28 ideas
| 6118 | Philosophy is logical analysis, followed by synthesis [Russell] |
| Full Idea: The business of philosophy, as I conceive it, is essentially that of logical analysis, followed by logical synthesis. | |
| From: Bertrand Russell (Logical Atomism [1924], p.162) | |
| A reaction: I am uneasy about Russell's hopes for the contribution that logic could make, but I totally agree that analysis is the route to wisdom, and I take Aristotle as my role model of an analytical philosopher, rather than the modern philosophers of logic. |
| 6116 | A logical language would show up the fallacy of inferring reality from ordinary language [Russell] |
| Full Idea: We are trying to create a perfectly logical language to prevent inferences from the nature of language to the nature of the world, which are fallacious because they depend upon the logical defects of language. | |
| From: Bertrand Russell (Logical Atomism [1924], p.159) | |
| A reaction: Wittgenstein seems to have rebelled against this idea, so that one strand of his later philosophy leads to 'ordinary language' philosophy, which is exactly what Russell is criticising. Wittgenstein seems to have seen 'logical language' as an oxymoron. |
| 6117 | Philosophy should be built on science, to reduce error [Russell] |
| Full Idea: We would be wise to build our philosophy upon science, because the risk of error in philosophy is pretty sure to be greater than in science. | |
| From: Bertrand Russell (Logical Atomism [1924], p.160) | |
| A reaction: If you do very little, it reduces the 'risk of error'. I agree that philosophers should start from the facts, and be responsive to new facts, and that science is excellent at discovering facts. But I don't think cognitive science is the new epistemology. |
| 18806 | Frege thought traditional categories had psychological and linguistic impurities [Frege, by Rumfitt] |
| Full Idea: Frege rejected the traditional categories as importing psychological and linguistic impurities into logic. | |
| From: report of Gottlob Frege (Function and Concept [1891]) by Ian Rumfitt - The Boundary Stones of Thought 1.2 | |
| A reaction: Resisting such impurities is the main motivation for making logic entirely symbolic, but it doesn't follow that the traditional categories have to be dropped. |
| 6110 | Subject-predicate logic (and substance-attribute metaphysics) arise from Aryan languages [Russell] |
| Full Idea: It is doubtful whether the subject-predicate logic, with the substance-attribute metaphysic, would have been invented by people speaking a non-Aryan language. | |
| From: Bertrand Russell (Logical Atomism [1924], p.151) | |
| A reaction: This is not far off the Sapir-Whorf Hypothesis (e.g. Idea 3917), which Russell would never accept. I presume that Russell would see true logic as running deeper, and the 'Aryan' approach as just one possible way to describe it. |
| 6107 | It is logic, not metaphysics, that is fundamental to philosophy [Russell] |
| Full Idea: I hold that logic is what is fundamental in philosophy, and that schools should be characterised rather by their logic than by their metaphysics. | |
| From: Bertrand Russell (Logical Atomism [1924], p.143) | |
| A reaction: Personally I disagree. Russell seems to have been most interested in the logical form underlying language, but that seems to be because he was interested in the ontological implications of what we say, which is metaphysics. |
| 6115 | Vagueness, and simples being beyond experience, are obstacles to a logical language [Russell] |
| Full Idea: The fact that we do not experience simples is one obstacle to the actual creation of a correct logical language, and vagueness is another. | |
| From: Bertrand Russell (Logical Atomism [1924], p.159) | |
| A reaction: The dream of creating a perfect logical language looks doomed from the start, but it is a very interesting project to try to pinpoint why it is unlikely to be possible. I say a perfect language cuts nature exactly at the joints, so find the joints. |
| 8490 | First-level functions have objects as arguments; second-level functions take functions as arguments [Frege] |
| Full Idea: Just as functions are fundamentally different from objects, so also functions whose arguments are and must be functions are fundamentally different from functions whose arguments are objects. The latter are first-level, the former second-level, functions. | |
| From: Gottlob Frege (Function and Concept [1891], p.38) | |
| A reaction: In 1884 he called it 'second-order'. This is the standard distinction between first- and second-order logic. The first quantifies over objects, the second over intensional entities such as properties and propositions. |
| 8492 | Relations are functions with two arguments [Frege] |
| Full Idea: Functions of one argument are concepts; functions of two arguments are relations. | |
| From: Gottlob Frege (Function and Concept [1891], p.39) | |
| A reaction: Nowadays we would say 'two or more'. Another interesting move in the aim of analytic philosophy to reduce the puzzling features of the world to mathematical logic. There is, of course, rather more to some relations than being two-argument functions. |
| 6109 | Some axioms may only become accepted when they lead to obvious conclusions [Russell] |
| Full Idea: Some of the premisses (of my logicist theory) are much less obvious than some of their consequences, and are believed chiefly because of their consequences. This will be found to be always the case when a science is arranged as a deductive system. | |
