It is widely held, to the point of being the received interpretation, that Frank Ramsey was the first to defend the so-called Redundancy Theory of Truth in his landmark article “Facts and Propositions” (hereafter “FP”) of 1927.1 For instance, A.J. Ayer2 cited this article in the context of arguing that saying that p is true is simply a way of asserting p and that truth is not a real quality or relation. Other holders of the received interpretation, such as George Pitcher,3 J.L. Mackie,4 Susan Haack,5 A.C. Grayling,6 Nils-Eric Sahlin,7 Richard Kirkham,8 Donald Davidson,9 and Michael Lynch10 credit Ramsey with having originated what they call “the Redundancy Theory.” Even an authoritative source such as The Encyclopedia of Philosophy11 attributes this theory to him. What’s more, Grover et al.,12 in defending their Prosentential Theory of Truth, claim that their theory is an improvement and development of the Redundancy Theory which they too attribute to Ramsey.13
Despite its currency, the received interpretation has been challenged by some dissenters. Brian Loar,14 Ulrich Majer,15 Hartry Field,16 and Jerome Dokic and Pascal Engel17 have argued that Ramsey held views on truth that went well beyond the Redundancy Theory as this theory is construed on the received interpretation.
In this paper, I too shall argue that Ramsey’s views on truth were richer than those typically attributed to him. However, what the aforementioned dissenters have not noticed is that, as I shall show,, as I shall argue, considerable light can be shed on Ramsey’s views on truth by contrasting them with those of his Cambridge colleague W.E. Johnson on the one hand, and the received interpretation on the other.
What will emerge from our discussion is a distinction, hitherto undrawn in the literature on theories of truth and their history, between three kinds of Redundancy Theory, and we will find that Ramsey’s kind may usefully be seen as one intermediate between the Redundancy Theory of the received interpretation at one pole and Johnson’s very different Redundancy Theory at the other.
In this connection, I should mention that this paper is primarily concerned with the question of what Ramsey actually held on the matter of truth, rather than with the question of whether he was right in holding it. But although the primary concern of this paper is interpretative, our discussion will bring to light a number of points concerning truth that should be of interest to contemporary truth-theorists.
The following structure informs the rest of this paper. In section 2, I summarize the received interpretation, and delineate what I call the “Deflationary and Non-Attributive Redundancy Theory.” In section 3, I explain W.E. Johnson’s conception of truth, most notably what I call his “Inflationary and Attributive Redundancy Theory.” In section 4, I adduce grounds for holding that Ramsey’s kind of Redundancy Theory is one intermediate between the two aforementioned; I shall call his the “Deflationary and Attributive Redundancy Theory.” In section 5, I end the paper by addressing a pertinent question, namely: If the received interpretation is incorrect, then why is it so widespread?
Gertrude Ezorsky18 nicely encapsulates the received interpretation. Taking Strawson’s theory19 to be a supplement to what she calls Ramsey’s “Assertive Redundancy” or “No Truth” theory of truth, Ezorsky writes:
Ramsey claimed that to say that a proposition is true means no more than to assert the proposition itself. “It is true that Caesar was murdered” means no more than “Caesar was murdered.” “It is false that Caesar was murdered” means no more than “Caesar was not murdered.” According to this view, “true” has no independent assertive meaning, and the traditional notion of truth as a property or relation is misguided. Ramsey suggested that “true” is used for purposes of emphasis or style, or to indicate the position of a statement in an argument.20 In a similar vein, Kirkham writes:
The claim of the redundancy theory is not that the predicate ‘is true’ repeats what has already been said in the sentence in which it appears.
(Ramsey does not mean that ‘is true’ is redundant in the sense that
‘“Caesar was murdered” is true’ says ‘Caesar was murdered’ twice.)
Rather, the claim is that ‘is true’ is vacuous, that it says nothing at all. (…) So ‘is true’ is redundant (‘gratuitous’ would be a better word) relative to the resources of English as a whole because anything that can be said with it can be said without it. Thus Ramsey seems to agree with Strawson that we are not saying something about a truth bearer when we make apparent truth ascriptions. However, unlike Strawson, he insists that when we ascribe truth, we are saying something, not merely performing an act. We are saying precisely what we would be saying if we simply uttered the proposition itself.21 To summarize the received interpretation, Ramsey putatively held the following:
(i) ‘It is true that p’ means no more than ‘p’ and ‘‘p’ is true’ means no more than
(ii) The expression ‘is true’ is redundant in the sense of being a vacuous, non-
(iii) Truth is not a property (or relation).
