiguities of common language; for, by translating the problematic statements into canonical notation, we can identify the precise locus in language of the ontological commitment we wish to make.²⁹ At the same time, this strategy

²⁷ Ibid., p. 254.

²⁸ Ibid., p. 257.

²⁹ In addition to this " advantage," Küng, in Ontology and the Logistic Analysis of Language, pp. 185-186, adds two others equally thin and dubious--the making possible of " a new kind of pragmatical evaluation " of ontological standpoints (an advantage which even Küng does " not wish to claim " as " of primary importance "), and a sharpening and increase of our ontological knowledge (which does not seem to me a defensible claim, for the reasons already given).

enables us to escape from the triviality, well-remarked by Quine,³⁰ of the ordinary language wing of analytic philosophy with its tacit claim that the ontological commitments apparent in common usage are the ontological commitments philosophy is bound to accept as its own. For what Russell achieved was a logical system based on the philosophical truth that, for a sufficiently informed observer--more exactly, for an omniscient one--it would be possible to so correct human discourse as to eliminate all mistaken, references resulting from the confusions of non-being with being.³¹

Thus the impasse: where phenomenology sees the inexplicability of the self-evident, analysis sees confusion and something to be explained away. But the ' self-evident ' phenomenon to which phenomenology appeals-- the relational character or " intentionality " of thought whereby it relates thinkers to objects thought of indifferent to the reality or unreality of those objects--is at least as puzzling as it is evident: for how can there be a relation between being and non-being, between a thought, which exists in a thinking subject (be it only as a brain state), and an object, which exists nowhere? How can there be a relation connecting what exists with something that does not exist, a relation whose term is not given at all in reality?

This is hardly a question about discourse which phenomenology is entitled to ignore. It will not do, without a word of explanation, to accept such a bizarre phenomenon as patently the case. A word of explanation is in order; yet phenomenology gives us only descriptive analyses which presuppose the validity of what is here up for question.

On the other hand, analysis, in seeking to explain away the problem of discursive reference to the non-existent, fails to provide a resolution any more satisfactory. For in denying

³⁰ W. V. O. Quine, " Philosophical Progress in Language Theory," Metaphilosophy, 1 (January, 1970), p. 18.

³¹ This truth, argued in a slightly altered formula, is one of the " theses " of classical Thomism: see John of St. Thomas (né Jean Poinsot), Cursus Philosophicus, ed. by Reiser (Turin: Mariett, 1930), Vol. I, Ars Logica, Part , Q. II, Art. V, " Utrum Deus Formet Entia Rationis."

that anything but confused discourse can appear to refer to what does not exist, the analysts fail to provide any standard for what clear discourse would be other than discourse which makes reference only to real objects. The elaborate program for such logically perfect discourse, called for by Frege, worked out in its fundamentals by Russell (with his theory of descriptions) and finally by Quine (with the elimination from Russell's theory of the last vestige of uncontrolled reference, the singular term³²), is little more than an elaborate circularity and proves in the end unworkable: for the program depends for its successful implementation on the prior judgments made by those who implement it concerning the reality of any given object of discourse, and yet it is just these prior judgments concerning the line between being and non-being that is the very point at issue between any two scientific or philosophical theories. Since the original problem of non-being for a philosophy centered on discourse concerns the capacity of discourse to deceive us in point of the reality of objects, the elaborate program of the logicians, so far as it holds any properly philosophical prospect or interest, proves to be a program that would only be realized after there was no longer a need for it. For if the problem is to decide why there is a problem as to which (and in what way) objects of discourse are real, the predicate calculus of modern logic provides no help whatever. All that it can do is to enable us to eliminate the references to the unreal objects once it has been decided what the real objects are. But how it is that unrealities appeared or seemed to appear among the objects of discourse in the first place remains as much in the dark as ever. The ex post facto account of non-being by the analysts, I conclude, is in its own way as unsatisfying as the silent appeal to self-evidence by the phenomenologists.

³² W. V. O. Quine, " Descriptions," and " Elimination of Singular Terms," in Methods of Logic (rev. ed.; New York: Holt, Rinehart, & Winston, 1959), pp. 215-219, and 220-224. See also Quine, " Notes on the Theory of Reference," in From a Logical Point of View (2nd ed., rev.; New York: Harper, 1961), pp. 130-138; and " Variables Explained Away," in Selected Logic Papers (New York: Random House, 1966), pp. 227-235.

Surely there must be yet a third way of coming to terms with the apparent capacity of discourse to refer to what does not exist, which will bridge the gap between these dominant contemporary approaches.

II. To Overcome or to Acquiesce in the Impasse?

We have found in the non-being of certain objects of discourse a problem common to any philosophy of discourse, and we have found further that the way of handling this problem in the currently dominant species of such philosophy has resulted in a kind of stand-off that leaves us as much in the dark as ever as to how it is that discourse has the appearance of indifference to the reality of its objects.

But perhaps it is our guiding question that leads us inevitably into blind alleys. Perhaps it is simply a mistake to seek for an explanation of the apparent indifference of discourse to the reality of its objects, a mistake based on a misunderstanding of the way language functions--the purpose it fulfills in human life. Within the analytic tradition a view just such as this was developed by Wittgenstein in reaction to the logical excesses in the analysis of language inspired by Russell's theory of descriptions--" as if," Wittgenstein remarked scornfully, " it took the logician to show people at last what a correct sentence looked like." ³³

It was the very requirement for a logically perfect language laid down, as we have seen, by Frege, that Wittgenstein held suspect. " The more narrowly we examine actual language," he noted, " the sharper becomes the conflict between it and our requirement."³⁴ Actual language is simply not tied down to any one concept of " reality " in the expressions of common life. Attention to this fact, thought Wittgenstein, should lead us to abandon " the preconceived idea of crystalline purity "

³³ Ludwig Wittgenstein, Philosophical Investigations, trans, by G. E. M. Anscombe (3rd rev. ed.; New York Macmillan, 1958), no. 81, p. 38 (reading " correct " for " richtiger," instead of Anscombe's " proper ").

³⁴ Ibid., no. 107, p. 46.

as logically required for authentic discourse, " by turning our whole examination around."³⁵ What is called for is neither an explanation showing how reference to what does not exist is possible, nor an explanation showing that reference to what does not exist is but a confused reference to what does exist. What is called for, rather, is the abandonment of the very aim of philosophy to explain discourse at all.

We must do away with all explanation, and description alone must take its place. And this description gets its light, that is to say its purpose, from the philosophical problems. These are, of course, not empirical problems; they are solved, rather, by looking into the workings of our language, and that in such a way as to make us realize these workings: in despite of an urge to misunderstand them. The problems are solved, not by giving new information, but by arranging what we have always known. Philosophy is a battle against the bewitchment of our intelligence by means of language.³⁶

This view of Wittgenstein has been widely adopted, and has resulted in a split of the analytic tradition into two factions or schools--the logical analysts, who look to mathematical logic for the tools of philosophical work; and the ordinary language analysts, who look to the actual workings of common speech as the proper object of philosophical consideration. Where the logical analysts think up " protocols " for translating this or that sort of discourse into logistic systems, Wittgenstein counsels the ordinary language philosophers to seek a different route: " don't think, but look! " ³⁷ And where logical analysts look for real objects in terms of which to implement their translation procedures, Wittgenstein counsels the ordinary language philosophers to subordinate such concern to a more fundamental goal--the elucidation of usage. " Philosophy simply puts everything before us, and neither explains nor deduces anything."³⁸ It " may in no way interfere with the actual use

³⁵ Ibid., no. 108, p. 46.

