Sunteți pe pagina 1din 17

A foundational perspective on the semantic and logical Paradoxes

Bhupinder Singh Anand Draft of April 10, 2012.


Abstract We argue that, from a foundational perspective, the paradoxical element in the familiar semantic and logical paradoxes is simply a reection of the attempt to ask of a language more than it is designed to deliver.
Keywords Algorithmic computability, algorithmic veriability, Aristotles particularisation, consistency, rst-order, -consistency, Peano Arithmetic PA, satisfaction, soundness, standard interpretation, Tarski.

The paradoxes

We presume familiarity with the semantic and logical paradoxes1 which involveeither implicitly or explicitlyquantication over an innitude. Where such quantication is not, or cannot be, explicitly dened in formal logical termseg. the classical expression of the Liar paradox as This sentence is a liethe paradoxes per se cannot be considered as posing serious linguistic or philosophical concerns. Of course, it would be a matter of serious concern if the word This in the English language sentence, This sentence is a lie, could be validly viewed as implicitly implying that: (i) there is a constructive innite enumeration of English language sentences; (ii) to each of which a truth-value can be constructively assigned by the rules of a two-valued logic; and, (iii) in which This refers uniquely to a particular sentence in the enumeration.
Commonly referred to as the paradoxes of self-reference, even though not all of them involve self-reference, e.g., the paradox constructed by Stephen Yablo [Ya93].
1

In 1931, Kurt Gdel used the above perspective in his seminal paper on o undecidable arithmetical propositions: (a) to show how the innitude of formulas, in a formally dened Peano Arithmetic P2 , could be constructively enumerated and referenced uniquely by natural numbers3 ; (b) to show how P-provability values could be constructively assigned to P-formulas by the rules of a two-valued logic4 ; and, (c) to construct a P-formula which interprets as an arithmetical proposition that could be viewedunder the standard interpretation of the Peano Arithmetic Pas expressing the sentence, This P-sentence is P-unprovable5 , without inviting a Liar type of contradiction. However, even where the quantication can be made expliciteg. Russells paradox or Yablos paradoxthe question arises whether such quantication is constructive or not. Russells paradox: Dene the set S by {All x : x S i x x}; / then S S i S S. / Yablos paradox: Dening the sentence Si for all i 0 as For all j > i, Sj is not true seems to lead to a contradiction6 . For instance, in Russells case it could be argued cogently that the contradiction itself establishes that S cannot be constructively dened over the range of the quantier. In Yablos case it could be argued that truth values cannot be constructively assigned to any sentence covered by the quantication since, in order to decide whether Si is true or not for any given i 0, we rst need to decide whether Si+1 is true or not. There are two issues involved herenot necessarily independent7 .
[Go31], pp.9-13. [Go31], p.13-14. 4 [Go31], p.13. 5 cf. [Go31], p.37, footnote 67. 6 [Ya93]. 7 A broader perspective on these issues is oered by W. T. Gowers in his talk, Does mathematics need a philosophy? , presented before the Cambridge University Society for the Philosophy of Mathematics and Mathematical Sciences, 2002, wherein he remarks that: If you ask a philosopher what the main problems are in the philosophy of mathematics, then the following two are likely to come up: what is the status of mathematical truth, and what is the nature of mathematical objects? That is, what gives mathematical statements their aura of infallibility, and what on earth are these statements about? .
3 2

