65 ideas
| 6420 | Only by analysing is progress possible in philosophy [Russell] |
| Full Idea: I remain firmly persuaded, in spite of some modern tendencies to the contrary, that only by analysing is progress possible, …for example, by analysing physics and perception, the problem of mind and matter can be completely solved. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.1) | |
| A reaction: I don't share his confidence in the second part of this, but I subscribe to the maxim that 'analsis is the path to wisdom'. It is a very western view, and lots of people (mostly of a mystical disposition) hate it, but I see no better path. |
| 6432 | Analysis gives new knowledge, without destroying what we already have [Russell] |
| Full Idea: It seems to me evident that, as in the case of impure water, analysis gives new knowledge without destroying any of the previously existing knowledge. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.11) | |
| A reaction: I agree. On the whole, opponents of analysis are sentimental mystics who are reluctant to think carefully about life. I'm not sure what careful and concentrated thought is capable of, apart from analysis. |
| 6437 | The theory of types makes 'Socrates and killing are two' illegitimate [Russell] |
| Full Idea: 'Socrates and killing are two' would be an illegitimate sentence according to the doctrine of types. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.14) | |
| A reaction: This nicely shows how Ryle's notion of a 'category mistake', although it is a commonsense observation of bogus reasoning, arises out of Russell's logical analysis of sets. Of course, the theory of types has its critics. |
| 6442 | Truth belongs to beliefs, not to propositions and sentences [Russell] |
| Full Idea: Truth and falsehood both belong primarily to beliefs, and only derivatively to propositions and sentences. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.15) | |
| A reaction: I'm not sure why a proposition which is date/place stamped ('it is raining, here and now') could not be considered a truth, even if no one believed it. Is not the proposition 'squares have four sides' true? |
| 21616 | Truth and falsity apply to suppositions as well as to assertions [Williamson] |
| Full Idea: The notion of truth and falsity apply to suppositions as well as to assertions. | |
| From: Timothy Williamson (Vagueness [1994], 7.2) | |
| A reaction: This may not be obvious to those who emphasise pragmatics and ordinary language, but it is self-evident to anyone who emphasises logic. |
| 21623 | True and false are not symmetrical; false is more complex, involving negation [Williamson] |
| Full Idea: The concepts of truth and falsity are not symmetrical. The asymmetry is visible in the fundamental principles governing them, for F is essentially more complex than T, by its use of negation. | |
| From: Timothy Williamson (Vagueness [1994], 7.5) | |
| A reaction: If T and F are primitives, controlled by axioms, then they might be symmetrical in nature, but asymmetrical in use. However, if forced to choose just one primitive, I presume it would be T. |
| 21602 | Many-valued logics don't solve vagueness; its presence at the meta-level is ignored [Williamson] |
| Full Idea: It is an illusion that many-valued logic constitutes a well-motivated and rigorously worked out theory of vagueness. ...[top] There has been a reluctance to acknowledge higher-order vagueness, or to abandon classical logic in the meta-language. | |
| From: Timothy Williamson (Vagueness [1994], 4.12) |
| 6436 | I gradually replaced classes with properties, and they ended as a symbolic convenience [Russell] |
| Full Idea: My original use of classes was gradually more and more replaced by properties, and in the end disappeared except as a symbolic convenience. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.14) | |
| A reaction: I wish I knew what properties are. On the whole, though, I agree with this, because it is more naturalistic. We may place things in classes because of their properties, and this means there are natural classes, but classes can't have a life of their own. |
| 21611 | Formal semantics defines validity as truth preserved in every model [Williamson] |
| Full Idea: An aim of formal semantics is to define in mathematical terms a set of models such that an argument is valid if and only if it preserves truth in every model in the set, for that will provide us with a precise standard of validity. | |
| From: Timothy Williamson (Vagueness [1994], 5.3) |
| 21606 | 'Bivalence' is the meta-linguistic principle that 'A' in the object language is true or false [Williamson] |
| Full Idea: The meta-logical law of excluded middle is the meta-linguistic principle that any statement 'A' in the object language is either truth or false; it is now known as the principle of 'bivalence'. | |
| From: Timothy Williamson (Vagueness [1994], 5.2) | |