³⁶ Ibid., no. 109, p. 47.

³⁷ Ibid., no. 66, p. 31.

³⁸ Ibid., no. 126, p. 50.

of language; it can in the end only describe it. For it cannot give language any foundation either." ³⁹

Wittgenstein's counsels here have prima facie plausibility, an intuitive soundness about them, as it were, in the face of the changes wrought in common speech by use of the theory of descriptions (changes that even Russell allowed to be " somewhat incredible "⁴⁰), and even in the face of the sometimes bizarre formulations of the phenomenologists. The earlier analysts and the phenomenologists latched onto the referential aspect of discourse as though this were the whole--the former, by making reference the key for logical translation; the latter, by making reference the key to consciousness itself as something irreducibly " mental " or " intentional." But the ordinary language philosophers make it clear from the start that referential use of words is only one feature of discourse; non-referential use, however, is another feature of discourse that must not be neglected. Why then subordinate the latter to the former? Why not rather simply regard them as equal partners, each instancing in different ways the primary feature of language--service in public life? The use of words to refer to what does not exist, then, is neither something to be explained, nor is it something to be explained away. It is simply a matter to be clarified through attention to the actual workings of language, according to the perfectly general formula, " the meaning of a word is its use in a language." There are words, and there are the uses to which they are put in discourse--that is all the philosopher of discourse knows or needs to know. When he knows this, he knows all there is for philosophy to know about language, save for details.

What goes on " inside the heads " of the users of language may be safely left to the psychologist and neurologist, just as " what there is " outside of language may be safely left to the other sciences. When the philosopher has mapped the usages of language, he has done all that he can or can be expected to do.

³⁹ Ibid., no. 124, p. 49.

⁴⁰ Russell, " On Denoting," loc. cit., p. 44.

On this view, the shift in interest from being to discourse that has taken place in modern times signals more than just a new perspective on ancient problems. It heralds a radically new age, a clean break with what has traditionally been known as philosophy. As Richard Rorty puts it, " analytic philosophers have in common the view that the pursuit of wisdom cannot be served by continuing the inquiries traditionally grouped together as ' philosophy'." ⁴¹ Russell and the logical analysts thought they had discredited traditional philosophies of being by the simple application of their new logical method to the old problems. But this opinion, we now see, was without foundation, inasmuch as the application of the new method to such an outcome is possible at all only to the extent that the outcome is presupposed in the very application. The ordinary language philosophers avoid this vicious circle by the simple expedient of eschewing all explanation. " Philosophy " becomes, as Wittgenstein put it, "what is possible before all new discoveries and inventions." ⁴²

From the point of view of our question, however, there is little to choose between (1) the logistic translations made by the logical analysts to conjure reference away from discourse, (2) the elucidations of usage made by the ordinary language analysts in precision from all questions concerning the reality or unreality of objects, and (3) the descriptions of consciousness or experience essayed by the phenomenologists on the postulate of intentionality: all three shed no light at all on how it can be that discourse seems to put us in a relation to what does not exist with as much facility as it brings us into relation with what does exist.

The position of ordinary language analysis alone confronts and does not conceal the question, even though it does so only to deny the legitimacy of the inquiry. This denial may seem arbitrary from the point of view of a traditional philosophy of

⁴¹ Richard Rorty, " Do Analysts and Metaphysicians Disagree? ", ACPA Proceeedings, XLI (1967), pp. 39-53.

⁴² Wittgenstein, Philosophical Investigations, no. 126, p. 50.

being. And it must be admitted that it is by no means clear either how empirical research of the sort that characterizes the sciences could ever provide the answer to our question, or how the question itself is based on a misunderstanding of the workings of language--particularly in view of the fact that it is the workings of ordinary speech itself that give rise to our question. Is it not the case that discourse does proceed with an apparent indifference to the reality of its objects? And if the possibility of so proceeding is not explained by any avenues of empirical inquiry, is it not arbitrary, or at least indecently hasty, to assert baldly that no explanation--no answer to our question--is possible or desirable?⁴³

Still, the stand of the later Wittgenstein and his school on the question of meaning does serve to make clear a key requirement any answer to our question must meet: it must establish a nexus between being and discourse, a point of contact or " perspectival overlap " between a philosophy of being transparent to itself and a philosophy of discourse aware of the capacity of speech to convey and designate reality.

If Wittgenstein is right, if the clarification of ordinary language in a manner that neither explains nor infers anything is all that philosophy could ever be, then there is no way out of the impasse reached over the issue of non-being by the phenomenologists and the analysts. And the reason why there is no exit is clear: to overcome the impasse, an explanation of how language really refers to objects would be necessary, and no explanation is possible within philosophy.

On the other hand, if an explanation of how language really refers to objects is possible, and if such explanation does not

⁴³ " Indecently hasty " is about as charitable a description as justice authorizes regarding Wittgenstein's pronouncements on the nature and history of philosophy. For in fact, Wittgenstein's knowledge of ancient philosophy consists almost solely in an acquaintance with Plato's Dialogues, while his knowledge of medieval thought consisted principally in a quite limited familiarity with St. Augustine. Read against the immediate background of Russell's work and his own Tractatus, Wittgenstein's Investigations have an authentic ring. Read as an adequate assessment of the philosophical enterprise, however, they have rather the ring of gratuitously assumed generality.

depend on the laboratory studies, field researches, or mathematical hypotheses of science, then a way out of the impasse is possible, the ordinary language analysts are mistaken in excluding all explanation from the domain of philosophy, and the opposition between the traditional philosophy of being and the modern philosophy of discourse will--in principle at least-- have been overcome. For whatever will really explain the capacity of discourse to refer equally to being and non-being, will by that very achievement establish a nexus or point of contact between the universe of being and the universe of discourse.

III. Overcoming the Impasse.

To the best of my knowledge, no ex professo account of the apparent indifference of discourse to the reality or unreality of its objects has ever been given by any philosopher. Of course, there may be such a treatise of which I am simply not aware; but what is certain is that, after four years of investigating the literature on language and philosophy--in the course of which this question only gradually took on clarity and central importance--I came across no full account of the relativity of discourse to objects and its indifference to being and non-being. The closest thing I found to the required treatment, curiously enough, is a 17th century Treatise on Signs, essayed in Latin by a singularly obscure Iberian philosopher, Jean Poinsot. The problem of explaining the referential capacity of discourse, according to Poinsot, is dependent upon a careful understanding of the nature of relation, more specifically, upon an understanding of the fact that relation is the only category or kind of being that need not be instantiated physically in order to really occur. Poinsot, for his part, gained this understanding of the peculiarity of relation from St. Thomas Aquinas. By taking Poinsot's application of St. Thomas's theory of relation to the theory of language, and by developing certain features of that application beyond Poinsot's own exposition, I find it is possible to arrive at an account of the apparent capacity of discourse to refer to what does not exist in

a way that meets, so far as I am aware of them, all the difficulties that led the later Brentano and both camps of analysts to eschew the relational or " intentional " view of discourse. In doing this, the account I am about to give overcomes the impasse reached by the phenomenologists and the analysts in their manner of coming to terms with non-being, by explaining precisely what the phenomenologists presuppose, namely, how it is possible for there to be a relation whose term is not given in reality apart from discourse. And finally, by providing such an explanation, my account establishes a nexus between the being-centered philosophies of the past and the discourse-centered philosophies of modern times, thereby restoring unity to the philosophical tradition and dignity to the philosophical enterprise as an avenue to understanding the world more and more clearly than is possible through the heritage of natural languages, even when this heritage is supplemented (as in the view of ordinary language analysis) or supplanted (as the logical analysts would have it) by knowledge acquired through the methods of experimental and mathematical science.