Comment: The practical signicance of the semantic and logical paradoxes is, of course, that they illustrate the absurd extent to which languages of common discourse need to tolerate ambiguity; both for ease of expression and for practicaleven if not theoretically unambiguous and eective communication in non-critical cases amongst intelligences capable of a lingua franca. Such absurdity is highlighted by the universal appreciation of Charles Dickens Mr. Bumbles retort that The law is an ass; a quote oft used to refer to the absurdities which sometimes surface8 in cases when judicial pronouncements attempt to resolve an ambiguity by subjective at that appeals to the powersand dutiesbestowed upon the judicial authority for the practical resolution of precisely such an ambiguity, even when the ambiguity may be theoretically irresolvable! In a thought-provoking Opinion piece, Desperately Seeking Mathematical Truth, in the August 2008 Notices of the American Mathematical Society, Melvyn B. Nathanson seeks to highlight the signicance for the mathematical sciences when similar authority is vested by societyalbeit tacitlyupon academic bosses (a reference, presumably, to the collective of reputedand respectedexperts in any eld of human endeavour): ... many great and important theorems dont actually have proofs. They have sketches of proofs, outlines of arguments, hints and intuitions that were obvious to the author (at least, at the time of writing) and that, hopefully, are understood and believed by some part of the mathematical community. But the community itself is tiny. In most elds of mathematics there are few experts. Indeed, there are very few active research mathematicians in the world, and many important problems, so the ratio of the number of mathematicians to the number of problems is small. In every eld, there are bosses who proclaim the correctness or incorrectness of a new result, and its importance or unimportance. Sometimes they disagree, like gang leaders ghting over turf. In any case, there is a web of semi-proved theorems throughout mathematics. Our knowledge of the truth of a theorem depends on the correctness of its proof and on the correctness of all of the theorems used in its proof. It is a shaky foundation. . . . [Na08]. Nathansons comments are intriguing, because addressing such ambiguity in critical casessuch as communication between mechanical artefacts, or a putative communication between terrestrial and extra-terrestrial intelligences is the very raison detre of mathematical activity! Such activity we view, rst, as the construction of richer and richer mathematical languages9 that can express those of our abstract concepts which can
8 9

See www.shazbot.com/lawass/. Such as, for instance, the rst order set theory ZFC.

be subjectively10 addressed unambiguously; and, thereafter, the study of the ability of the mathematical languages11 to precisely express and objectively12 communicate these concepts eectively13 .

Truth: Is quantication currently interpreted constructively over an innite domain?

The rstand more signicantissue is whether the currently accepted interpretations of formal quanticationessentially as dened by David Hilbert (see Section 5) in his formalisation of Aristotles logic of predicates in terms of his -function 14 can be treated as constructive over an innite domain. Now, classical theory follows the 2000-year old philosophical perspective of Greek philosophers such as Aristotle, and asks only that the core axioms of a formal language, and its rules of inference, be interpretableunder Alfred Tarskis inductive denitions of the satisability, and truth, of the formulas of a formal language under an interpretation15 as self-evidently sound16 propositions. The subjectiveand essentially irresolvableelement in the soundness of such a viewpoint is self-evident.
In what follows, we use the words subjective and non-constructive inter-changeably. Such as, for instance, the rst order Peano Arithmetic PA. 12 In what follows, we use the words objective and constructive inter-changeably. 13 Our perspective on the expressions subjectively addressed unambiguously and objectively communicate . . . eectively is broadly that: By subjectively addressed unambiguously we intend in this context that there is essentially a subjective acceptance of identity by us between an original abstract concept in our mind that we intended to express symbolically in a language, and the interpreted abstract concept created in our mind each time we subsequently attempt to understand the import of the expression. By objectively communicate . . . eectively we intend in this context that there is essentially, rst, an objective acceptance of identity by another mind between the original abstract concept created in the other mind when rst attempting to understand the import of what we have expressed symbolically in a language, and the interpreted abstract concept created in the other mind each time it subsequently attempts to understand the import of the expression; and, second, an objective acceptance of functional identity between abstract concepts that can be objectively communicated eectively based on the evidence provided by a commonly accepted agency (as, for instance the viewgaining currency today ([Mu91])that a simple functional language can be used for specifying evidence for propositions in a constructive logic). 14 [Hi27]. 15 cf. [Me64], p.50. 16 The word sound is occasionally used informally as a synonym for reliable, but mostly as a technical term of mathematical logic (cf. [Me64], p.51). The distinction is assumed to be obvious from the context in which it is used.
11 10