The conjunction of (i)-(iii) I call the “Deflationary and Non-Attributive Redundancy
Theory.” By “deflationary,” I mean that it deflates or reduces the meaning of ‘It is true that p’ and of ‘‘p’ is true’ to (no more than) that of ‘p’; by “non-attributive,” I mean that it takes the truth-predicate to not attribute any property (or relation) of truth, there being no such property or relation on this view.
In the next section, we turn to a remarkably different kind of Redundancy Theory, namely, W.E. Johnson’s.
3. W.E. Johnson’s Conception of Truth
It’s no mere accident that Ramsey knew W.E. Johnson and his work. Ramsey attended Cambridge University, taking a degree in mathematics in 1923, became a fellow of King’s College in 1924 at the age of 21, and was made lecturer in mathematics in 1926 at Cambridge.22 There his colleagues included such philosophers as W.E. Johnson, G.E. Moore, Bertrand Russell, A.N. Whitehead, and later Wittgenstein, whose Tractatus Ramsey translated.23
Although Moore, Russell, Whitehead, and Wittgenstein are widely regarded as among the greatest philosophers of the 20th century, Johnson, an important figure at Cambridge during Ramsey’s time there, is largely forgotten today. In any case, it’s worthy of note that Ramsey knew Johnson’s work extremely well.24 In a paper entitled “The Nature of Propositions” (hereafter “NP”)25 that he read as an undergraduate to Cambridge’s Moral Sciences Club on 18 November 1921, Ramsey mentioned Johnson no fewer than five times and showed a keen understanding of the latter’s views on the nature of propositions, as well as of those of Russell’s and Whitehead’s.26 What’s more, in a 1922 review of the second part of Johnson’s Logic, Ramsey characterized this book as “likely to be the most important work on Logic that has appeared for many years. It is full
of clever, original points and throws new light on almost every topic with which it deals.”27 Furthermore, in his article “Universals”28 of 1925, in which he forcefully criticized the contrasting views of Russell and Johnson on the nature of universals, Ramsey demonstrated a close reading of Parts I and II of Johnson’s Logic.29 Given Ramsey’s familiarity with Johnson’s work, let’s consider what the latter wrote about truth.
Johnson begins his Logic (Part I) with a discussion of propositions, holding that it is the proposition or assertum “of which truth and falsity can be significantly predicated.”30 The adjectives ‘true’ and ‘false’ predicated of propositions “derive their significance from the fact that the proposition is not so to speak a self-subsistent entity, but only a factor in the concrete act of judgment.”31 For instance, though we may predicate truth or falsity of the proposition ‘matter exists’, this predication, according to Johnson, means
that any and every thinker who might at any time assert the proposition would either be exempt from error or not exempt from error. In other words, the criticism which reason may offer is directed—not to the proposition—but to the asserting of the proposition; and hence the
means that anyone’s assertion of the proposition would be erroneous.32
It should be noted here that Johnson defined assertion as conscious belief (or judgment); hence, to assert in his sense “does not mean to utter (without belief)”, and “merely to believe unconsciously is not to assert.”33 On his view, ‘true’ and ‘false’ only derive their meaning “from the point of view of criticizing a certain possible mental
attitude,” and coincide with the application of the imperatives ‘to be accepted’ and ‘to be rejected’ respectively.34 However, in maintaining this coincidence between these two imperatives on the one hand and the two adjectives ‘true’ and ‘false’ on the other,
Johnson held that “it must not be taken that we are able thus to define the adjectives true and false. On the contrary, we are forced to insist that they are indefinable.”35
But though he insisted (like Frege) that ‘true’ and ‘false’ are indefinable, he also held that the truth of a judgment could be understood as its proposition being in accordance with a certain fact:
the truth of a judgment (expressed in a proposition) may be said to mean that the proposition is in accordance with a certain fact, while any proposition whose falsity would necessarily follow from the truth of the former is in discordance with that fact. In this way the somewhat vague conception of the correspondence between thought and reality is replaced by the relation of accordance with a certain fact attributed to the true proposition, and of discordance with the same fact attributed to the associated false proposition.36 After extensively discussing simple and compound propositions, Johnson made an interesting observation about truth and falsity in the context of discussing secondary propositions (that is, propositions predicating some characteristic of some primary proposition). His observation is worth quoting in its entirety:
[A] question of some interest arises as to whether the two elementary cases ‘p is true’ and ‘p is false’ where p is a proposition are legitimate illustrations of secondary propositions. It may be held that the proposition ‘p is true’ is in general reducible to the simple proposition p; so that, if this were so, ‘p is true’ would only have a semblance of a secondary proposition, and would be equivalent for all purposes to the primary proposition p. It appears to me futile to enter into much controversy on this point, because it will be universally agreed that anyone who asserts the proposition p is implicitly committing himself to the assertion that p is true. And again the consideration of the proposition p is indistinguishable from the consideration of the proposition p as being true; or the attitude of doubt in regard to proposition p simply means that attitude of doubt as regards p being true. These illustrations, in my view, show that we may say strictly that the adjective true is redundant as applied to the proposition p; which illustrates the principle which I have put forward, that a proposition by itself is, in a certain sense, incomplete and requires to be supplemented by reference to the assertive attitude. Thus the assertion of p is equivalent to the assertion that p is true though of course the assertum p is not the same as the assertion that p is true. The adjective true has thus an obvious analogy to the multiplier one in arithmetic: a number is unaltered when multiplied by unity, and therefore in multiplication the factor one may be dropped; and in the same way the introduction of the adjective true may be dropped without altering the value or significance of the proposition taken as asserted or considered.37
Notice that Johnson did not take ‘is true’ to be redundant in the sense of attributing no property of truth to a proposition of which it is predicated. Rather, predicating ‘is true’ of a proposition is redundant for him in the sense of being superfluous, for anyone who asserts pimplicitly commits himself to asserting that p is true. Thus, asserting p and asserting that p is true are equivalent not because asserting that p is true is reducible to merely asserting that p, where ‘is true’ in the former assertion is vacuous, but rather because asserting that p involves asserting that p is true.
Accordingly, adding ‘is true’ to p only makes explicit what is already implicit, and that is why its addition is redundant (i.e., superfluous) and does not alter the meaning of p. His analogy with multiplying by unity is telling: though multiplying by unity does not alter a numeric value, we are still multiplying by a number; likewise, though predicating ‘is true’ of p does not alter its meaning, we are still attributing to p the property of truth.
To summarize, Johnson held the following concerning truth:
(i) any proposition p must be understood as a factor in an act of assertion or
judgment, and to assert or judge that p is no less than to assert or judge
that p is true.
(ii) The predicate ‘is true’ is redundant in the sense of being superfluous, but not
in the sense of being a vacuous, non-attributive predicate.
(iii) Truth is a property.
(iv) The adjective ‘true’ is indefinable.
(v) The truth of a judgment may be understood as its accordance with a certain
The conjunction of (i)-(v) I call the “Inflationary and Attributive Redundancy Theory.” By “inflationary” I mean that it takes the assertion or judgment that p to be no less than the assertion or judgment that p is true. By “attributive” I mean that it takes the truth-predicate to attribute the property of truth to that of which it is predicated.
Having explicated Johnson’s conception of truth, we turn next to Ramsey’s.
4. Ramsey’s Conception of Truth
I shall argue that Ramsey endorsed a kind of Redundancy Theory intermediate between the Deflationary and Non-Attributive Redundancy Theory (DNRT) of the received interpretation on the one hand, and the Inflationary and Attributive Redundancy Theory (IART) of Johnson’s on the other. I will begin by contrasting Ramsey’s Redundancy Theory with the DNRT.