Let me develop my account of the referential capacity of discourse from a point that can be agreed to by all philosophers interested in discourse, be they analysts, phenomenologists, or whatever; then, in a separate section, I will show how my account owes its chief inspiration to the theory of the sign Poinsot developed on the basis of St. Thomas's insight into the peculiar nature of relation as a kind of being. The point of common agreement, which I take as providing a starting point for a new attack on the problem of non-being generated by ordinary discourse, has been satisfactorily formulated by A. J. Ayer. " Certainly there is a difference," he writes, " between understanding what another person says and merely hearing the noises that he makes."⁴⁴

No more than this need be admitted in order to get onto the way out of the impasse. Let us re-state Ayer's observation in

⁴⁴ A. J. Ayer, Thinking and Meaning (London: H. K. Lewis, 1947), p. 21. I am indebted to Mortimer Adler for the suggestion of this as a starting point.

the form of a question: In the case of a mark, sound, or movement--be it a word or a sentence, and be it verbally or graphically manifested--actually used to refer to an object, what is it that transforms the in itself physical mark, sound, or movement into a linguistic occurrence? My thesis is that in any case where discourse concerns an object, that is, in any case where heard or seen noises are understood linguistically, it is the difference between the mark, sound, or movement as such and as conveying the linguistic reference that is the key to understanding the nature and function of discourse, and that is, therefore, de jure, the fundamental datum explanandum for any philosophy of discourse that wishes to be grounded in principle. The apparent capacity of discourse to refer to what does not exist will be adequately judged as merely apparent or as real only to the extent that such a judgment is based on an understanding of this difference. To overcome the impasse over non-being, therefore, and to solve the problem of the difference between the physical as such and as making a linguistic reference, are one and the same.

Let us state the factors involved in this problem in neutral terms, i. e., terms that should be acceptable to any philosopher who accepts as a descriptive definition of meaning the difference--whatever it may be--between a given mark, sound, or movement as such and as sign of something other than itself, and accepts also (even if only tentatively) our proposal to make the explanation of meaning so defined the touchstone for deciding whether the apparent capacity of discourse to refer to what does not exist is real or merely apparent.

Whenever a mark, sound, or movement is employed successfully as a linguistic sign (i. e., to make a linguistic reference), whether that sign be a word or a sentence (i. e., whatever it is that one takes the unit of linguistic reference to be--word or phrase, subject of a sentence, whole sentence: whatever), it seems clear that its status as a piece of discourse involves three factors. First, there is the linguistic expression itself, which occurs outside the organism: let us call this the extraorganismic

factor X. Second, there is the object that X is a sign of, the significate of X, what X refers to or is about: let us call this the objective or signified factor O. Third, there is the factor within the organism using or apprehending X as a sign of O, the organism's " understanding," on the basis of which factor precisely X is, over and above a mere mark, sound, or movement occurring extraorganismically, a linguistic mark, sound or movement signifying or referring to O: let us call this the intraorganismic factor C, inasmuch as it is the cause, here and now, of X's actual functioning status as sign of the object O. For, if C does not occur in connection with the perception of X, clearly, X will not be perceived by the organism in relation to O as a sign thereof. It is thus C, the intraorganismic factor, that holds, in final analysis, the key to the analysis of meaning. I say, " in the final analysis," for it is clear that there are many ancillary considerations that may be pursued in connection with C--for example, the analysis of the conventions of use surrounding the occurrence of X as evocative of C (the province of ordinary language analysis) or the analysis of the behavior of the organism consequent upon the occurrence of C (the province of psycholinguistics and behavioral psychology) . But just as clearly, such pursuits are ancillary, and the heart of the problem lies in C itself. X functions as a sign of O on the basis of C, the intraorganismic factor whereby the language user apprehends O as object signified by X.

Notice here that the being of O as an object for the user of X is dependent upon the factor C. Even supposing that O exists in the world independently of C, it exists here and now as something actually signified by X dependently upon C. As an object of discourse, therefore, as something signified by X, O has its being dependently upon C. In the analytic tradition, the tendency has been to give physical objects pride of place precisely because here, at least, we tend to have confidence in the reality of that about which we discourse. But, from the standpoint of C, this tendency suppresses a fundamental insight. Whenever an object exists as an object of discourse, it is not its

supposed physical reality that is decisive, for the physically real object as physically real is thought to be such on the grounds that it seems to have being independently of discourse. But an object of discourse as such has its being dependently upon discourse, regardless of whether it also has an independent or physical being.

The dependence of O on C for its being an object of discourse raises immediately a problem whose importance cannot, I think, be exaggerated. Given that the factor C, regardless of whether it be called an " idea " or a brain state or a muscular disposition of some sort, is--as intraorganismic--a subjective or "private" factor; and given that C is the raison d'être of O so far as O actually enters into discourse, it must be asked at once: how is it that O does not participate entirely in the subjectivity of C? How is it that two language users of X, each of which has his own C, can have or seem to have in common the object O? They do not have the same C. They have or seem to have the same O. But O exists in discourse as an effect of C. Therefore it would seem that there are as many O s as there are C s, and that each O is private in just the way that each C is.

Yet such a conclusion flies in the face of common experience. We live in a public world, not a merely private one. Language is an instrument for sharing the world. Here all the arguments of Wittgenstein against the possibility of a private language are appropriate. We may take it as a touchstone for any sound account of discourse that it give an account of meaning as public and common, not as private and subjective. The existence of C and its necessity for actual discourse here and now, I think, cannot be gainsaid. But some account must be given which reconciles the privacy of C with the publicy of O, and of X as sign of O. Otherwise, our theory falls prey to all the standard and valid objections that have led analysts--both " ordinary " and " logistic " ones--to eschew any role for psychology and mental entities in the philosophy of discourse. How are we to reconcile the dependence of O (for its being

as object) and of X (for its being as sign of O) upon C, with the fact that O and X function publicly, while C functions privately? ⁴⁵

To get clear about what is at stake here, let us proceed for the moment on the assumption that the object is a really existing thing factually given in the physical world quite apart from discourse. The advantage gained by making this assumption is simply that in the stipulated case--the case where O exists in fact independently of discourse--it is clear that not only C and X are really distinct as well as really related-- (really related, that is, inasmuch as it is by virtue of C that the organism is aware of O), but also that distinction between O as something existing in the physical world--O as thing, let us say--and O as something apprehended by a given organism--O as object, let us say--is precisely the difference, for purposes of discourse, between O as related in its existence to a given organism and O as unrelated to that organism, a difference made by the presence or absence in the organism of C. When C exists, O exists as apprehended as well as in fact; when C does not exist, O exists in fact but not as apprehended. C, therefore, is the basis of the cognitive relation R between an organism A and an object O. Here we have a preliminary answer to the problem posed by the privacy of C