L. E. J. Brouwer17 emphaticallyand justiably so far as number theory was concerned (see Section 5)objected to such subjectivity, and asserted that Hilberts interpretations of formal quantication were non-constructive. Although Hilberts formalisation of the quantiers (an integral part of his formalisation of Aristotles logic of predicates) appeared adequate, Brouwer rejected Hilberts interpretations of them on the grounds that the interpretations were open to ambiguity18 . However, Brouwers rejection of the Law of the Excluded Middle as a resolution of the objection was seenalso justiably19 as unconvincingly rejecting a comfortable interpretation thatdespite its Platonic overtones appeared intuitively plausible to the larger body of academics that was increasingly attracted to, and inuenced by, the remarkably expressive powers provided by Cantor-inspired set theories such as ZF. Since Brouwers seminal work preceded that of Alan Turing, it was unable to oer his critics an alternativeand intuitively convincingconstructive denition of quantication based on the viewgaining currency todaythat a simple functional language can be used for specifying evidence for propositions in a constructive logic20 . Moreover, since Brouwers objections did not gain much currency amongst mainstream logicians, they were unable to inuence Turing who, it is our contention, could easily have provided the necessary constructive interpretations21 sought by Hilbert for number theory, had Turing not been inuenced by Gdels powerful presentationand Gdels persuasive Platonic, albeit o o (contrary to accepted dogma) logically rooted22 , interpretation of his own
[Br08]. They could not, therefore, be accepted as admitting eective communication. 19 Reason: Although Hilberts formalisation and interpretation of the quantiers implies that the Law of the Excluded Middle must hold in any model of the formalisation, the converse is not true. 20 [Mu91]. 21 Detailed in [An12]. 22 Although meriting a more complete discussion than is appropriate to the intent of this paper, it is worth noting that the roots of of Gdels Platonism can be cogently o argued as lyingcontrary to generally held opinionspurely in an axiomatic logical, rather than philosophical, presumption: more specically in Gdels implicit (but, cuo riously, explicitly suggested! ) presumption that Peano Arithmetic is (unquestionably? ) -consistent ([Go31], p.28). The presumption is unwittingly shared universally even by those who (cf. [Pa95], [Ff02]) accept Gdels formal arguments in [Go31] but claim to o reject Gdels Platonic interpretations of them. Reason: Gdels presumption is equivo o alent to the premisealmost universally unchallenged hitherto in mainstream literature, and excepted to largely by ultra-nitists such as Alexander Yessenin-Volpin (as outlined in [He04] )that the standard interpretation of rst-order PA may be assumed to yield a valid model for number theory. Thus (as we detail in the unpublished manuscript
18 17

formal reasoning in his (Gdels) seminal 1931 paper on formally undecido able arithmetical propositions23 . Thus, in his 1939 paper24 on ordinal-based logics, Turing applied his computational methodwhich he had developed in his 1936 paper25 in seeking partial completeness in interpretations of Cantors ordinal arithmetic (as dened in a set theory such as ZF)26 rather than in seeking a categorical interpretation of PA27 . Turing perhaps viewed his 1936 paper28 as complementing and extending Gdels and Cantors reasoning. o It is our contention that Turing thus overlooked the fact that his 1936 paper29 actually conicts with Gdels and Cantors interpretations of their o own, formal, reasoning by admitting an objective denition of satisfaction that yields a sound, nitary, interpretation IP A(N , Algorithmic) of PA30 . Specically, whereas Gdels and Cantors reasoning implicitly presumes o that satisfaction under the standard interpretation IP A(N , Standard) of PA can only be dened non-constructively in terms of subjectively veriable truth31 , it can be cogently argued that satisfaction under both IP A(N , Standard) and
http://alixcomsi.com/30 Finitary Logic PA Update.pdf) Gdels Platonism essentially reects the o fact that PA is -consistent if, and only if, Aristotles particularisation is presumed to always hold under any interpretation of the associated logic, and that the standard interpretation of PA is a model of PA if, and only if, PA is -consistent. In this respect, Gdel was only endorsing Hilberts essentially Platonic formalisation of Aristotles Logic o of Predicates ([Hi27], p.466.) as valid. 23 [Go31]. 24 [Tu39]. 25 [Tu36]. 26 [Tu39], pp.155-156 . . . The well-known theorem of Gdel shows that every system of o logic is in a certain sense incomplete, but at the same time it indicates means whereby from a system L of logic a more complete system L may be obtained. By repeating the process we get a sequence of L, L1 = L , L2 = L1 , . . . each more complete than the preceding. . . . Proceeding in this way we can associate a system of logic with any constructive ordinal. It may be asked whether a sequence of logics of this kind is complete in the sense that to any problem A there corresponds an ordinal such that A is solvable by means of the logic L . I propose to investigate this question in a more general case, and to give some other examples of ways in which systems of logic may be associated with constructive ordinals. 27 Perhaps since such a search would involve questioning either the validity of Gdels o belief that systems of Arithmetic such as PA are -consistent (as hinted at in [Go31], p.28), or Gdels interpretation of his argument in [Go31] as having meta-mathematically o proven that systems of Arithmetic such as PA are essentially incomplete! 28 [Tu36]. 29 [Tu36]. 30 Detailed in the unpublished manuscript http://alixcomsi.com/30 Finitary Logic PA Update.pdf. 31 A view (such as in http://en.wikipedia.org/wiki/Tarskis undenability theorem) reected in the not uncommon interpretation of Tarskis Theorem ([Me64], p.151) as establishing the formal undenability of arithmetical truth in arithmetic. We show in [An07] why such an interpretation is untenable.

