What is Truth?

Edo Pivcevic

• Publisher: Routledge
• Year: 2018
• Print ISBN: 9781138359048
• eBook ISBN: 9780429783388
• Edition: 1
• Topic: Filosofia
• Subtopic: Lingua in filosofia

First published in 1997, this volume advances the view that the nature of truth, in so far as truth has a 'nature', lies in the manner of its occurrence. Edo Piv?evicì argues that truth is an vent, i.e. it does not exist until it occurs, and survives only as long as the requisite conditions for its occurrence are in place. Positing that language sets traps for the unwary, Piv?evicì states that calling 'x' true involves a property ascription does so only in the sense that 'x' enters into truth and is part of a 'truth event'.

1 The problem of truth, or how does thought hook onto the world?

To make a truth claim, either by asserting a proposition or indirectly by characterising a proposition as 'true', is to say that things are as they are said to be. It is to signal or affirm that a certain situation or constellation of circumstances obtains in reality. This surely is as plain as day —— hat more might there be to add? Yet on closer inspection this produces more bafflement than light, and in particular it does nothing to explain why the concept of truth has such a powerful hold on our imagination.

What we need to know is not just that an assertion, if true, conveys what obtains in actual fact. This comes uncomfortably close to trading synonyms for synonyms. For how do we explain what facts are except by saying that certain propositions are true? We need to know just what such conveying involves and under what conditions it is effected, or can be effected. Without an account of such conditions to give substance to our claims, the search for truth reduces to a mere groping in the void.

Consider what it is to explain a concept. To explain a concept is to set out how that concept may be legitimately applied, and in so doing circumscribe its extension; it is to specify the criteria for sorting out the items that fall under it from those that fall outside it. It follows that part of an explanation of truth will have to be an account of the conditions under which a proposition might legitimately qualify as true. Not all truth claims can be validly made. Sometimes what is said or written will fail as a vehicle for conveying facts as a matter of principle on account of some semantic incongruities or bad syntax. The random anagrammatic sequence 'Book in sentence this false is every' does not add up to an English sentence, whereas its unscrambled version 'Every sentence in this book is false', though syntactically impeccable, fails a consistency test: if it is true, then it is false. It has no unequivocally specifiable content, and by definition cannot qualify as a truth candidate.

Again, in other cases what looks like a semantically as well as syntactically sound declarative sentence will produce an inconsistency merely by virtue of being asserted. Thus I cannot assert 'I am not now uttering a sound' without contradicting, by my act of utterance, what I claim to be the case. Nevertheless the sentence itself is both syntactically and semantically unimpeachable, and if it were to be written down by me rather than uttered, or if it should be transposed into a third person idiom and asserted by any old onlooker, that would cause no difficulty, and might well be true.

Nor is this a rare occurrence. It is well known that the thought a sentence carries, or what is sometimes called its 'prepositional content', can fluctuate and change depending upon the circumstances of utterance; with verbally and syntactically identical sentences, as a result, transmitting different messages. Conversely, it is sometimes possible deliberately to alter the sentence, and moreover alter it extensively in respect both of meaning and syntax, without thereby necessarily altering the state of affairs conveyed in the given instance. But if so, then the inevitable question is, exactly what elements or aspects of the meaning content have a bearing upon the truth or falsity of what is said? And above all, what factors are instrumental in determining such a content?

Truth-relevant content and extensional isomorphism

At first glance, the answer to this may not seem difficult. The clue, surely, lies in the extensional properties of what is said or claimed to be true. After all, we all have a good intuitive grasp of what it is for two different declarative sentences to carry the same factual freight; for example, when two non-synonymous but co-referential singular expressions, i.e. expressions sharing the same designation, are substituted for each other; or, perhaps, when a general term appearing in a subject position in a general proposition is replaced with another of the same extension. Thus if 'The Morning Star' designates the same object as 'The Evening Star', then the truth-relevant content of 'The Morning Star moves around the sun' and 'The Evening Star moves around the sun' is surely the same. And so is that of 'Whales are not fish' and 'The largest marine animals are not fish'. There is nothing arcane or startling about that. It is common practice to make use of different but contextually co-extensive expressions, either in order to help the message on its way, or deliberately to highlight things from a different angle, or simply to avoid repetition or relieve tedium. If fly to Boston, I go there by aeroplane. These are not synonyms. Nevertheless contextually they amount to a description of the same event. What matters in such and similar cases is that the contextual isomorphism of the terms used ensures that the two sentences, for all their differences, are both true if either of them is. Provided the extensional properties of their terms do not change, their truth-relevant content will remain the same.¹