| A reaction: [He cites Henryk Mehlberg 1958] See also Idea 21605. Without this way of distinguishing bivalence from excluded middle, most discussions of them strikes me as shockingly lacking in clarity. Personally I would cut the normativity from this one. |
| 21605 | Excluded Middle is 'A or not A' in the object language [Williamson] |
| Full Idea: The logical law of excluded middle (now the standard one) is the schema 'A or not A' in the object-language. | |
| From: Timothy Williamson (Vagueness [1994], 5.2) | |
| A reaction: [He cites Henryk Mehlberg 1958] See Idea 21606. The only sensible way to keep Excluded Middle and Bivalence distinct. I would say: (meta-) only T and F are available, and (object) each proposition must have one of them. Are they both normative? |
| 7528 | Leibniz bases everything on subject/predicate and substance/property propositions [Russell] |
| Full Idea: The metaphysics of Leibniz was explicitly based upon the doctrine that every proposition attributes a predicate to a subject and (what seemed to him almost the same thing) that every fact consists of a substance having a property. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.5) | |
| A reaction: I think it is realised now that although predicates tend to attribute properties to things, they are far from being the same thing. See Idea 4587, for example. Russell gives us an interesting foot in the door of Leibniz's complex system. |
| 6439 | Names are meaningless unless there is an object which they designate [Russell] |
| Full Idea: Unlike descriptions, names are meaningless unless there is an object which they designate. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.14) | |
| A reaction: This interests Russell because of its ontological implications. If we reduce language to names, we can have a pure ontology of 'objects'. We need a system for saying whether a description names something - which is his theory of definite descriptions. |
| 21612 | Or-elimination is 'Argument by Cases'; it shows how to derive C from 'A or B' [Williamson] |
| Full Idea: Argument by Cases (or or-elimination) is the standard way of using disjunctive premises. If one can argue from A and some premises to C, and from B and some premises to C, one can argue from 'A or B' and the combined premises to C. | |
| From: Timothy Williamson (Vagueness [1994], 5.3) |
| 21599 | A sorites stops when it collides with an opposite sorites [Williamson] |
| Full Idea: A sorites paradox is stopped when it collides with a sorites paradox going in the opposite direction. That account will not strike a logician as solving the sorites paradox. | |
| From: Timothy Williamson (Vagueness [1994], 3.3) |
| 6423 | We tried to define all of pure maths using logical premisses and concepts [Russell] |
| Full Idea: The primary aim of our 'Principia Mathematica' was to show that all pure mathematics follows from purely logical premisses and uses only concepts definable in logical terms. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.7) | |
| A reaction: This spells out the main programme of logicism, by its great hero, Russell. The big question now is whether Gödel's Incompleteness Theorems have succeeded in disproving logicism. |
| 6425 | Formalism can't apply numbers to reality, so it is an evasion [Russell] |
| Full Idea: Formalism is perfectly adequate for doing sums, but not for the application of number, such as the simple statement 'there are three men in this room', so it must be regarded as an unsatisfactory evasion. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.10) | |
| A reaction: This seems to me a powerful and simple objection. The foundation of arithmetic is that there are three men in the room, not that one plus two is three. Three men and three ties make a pattern, which we call 'three'. |
| 6424 | Formalists say maths is merely conventional marks on paper, like the arbitrary rules of chess [Russell] |
| Full Idea: The Formalists, led by Hilbert, maintain that arithmetic symbols are merely marks on paper, devoid of meaning, and that arithmetic consists of certain arbitrary rules, like the rules of chess, by which these marks can be manipulated. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.10) | |
| A reaction: I just don't believe that maths is arbitrary, and this view pushes me into the arms of the empiricists, who say maths is far more likely to arise from experience than from arbitrary convention. The key to maths is patterns. |
| 6426 | Intuitionism says propositions are only true or false if there is a method of showing it [Russell] |
| Full Idea: The nerve of the Intuitionist theory, led by Brouwer, is the denial of the law of excluded middle; it holds that a proposition can only be accounted true or false when there is some method of ascertaining which of these it is. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.2) | |
| A reaction: He cites 'there are three successive sevens in the expansion of pi' as a case in point. This seems to me an example of the verificationism and anti-realism which is typical of that period. It strikes me as nonsense, but Russell takes it seriously. |