IP A(N , Algorithmic) is denable constructively in terms of objectively veriable Turing-computability32 . As a result, current theory continuedand continues to this dayto essentially follow Hilberts Platonically-inuenced (hence, subjective) denitions and interpretations of the quantiers (based on Aristotles logic of predicates) when dening them under the standard interpretation33 IP A(N , Standard) of formal number theory. Now, the latter denitions and interpretations34 are, in turn, founded upon Tarskis analysis of the inductive denability of the truth of compound expressions of a symbolic language under an interpretation in terms of the satisfaction of the atomic expressions of the language under the interpretation35 . Tarski denes the formal sentence P as True if and only if pwhere p is the proposition expressed by P . In other words, the sentence Snow is white is True if, and only if, it is subjectively true in all cases; and it is subjectively true in a particular case if, and only if, it expresses the subjectively veriable fact that snow is white in that particular case. Thus, for Tarski the commonality of the satisfaction of the atomic formulas of a language under an interpretation is axiomatic36 . We highlight the limitations of such subjectivity elsewhere37 and, in the case of the standard interpretation of the Peano Arithmetic PA, show how to avoid them by requiring that the core axioms of PA, and its rules of inference, be interpretable as algorithmically (and, ipso facto, objectively) veriable sound propositions.

Mathematical Objects: When is the concept of a completed innity consistent with a formal language?

The second issue is when, and whether, the concept of a completed innity is consistent with the interpretation of a formal language.
See Section 5.2. cf. [Me64], p.107; [Sh67], p.23, p.209; [BBJ03], p.104. For purposes of this paper, we treat [Me64] as a reliable and representative expositionwhere citedof the standard logical concepts and theory that are addressed at that point. 34 eg. [Me64], pp.49-53. 35 [Ta33]. 36 cf. [Me64], p.51(i). 37 In the unpublished manuscript http://alixcomsi.com/30 Finitary Logic PA Update.pdf
33 32

Clearly, the consistency of the concept would follow immediately in any sound interpretation38 of the axioms (and rules of inference) of a set theory such as ZF (whether such an interpretation exists at all is, of course, another question). In view of the perceived power of ZF39 as an unsurpassed language of rich and adequate expression of mathematically expressible abstract concepts precisely, it is not surprising that many of the semantic and logical paradoxes depend on the implicit assumption that the domain over which the paradox quanties can always be treated as a well-dened mathematical object that can be formalised in ZF, even if this domain is not explicitly dened settheoretically. This assumption is rooted in the questionable belief that ZF can express all mathematical truths40 . From this it is but a short step to the non-constructive argumentrooted in Gdels Platonic interpretation of his own formal reasoning in his 1931 o paper41 that PA must have non-standard models. However, it is our contention that both of the above foundational issues need to be reviewed carefully, and that we need to recognize explicitly the limitations on the ability of highly expressive mathematical languages such as ZF to communicate eectively; and the limitations on the ability of eectively communicating mathematical languages such as PA to adequately express abstract conceptssuch as those involving Cantors rst limit ordinal 42 . Prima facie, the semantic and logical paradoxesas also the seeming paradoxes associated with fractal constructions such as the Cantor ternary set, and the constructions described belowseem to arise out of a blurring of this distinction, and an attempt to ask of a language more than it is designed to deliver.
We dene an interpretation IS of a formal system S as sound if, and only if, every provable formula [F ] of S translates as a true proposition F under the interpretation (cf. [BBJ03], p.174). 39 More accurately, ZFC. 40 In the unpublished manuscript http://alixcomsi.com/10 Goodstein case against 1000.pdf we show howin the case of Goodsteins Theoremsuch a belief leads to a curious argument. 41 [Go31]. 42 In the unpublished manuscript http://alixcomsi.com/10 Goodstein case against 1000.pdf we show why we cannot add a symbol corresponding formally to the concept of an innite mathematical entity (such as is referred to symbolically in the literature by or ) to the rst-order Peano Arithmetic PA without inviting inconsistency.
38