⁴⁵ The denunciation of " psychologism " has been a rallying cry for both the analytic and the phenomenological traditions from their very beginnings. Yet neither tradition offers any clear account reconciling what I am here calling the privacy of C versus the publicy of O. The phenomenologists, to their credit, have from the first recognized that O necessarily depends upon C, whatever the difficulties in explaining this fact. The analysts, to their discredit, have proceeded as though the mere eschewing of any consideration of the intraorganismic factor C were by itself sufficient to avoid all difficulties in the explanation of objective discourse. This ploy is most explicit, perhaps, in Quine's distinction between the theory of reference and the theory of meaning, drawn in the spirit of what we have seen in the first part of this article is the mistaken belief, inherited from Russell, to the effect that, as Quine puts it: " There are no ultimate philosophical problems concerning terms and their references, but only concerning variables and their values; and there are no ultimate philosophical problems concerning existence except insofar as existence is expressed by the quantifier '($ x)'." (The Methods of Logic, p. 224).

vis-à-vis the publicy of O: just as one and the same thing can serve as the term of several different relations, so one and the same object O can be the term of relations founded on the C's existing in diverse individuals. One and the same object can stand at the term of several different relations R founded on the different C's existing in different individuals. Thus the publicy of O is reconciled with the privacy of C by the diversity of status between and C: O exists as the term of the relation R, whereas C exists as the fundament or basis of that same relation R. C and O differ as fundament and term of a relation differ. The fact that C and O are located in physically disparate subjects, and that several subjects can be related to one and the same term, provides a preliminary resolution of our problem, though we shall soon have occasion to add some essential refinements.

Let us now reintroduce into the picture the linguistic factor-- word or sentence or whole treatise--X, which, in conjunction with C, refers to O. Our preliminary assumption about O as a thing need not be altered, for X, obviously, insofar as it is a mark, sound, or movement (or a series of marks, sounds and movements), is also a thing existing in fact independently of any C here and now. This is an important point, whose significance has generally not been appreciated by philosophers of language.

Up to now, we have spoken of C and X as co-occurrent, the former intraorganismically and the latter extraorganismically, whenever X actually refers, here and now, to O. From this point of view it seems sufficient to say that X is able to be referred to O owing to the fact that O is apprehended by the user of X thanks to C. But this appearance of sufficiency is an illusion, for it glosses over the fact that X is every bit as much an object--a term of a C-based relation--as O itself is. C and X, as the thought and the word, do not belong on one side of discourse, and O, object and thing, on the other side. On the contrary, C belongs on one side and X and O belong on the other, for X and O have in common the character of being objects, while C is known to exist only analytically, as the

basis that must be posited for the difference between one and the same thing now existing only in fact, now existing also in apprehension. Here we arrive at a conclusion whose consequences are definitive: the difference between X as physical and X as significant or linguistic, is the same as the difference between O as thing and O as object of discourse.

The importance of the point demands that we spell it out in detail. Whenever a given individual, A, let us say, uses X, the X too is something that he is aware of, and so is an object of apprehension--something A is aware of--in its own right, as well as and as a necessary condition for X's being a sign of O. Both X and O must be apprehended as objects in order for either to function as a sign of the other. This means that, since C is posited precisely and only as the intraorganismic factor making the difference between A's actually being aware or not being aware of a given object here and now, before (logically, not always temporally) X is used to refer to O, there are not one but two C's involved, one whereby X exists as apprehended and one whereby O exists as apprehended. Strictly speaking, then, the sign character of X relative to O is due to the establishment of an association not between C and X but rather between X as one object and O as another object, which association in turn results in the formation of a third C, which is the " idea " or " concept " neither of X nor of O disjunctively but of X and O conjunctively. It is this C, properly speaking, that X evokes when it functions linguistically as a sign of O.

The situation can be clarified, perhaps, by a sequence of diagrams:

Diagram I: A is aware of O by virtue of C_1; the idea of O, prescinding from the question of how C₁ was formed.

( C₁ ) ------- O

   A

page 284

Diagram II: A is aware of X by virtue of C₂, the idea whereby X exists as something A is aware of.

( C₂ ) ------- X

   A

Diagram III: A is aware of X and O together, making possible the formation of an idea of the two as a unity--a unity of thought.

( C₁ ) ------- O

( C₂ ) ------- X

   A

Diagram IV: A is aware of X and not only together, but also as connected, by a connection consequent upon the idea, C₃, whereby X and O exist as something A is aware of conjunctively.

( C₃ ) ------- X

   A            \ O

 

Viewed in this way, there is no chance of committing the blunder of thinking that thoughts and words are two separate processes, each of which would be just what it is if one or the other were to be removed. On the contrary, discourse--the whole of thought, language, and object--is profoundly one, so unified that the intraorganismic thoughts and the extraorganismic words and objects are what they are precisely and only to the extent that each is simultaneously. For, while the object of cognition exists dependently on the thought from the standpoint of being, object and thought co-exist in perfect proportion to one another from a temporal standpoint. The object cannot be as apprehended save when and as an idea of it exists, and an idea as idea cannot be save as giving an existence of presence to an object. In a certain sense, therefore, only upon the formation of C₃ in our sequence of diagrams do we have the existence of discourse properly speaking.

This reveals a profound import to the contemporary expression, " universe of discourse." For, from the standpoint of the

concept or idea, words are not opposed to objects as words are to things. On the contrary, words and things known are equally objects, but objects differing in status primarily as regards their manageability. Thus, words and sentences call things to mind and order our thoughts about things, but, equally, things suggest words and patterns of words. For both words and things exist, in discourse, in an objective unity derived from the concepts or ideas which give them, their being--which " constitute them," the phenomenologists say--as cognized or known.

A linguistic remark can fail to be understood--Q, on hearing X, may think of N rather than O. Or a perceived object may suggest no words to the perceiver but present itself, as it were, mutely. But in any case, whenever anything enters our consciousness, it does so by virtue of an intraorganismic occurrence C.

These remarks bring out the utter peculiarity of the intraorganismic factor C, the concept or idea: unlike objects, including words, that are also signs, ideas cannot fail to give presence to something besides themselves. Inasmuch as a sign is anything that makes present in awareness something besides itself, ideas are pure signs: they do nothing but signify, i. e., make present the objects of awareness--be they words or things or whatnot--that the ideas themselves are not. Unlike the objects which are also signs, ideas cannot fail to signify. They alone, among all the furniture of the world, signify by necessity. For them, to be and to signify are simply one. The object of which we are directly aware, be it word or thing apprehended disjunctively or conjunctively, is, in every case, just what C qua C is not. If we were or could be directly aware of C, it would not be C, but something else, for C is just that factor that makes us aware of something that it itself is not. No matter how the matter is approached, careful attention to the function of C in relation to X and O reveals that, while C can be known to exist by a reflexive analysis of discourse, it can under no conditions be directly observed or apprehended. This is a point of some importance that was well understood by the

older scholastics and by a few recent philosophers--notably Jacques Maritain and F. H. Bradley--but that seems never to have been adverted to in a systematic way by anyone in the analytic or phenomenological traditions excepting Heidegger. Substituting the word " physical " for " psychical," the following text can be cited from Bradley's Logic with even more timeliness, in many respects, than when Bradley wrote:

An idea, within my head, and as a state of my mind, is as stubborn a fact as any outward object. . . . but, intent on this, we have as good as forgotten the way in which logic uses ideas. We have not seen that in judgment no fact ever is just that which it means, or can mean what it is; and we have not learnt that, wherever we have truth or falsehood, it is the signification we use, and not the existence. We never assert the fact in our heads, but something else which that fact stands for. And if an idea were treated as a physical reality, then it would not represent either truth or falsehood. When we use it in judgment, it must be referred away from itself. If it is not the idea of some existence, then, despite its emphatic actuality, . . . it is a something which, in relation to the reality we mean, is nothing at all.⁴⁶

Or again, substituting " discourse " for " judge " :

Not only are we unable to discourse before we use ideas, but, strictly speaking, we can not discourse till we use them as ideas. We must have become aware that they are not realities, that they are mere ideas, signs of an existence other than themselves. Ideas are not ideas until they are symbols, and, before we use symbols, we can not discourse.⁴⁷

So far, then, we have given a preliminary explanation of how the privacy of C reconciles with the publicy of X and O. C, as subjective or private, is not something experienced or known but the basis or fundament of all experience and knowledge. What are experienced and known are words and things, O's and X's insofar as they enter into relations R with us on the

⁴⁶ F. H. Bradley, The Principles of Logic (2nd ed., rev.; London: Oxford University Press, 1922), Vol. I, p. 2.

⁴⁷ Ibid.

basis of C's. And O's and X's have a common or public character, precisely because they exist at the term of the diverse relation R's founded on the several C's.

But so far we have also been proceeding on the assumption that the O's in question exist in fact as well as in discourse. How does the foregoing analysis clarify or apply to the case where O does not exist in fact and where, accordingly, Quine, Russell, and the logisticians generally assert--not without plausibility--that the apparent reference of the discourse is merely apparent or at least thoroughly confused?

It is in the systematic clarification of the privileged sign status of the intraorganismic factor C that the answer to this question, and the explanation of the apparent indifference of discourse to being and non-being, is found. Clearly, in calling ideas " intentional," and in assigning them the property of intentionality, Brentano and the phenomenologists after him were getting at the fact that ideas cannot be save as giving the being of presence in cognition to objects. Brentano, in later rejecting the doctrine of intentionality as a truly relational property on the grounds that the object terminating an intentional relation is often unreal, seems to be getting at the fact that the notion of a relation without a term is unintelligible. But what he failed to see, and what Husserl glimpses in his doctrine of ideas as constituting their objects, and even in his perverse doctrine of the époché, is the fact that the term of a relation, as it is a term, has its being from the relation; and hence there can be a true relation between idea and object even when the object has no further being than that of a term.

Consider again the unique status of C, the idea or concept. In its function as C, it is neither experienced nor experienceable but has its whole being in provenating--" dimanating," as Poinsot would say--the relation R terminating in O. If C were to exist without engendering this relation R, it would not be as C that it existed, for as C it is precisely the fundament of R. C as C gives rise to R by necessity. This is what constitutes its uniqueness among the furnishings of the world. This is also

what endows discourse with its indifference to the being and non-being of its objects, for the following reason: the term of any relation, inasmuch as it is a term, owes its being to the relation it terminates.

To see how this is so, consider the case of a non-cognitive relation, say, the relation of " larger than." A, for example, is something larger than B, only so long as both A and B exist. A's size as a natural thing provides a basis for its being related to B, but only on condition that B actually exist. Given the existence of B in itself, then B will also have an existence as term of A's relation to it based on A's size, and vice-versa. Thus the existence of B relative to A as something smaller than A is owing to or based upon A's size, just as the existence of A relative to B as something larger than B is owing to or based upon B's size. The point this proves is perfectly general: even when the term of a given relation is something existing in fact, it owes its being as term to the relation that it terminates.

A term, insofar as it is a term formally, is the term of something; for nothing terminates except another. The term of a relation, therefore, is something of the relation; if the relation is real, its term is a term purely, that is, it does not have other than to terminate or be opposed to the relation and to be something of the relation itself as of the respecting. In this it differs from the fundament, because it is necessary for the fundament to give existence to the relation according to inherence, in which existence the relation coincides with an absolute determination of the subject of the relation. The term, however, does not give existence to the relation, but the opposition of termination. Therefore the formality of the term is not something absolute.⁴⁸

Applying this perfectly general point to the case of C, we see that the dependence of O as object upon C is but a special case of the dependence of any term as such upon the relation of which it is the term. The case of CRO differs from the case of A>B only in this, that where A functions as fundament contingently upon the existence of B in fact, C functions as a

⁴⁸ Jean Poinsot, Cursus Philosophicus, ed. B. Reiser (Turin: Marietti, 1930), Vol. I, p. 596a46-b15.

fundament by virtue of what C in itself is, that is, C functions as a fundament necessarily. Hence A will generate B's status as term only if B also exists in fact. But C will generate O's existence as term regardless of whether O also exists in fact. This explains the capacity of discourse to refer to what is not. All objects of discourse, as such, exist as terms of the relation R generated by C. Since C cannot be without generating R, and O exists qua object as term of R, O will exist as an object whenever C exists as a fundament, regardless of whether O also exists as a thing in fact. Hence the objects of discourse need not be independently of discourse in order to be really referred to.

Here we must refine our preliminary solution to the publicy of objects. If every term of a relation qua term has its being from the relation, and if every object of discourse exists as such in the capacity of term, will there not be as many terms as there are relations founded on C's? And if there are as many terms as there are C's, is not the common or public status of the object destroyed after all? Does it not merely become the extraorganismic correlative of the intraorganismic C, as private in its own way as the C founding the relation R that gives O its being as term?

The answer to this difficulty involves some unavoidable subtlety but continues to be a firm No. In every case of an actual relation the three factors--fundament, relation, and term--are existentially inseparable but really distinct. For example, the A that is larger than B differs from the A that continues to exist when B exists no longer, only by a difference in mode. It is one and the same A existing in both cases but once with an added dimension or mode--fundament of the relation "larger than." Similarly with the term of any relation. Apart from the relation it does not exist as term; yet if, besides existing as term, it also exists in its own right, those two existences differ in mode only.

Suppose then two persons, A and B, each having his own concept--C₁ and C₂, respectively--of an object O which happens to exist in fact. The object O, existing in relation to C₁

and C₂, differs from the thing O, existing in fact, only modally-- just as a point qua terminus of a line has a mode of being super-ordinate to the being of the point as such. Hence the object O is identical with the thing O but adds to it the mode of being as term--something O has, not from itself but from C. C₁ and C₂ generate R₁ and R₂ both terminating at the thing O in its mode as object. But just as O as object is also O as thing, so also O as term of R₂ is also O as term of R₂. All three-- as thing, as term of R₁, O as term of R₂--coincide in being, though they are also modally distinct. Similarly, one and the same point can terminate two different lines, even though that point, considered as terminating, exists dependently on each of the lines it terminates and as modally distinct from each of them. More concretely, two individuals, each perceiving the same cloud, give the very cloud floating in the world a new mode of being as object perceived. Each perceiver has his own intraorganismic factor C founding the relation terminating at the cloud as perceived. Each, therefore, gives a modally distinct objectivity to the cloud: yet it is one and the same cloud that is objectified in the two cases. C₁ and C₂, located in spatially diverse subjects, terminate in the numerically same space with a difference only in the mode of the termination.