| From: Bertrand Russell (Logical Atomism [1924], p.145) | |
| A reaction: We shouldn't assume the model of self-evident axioms leading to surprising conclusions, which is something like the standard model for rationalist foundationalists. Russell nicely points out that the situation could be just the opposite |
| 8487 | Arithmetic is a development of logic, so arithmetical symbolism must expand into logical symbolism [Frege] |
| Full Idea: I am of the opinion that arithmetic is a further development of logic, which leads to the requirement that the symbolic language of arithmetic must be expanded into a logical symbolism. | |
| From: Gottlob Frege (Function and Concept [1891], p.30) | |
| A reaction: This may the the one key idea at the heart of modern analytic philosophy (even though logicism may be a total mistake!). Logic and arithmetical foundations become the master of ontology, instead of the servant. The jury is out on the whole enterprise. |
| 6108 | Maths can be deduced from logical axioms and the logic of relations [Russell] |
| Full Idea: I think that no one will dispute that from certain ideas and axioms of formal logic, but with the help of the logic of relations, all pure mathematics can be deduced. | |
| From: Bertrand Russell (Logical Atomism [1924], p.145) | |
| A reaction: It has been said for a long time that Gödel's Incompleteness Theorems of 1930 disproved this claim, though recently there have been defenders of logicism. Beginning with 'certain ideas' sounds like begging the question. |
| 18899 | Frege takes the existence of horses to be part of their concept [Frege, by Sommers] |
| Full Idea: Frege regarded the existence of horses as a property of the concept 'horse'. | |
| From: report of Gottlob Frege (Function and Concept [1891]) by Fred Sommers - Intellectual Autobiography 'Realism' |
| 10968 | Russell gave up logical atomism because of negative, general and belief propositions [Russell, by Read] |
| Full Idea: Russell preceded Wittgenstein in deciding that the reduction of all propositions to atomic propositions could not be achieved. The problem cases were negative propositions, general propositions, and belief propositions. | |
| From: report of Bertrand Russell (Logical Atomism [1924]) by Stephen Read - Thinking About Logic Ch.1 |
| 6113 | To mean facts we assert them; to mean simples we name them [Russell] |
| Full Idea: The way to mean a fact is to assert it; the way to mean a simple is to name it. | |
| From: Bertrand Russell (Logical Atomism [1924], p.156) | |
| A reaction: Thus logical atomism is a linguistic programme, of reducing our language to a foundation of pure names. The recent thought of McDowell and others is aimed at undermining any possibility of a 'simple' in perception. The myth of 'The Given'. |
| 6114 | 'Simples' are not experienced, but are inferred at the limits of analysis [Russell] |
| Full Idea: When I speak of 'simples' I am speaking of something not experienced as such, but known only inferentially as the limits of analysis. | |
| From: Bertrand Russell (Logical Atomism [1924], p.158) | |
| A reaction: He claims that the simples are 'known', so he does not mean purely theoretical entities. They have something like the status of quarks in physics, whose existence is inferred from experience. |
| 21722 | Better to construct from what is known, than to infer what is unknown [Russell] |
| Full Idea: Whenever possible, substitute constructions out of known entities for inferences to unknown entities. | |
| From: Bertrand Russell (Logical Atomism [1924], p.161), quoted by Bernard Linsky - Russell's Metaphysical Logic 7 | |
| A reaction: In 1919 he said that the alternative, of 'postulating' new entities, has 'all the advantages of theft over honest toil' [IMP p.71]. This is Russell's commitment to 'constructing' everything, even his concept of matter. Arithmetic as PA is postulation. |
| 6111 | As propositions can be put in subject-predicate form, we wrongly infer that facts have substance-quality form [Russell] |
| Full Idea: Since any proposition can be put into a form with a subject and a predicate, united by a copula, it is natural to infer that every fact consists in the possession of a quality by a substance, which seems to me a mistake. | |
| From: Bertrand Russell (Logical Atomism [1924], p.152) | |
| A reaction: This disagrees with McGinn on facts (Idea 6075). I approve of this warning from Russell, which is a recognition that we can't just infer our metaphysics from our language. I think of this as the 'Frege Fallacy', which ensnared Quine and others. |
| 4028 | Frege allows either too few properties (as extensions) or too many (as predicates) [Mellor/Oliver on Frege] |
| Full Idea: Frege's theory of properties (which he calls 'concepts') yields too few properties, by identifying coextensive properties, and also too many, by letting every predicate express a property. | |
| From: comment on Gottlob Frege (Function and Concept [1891]) by DH Mellor / A Oliver - Introduction to 'Properties' §2 | |
| A reaction: Seems right; one extension may have two properties (have heart/kidneys), two predicates might express the same property. 'Cutting nature at the joints' covers properties as well as objects. |
| 8489 | The concept 'object' is too simple for analysis; unlike a function, it is an expression with no empty place [Frege] |