Ramsey certainly held tenet (i) of the DNRT as defenders of the received interpretation invariably point out; that is, on Ramsey’s view, ‘it is true that p’ means no more than ‘p’ and ‘‘p’ is true’ means no more than ‘p’. In fact, his endorsement of this tenet proves to be a recurring theme in his writings on truth.38 In his most widely known writing on truth in FP, he claimed that the proposition ‘It is true that p’ (e.g., ‘It is true that Caesar was murdered’) means no more than ‘p’ (‘Caesar was murdered’).39 When a proposition is not explicitly given but described—for instance, ‘He is always right’—Ramsey noted that it is more difficult to eliminate ‘true’ or ‘false’ in ordinary language; however, if we take ‘He is always right’ to mean ‘For all p, if he asserts p, p is true’, then “we see that the propositional function pis true is simply the same as p, as e.g. its value ‘Caesar was murdered is true’ is the same as ‘Caesar was murdered’.40
As a result of such claims, defenders of the received interpretation have taken Ramsey to have endorsed not only tenet (i) of the DNRT, but also tenets (ii) and (iii), namely, that ‘is true’ is redundant in the sense of being a vacuous, non-attributive predicate, and that truth is not a property (or relation). However, nowhere in FP, nor in any other of his writings, did Ramsey endorse (ii) or (iii). In fact, in FP itself, Ramsey explicitly wrote: “Truth and falsity are ascribed primarily to propositions”41 near the beginning of the very paragraph where he discusses the redundancy of ‘is true’; and in his introduction to his manuscript On Truth, Ramsey wrote: “Truth is an attribute of opinions, statements or propositions, what exactly it means we shall discuss later, but in a preliminary way we can explain it as accordance with fact.”42
On the assumption that (i) implies (ii) and (iii), it may seem obvious to some that Ramsey should have endorsed (ii) and (iii) given his endorsement of (i). However, on his conception of truth, this assumption is incorrect. To see why, we must remember that for Ramsey, there is no problem of truth separate from the analysis of judgment (or belief), and that it is “immediately obvious that if we have analysed judgment we have solved the problem of truth.”43 So let us consider more closely Ramsey’s analysis of judgment. I will focus mostly on its articulation in FP, since defenders of the received interpretation always cite this article as evidence of his adherence to the DNRT, and I will argue that even there Ramsey did not hold that ‘is true’ is a vacuous, non-attributive predicate.
Ramsey in FP distinguished between what he called the “objective factor or factors” and the “mental factor or factors” in judgment. In, for instance, the judgment that Caesar was murdered, we may distinguish between the mind, or the present state of mind, or words or images in the mind on one side (the mental factors); and on the other side, Caesar, or Caesar’s murder, or Caesar and Murder or the proposition that Caesar was murdered, or the fact that Caesar was murdered on the other (the objective factors).44
Beginning with the objective factors, Ramsey endorsed Russell’s Multiple Relation Theory, according to which judgment “has no single object, but is a multiple relation of the mind or mental factors to many objects, those, namely, which
we would ordinarily call constituents of the proposition judged.”45 For instance, in the judgment that Caesar was murdered, the objective factors are the individual Caesar and the universal Murder. In endorsing this theory, Ramsey rejected Single Objective Factor Theories; that is, theories according to which a proposition, or alternatively a fact, is the single objective factor of judgment. As did Russell, he rejected propositions as single objective factors on the ground of the “incredibility of the existence of such objects as ‘that Caesar died in his bed’, which could be described as objective falsehoods, and the mysterious nature of the difference, on this theory, between truth and falsehood.”46 He rejected facts as single objective factors on the grounds of the mysteriousness of postulating negative facts as the objective factors of false judgments,47 and on the ground that the analysis of facts must ultimately be done in terms of individuals and universals, which drives us back to there being a multiple relation between the mind and many objective factors in judgment.48
Now for his account of the mental factors. In the case of belief not expressed in words—e.g., a chicken’s belief that a certain sort of caterpillar is poisonous—the mental factors are parts of the chicken’s behaviour “related to the objective factors, viz. the kind of caterpillar and poisonousness.”49 And here he endorsed a kind of pragmatism:
the relation between the chicken’s behaviour and the objective factors was that the actions were such as to be useful if, and only if, the caterpillars were actually poisonous. Thus any set of actions for whose utility p is a necessary and sufficient condition might be called a belief that p, and so would be true if p, i.e. if they are useful.50
He added in a footnote: “It is useful to believe aRb would mean that it is useful to do things which are useful if, and only if, aRb; which is evidently equivalent to aRb.”51 Ramsey thus endorsed the following equivalences:
(E1) The belief that p is true iff p (e.g., believing that individual a bears
relation R to individual b is true iff a actually bears R to b).