| 6419 | In 1899-1900 I adopted the philosophy of logical atomism [Russell] |
| Full Idea: In the years 1899-1900 I adopted the philosophy of logical atomism. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.1) | |
| A reaction: This is interesting (about Russell) because he only labelled it as 'logical atomism' in about 1912, and only wrote about it as such in 1918. It is helpful to understand that the theory of definite descriptions was part of his logical atomism. |
| 6438 | Complex things can be known, but not simple things [Russell] |
| Full Idea: I have come to think that, although many things can be known to be complex, nothing can be known to be simple. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.14) | |
| A reaction: This appears to be a rejection of his logical atomism. It goes with a general rebellion against foundationalist epistemology, because the empiricists foundations (e.g. Hume's impressions) seem devoid of all content. |
| 6434 | Facts are everything, except simples; they are either relations or qualities [Russell] |
| Full Idea: Facts, as I am using the word, consist always of relations between parts of a whole or qualities of single things; facts, in a word, are whatever there is except what (if anything) is completely simple. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.13) | |
| A reaction: This is the view that goes with Russell's 'logical atomism', where the 'completely simple' is used to build up the 'facts'. If World War One was a fact, was it a 'relation' or a 'quality'. Must events then be defined in terms of those two? |
| 21601 | A vague term can refer to very precise elements [Williamson] |
| Full Idea: Both 30° and 60° are clearly acute angles. 'Acute' is precise in all relevant respects. Nevertheless, 30° is acuter than 60°. | |
| From: Timothy Williamson (Vagueness [1994], 4.11) | |
| A reaction: A very nice example of something which is vague, despite involving precise ingredients. But then 'bald' is vague, while 'this is a hair on his head' is fairly precise. |
| 21589 | When bivalence is rejected because of vagueness, we lose classical logic [Williamson] |
| Full Idea: The principle of bivalence (that every statement is either true or false) has been rejected for vague languages. To reject bivalence is to reject classical logic or semantics. | |
| From: Timothy Williamson (Vagueness [1994], Intro) | |
| A reaction: His example is specifying a moment when Rembrandt became 'old'. This is the number one reason why the problem of vagueness is seen as important. Is the rejection of classical logic a loss of our grip on the world? |
| 21596 | Vagueness undermines the stable references needed by logic [Williamson] |
| Full Idea: Logic requires expressions to have the same referents wherever they occur; vague natural languages violate this contraint. | |
| From: Timothy Williamson (Vagueness [1994], 2.2) | |
| A reaction: This doesn't mean that logic has to win. Maybe it is important for philosophers who see logic as central to be always aware of vagueness as the gulf between their precision and the mess of reality. Precision is worth trying for, though. |
| 21629 | Equally fuzzy objects can be identical, so fuzziness doesn't entail vagueness [Williamson] |
| Full Idea: Fuzzy boundaries do not in any way require vague identity. Objects are identical only if their boundaries have exactly the same fuzziness. | |
| From: Timothy Williamson (Vagueness [1994], 9.2) | |
| A reaction: This all rests on the Fregean idea that determinate existence requires the ability to participate in an identity statement. |
| 21591 | Vagueness is epistemic. Statements are true or false, but we often don't know which [Williamson] |
| Full Idea: My thesis is that vagueness is an epistemic phenomenon. In cases of unclarity, statements remain true or false, but speakers of the language have no way of knowing which. Higher-order vagueness consists in ignorance about ignorance. | |
| From: Timothy Williamson (Vagueness [1994], Intro) | |
| A reaction: He has plumped for the intuitively least plausible theory. It means that a hair dropping out of someone's head triggers a situation where they are 'bald', but none of us know when that was. And Rembrandt became 'old' in an instant. |
| 21619 | If a heap has a real boundary, omniscient speakers would agree where it is [Williamson] |
| Full Idea: If, in judging a heap as grains are removed, omniscient speakers all stop at the same point, it must does mark some sort of previously hidden boundary. ...If there is no hidden boundary, then different omniscient speakers would stop at different points. | |
| From: Timothy Williamson (Vagueness [1994], 7.3) | |
| A reaction: A very nice thought experiment, which obviously won't settle anything, but brings out nicely the view the vagueness is a sort of ignorance. God is never vague in the application of terms (though God might withhold the application if there is no boundary). |
