26 ideas
17641 | Discoveries in mathematics can challenge philosophy, and offer it a new foundation [Russell] |
Full Idea: Any new discovery as to mathematical method and principles is likely to upset a great deal of otherwise plausible philosophising, as well as to suggest a new philosophy which will be solid in proportion as its foundations in mathematics are securely laid. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.283) | |
A reaction: This is a manifesto for modern analytic philosophy. I'm not convinced, especially if a fictionalist view of maths is plausible. What Russell wants is rigour, but there are other ways of getting that. Currently I favour artificial intelligence. |
17638 | If one proposition is deduced from another, they are more certain together than alone [Russell] |
Full Idea: Two obvious propositions of which one can be deduced from the other both become more certain than either in isolation; thus in a complicated deductive system, many parts of which are obvious, the total probability may become all but absolute certainty. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.279) | |
A reaction: Thagard picked this remark out, in support of his work on coherence. |
17632 | Non-contradiction was learned from instances, and then found to be indubitable [Russell] |
Full Idea: The law of contradiction must have been originally discovered by generalising from instances, though, once discovered, it was found to be quite as indubitable as the instances. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.274) |
10147 | The Axiom of Choice is consistent with the other axioms of set theory [Feferman/Feferman] |
Full Idea: In 1938 Gödel proved that the Axiom of Choice is consistent with the other axioms of set theory. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int I) | |
A reaction: Hence people now standardly accept ZFC, rather than just ZF. |
10148 | Axiom of Choice: a set exists which chooses just one element each of any set of sets [Feferman/Feferman] |
Full Idea: Zermelo's Axiom of Choice asserts that for any set of non-empty sets that (pairwise) have no elements in common, then there is a set that 'simultaneously chooses' exactly one element from each set. Note that this is an existential claim. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int I) | |
A reaction: The Axiom is now widely accepted, after much debate in the early years. Even critics of the Axiom turn out to be relying on it. |
10149 | Platonist will accept the Axiom of Choice, but others want criteria of selection or definition [Feferman/Feferman] |
Full Idea: The Axiom of Choice seems clearly true from the Platonistic point of view, independently of how sets may be defined, but is rejected by those who think such existential claims must show how to pick out or define the object claimed to exist. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int I) | |
A reaction: The typical critics are likely to be intuitionists or formalists, who seek for both rigour and a plausible epistemology in our theory. |
10150 | The Trichotomy Principle is equivalent to the Axiom of Choice [Feferman/Feferman] |
Full Idea: The Trichotomy Principle (any number is less, equal to, or greater than, another number) turned out to be equivalent to the Axiom of Choice. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int I) | |
A reaction: [He credits Sierpinski (1918) with this discovery] |
10146 | Cantor's theories needed the Axiom of Choice, but it has led to great controversy [Feferman/Feferman] |
Full Idea: The Axiom of Choice is a pure existence statement, without defining conditions. It was necessary to provide a foundation for Cantor's theory of transfinite cardinals and ordinal numbers, but its nonconstructive character engendered heated controversy. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int I) |
10158 | A structure is a 'model' when the axioms are true. So which of the structures are models? [Feferman/Feferman] |
Full Idea: A structure is said to be a 'model' of an axiom system if each of its axioms is true in the structure (e.g. Euclidean or non-Euclidean geometry). 'Model theory' concerns which structures are models of a given language and axiom system. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int V) | |
A reaction: This strikes me as the most interesting aspect of mathematical logic, since it concerns the ways in which syntactic proof-systems actually connect with reality. Tarski is the central theoretician here, and his theory of truth is the key. |
10162 | Tarski and Vaught established the equivalence relations between first-order structures [Feferman/Feferman] |
Full Idea: In the late 1950s Tarski and Vaught defined and established basic properties of the relation of elementary equivalence between two structures, which holds when they make true exactly the same first-order sentences. This is fundamental to model theory. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int V) | |
A reaction: This is isomorphism, which clarifies what a model is by giving identity conditions between two models. Note that it is 'first-order', and presumably founded on classical logic. |
10160 | Löwenheim-Skolem says if the sentences are countable, so is the model [Feferman/Feferman] |
Full Idea: The Löwenheim-Skolem Theorem, the earliest in model theory, states that if a countable set of sentences in a first-order language has a model, then it has a countable model. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int V) | |
A reaction: There are 'upward' (sentences-to-model) and 'downward' (model-to-sentences) versions of the theory. |
10159 | Löwenheim-Skolem Theorem, and Gödel's completeness of first-order logic, the earliest model theory [Feferman/Feferman] |