Yet how do we decide that two sentences are 'extensionally isomorphic' in the sense just described — i.e. convey the same state of affairs — except by considering their respective meaning contents? It is only through their meaning that we can access their 'extensional properties'. But if it is necessary first to sort out those aspects of the meaning content which account for their extensional isomorphism, then how can we appeal to their extensional isomorphism in order to account for their joint truth-relevant content? Surely a different approach is required if we are to keep well clear of any vicious circularity. In point of fact, I shall argue that whilst it is possible for non-synonymous sentences to target, and convey, the same state of affairs, conveying the same state of affairs is not the same as expressing literally the same truth. Truth, as we shall see, is a much more complex phenomenon than that. It follows that any radically 'reductionist' treatment of truth-relevant content will have to be rejected as inappropriate. The reductionist policy is propelled by the belief that only some, not all features of the content are relevant to a consideration of truth or falsity. Sentences, it is argued, may differ in meaning as well as 'surface' syntax while nevertheless sharing the same truth-relevant core, and this common core, which represents their 'prepositional substance' — or quite simply the relevant proposition — can be isolated from the surrounding frippery and rendered explicit by a suitable method of logical paraphrase. There is always a quantity of meaning which can safely be axed without altering the substance of the message, or detracting from its truth.

But the problem is that we need a criterion for deciding which aspects of the content are essential, and which are not essential, and can therefore be safely left to one side; and any such criterion is likely to be susceptible to critical challenge. Above all, before we can begin a process of selective pruning we need to understand the meaning of a sentence in its entirety. Nothing can be dismissed a priori as irrelevant. But if the meaning in its entirety is thus instrumental in determining the truth-relevant content, then that is what the truth-relevant content is. And if that is the case, i.e. if the truth-relevant content embraces all the aspects of the meaning content (of that which can be self-consistently uttered, at any rate), then how can it be justifiably claimed that two non-synonymous sentences, even if they do happen to convey the 'same' state of affairs, have exactly the same truth-conditions? And if they don't have the same truth-conditions, how can they 'express' the same truth?

Access to truth-relevant content via predicates and variables of quantification

These questions cannot be shirked. Nevertheless the temptation to adopt a 'reductionist' approach with regard to the truth-relevant content remains strong and persistent, for obvious reasons. It has to do with the natural desire to eliminate any fuzziness or ambiguity from truth claims, and make their content precise. We saw that such claims cannot always be taken at their face value; that they sometimes carry a different message to what appears on the surface, or are simply muddled. It might seem, therefore, that the best — perhaps the only — way to bring out with sufficient clarity what truth-relevant content they may carry in a given instance is to transpose what is said into a more rigorous albeit more austere language of predicates and variables', and, having accomplished this task, carry the whole process a step further, perhaps, by undertaking a thoroughgoing analysis of predicates and their logical relationships in the light of their extensional properties.

In order to illustrate what is at stake here, consider sentences about fictional entities. What, for example, might be the truth-relevant content of 'The philosophers' stone is brittle'? As it stands, the sentence cannot significantly be characterised as either true or false, for the philosophers' stone does not exist. Should we then say that the sentence in reality does not have a truth-relevant content? This seems much too extreme, for the sentence, on the face of it at least, is semantically sound as well as syntactically well formed. Perhaps all that is needed is to spell out more clearly its underlying essentials. It might accordingly be felt that the appropriate way to proceed is to rephrase the whole sentence into an existentially quantified form, signalling in a general way that the relevant predicate (or concept) does indeed apply to something, or that there is something answering the given description. The idea, in short, is to show that the sentence at root says something about a predicate, not about any specific entity. On this interpretation, our example above will unfold into: There is something which uniquely has the properties of the philosophers' stone, and is brittle; in which form it evidently can be asserted or denied without any problems. That, at any rate, is how things seemed to Russell.²

This method of paraphrase, as I have already indicated, can be applied, and indeed has on occasion been applied, universally, with farreaching implications. Thus some philosophers (e.g. W.V. Quine) have taken the view, in effect, that all significant assertions without exception can be expressed in terms of predicates and variables along these lines. Moreover the tendency often has been to use this kind of logical manipulation, designed with the explicit purpose of highlighting the truth-relevant content and making it precise, as a starting-off point for some sweeping reductionist epistemological operations on predicates, by subjecting them to rigorous phenomenalist or physicalist tests.

But just how necessary, one might wonder, is it to go to such lengths in order to extract the truth-relevant content from any old significant sentence, and ensure that it can be asserted free from any ambiguity? One — the most familiar — objection to the above method of logical paraphrase comes from Russell's critics, like P.F. Strawson, who claim that the attempt to make all significant sentences without exception accessible to truth value in the way described suggests a confusion of existential presuppositions with existential assertions. Thus, to take our example again, 'The philosophers' stone is brittle' presupposes rather than asserts the existence of the relevant entity, and since the presupposition is incorrect, the question of truth or falsity does not arise. Given that the vital existential prerequisite is not fulfilled, the sentence is automatically disqualified as a legitimate vehicle of a truth claim. At the same time, this manifestly does not render the sentence meaningless. The sentence, as it stands, may fail to satisfy the conditions of 'assertibility', but it does not cease to be intelligible on that account. And this, it is argued, only goes to prove that if a sentence makes good sense, this has nothing to do with its capability, or otherwise, of having a truth value; rather the sources of its meaningfulness should be sought elsewhere — more specifically, in the customs and practices that govern our use of language.