This unavoidable subtlety, actually, saves the public character of discourse in a manner that is more genuinely satisfying than the more facile formulation of our preliminary solution could provide. For the modal distinctness of one and the same object as cognized by two or more individuals (or by one and the same individual at two separate times), together with the modal distinctness between things as such and things become object, fits very well with our experience of the " slipperiness of things " and of the great difficulty with which any deep agreement is reached in human discourse; for it is precisely through these modal distinctions that differences in experience slip in between man and man, and through them too that history insinuates itself between man and the physical world.

We have seen now how it is that the objects of discourse

can also be things of the world, and why they need not be; and we have also seen how it is that discourse has a public character, even though it is based on an intraorganismic or " private " factor C, which I have repeatedly referred to as an " idea " or " concept."

This choice of terms bears closer scrutiny, for according to a celebrated theory widely held in current scientific and philosophical circles--the so-called " identity hypothesis "--the intraorganismic factor C is in fact a state of the brain of the organism A which uses X to discourse about O. What I have called an " idea " or " concept," according to this view, is more properly called an event or state of the central nervous system--an identification that must be made because, if the postulation of such intraorganismic factors " is going to prove fruitful," as Charles Osgood remarks, " and serve as anything more than a label for ignorance, properties must be attributed to them." ⁴⁹

We can agree with Osgood on the importance of assigning properties for C, but to ascribe to C the properties of a nervous event or state is possible only to the extent that the above-described manner in which C functions relative to X and O has been systematically misunderstood. If what I have said is the function of C in discourse be admitted, then the view that " mental states are brain states " must be false· A brain state is something that is observable in principle. Therefore C as such is not a brain state but must be other than and super-ordinate to any state of the central nervous system; for, as we have seen, since an idea (C as C) cannot be safe as giving an existence of presence to something which it itself is not, an idea as idea is intrinsically unobservable. It can be objectified only by inference, never by observation.⁵⁰ To repeat an earlier remark:

⁴⁹ Chares E. Osgood, Method and Theory in Experimental Psychology (New York: Oxford University Press, 1953), p. 410. Cf. p. 681: "We must postulate material events for meaning and then investigate the theoretical consequences of this postulation." (Osgood's emphasis)

⁵⁰ Cf. Poinsot, Cursus Philosophicus, III, 185a33-b25.

" careful attention to the function of C in relation to X and O reveals that, while C can be known to exist by a reflexive analysis of discourse, it can under no conditions be directly observed or apprehended."

Physical marks, sounds or movements, when functioning in discourse, undergo--thanks to C--a singular and mysterious " elevation " (as Cajetan puts it ⁵¹), during which they exist in a higher way than is proper to them as observable, physical occurrences; and they do so inasmuch as they are the objective effects of the intraorganismic factor C--the concept or idea-- within the speaker and hearer of language, a factor which, as it functions in discourse, is itself no more directly inspectable than the significance it causes. This is the point of view proper to a would-be philosophy of language; it is on this point that an account of meaning can properly turn. Here we may apply another of Cajetan's remarks:

From this it will appear how crude is the thinking of those who treat of sense and the sensible, of understanding and the understandable, as also of the processes of sensation and understanding, according to the canons of judgment applicable to material events. Et disces elevare ingenium, aliumque rerum ordinem ingredi--you must learn to raise up your mind, and enter into quite a different order of occurences.⁵²

The point is not to deny that brain states are somehow correlative with and indeed necessary conditions for the existence of ideas. It is simply to point out the error of reductively identifying C, as that which is conditioned, with a brain state, as that which doubtless conditions C. ⁵³ Whatever difficulties one may have with the terminology of ideas and concepts, they are as nothing compared to the difficulties consequent upon the failure to grasp this principle: the conditioned as such is always other than its necessary and even sufficient conditions.

⁵¹ Cajetan, Commentaria in summam theologicam S. Thomae, I, q. 14, art. 2, nn. 4 and 7; q. 79, art. 2, n. 14.

⁵² Ibid., I, q. 14, art. 1, n. 7.

⁵³ Cf. Poinsot, Cursus Philosophicus, III, 185b26-186b40, esp. 186b3-16.

IV. A Thomistic Perspective on Discourse (The Classical Doctrine of Intentionality) .

Brentano, in introducing the concept of intentionality into modern discussion, contented himself with a vague reference to the origins of the notion among " the scholastics of the Middle Ages." ⁵⁴ To judge from the manner in which he formulates the doctrine, and from the reasons that led him finally to evacuate the doctrine of its relational content, it seems fair to say that Brentano was rather poorly informed as to what the classical scholastic doctrine of intentionality actually contained.⁵⁵

The notion of an intentional mode of existence, or at least the invention of the terminology, seems to have come from Averroes, and in particular, from his remarks in commentary on Aristotle's discussion in the de Anima of the character of the stimulus engendering the sense impressions that form the basis of our perceptions of the physical world·⁵⁶ In St. Thomas himself, this notion of an intentional determination of being, or " species," as distinct from the natural or entitative determinations of being--the " formae naturales "--became the organizing concept for the account of the genesis of concepts, beginning from the " species impressae sensuum externorum "-- the intentional stimuli of sensation--proceeding up through the workings of the internal senses to the culmination finally in the " species expressa intellectus possibilis "--the concept or intentional form made by the understanding itself as the medium in which the world exists as understood.

It was the privilege of Cajetan to clarify, principally in his Commentary on Aristotle's de Anima, the exact basis of the distinction between the formae and the species as two typically

⁵⁴ Brentano, " The Distinction Between Mental and Physical Phenomena," loc. cit., p. 50.

⁵⁵ I have given some detailed evidence for this judgment in " The Ontological Status of Intentionality," The New Scholasticism, XLVI (Spring, 1972), pp. 220-233.

⁵⁶ Averroes, Commentarium Magnum in Aristotelis de Anima Libros, ed. by F. S. Crawford (Cambridge, Mass.: The Mediaeval Academy of America, 1953), II, sec. 60, pp. 219-221.

diverse ontological realities, the one (the formae) by which the affected subject is primarily altered or changed, the other (the species) by which the affected subject is primarily identified with something that it itself is not.⁵⁷

However, the nature of this intentional union and the mechanism, so to speak, that underlies its possibility, was not systematically clarified until the 17th century, when Jean Poinsot, called by Yves Simon the last Commentator of genius on the work of St. Thomas, took up the doctrine of the species in order to develop a comprehensive theory of the sign and of cognition.⁵⁸

In Poinsot's theory, unlike most other versions of the nature of discourse essayed in recent times along Thomistic lines, the doctrine of the intentional form is not deployed in an opaque fashion as a fundamental datum not susceptible of further elucidation. On the contrary, the capacity and function of the intentional form is shown to be possible as a direct consequence of the peculiar status of relation among all the possible modes of being. This peculiarity, which Poinsot exhibits as constituting the prior possibility of signs in general and of human discourse in particular, Poinsot learned from St. Thomas:

Relation is said to be in a way different than any other kind of being. For in the case of other kinds of being, each one is said to be in two ways, both as regards its existence, and as regards the character of its essence. . . . But relation is something according to the existence it has in a subject, while according to its essential character it has not to be something, but only to be referred to another; whence according to its essential character it does not posit anything in a subject. . . . Thence also is it that something is found to be related in which there is only a mind-

⁵⁷ Cajetan, Commentaria in de anima, Aristotelis, ed. by P. I. Coquelle (Rome: Angelicum, 1939), Vol. II, sees. 264-267, pp. 251-255.