(E2) The belief that p is useful iff p (e.g., believing that individual a bears
relation R to individual b is useful iff a actually bears R to b).
From (E1) and (E2), it follows that
(E3) The belief that p is true iff the belief that p is useful.
But it no more follows from his endorsement of (E1) that Ramsey held that the predicate ‘is true’ does not attribute the property of being true than it follows from his endorsement of (E2) that he held that the predicate ‘is useful’ does not attribute the property of being useful. What he did is take the property of being true and the property of being useful to be co-extensive not only with each other but with the objective factors of belief being as they are believed to be. And in this sense, he married a Pragmatist conception of truth to an Aristotelian correspondence conception of truth according to which a belief is true when things are as they are believed to be.52
In the case of belief expressed in words, he takes the mental factors “to be words, spoken aloud or to oneself or merely imagined, connected together by a feeling or feelings of belief or disbelief.”53 Feeling belief towards the words in ‘p’ (where the latter is a sentence) has causal properties and therein lies its importance:
It is evident that the importance of beliefs and disbeliefs lies not in their intrinsic nature, but in their causal properties, i.e. their causes and more
especially their effects. For why should I want to have the feeling of belief towards names ‘a’, ‘R’, and ‘b’, when aRb, and of disbelief when not-aRb, except because the effects of these feelings are more often satisfactory than those of the alternative ones.54
For Ramsey, to believe sentence p is to express agreement with the truth-possibilities specified in p. Moreover, to “say that feeling belief towards a sentence expresses such an attitude is to say that it has certain causal properties which vary with the attitude, i.e. with which possibilities are knocked out and which, so to speak, are still left in. Very roughly the thinker will act in disregard of the possibilities rejected.”55 The meaning of a sentence is accordingly “agreement and disagreement with such and such truth-possibilities, meaning that one who asserts or so believes the sentence so agrees and disagrees.”56
In contemporary terminology, Ramsey endorsed the view that the meaning of a sentence consists in its truth-conditions. Consequently, he held that ‘It is true that p’ and ‘‘p’ is true’ mean no more than ‘p’, because they have identical truth-conditions. So adding ‘is true’ or ‘It is true that’ to ‘p’ is redundant in the sense of being superfluous. But it does not follow therefrom that ‘is true’ or ‘It is true that’ do not attribute the property of being true or that there is no such property, for to have identical truth-conditions is to have the property of being true in the same conditions. So though Ramsey accepted tenet (i) of the DNRT, his conception of meaning committed him to rejecting tenets (ii) and (iii).
Thus, when Ramsey’s remarks about the redundancy of ‘is true’ and ‘It is true that’ are read in the proper context of his analysis of judgment (from which flowed his account of meaning), it is clear that, contra the received interpretation, Ramsey cannot be properly interpreted as endorsing the DNRT. Was he then a Johnsonian in accepting the IART? To this question we turn next.
Ramsey’s and Johnson’s views were similar to the extent that both of them took ‘is true’ to be a genuinely attributive predicate, and truth to be a genuine property. Thus neither of them accepted the DNRT. There is also a superficial sense in which they both agreed that ‘is true’ and ‘It is true that’ are redundant, and also a superficial sense in which they both agreed that the truth of a judgment may be understood as its accordance with a certain fact. But, however influenced he may have been by Johnson, Ramsey was by no means a mere epigone of the latter. To see why, let’s consider some of the key ways in which their respective views differed.