| 21620 | The epistemic view says that the essence of vagueness is ignorance [Williamson] |
| Full Idea: The epistemic view is that ignorance is the real essence of the phenomenon ostensively identified as vagueness. ...[203] According to the epistemic view, I am either thin or not thin, ...and we have no idea how to find out out which. | |
| From: Timothy Williamson (Vagueness [1994], 7.4) | |
| A reaction: Presumably this implies that there is often a real border (of which we may be ignorant), but it doesn't seem to rule out cases where there just is no border. Where does the east Atlantic meet the west Atlantic? |
| 21622 | If there is a true borderline of which we are ignorant, this drives a wedge between meaning and use [Williamson] |
| Full Idea: A common complaint against the epistemic view is that to postulate a matter of fact in borderline cases is to suppose, incoherently, that the meanings of our words draw a line where our use of them does not. | |
| From: Timothy Williamson (Vagueness [1994], 7.5) | |
| A reaction: This doesn't necessarily seem to require the view that the meaning of words is their usage. Just that if there is one consensus on usage, it seems unlikely that there is a different underlying reality about the true meaning. Externalist meanings? |
| 9120 | Vagueness in a concept is its indiscriminability from other possible concepts [Williamson] |
| Full Idea: Vagueness in a concept is its indiscriminability from other possible concepts; this can be reconciled with our knowledge of vague terms. | |
| From: Timothy Williamson (Vagueness [1994], 8.1) | |
| A reaction: Sorensen objects that this makes vagueness too relative to members of a speech community. He prefers 'absolute borderline cases'. If you like the epistemic view, then Williamson seems more plausible. My 'vague' might differ from yours. |
| 21614 | The 'nihilist' view of vagueness says that 'heap' is not a legitimate concept [Williamson] |
| Full Idea: The 'nihilist' view is that no genuine distinction can be vaguely drawn; since vague expressions are not properly meaningful, there is nothing for sorites reasoning to betray; they are empty. | |
| From: Timothy Williamson (Vagueness [1994], 6.1) | |
| A reaction: He cites Frege as holding this view. The thought is that 'heap' is not a legitimate concept, so fussing over what qualifies as one is pointless. This seems to be a semantic view of vagueness, of which the main rival is the contextual view. |
| 21617 | We can say propositions are bivalent, but vague utterances don't express a proposition [Williamson] |
| Full Idea: A philosopher might endorse bivalence for propositions, while treating vagueness as the failure of an utterance to express a unique proposition. | |
| From: Timothy Williamson (Vagueness [1994], 7.2) | |
| A reaction: This idea jumps at out me as an extremely promising approach to vagueness, because I am a fan of propositions (and have written a paper on them). The whole point of propositions is that they are not ambiguous (and probably not vague). |
| 21618 | If the vague 'TW is thin' says nothing, what does 'TW is thin if his perfect twin is thin' say? [Williamson] |
| Full Idea: If vague utterances in borderline cases fail to say anything, then if 'TW is thin' is vague, and TW has a twin of identical dimensions, it still seems that 'If TW is thin then his twin is thin' must be true, and so it must have said something. | |
| From: Timothy Williamson (Vagueness [1994], 7.2 (d)) | |
| A reaction: This an objection to the Fregean 'nihilistic' view of Idea 21614. I am inclined to a solution based on the proposition expressed, rather than the sentence. The first question is whether you are willing to assert 'TW is thin'. |
| 21625 | The vagueness of 'heap' can remain even when the context is fixed [Williamson] |
| Full Idea: Vagueness remains even when the context is fixed. In principle, a vague word might exhibit no context dependence whatsoever. ...For example, a dispute over whether someone has left a 'heap' of sand on the floor. | |
| From: Timothy Williamson (Vagueness [1994], 7.7) | |
| A reaction: A fairly devastating rebuttal of what seems to be David Lewis's view. He talks of something being 'smooth' depending on context. |
| 21590 | Asking when someone is 'clearly' old is higher-order vagueness [Williamson] |
| Full Idea: Difficulties of vagueness are presented by the question 'When did Rembrandt become clearly old?', and the iterating question 'When did he become clearly clearly old?'. This is the phenomenon of higher-order vagueness. The language of vagueness is vague. | |
| From: Timothy Williamson (Vagueness [1994], Intro) | |
| A reaction: [compressed] I presume the bottom level is a question about Rembrandt, the second level is about this use of the word 'old', and the third level is about this particular application of the word 'clearly'. Meta-languages. |