Full Idea: Before Tarski's work in the 1930s, the main results in model theory were the Löwenheim-Skolem Theorem, and Gödel's establishment in 1929 of the completeness of the axioms and rules for the classical first-order predicate (or quantificational) calculus. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int V) |
17629 | Which premises are ultimate varies with context [Russell] |
Full Idea: Premises which are ultimate in one investigation may cease to be so in another. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.273) |
17630 | The sources of a proof are the reasons why we believe its conclusion [Russell] |
Full Idea: In mathematics, except in the earliest parts, the propositions from which a given proposition is deduced generally give the reason why we believe the given proposition. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.273) |
17640 | Finding the axioms may be the only route to some new results [Russell] |
Full Idea: The premises [of a science] ...are pretty certain to lead to a number of new results which could not otherwise have been known. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.282) | |
A reaction: I identify this as the 'fruitfulness' that results when the essence of something is discovered. |
10161 | If a sentence holds in every model of a theory, then it is logically derivable from the theory [Feferman/Feferman] |
Full Idea: Completeness is when, if a sentences holds in every model of a theory, then it is logically derivable from that theory. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int V) |
10156 | 'Recursion theory' concerns what can be solved by computing machines [Feferman/Feferman] |
Full Idea: 'Recursion theory' is the subject of what can and cannot be solved by computing machines | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Ch.9) | |
A reaction: This because 'recursion' will grind out a result step-by-step, as long as the steps will 'halt' eventually. |
10155 | Both Principia Mathematica and Peano Arithmetic are undecidable [Feferman/Feferman] |
Full Idea: In 1936 Church showed that Principia Mathematica is undecidable if it is ω-consistent, and a year later Rosser showed that Peano Arithmetic is undecidable, and any consistent extension of it. | |
From: Feferman / Feferman (Alfred Tarski: life and logic [2004], Int IV) |
17627 | It seems absurd to prove 2+2=4, where the conclusion is more certain than premises [Russell] |
Full Idea: It is an apparent absurdity in proceeding ...through many rather recondite propositions of symbolic logic, to the 'proof' of such truisms as 2+2=4: for it is plain that the conclusion is more certain than the premises, and the supposed proof seems futile. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.272) | |
A reaction: Famously, 'Principia Mathematica' proved this fact at enormous length. I wonder if this thought led Moore to his common sense view of his own hand - the conclusion being better than the sceptical arguments? |
17628 | Arithmetic was probably inferred from relationships between physical objects [Russell] |
Full Idea: When 2 + 2 =4 was first discovered, it was probably inferred from the case of sheep and other concrete cases. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.272) |
17637 | The most obvious beliefs are not infallible, as other obvious beliefs may conflict [Russell] |
Full Idea: Even where there is the highest degree of obviousness, we cannot assume that we are infallible - a sufficient conflict with other obvious propositions may lead us to abandon our belief, as in the case of a hallucination afterwards recognised as such. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.279) | |
A reaction: This approach to fallibilism seems to arise from the paradox that undermined Frege's rather obvious looking axioms. After Peirce and Russell, fallibilism has become a secure norm of modern thought. |
17639 | Believing a whole science is more than believing each of its propositions [Russell] |
Full Idea: Although intrinsic obviousness is the basis of every science, it is never, in a fairly advanced science, the whole of our reason for believing any one proposition of the science. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.279) |
17631 | Induction is inferring premises from consequences [Russell] |
Full Idea: The inferring of premises from consequences is the essence of induction. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.274) | |
A reaction: So induction is just deduction in reverse? Induction is transcendental deduction? Do I deduce the premises from observing a lot of white swans? Hm. |
13304 | Learned men gain more in one day than others do in a lifetime [Posidonius] |
Full Idea: In a single day there lies open to men of learning more than there ever does to the unenlightened in the longest of lifetimes. | |
From: Posidonius (fragments/reports [c.95 BCE]), quoted by Seneca the Younger - Letters from a Stoic 078 | |
A reaction: These remarks endorsing the infinite superiority of the educated to the uneducated seem to have been popular in late antiquity. It tends to be the religions which discourage great learning, especially in their emphasis on a single book. |
17633 | The law of gravity has many consequences beyond its grounding observations [Russell] |
Full Idea: The law of gravitation leads to many consequences which could not be discovered merely from the apparent motions of the heavenly bodies. | |
From: Bertrand Russell (Regressive Method for Premises in Mathematics [1907], p.275) |
20820 | Time is an interval of motion, or the measure of speed [Posidonius, by Stobaeus] |
Full Idea: Posidonius defined time thus: it is an interval of motion, or the measure of speed and slowness. | |
From: report of Posidonius (fragments/reports [c.95 BCE]) by John Stobaeus - Anthology 1.08.42 | |
A reaction: Hm. Can we define motion or speed without alluding to time? Looks like we have to define them as a conjoined pair, which means we cannot fully understand either of them. |