One key difference is that Johnson, like Frege, took ‘true’ and ‘false’ to be indefinable, whereas Ramsey took ‘true’ and ‘false’ to be definable. More precisely, Ramsey, unlike Johnson, held that we could explicate the meaning of ‘true’ in the context of defining true belief as follows. Where “any belief whatever we may symbolise as a belief that p, where ‘p’ is a variable sentence, (…) [w]e can then say that a belief is true if it is a belief that p, and p.”57 Or as he put it, comparing his view to Aristotle’s:
Our definition that a belief is true if it is a “belief that p” and p, but false if it is a “belief that p” and ~p is, it may be remarked, substantially that of Aristotle who, considering only the forms “A is” and “A is not”, declared that “To say of what is, that it is not, or of what is not, that it is, is false, while to say of what is that it is, and of what is not that it is not, is true.”58
Another key difference between Johnson and Ramsey concerns their respective views on the nature of judgments and propositions. Johnson countenanced the existence of propositions as bearers of truth or falsity, taking them to be concrete factors in judgment, and held a Single Objective Factor Theory according to which a proposition is the single objective factor in judgment, where judgment involves a relation of the mind to this proposition. As Ramsey noted:
Mr. Johnson supposes that the relation in question is of the kind expressed by saying that the proposition is a factor in the belief. This way of analysing references is undoubtedly favored by the English language, for we do talk of people believing or doubting propositions. But this is a weak argument, clearly inadequate as a basis for supposing there to be propositions, so unlike anything else in the world; and careful consideration of the implications of this linguistic custom shows that no good argument can be based upon it.59
As we noted earlier in this paper, Ramsey himself endorsed by contrast Russell’s Multiple Relation Theory. As such, he held that, in judgment, the mind is not related to a proposition, but multiply related to individuals and universals: “when I believe p, my belief is multiply related to the things ordinarily called the constituents of p; if then I write ‘p’, the symbol stands not for a proposition which is before my mind when I write, but for the property of my belief, expressed by saying that it is a belief that p.60 Though Ramsey often sounded (for instance in FP) as if he countenanced the existence of propositions, we must remember that his talk of propositions (for instance, of propositions being true or false) was just a facon de parler, and that strictly speaking he, unlike Johnson, did not posit their existence.61
Yet another ground of disagreement concerned the existence of facts as that which true propositions correspond to or accord with. Johnson posited the existence of such facts, holding that the truth of a judgment could be understood as its proposition being in accordance with a certain fact. Ramsey countenanced such talk, but only again as a facon de parler. For instance, with respect to the judgment that a has R to b, Ramsey observed that we “can, if we like, say that it is true if there exists a corresponding fact that a has R to b, but this is essentially not an analysis but a periphrasis, for ‘The fact that a has R to b exists’ is no different from ‘a has R to b’.”62
Finally, though Ramsey echoed Johnson in taking ‘is true’ and ‘It is true that’ to be redundant (that is, superfluous), Ramsey’s conception of their redundancy was what we might call “deflationary” whereas Johnson’s was what we might call “inflationary.” That is, for Ramsey, since they have the same truth-conditions and hence the same meaning, ‘‘p’ is true’ and ‘It is true that p’ mean no more than ‘p’, and hence the addition of ‘is true’ or ‘It is true that’ to ‘p’ is redundant; by contrast, for Johnson, the judgment or
assertion that p is no less than the judgment or assertion that p is true, and hence the
addition of ‘is true’ or ‘It is true that’ to ‘p’ only makes explicit what is already implicit.
To sum up, Ramsey defended a Redundancy Theory of truth intermediate between the Deflationary and Non-Attributive Redundancy Theory of the received interpretation at one pole, and Johnson’s Inflationary and Attributive Redundancy Theory at the other. This intermediate theory we may call the “Deflationary and Attributive Redundancy Theory.” Its key tenets may be encapsulated as follows:
(i) ‘‘p’ is true’ and ‘It is true that p’ mean no more than ‘p’ because they
have the same truth conditions.
(ii) So adding ‘is true’ or ‘It is true that’ to ‘p’ is redundant in the sense
of being superfluous.
(iii) But ‘is true’ and ‘It is true that’ are still genuinely attributive.
(iv) Truth is a genuine property, one which is co-extensive with the
property of usefulness.
(v) The truth of a belief that p may be defined thus: a belief that p is true
if and only if p, i.e., when the objective factors of belief (individuals
and universals) are as one believes them to be.