⁵⁸ Poinsot's theory of the sign is in the Cursus Philosophicus, Vol. I, Part II, Qq. XXI-XXIII, pp. 646a14-749b47, but draws also on other parts of the work. This entire matter is discussed at length in my article, " The Two Approaches to Language," The Thomist, XXXVIII (October, 1974), pp. 856-907.

Poinsot's treatment of cognition runs throughout in Vol. Ill of the Cursus Philosophicus.

dependent or mental relation, and the relation is not posited there according to physical being, as when the knowable is referred to knowledge.⁵⁹

Following out this clue, Poinsot was able to elucidate the nature of signs and the function of concepts or ideas in discourse with an unrivalled profundity. From the standpoint of their role in discourse Poinsot divided the sign into two classes, which he called formal and instrumental. Formal signs are all signs that correspond to what we identified above as the intraorganismic factor C, that is, they are the signs whose whole being is exhausted in the function of signifying or presenting to mind objects other than themselves. These signs, since they are never objects directly apprehended (this is precluded by their mode of being, as we have seen), are known only reflexively and by inference,⁶⁰ from the fact that, if there were no such signs, the interpretation of words would involve us in an infinite regress, as would the awareness of any object.⁶¹

Instrumental signs, by contrast, are all signs, including, therefore (and principally), words and sentences, that are perceived directly as objects, and whose functioning as signs depends on their being objectively perceptible.

The class of formal signs, thus, includes the whole of what traditional philosophers have variously called ideas, concepts, images, imaginations, etc. They exist as such only in the cognitive act ⁶² and only as presenting objects that they themselves are not--which objects, in turn, often (indeed, normally), by the associative processes so familiar to modern psychology, become signs in their own right (instrumental signs) as well as objects. Since the class of instrumental signs is comprised of

⁵⁹ 1 Sent., dist. 20, q. 1, art. 1. St. Thomas makes this point in numerous other passages as well--e. g., Quodlib. I, a. 2; Quodlib. IX, a. 4; de Verit., q. 1, a. 5 ad 16; Summa Theol., I, q. 28, aa. 1 and 2; et alibi. I have cited the text from the Sentences only for reasons of convenience in the present context.

⁶⁰ See Poinsot, Cursus Philosophicus, III, 185 a 33-b 10.

⁶¹ See John A. Oesterle, "Another Approach to the Problem of Meaning," The Thomist, VII (April, 1944), pp. 258-260; and John N. Deely, " The Ontological Status of Intentionality " (cited in fn. 55 above), pp. 229-230.

⁶² Poinsot, Cursus Philosophicus, I, 303b29-38; III, 185b26-186a15.

signs that are apprehensible objects besides, and since objects as apprehended depend upon formal signs, if there were no formal signs, there would be no signs of any kind. " The formal sign," as Poinsot put it, " is a sign simply and absolutely " ;⁶³ the instrumental sign is a sign only in a certain respect. Or as Bradley put it, " in the end, there are no signs save ideas." ⁶⁴

Formal signs were also called by Poinsot (here following Cajetan, Aquinas, and Averroes) intentional forms, in order to contrast them with the natural forms or determinations of things. The point of the contrast lies in the fact that an intentional form functions to generate a relation necessarily, by virtue of what it is. Natural forms, by contrast, though they may indeed serve as the basis or fundament generating a relation, do not do so necessarily, but only contingently--contingent, that is, upon there being in fact an entity to which the mode of being as term can be added. The further point, that every term qua term owes its being to the relation it terminates--" is something of the relation," as Poinsot says ⁶⁵--even when that term (as is normally the case outside of relations of cognition) is also something existing in reality outside of the relation, becomes, as we have seen, the key to interpreting the apparent indifference of discourse to the real being or non-being of its objects. Poinsot himself does not make this point an explicit element in his treatise on the sign; but he does devote an entire article to it in his earlier treatise on relation,⁶⁶ of which treatise he says the theory of the sign is but an extension and particular application.⁶⁷

With the addition of this explicit element to Poinsot's theory, then, it becomes perfectly clear why the objects of discourse need not be independently of discourse in order to be truly and really referred to. Given the nature of relation with its necessary

⁶³ Poinsot, Cursus Philosophicus, I, 694 b 23-69.

⁶⁴ Bradley, The Principles of Logic, Vol. I, p. 5.

⁶⁵ Cursus Philosophicus, I, 596 b 2-3.

⁶⁶ Cursus Philosophicus, I, Part II, Q. XVII, Art. V, " Utrum relatio formaliter terminetur ad absolutum vel ad relativum."

⁶⁷ Cursus Philosophicus, I, 642 a 25-37.

elements of fundament, connection, and term, the indifference of discourse to being and non-being becomes the direct consequence of the difference between intentional and natural forms--between, if you like, the " mental " and the " physical." Both an intentional and a natural form can function as fundament of a relation, but only the intentional form does so by necessity. Hence, any form (any determination of being), insofar as it is the fundament of a relation, is mediately the cause of the existence of the term regarded formally as term of the relation in question. But a natural form gives being to the term of the relations it generates contingently--contingent, that is, upon the real existence of something which, besides existing in relation to another as term, also exists in itself as more than a mere term; while an intentional form, an idea, gives being to the term of the relation it generates necessarily--necessarily, that is, regardless of whether or not there is a real existent which, besides being a term in relation to the idea, has a being in fact independently of discourse. When my uncle dies, he continues to exist as an object of discourse, for this existence he has as term of the cognitive relations my ideas engender.

The crucial point to be noted here is that, while every object of discourse as such exists necessarily as term of the relation generated when, as, and while an idea exists, not every object of discourse need exist only as such, i. e., only as term of cognition. In other words, it can perfectly well happen, as common experience suggests does often happen, that the very object existing in discourse as an object also exists as a thing in the world independently of discourse here and now. Objects existing only as objects Poinsot, following an ancient tradition, calls "beings of the mind" (entia rationis). Objects also existing as things or "in fact," Poinsot calls "real beings" (entia realia). Real beings and beings of the mind, as objects of discourse, have in common the existence as terms of cognitive relations. For this existence as " objective beings," both alike depend upon being known, " since indeed it is from the actual termination of a relation that it results that a given thing is said to be the term of that

relation." ⁶⁸ But real beings, in addition to their being known-- i.e., in addition to their being as terms of relations founded on ideas--also have being independently of being known, and in this consists their reality in point of fact. Beings of the mind, by contrast, have no being in addition to being known--i.e., have no being in addition to their being as terms of relations founded on ideas--and in this consists their unreality in point of fact. Beings of the mind are precisely non-beings relative to the beings of nature--the physical or " real " beings- "A being is properly said to be of the mind," writes Poinsot, " because it has no being independent of the understanding, but is said to exist in the understanding only objectively, and so is opposed to real being." ⁶⁹

Being and non-being, thus considered, belong to objects in point of fact according as they are or are not something more than pure terms of idea-based relations. The objects of discourse, however, as such, are nothing more than pure terms of idea-based relations. These relations have the ontological character of relations both when the objects at which they terminate are and when those objects are not also things in the world. But the objects are considered by " common sense " to be realities only to the extent they are or are believed to be things in the world besides being objects; to the extent the objects are or are believed to be only objects (cognized terms) and nothing more, they are considered unrealities. The crucial point for a philosophy of discourse, however, is that in point of the existence they receive from ideas, objects, real and unreal, are on an equal footing and are equally public. Once this is understood, it is perfectly understandable also why discourse should really be as it appears to be--endowed with a certain internal indifference to the being and non-being of its objects, not, indeed, insofar as they are objects but insofar as they are things in addition to being objects.