| Full Idea: I regard a regular definition of 'object' as impossible, since it is too simple to admit of logical analysis. Briefly: an object is anything that is not a function, so that an expression for it does not contain any empty place. | |
| From: Gottlob Frege (Function and Concept [1891], p.32) | |
| A reaction: Here is the core of the programme for deriving our ontology from our logic and language, followed through by Russell and Quine. Once we extend objects beyond the physical, it becomes incredibly hard to individuate them. |
| 16614 | Matter and form give true unity; subject and accident is just unity 'per accidens' [Duns Scotus] |
| Full Idea: From matter and form comes one thing per se. This is not so for subject and accident. Matter and form are instrinsic causes of a composite being, but whiteness and a human being are not. Humans can exist without whiteness, so it is one thing per accidens. | |
| From: John Duns Scotus (Oxford Commentary on Sentences [1301], II.12.1.14), quoted by Robert Pasnau - Metaphysical Themes 1274-1671 | |
| A reaction: This isn't much of a theory, but at least it is focusing on an interesting question, and the distinction between genuinely unified, and unified by chance. Compare a loving couple with siblings who hate each other. |
| 9947 | Concepts are the ontological counterparts of predicative expressions [Frege, by George/Velleman] |
| Full Idea: Concepts, for Frege, are the ontological counterparts of predicative expressions. | |
| From: report of Gottlob Frege (Function and Concept [1891]) by A.George / D.J.Velleman - Philosophies of Mathematics Ch.2 | |
| A reaction: That sounds awfully like what many philosophers call 'universals'. Frege, as a platonist (at least about numbers), I would take to be in sympathy with that. At least we can say that concepts seem to be properties. |
| 10319 | An assertion about the concept 'horse' must indirectly speak of an object [Frege, by Hale] |
| Full Idea: Frege had a notorious difficulty over the concept 'horse', when he suggests that if we wish to assert something about a concept, we are obliged to proceed indirectly by speaking of an object that represents it. | |
| From: report of Gottlob Frege (Function and Concept [1891], Ch.2.II) by Bob Hale - Abstract Objects | |
| A reaction: This sounds like the thin end of a wedge. The great champion of objects is forced to accept them here as a façon de parler, when elsewhere they have ontological status. |
| 8488 | A concept is a function whose value is always a truth-value [Frege] |
| Full Idea: A concept in logic is closely connected with what we call a function. Indeed, we may say at once: a concept is a function whose value is always a truth-value. ..I give the name 'function' to what is meant by the 'unsaturated' part. | |
| From: Gottlob Frege (Function and Concept [1891], p.30) | |
| A reaction: So a function becomes a concept when the variable takes a value. Problems arise when the value is vague, or the truth-value is indeterminable. |
| 9948 | Unlike objects, concepts are inherently incomplete [Frege, by George/Velleman] |
| Full Idea: For Frege, concepts differ from objects in being inherently incomplete in nature. | |
| From: report of Gottlob Frege (Function and Concept [1891]) by A.George / D.J.Velleman - Philosophies of Mathematics Ch.2 | |
| A reaction: This is because they are 'unsaturated', needing a quantified variable to complete the sentence. This could be a pointer towards Quine's view of properties, as simply an intrinsic feature of predication about objects, with no separate identity. |
| 6112 | Meaning takes many different forms, depending on different logical types [Russell] |
| Full Idea: There is not one relation of meaning between words and what they stand for, but as many relations of meaning, each of a different logical type, as there are logical types among the objects for which there are words. | |
| From: Bertrand Russell (Logical Atomism [1924], p.153) | |
| A reaction: This might be a good warning for those engaged in the externalist/internalist debate over the meaning of concepts such as natural kind terms like 'water'. I could have an external meaning for 'elms', but an internal meaning for 'ferns'. |
| 4972 | I may regard a thought about Phosphorus as true, and the same thought about Hesperus as false [Frege] |
| Full Idea: From sameness of meaning there does not follow sameness of thought expressed. A fact about the Morning Star may express something different from a fact about the Evening Star, as someone may regard one as true and the other false. | |
| From: Gottlob Frege (Function and Concept [1891], p.14) | |
| A reaction: This all gets clearer if we distinguish internalist and externalist theories of content. Why take sides on this? Why not just ask 'what is in the speaker's head?', 'what does the sentence mean in the community?', and 'what is the corresponding situation?' |
| 8491 | The Ontological Argument fallaciously treats existence as a first-level concept [Frege] |
| Full Idea: The ontological proof of God's existence suffers from the fallacy of treating existence as a first-level concept. | |
| From: Gottlob Frege (Function and Concept [1891], p.38 n) | |
| A reaction: [See Idea 8490 for first- and second-order functions] This is usually summarised as the idea that existence is a quantifier rather than a predicate. |