In closing, let me address a pertinent question, namely: If the received interpretation is incorrect, then why is it so widespread? Though it is difficult to give a definitive answer, I submit that at least five main factors have contributed to the prevalence of the received interpretation.
First, A.J. Ayer’s Language, Truth, and Logic63 (LTL) proved enormously influential in Anglo-American analytic philosophy in the first half of the last century as an articulation of logical positivist tenets.64 In this work, Ayer advanced the DNRT, and in a footnote on page 89 wrote: “cf. F.P. Ramsey on ‘Facts and Propositions’ The Foundations of Mathematics, pp. 142-3.” This suggested that Ramsey held the DNRT as well, though, as I have argued here, he did not. Since Ramsey died at the premature age of 26 in 1930, he did not have the chance to correct this misinterpretation, and since Ayer’s LTL was far more widely read than Ramsey’s works, the misinterpretation spread.
Second, in connection with Ramsey’s premature death, it should be noted that Ramsey never completed his book manuscript On Truth and Probability where he intended (among other things) to develop in depth the points made in his brief remarks on truth (most notably in FP). The first part of this manuscript was only published sixty-one years after Ramsey’s death under the title On Truth. Reading it makes clear that Ramsey did not endorse the DNRT,65 but its publication came far after the received interpretation took root.
Third, Ramsey’s remarks on truth outside of On Truth were quite brief. For instance, his most often quoted remarks in FP amounted to no more than two paragraphs. This very brevity lent itself to misinterpretation.
Fourth, these brief remarks on truth in FP have often been read out of the broader context of his analysis of judgment, despite the fact that Ramsey himself claimed that “if we have analysed judgment we have solved the problem of truth.”66 For instance, George Pitcher’s anthology Truth does not include all of FP, but only the two aforementioned paragraphs in complete isolation from the rest of the text where Ramsey offers an analysis of judgment. Holders of the received interpretation such as A.C. Grayling67 and Richard Kirkham68 cite this brief excerpt as their only evidence that Ramsey held the DNRT, and thereby neglect the larger context which makes clear that he did not endorse this theory.
Fifth, as I have argued in this paper, Johnson’s views on truth shed considerable light on Ramsey’s. However, given that Johnson’s work has been little read and largely forgotten since the 1920s, the influence of Johnson’s work on Ramsey has tended to go unnoticed. Understanding how Ramsey’s conception of truth resembled and differed from Johnson’s enables us to grasp why Ramsey did not endorse the DNRT.
These five factors, I submit, have contributed to the prevalence of the received interpretation. The aim of this paper has been to set the record straight on what Ramsey actually held concerning truth.
1 Frank P. Ramsey, “Facts and Propositions” (1927), reprinted in F.P. Ramsey: Philosophical Papers, edited by D.H. Mellor (Cambridge, 1990) 34-51.
2 A.J. Ayer, Language, Truth and Logic (New York, 1936) 89.
3 George Pitcher, editor. Truth (Englewood Cliffs, New Jersey, 1964) 14.
4 J.L. Mackie, Truth, Probability and Paradox (Oxford, 1973) 51-2.
5 Susan Haack, Philosophy of Logics (Cambridge, 1978) 127-8; “Is Truth Flat or Bumpy?” in Prospects for Pragmatism, edited by D.H. Mellor (Cambridge, 1980) 17-8.
6 A.C. Grayling, Introduction to Philosophical Logic (London, 1982) 153-4.
7 Nils-Eric Sahlin, The Philosophy of F.P. Ramsey (Cambridge: 1990) 57-9.
8 Richard Kirkham, Theories of Truth (Cambridge, Massachusetts, 1992) 317.
9 Donald Davidson, Inquiries into Truth and Interpretation (Oxford, 2001) 40.
10 Michael Lynch, “Deflationary Views and their Critics” in his The Nature of Truth (Cambridge, Massachusetts, 2001) 422.
11 Gertrude Ezorsky, “Strawson’s Performative Theory of Truth” in The Encyclopedia of Philosophy (New York, 1967) Volume 5, pp. 88-90.
12 Dorothy Grover, Joseph Camp, and Nuel Belnap, “A Prosentential Theory of Truth,” Philosophical Studies, 27 (1975): 73-4.