A paradoxical fractal construction

For instance, consider the claim that fractal43 constructionssuch as the Cantor ternary set44 yield (presumably in some interpretable sense) valid mathematical objects (sets) in the limit. We consider an equilateral triangle BAC of height h and side s. Divide the base BC in half and construct two isosceles triangles of height h.d and base s/2 on BC, where 1 d > 0. Iterate the construction on each constructed triangle ad innitum. Thus, the height of each of the 2n triangles on the base BC at the nth construction is h.dn , and the base of each triangle s/2n . Hence, the total area of all these triangles subtended by the base BC is s.h.dn /2. Now, if d = 1, the total area of all the constructed triangles after each iteration remains constant at s.h/2, although the total length of all the sides opposing the base BC increases monotonically. Moreover, if 1 > d > 0, it would appear that the base BC of the original equilateral triangle will always be the limiting conguration of the sides opposing the base BC.
[Ba88]. Dened as a set-theoretical limit of an iterative process in the completion of a metric space; see, for instance, [Ru53], p54.
44 43

A
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J B C Fig. 1: ln s if d < 1/2

This is indeed so if 0 < d < 1/2, since (see g. 1 above) the total length of all the sides opposing the base BC at the nth iterationsay ln yields a Cauchy sequence whose limiting value is, indeed, the length s of the base BC.

10

A
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J J
J

J
J
J
J
J
J

J J
J
J

J J
J J
J
J
J



J
J
J
J
J
J
J
J
J J J
J
J
J

J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J J
J
J
J
J
J
J
J
J
J
J
J
J B C Fig. 2: ln = 2s if d = 1/2

However, if d = 1/2, the total length (see g. 2 above) of all the sides opposing their base on BC is always 2s!

11

A
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J B C Fig. 3: ln if d > 1/2

Finally, if d > 1/2, the total length (see g. 3 above) of all the sides opposing their base on BC is a monotonically increasing value. Consider now: Case 1 Treating the above iteration as the fragmentation of a land-holding, how would one interpret the limit of such an interpretation (which is postulated as existing in a putative completion of Euclidean space)? Case 2 Let s be one light-year and consider how long it would take a light signal to travel from B to C along the sides opposing the base in each of the above cases. Case 3 Let the area BAC denote the population size of a virus cluster, where each virus cell has a virulence measure h/s. Let each triangle at the nth iteration denote a virus clusterwith a virulence factor h.dn /(s/2n )that reacts to the next generation anti-virus by splitting into two smaller clusters with inherited virulence h.dn+1 /(s/2n+1 ). If d < 1/2, the eects of the virus canin a sensebe contained and eventually eliminated. If d = 1/2, the

12

eects of the virus can be contained, but never eliminated. However, if d > 1/2, the eects of the virus can neither be contained nor eliminated. The question arises: Since the raison detre of a mathematical language isor should be45 to express our abstractions of natural phenomena precisely and communicate them unequivocally, in what sense can we sensibly admit an interpretation of a mathematical language that asserts all the above cases as having limiting congurations in a putative completion of Euclidean Space?

Appendix: The -function


IV. The logical -axiom 13. A(a) A((A)) Here (A) stands for an object of which the proposition A(a) certainly holds if it holds of any object at all; let us call the logical -function. . . . 1. By means of , all and there exists can be dened, namely, as follows: (i) (a)A(a) A((A)) (ii) (a)A(a) A((A)) . . . On the basis of this denition the -axiom IV(13) yields the logical relations that hold for the universal and the existential quantier, such as: (a)A(a) A(b) . . . (Aristotles dictum), and: ((a)A(a)) (a)(A(a)) . . . (principle of excluded middle).