| 21609 | Supervaluationism defines 'supertruth', but neglects it when defining 'valid' [Williamson] |
| Full Idea: Supervaluationists identify truth with 'supertruth'; since validity is necessary preservation of truth, they should identify it with necessary preservation of supertruth. But it plays no role in their definition of 'local' validity. | |
| From: Timothy Williamson (Vagueness [1994], 5.3) | |
| A reaction: [See text for 'local'] Generally Williamson's main concern with attempts to sort out vagueness is that higher-order and meta-language issues are neglected. |
| 21610 | Supervaluation adds a 'definitely' operator to classical logic [Williamson] |
| Full Idea: Supervaluation seems to inherit the power of classical logic, ...but also enables it to be extended. It makes room for a new operator 'definitely' to express supertruth in the object-language. | |
| From: Timothy Williamson (Vagueness [1994], 5.3) | |
| A reaction: Once you mention higher-order vagueness you can see a regress looming over the horizon. 'He is definitely definitely definitely bald'. [p.164 he says 'definitely' has no analysis, and is an uninteresting primitive] |
| 21613 | Supervaluationism cannot eliminate higher-order vagueness [Williamson] |
| Full Idea: Supervaluationism cannot eliminate higher-order vagueness. It must conduct its business in a vague meta-language. ...[162] All truth is at least disquotational, and supertruth is not. | |
| From: Timothy Williamson (Vagueness [1994], 5.6) | |
| A reaction: This is Williamson's final verdict on the supervaluation strategy for vagueness. Intuitively, it looks as if merely narrowing down the vagueness (by some sort of consensus) is no solution to the problem of vagueness. |
| 21592 | Supervaluation keeps classical logic, but changes the truth in classical semantics [Williamson] |
| Full Idea: Supervaluationism preserves almost all of classical logic, at the expense of classical semantics, but giving a non-standard account of truth. I argue that its treatment of higher-order vagueness undermines the non-standard account of truth. | |
| From: Timothy Williamson (Vagueness [1994], Intro) |
| 21603 | You can't give a precise description of a language which is intrinsically vague [Williamson] |
| Full Idea: If a vague language is made precise, its expressions change in meaning, so an accurate semantic description of the precise language is inaccurate as a description of the vague one. | |
| From: Timothy Williamson (Vagueness [1994], 5.1) | |
| A reaction: Kind of obvious, really, but it clarifies the nature of any project (starting with Leibniz) to produce a wholly precise language. That is usually seen as a specialist language for science. |
| 21604 | Supervaluation assigns truth when all the facts are respected [Williamson] |
| Full Idea: 'Admissible' interpretations respect all the theoretical and ostensive connections. ...'Supervaluation' is the assignment of truth to the statements true on all admissible valuations, falsity to the false one, and neither to the rest. | |
| From: Timothy Williamson (Vagueness [1994], 5.2) | |
| A reaction: So 'he is bald' is true if when faced with all observations and definitions it is acceptable. Prima facie, that doesn't sound like a solution to the problem. Supervaluation started in philosophy of science. [p.156 'Admissible seems vague'] |
| 21607 | Supervaluation has excluded middle but not bivalence; 'A or not-A' is true, even when A is undecided [Williamson] |
| Full Idea: The supervaluationist denies bivalence but accepts excluded middle. The statement 'A or not-A' is true on each admissible interpretation, and therefore true, even if 'A' (and hence 'not-A') are true and some and false on others, so neither T nor F. | |
| From: Timothy Williamson (Vagueness [1994], 5.2) | |
| A reaction: See Ideas 21605 and 21606 for the distinction being used here. Denying bivalence allows 'A' to be neither true nor false. It seems common sense that 'he is either bald or not-bald' is true, without being sure about the disjuncts. |
| 21608 | Truth-functionality for compound statements fails in supervaluation [Williamson] |
| Full Idea: A striking fearure of supervaluations is the failure of truth-functionality for compound statements. | |
| From: Timothy Williamson (Vagueness [1994], 5.3) | |
| A reaction: Supervaluations has the initial appearance of enhancing classical logic, but turns out to somewhat undermine it. Hence Williamson's lack of sympathy. But see Idea 21610. |
| 21633 | Nominalists suspect that properties etc are our projections, and could have been different [Williamson] |
| Full Idea: The nominalist suspects that properties, relations and states of affairs are mere projections onto the world of our forms of speech. One source of the suspicion is a sense that we could just as well have classified things differently. | |
| From: Timothy Williamson (Vagueness [1994], 9.3) | |