13 According to the Prosentential Theory, ‘is true’ does not attribute the property of being true, but is used to construct prosentences such as ‘it is true’ or ‘that is true’ whose role is to allow one to express agreement, make connections in discourse, and generalize.
14 Brian Loar, “Ramsey’s Theory of Belief and Truth” in Prospects for Pragmatism, edited by D.H. Mellor (Cambridge, 1980) 49-69.
15 Ulrich Majer, “Ramsey’s Theory of Truth and the Truth of Theories,” Theoria, 57(3) (1991) 162-195.
16 Hartry Field, “Correspondence Truth, Disquotational Truth, and Deflationism” in The Nature of Truth, edited by Michael P. Lynch (Cambridge, Massachusetts, 2001) 483-504.
17 Jerome Dokic and Pascal Engel, Ramsey: Verite et succes (Paris, 2001).
18 Ezorsky, op. cit., 88-90.
19 According to Strawson’s Performative Theory, the truth predicate is primarily used as a performative expression; it is not used to make a statement about a statement, but to perform an action of agreeing with, accepting, or endorsing a statement. On this view, ‘is true’ is non-attributive.
20 Ezorsky, op. cit., 88.
21 Kirkham, op. cit., 317.
22 Sahlin, op. cit., 222.
23 Ibid., 226.
24 See Ibid., 57f.
25 Frank P. Ramsey, “The Nature of Propositions” (1921), reprinted in his On Truth, edited by Nicholas Rescher and Ulrich Majer (Boston, 1991) 107-19.
26 See Ibid.
27 Frank P. Ramsey, ‘Review of W.E. Johnson’s Logic Part II,’ The New Statesman, 19 (July 29, 1922) 469.
28 Frank P. Ramsey, “Universals” (1925), reprinted in F.P. Ramsey: Philosophical Papers, edited by D.H. Mellor (Cambridge, 1990) 8-30.
29 See Ibid. See W.E. Johnson, Logic Part I (Cambridge, 1921); Logic Part II (Cambridge, 1922); Logic Part III (Cambridge, 1924).
30 Johnson, Logic Part I, 1.
31 Ibid., 3.
32 Ibid., 4.
33 Ibid., 6.
34 Ibid., 7-8.
35 Ibid., 8.
36 Ibid., 16-7.
37 Ibid., 52-3.
38 In NP, he claimed that “[t]he most certain thing about truth is that ‘p is true’ and ‘p’, if not identical, are equivalent.” (118) Cf. Ramsey, On Truth, 6-11.
39 FP, 39.
41 Ibid., 38.
42 On Truth, 3.
43 FP, 39.
44 Ibid., 34.
45 Ibid., 35.
46 Ibid., 34.
47 Ibid., 35-36.
48 Ibid., 36-38.
49 Ibid., 40.
51 Ibid., n. 1.
52 See On Truth, 11 and FP, 51.
53 FP, 40.
54 Ibid., 44.
55 Ibid., 46.
57 On Truth, 9.
58 Ibid., 11.
59 NP, 111.
60 Ibid., 114.
61 See his reasons for rejecting the existence of propositions in NP (107-114) and in On Truth (6-14).
62 FP, 39.
63 A.J.Ayer, Language, Truth and Logic (New York, 1936).
64 As Hartry Field notes: “A prevalent view in the early days of the Vienna circle was that the notions of truth and of truth conditions are a piece of useless metaphysics that we ought to abandon. For instance, legend has it that Neurath put ‘true’ on an ‘Index of Prohibited Words’. Ayer’s view in chapter 5 of Language, Truth and Logic was slightly less radical, but quite similar in spirit: he proposed that we allow talk of our utterances being true and having truth conditions, but that we give a deflationary account of such talk. Indeed, this feature of Ayer’s view is closely tied to his verifiability theory of meaning: it is essential to a verifiability theory of meaning and thought to dethrone truth-conditions from the central place in the theory of meaning and theory of thought that they had in the work of Frege, Russell, early Wittgenstein and Ramsey.” (2001, 484-485).
65 See in particular On Truth “Chapter 1: The Nature of Truth.”: 6-16.