Hilbert interpreted quantication in terms of his -function as follows46 :

Thus, Hilberts interpretation of universal quantication dened in (i) is that the sentence (x)F (x) holds (under a consistent interpretation I) if, and only if, F (a) holds whenever F (a) holds for any given a (in I); hence
See, for instance, Richard P. Feynmans delightful commentary ([Fe85]) on his horrifying experiences of how the teaching of mathematical languages was actually introduced into the schools curricula in the State of California in 1964. 46 [Hi27], p.466.
45

13

F (a) does not hold for any a (since I is consistent), and so F (a) holds for any given a (in I). Further, Hilberts interpretation of existential quantication dened in (ii) is that (x)F (x) holds (in I) if, and only if, F (a) holds for some a (in I). Brouwers objection to such an unqualied interpretation of the existential quantier was that, for the interpretation to be considered sound when the domain of the quantiers under an interpretation is innite, the decidability of the quantication under the interpretation must be constructively veriable in some intuitively and mathematically acceptable sense of the term constructive47 . Two questions arise: (a) Is Brouwers objection relevant today? (b) If so, can we interpret quantication constructively ?

5.1

The standard interpretation of PA

We consider the standard interpretation, say IP A(N , Standard) , of rst-order Peano Arithmetic (PA). Now, if [(x)F (x)] and [(x)F (x)] are PA-formulas, and the relation F (x) is the interpretation under IP A(N , Standard) of the PA-formula [F (x)], then, in current literature: (1a) [(x)F (x)] is dened as true under IP A(N , Standard) if, and only if, for any given natural number n, the assertion denoted by F (n) holds under IP A(N , Standard) ; (1b) [(x)F (x)] is an abbreviation of [(x)F (x)], and is dened as true under IP A(N , Standard) if, and only if, it is not the case that, for any given natural number n, the assertion denoted by F (n) holds under IP A(N , Standard) ; (1c) F (n) holds under IP A(N , Standard) for some natural number n if, and only if, it is not the case that, for any given natural number n, the assertion denoted by F (n) holds under IP A(N , Standard) . Since (1a), (1b) and (1c) together interpret [(x)F (x)] and [(x)F (x)] under IP A(N , Standard) as intended by Hilberts -function, they attract Brouwers objection. This answers question (a).
47

[Br08].

14

5.2

A nitary model of PA

Clearly, the specic target of Brouwers objection is (1c), which appeals to Platonically non-constructive, rather than intuitively constructive, plausibility. We can thus re-phrase question (b) more specically: Can we dene an interpretation of PA over N that does not appeal to (1c)? Now, it follows from Turings seminal 1936 paper on computable numbers that every quantier-free arithmetical function (or relation, when interpreted as a Boolean function) F denes a Turing-machine TMF 48 Consequently we can, in principle, dene another interpretation IP A(N , Algorithmic) over the structure N , where: (2a) [(x)F (x)] is dened as true under IP A(N , Algorithmic) if, and only if, the Turing-machine TMF evidences 49 the assertion denoted by F (n) as always true (i.e., as true for any given natural number n) under IP A(N , Algorithmic) ; (2b) [(x)F (x)] is an abbreviation of [(x)F (x)], and is dened as true under IP A(N , Algorithmic) if, and only if, it is not the case that the Turing-machine TMF evidences the assertion denoted by F (n) as always false (i.e., as false for any given natural number n) under IP A(N , Algorithmic) . IP A(N , Algorithmic) is a nitary model of PA since we argue elsewhere50 thatwhen interpreted suitablyall theorems of rst-order PA are constructively true under IP A(N , Algorithmic) . This answers question (b).

References
[Ba88] Michael Barnsley. 1988. Fractals Everywhere. Academic Press, Inc. San Diego.

[BBJ03] George S. Boolos, John P. Burgess, Richard C. Jerey. 2003. Computability and Logic (4th ed). Cambridge University Press, Cambridge. [Br08]
48 50

L. E. J. Brouwer. 1908. The Unreliability of the Logical Principles. English translation in A. Heyting, Ed. L. E. J. Brouwer: Collected

[Tu36], pp. 138-139. In the unpublished manuscript http://alixcomsi.com/30 Finitary Logic PA Update.pdf.