| A reaction: I know it is very wicked to say so, but I'm afraid I have some sympathy with this view. But I like the primary/secondary distinction, so there is more 'projection' in the latter case. Classification is not random; it is a response to reality. |
| 6440 | Universals can't just be words, because words themselves are universals [Russell] |
| Full Idea: Those who dislike universals have thought that they could be merely words; the trouble with this view is that a word itself is a universal. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.14) | |
| A reaction: Russell gradually lost his faith in most things, but never in universals. I find it unconvincing that we might dismiss nominalism so easily. I'm not sure why the application of the word 'cat' could not just be conventional. |
| 21630 | If fuzzy edges are fine, then why not fuzzy temporal, modal or mereological boundaries? [Williamson] |
| Full Idea: If objects can have fuzzy spatial boundaries, surely they can have fuzzy temporal, modal or mereological boundaries too. | |
| From: Timothy Williamson (Vagueness [1994], 9.2) | |
| A reaction: Fair point. I think there is a distinction between parts of the thing, such as its edges, being fuzzy, and the whole thing being fuzzy, in the temporal case. |
| 21632 | A river is not just event; it needs actual and counterfactual boundaries [Williamson] |
| Full Idea: A river is not just an event. One would need to specify counterfactual as well as actual boundaries. | |
| From: Timothy Williamson (Vagueness [1994], 9.3) | |
| A reaction: In other words the same river can change its course a bit, but it can't head off in the opposite direction. |
| 21621 | We can't infer metaphysical necessities to be a priori knowable - or indeed knowable in any way [Williamson] |
| Full Idea: The inference from metaphysical necessity to a priori knowlability is, as Kripke has emphasized, fallacious. Indeed, metaphysical necessities cannot be assumed knowable in any way at all. | |
| From: Timothy Williamson (Vagueness [1994], 7.4) | |
| A reaction: The second sentence sounds like common sense. He cites Goldbach's Conjecture. A nice case of the procedural rule of keeping your ontology firmly separated from your epistemology. How is it? is not How do we know it? |
| 6430 | In epistemology we should emphasis the continuity between animal and human minds [Russell] |
| Full Idea: It seems to me desirable in the theory of knowledge to emphasise the continuity between animal and human minds. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.11) | |
| A reaction: I strongly agree with this, mainly because it avoids overemphasis on language in epistemology. It doesn't follow that animals know a lot, and there is a good case for saying that they don't actually 'know' anything, despite having true beliefs. |
| 21627 | We have inexact knowledge when we include margins of error [Williamson] |
| Full Idea: Inexact knowledge is a widespread and easily recognised cognitive phenomenon, whose underlying nature turns out to be characterised by the holding of margin of error principles. | |
| From: Timothy Williamson (Vagueness [1994], 8.3) | |
| A reaction: Williamson is invoking this as a tool in developing his epistemic view of vagueness. It obviously invites the question of how it can be knowledge if error is a possibility. A very large margin of error would obviously invalidate it. |
| 6441 | Pragmatism judges by effects, but I judge truth by causes [Russell] |
| Full Idea: Pragmatism holds that a belief is to be judged if it has certain effects, whereas I hold that an empirical belief is to be judged true if it has certain kinds of causes. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.15) | |
| A reaction: I'm with Russell here, and this seems to me a convincing objection to pragmatism. The simple problem is that falsehoods can occasionally have very beneficial effects. Beliefs are made true by the facts, not by their consequences. |
| 6431 | Empiricists seem unclear what they mean by 'experience' [Russell] |
| Full Idea: When I began to think about theory of knowledge, I found that none of the philosophers who emphasise 'experience' tells us what they mean by the word. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.11) | |
| A reaction: A very significant comment about empiricism. Hume does not seem very clear about what an 'impression' is. Russell's problem has been dealt with intensively by modern empiricists, who discuss 'the given', and conceptualised perception. |
| 21626 | Knowing you know (KK) is usually denied if the knowledge concept is missing, or not considered [Williamson] |
| Full Idea: The failure of the KK principle is not news. The standard counterexamples involve knowing subjects who lack the concept of knowledge, or have not reflected on their knowledge, and therefore do not know that they know. | |
| From: Timothy Williamson (Vagueness [1994], 8.2) | |