It is this last point--that nothing prevents some objects of

⁶⁸ Ibid., 596 a 36-39.

⁶⁹ Ibid., 285 a 39-43.

discourse, existing as such as terms of relations founded on formal signs, from also existing in themselves as things of nature independently of discourse--that Husserl and the practitioners of his époché (suspension of belief in the factuality of any objects) systematically neglect, but that seems to be allowed for in the development of later trends known as " existential phenomenology."

In any event, this application by Poinsot of the theory of relation found in St. Thomas to the explanation of the sign does seem to provide a way out of the impasse reached by the phenomenologists and the analysts over the apparent indifference of discourse to being and non-being. This indifference is a consequence of the very nature of discourse as relational, for, as Cajetan put it, since what is essential to relation as such is a being toward, and not a being mental or physical, " the consequence is that the toward as such is neither physical nor mental by necessity, but either permissively," ⁷⁰ depending on the conditions concerning the nature of the fundament (is it a natural or an intentional form?) and--primarily--of the term (is it a pure term and nothing more, or is it a thing of nature besides being a term?). When the term of a cognitive relation is a pure term and nothing more--when it is an object only, let us say, and not also a thing existing in fact as well as and while it is apprehended--it lacks a positive character possessed in its own right. But when the cognitive relation is merely cognitive and nothing more, that is, when the cognitive relation is between an idea in the knower's mind and an object having no other existence than the being of term given by the relation and its fundament (the idea) ; when, thus, the cognitive relation obtains between a physical reality or being in fact--the knower--and a physical unreality or non-being in fact--the object as mere term and nothing more--it is still a true and genuine relation possessing the positive or ontological character of relation in its own right. In the case of relation alone, therefore, as Cajetan puts it, " to be in the mind is not a diminishing

⁷⁰ Cajetan, in summam theologicam, I, q. 28, a. 1.

condition, as in the case of all other modes of being that admit of physical instantiation. For a rose existing only in the mind is not a genuine rose, any more than the Homer existing in belief is genuinely Homer. But a relation in the mind is a genuine relation." ⁷¹

The doubts of Brentano and the analysts concerning the relational character of discourse, and the silence of the phenomenologists when asked to explain how it is possible for discourse to really refer to what does not exist, alike stem, it would seem, from the failure to articulate the peculiar status of relation as a mode of being--the failure to understand that, as Poinsot learned from Aquinas, relation alone among the ontological categories retains its positive content whether its existence is physical or merely mental, because

only relation has to be a being and toward a being, and from the side whence it has a being toward it exists positively, and yet it does not have thence the character of being real. But the real being of a relation originates from one place, namely, from the fundament; the positive character of [the relation as a] being toward from another place, namely, from the term, whence the relation does not have to be a being, but toward being. . . . That therefore something could be considered ontologically, or positively, even if not really in an entitative or physical way, is something peculiar to relation.⁷²

⁷¹ Ibid.

⁷² Poinsot, Cursus Philosopicus, I, 581 b 1-13. One caveat must be entered before concluding this discussion of the " classical doctrine." Throughout this and the preceding Section, I have spoken about objects of discourse being, as such, terms of idea-based relations. In speaking thus, I have been viewing objects under a restricted and, to tell the truth, " improper " formality--to wit, solely and wholly from the standpoint of the here and now cognized. It would be a mistake if the reader took from this the impression that such a notion of object is a fully adequate one, in need essentially of no further analysis. Since such further analysis lies outside the scope of the present argument, however, suffice it to note here that the notion of object adequately considered reduces to that of a " cause " in the order of extrinsic formal specification, and applies not only to cognition. These points are expressly made by Poinsot in his theory of signs (which intentional forms are in the line of), and the interested reader is well advised to have a close look at Poinsot's adequate and formal consideration of objects in the Cursus Philosophicus, I, 670a11-679b5, esp. 673b50-674a4, 677b5-ll, and 678a7-32, where it is asserted and explained why the consideration of objects as here and now cognized, i. e., as terms of idea-based relations, is not the formality most proper to objects as such

V. Conclusion.

The analytic tradition in philosophy began when Bertrand Russell turned to mathematical logic as the instrument for vindicating the standpoint of pluralistic realism that G. E. Moore had adopted around the turn of the century in order to escape from the pseudo-Hegelian idealism of Bradley and Bosanquet then dominating British and American thought. The standpoint of pluralistic realism, the standpoint that lies at the base of analysis, lies also at the base of Thomistic philosophy. But what about Russell's appeal, followed, moreover, by that of the Wittgenstein of the Tractatus and the so-called " logical analysts " generally, to mathematical logic as the means of justifying this standpoint?

Russell turned to mathematical logic because he found in the theory of polyadic functions, developed by mathematicians since the middle of the 19th century, a pure logic of relations that could express relational structure without reducing the relative element to a mere attribute of a subject conceived in a certain way by the mind. In other words, Russell turned to mathematical logic because, unlike the traditional logic with which he was acquainted, mathematical logic allowed for the reality of relations as something external to the related subjects and so something existing over and above those subjects and their respective attributes.

Two things must be said about this appeal from the standpoint of Thomistic philosophy. First, Russell, in the eyes of St. Thomas, is unequivocally correct in rejecting the idealistic view that relations are not real, either in the sense that all relations are reducible to inherent attributes of subjects (esse in), or in the sense that all relations are products of the requirement that certain aspects of subjects, which in themselves are not relations, nonetheless must be conceived by the mind through a comparison with elements other than themselves (relativa secundum dici--things relative according to the terms

in which the understanding must give expression to them). This view that the being of relations as relations is something unique (esse ad), reducible neither to the inherent attributes of a substance (esse in), nor to the activity of the mind (secundum dici ,⁷³ is fundamental to Thomism and a direct consequence, for Thomas as for Russell, of the pluralistic character of the world.

Second, there is little room for doubt that the modern logic allows for a purer and more straightforward expression of relational facts than does any developed technique of traditional logic and has the added advantage of making explicit, in the quantification of predicates, the fact that our knowledge of the world is a highly constructional, historical achievement. These advantages, however, are not sufficient to support the hopes Russell entertained for the new logic--hopes that continue to be vainly sustained in a large part of the analytic community.

For it is simply not true that the whole of logical development prior to modern times was based on a system of thought committed to denial of the extramental reality of external relations. Nor is it true that the analysis of relational facts admittedly real is impossible to achieve according to protocols compatible with the general framework of a logic of the Aristotelian type. The relational analysis of the sign essayed by Poinsot, for example, is embedded completely in one of the purest and most complete treatments of the pre-modern logical tradition, a massive volume titled simply the Ars Logica.

No man knows the whole of any tradition, however, and when a learned man such as Russell is actively engaged in the creative solution of distinctive problems, he may be excused for lapses of historical knowledge. A man is justified in not relying overmuch on the thoughts of men long dead to the extent that their systematization was not achieved with an eye to encompassing the novelties and difficulties consequent