What is more, the problem of 'assertibility', it is pointed out, is not confined solely to sentences about fictional entities. There are countless other sentences which cannot legitimately be asserted or denied, although they are not only syntactically and semantically sound but target actually existing entities; the simple reason being that certain other vital truth conditions are not fulfilled. Thus given that Kant never wrote poetry, the sentence to the effect that he was a dismal poet clearly cannot be appropriately asserted. At the same time, any attempt to make such a sentence accessible to truth value by translating it into an existentially quantified sentence, saying that there was 'someone or other' answering Kant's description who wrote bad verse, in accordance with the method of paraphrase suggested by Russell³, is surely absurd, for the sentence is intrinsically singular, not existential.

The substance of the above criticism of the reductionist paraphrase in terms of predicates and variables à la Russell is thus: a) that a declarative sentence may have a meaning without necessarily satisfying the conditions needed for it to have a truth-value; and b) that in such a case the sentence cannot properly be said to possess the kind of truthrelevant content conveyed by the paraphrase in question (which, therefore, becomes superfluous). In short, the question of meaningfulness of declarative sentences and the question of their truth-value are not necessarily linked, and have to be treated separately. A declarative sentence may have a meaning without qualifying for truth value, and it is only those that do so qualify that can legitimately be asserted.

Conditions of assertibility

But plausible though this solution may seem at first sight, it creates serious problems of its own. What is suggested, basically, is that the subject 'talked about' in the given instance must exist, and in some cases, moreover, has to have certain specified attributes (such as Kant writing poetry, in our example above), before the relevant truth claim, or claims, can validly be made. But does this mean that we should suspend our truth claims until we are quite sure that the required truth conditions are satisfied? On the above view, a decision as to whether a given truth claim should be allowed to stand or be dismissed as invalid presupposes an agreed ontology. More than that: there must be an agreed decision procedure, enabling us to establish whether such conditions are fulfilled in the given case. But there might be no such agreement. It is conceivable that — with the exception, possibly, of some simple sentences such as e.g. 'It is cold here', or some first person statements — we may never be in a position to know with certainty if the required assertibility conditions are fulfilled. And if so, how then are we to decide what can and what cannot be appropriately claimed to be truel

In a sense, it was this very difficulty that the 'predicates and variables' theory referred to earlier was, in part at least, designed to meet. Its underlying motive was to pre-empt any disagreements over which significant sentences can, and which cannot, appropriately be asserted, or said to be true or false, by suggesting a way of paraphrasing their truth-relevant content in such a way that they automatically become accessible to truth value. There was thus no need to wait for an agreement as to whether the entity, or entities, 'talked about' in a given instance did actually exist. The paraphrase in terms of predicates and (bound) variables guaranteed instant assertibility. This greatly simplified things, and moreover offered palpable advantages in terms of enhanced clarity and precision, But — as always — there was a price to pay. For such a paraphrase inevitably entailed a loss of certain important features of sentential meaning. Indeed the more assiduously such a paraphrase was pursued the more its reductionist consequences became apparent. Thus it entailed dispensing not only with singular expressions (such as 'The philosophers' stone'), but also with proper names, personal pronouns, indeed with most 'indexical' expressions, which pervade ordinary speech in such glorious profusion, and which, I shall argue, so far from simply adding charm and episodic colour to our utterances represent important constituent elements of their truth-relevant content.

There is, however, a third option, which avoids such undesirable consequences, without at the same time tying us in advance to any specific choice of entities. This is to accept that sentences like 'The philosophers' stone is brittle' and 'Kant was a dismal poet' can act as vehicles of genuine truth claims, although not (as Russell would have it) because they reduce to existentially quantified sentences whereby a property is attributed to certain predicates or concepts (viz. that such predicates apply to something, or that the concepts in question are not empty), but because they may be taken as saying that an irreducibly unique individual with certain attributes is part of the actual world. If anything is 'presupposed', then, it is the existence of a world as a totality of what is the case. The problem of assertibility can thus be solved without engaging in doubtful reductionist manoeuvres to which the 'predicates and variables' theory seems to be irrevocably committed, and, at the same time, without waiting for an agreement about ontology.

On this view, 'The philosophers' stone is brittle' effectively unfolds into: 'The philosophers' stone is part...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• Preface
• Outline of argument
• 1. The problem of truth, or how does thought hook onto the world?
• 2. The strategy of austerity
• 3. The naturalistic approach to truth conditions
• 4. Truth and certainty
• 5. Objectivity and the social construction of time
• 6. Truth as a species of good
• 7. Truth and philosophical explanation
• 8. Truth as event
• Index