15

Works 1: Philosophy and Foundations of Mathematics. Amsterdam: North Holland / New York: American Elsevier (1975): pp. 107-111. [Fe85] Richard P. Feynman. 1985. Judging Books by Their Covers in Surely Youre Joking, Mr. Feynman! (Adventures of a curious character). Norton, New York. (Extracts: Judging Books by Their Covers.) Solomon Feferman. 2002. Predicativity. Source: http://math.stanford.edu/ feferman/papers/predicativity.pdf .

[Ff02]

[Go31] Kurt Gdel. 1931. On formally undecidable propositions of Principia o Mathematica and related systems I. Translated by Elliott Mendelson. In M. Davis (ed.). 1965. The Undecidable. Raven Press, New York. [He04] Catherine Christer-Hennix. 2004. Some remarks on Finitistic Model Theory, Ultra-Intuitionism and the main problem of the Foundation of Mathematics. ILLC Seminar, 2nd April 2004, Amsterdam. David Hilbert. 1927. The Foundations of Mathematics. Text of an address delivered in July 1927 at the Hamburg Mathematical Seminar. In Jean van Heijenoort. 1967. Ed. From Frege to Gdel: A o source book in Mathematical Logic, 1878 - 1931. Harvard University Press, Cambridge, Massachusetts.

[Hi27]

[Me64] Elliott Mendelson. 1964. Introduction to Mathematical Logic. Van Norstrand. pp.145-146. [Mu91] Chetan R. Murthy. 1991. An Evaluation Semantics for Classical Proofs. Proceedings of Sixth IEEE Symposium on Logic in Computer Science, pp. 96-109, (also Cornell TR 91-1213), 1991. [Na08] Melvyn B. Nathanson. 2008. Desperately Seeking Mathematical Truth. Opinion in the August 2008 Notices of the American Mathematical Society, Vol. 55, Issue 7. Charles Parsons. 1995. Platonism and Mathematical Intuition in Kurt Gdels Thought. The Bulletin of Symbolic Logic, Volume 1, o Number 1, March 1995, pp. 44-74.

[Pa95]

[Ru53] Walter Rudin. 1953. Principles of Mathematical Analysis. McGraw Hill, New York. [Sh67] Joseph R. Shoeneld. 1967. Mathematical Logic. Reprinted 2001. A. K. Peters Ltd., Massachusetts. 16

[Ta33]

Alfred Tarski. 1933. The concept of truth in the languages of the deductive sciences. In Logic, Semantics, Metamathematics, papers from 1923 to 1938 (p152-278). ed. John Corcoran. 1983. Hackett Publishing Company, Indianapolis. Alan Turing. 1936. On computable numbers, with an application to the Entscheidungsproblem. In M. Davis (ed.). 1965. The Undecidable. Raven Press, New York. Reprinted from the Proceedings of the London Mathematical Society, ser. 2. vol. 42 (1936-7), pp.230265; corrections, Ibid, vol 43 (1937) pp. 544-546. Alan Turing. 1939. Systems of logic based on ordinals. In M. Davis (ed.). 1965. The Undecidable. Raven Press, New York. Reprinted from the Proceedings of the London Mathematical Society, ser. 2. vol. 45 (1939), pp.161-228. Stephen Yablo. 1993. Paradox without self-reference. Analysis, 53(4), pp. 251-252.

[Tu36]

[Tu39]

[Ya93]

[An07] Bhupinder Singh Anand. 2007. Why we shouldnt fault Lucas and Penrose for continuing to believe in the Gdelian argument against o computationalism - I . The Reasoner, Vol(1)6 p3-4. [An12] Bhupinder Singh Anand. 2012. Evidence-Based Interpretations of PA. Updated draft of the paper accepted for the Symposium on Computational Philosophy at the AISB/IACAP World Congress 2012-Alan Turing 2012, 2-6 July 2012, University of Birmingham, Birmingham, UK.

Authors Postal address: 32 Agarwal House, D Road, Churchgate, Mumbai - 400 020, Maharashtra, India. Email: re@alixcomsi.com, anandb@vsnl.com.

50

17