| A reaction: There is also the timid but knowledgeable pupil, who can't believe they know so much. The simplest case would be if we accept that animals know lots of things, but are largely devoid of any metathinking. |
| 6444 | True belief about the time is not knowledge if I luckily observe a stopped clock at the right moment [Russell] |
| Full Idea: Not all true beliefs are knowledge; the stock example to the contrary is that of a clock which has stopped by which I believe to be going and which I happen to look at when, by chance, it shows the right time. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.15) | |
| A reaction: [in his 1948:112] Russell had spotted Gettier-type problems long before Gettier. The problem of lucky true beliefs dates back to Plato (Idea 2140). This example is also a problem for reliabilism, if the clock is usually working fine. |
| 3102 | Why don't we experience or remember going to sleep at night? [Magee] |
| Full Idea: As a child it was incomprehensible to me that I did not experience going to sleep, and never remembered it. When my sister said 'Nobody remembers that', I just thought 'How does she know?' | |
| From: Bryan Magee (Confessions of a Philosopher [1997], Ch.I) | |
| A reaction: This is actually evidence for something - that we do not have some sort of personal identity which is separate from consciousness, so that "I am conscious" would literally mean that an item has a property, which it can lose. |
| 6433 | Behaviourists struggle to explain memory and imagination, because they won't admit images [Russell] |
| Full Idea: Behaviourists refuse to admit images because they cannot be observed from without, but this causes them difficulties when they attempt to explain either memory or imagination. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.13) | |
| A reaction: This is a striking objection to behaviourism, and it is rarely mentioned in modern discussions of the topic. They might try denying the existence of private 'images', but that wouldn't be very plausible. |
| 21631 | To know, believe, hope or fear, one must grasp the thought, but not when you fail to do them [Williamson] |
| Full Idea: To know, believe, hope, or fear that A, one must grasp the thought that A. In contrast, to fail to know, believe, hope or fear that A, one need not grasp the thought that A. | |
| From: Timothy Williamson (Vagueness [1994], 9.3 c) | |
| A reaction: A simple point, which at least shows that propositional attitudes are a two-stage operation. |
| 6443 | Surprise is a criterion of error [Russell] |
| Full Idea: Surprise is a criterion of error. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.15) | |
| A reaction: Russell is not too precise about this, but it is a nice point. Surprise is thwarted expectation, which implies prior misjudgement. |
| 21600 | 'Blue' is not a family resemblance, because all the blues resemble in some respect [Williamson] |
| Full Idea: 'Blue' is vague by some standards, for it has borderline cases, but that does not make it a family resemblance term, for all the shades of blue resemble each other in some respect. | |
| From: Timothy Williamson (Vagueness [1994], 3.3) | |
| A reaction: Presumably the point of family resemblance is that fringe members as still linked to the family, despite having lost the main features. A bit of essentialism seems needed here. |
| 6427 | Unverifiable propositions about the remote past are still either true or false [Russell] |
| Full Idea: There is no conceivable method by which we can discover whether the proposition 'It snowed on Manhattan Island on the 1st January in the year 1 A.D.' is true or false, but it seems preposterous to maintain that it is neither. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.10) | |
| A reaction: I love this example, which seems so simple and so clear-cut. It criticises verificationism, and gives strong intuitive support for realism, and supports the law of excluded middle. |
| 21615 | References to the 'greatest prime number' have no reference, but are meaningful [Williamson] |
| Full Idea: The predicate 'is a prime number greater than all other prime numbers' is necessarily not true of anything, but it is not semantically defective, for it occurs in sentences that constitute a sound proof that there is no such number. | |
| From: Timothy Williamson (Vagueness [1994], 6.2) | |
| A reaction: One might reply that the description can be legitimately mentioned, but not legitimately used. |
| 18038 | The 't' and 'f' of formal semantics has no philosophical interest, and may not refer to true and false [Williamson] |
| Full Idea: In a formal semantics we can label two properties 't' and 'f' and suppose that some sentences have neither (or both). Such a manoeuvre shows nothing of philosophical interest. No connection has been made between 't' and 'f' and truth and falsity. | |
| From: Timothy Williamson (Vagueness [1994], 7.2) | |
| A reaction: This is right, and means there is a huge gulf between 'formal' semantics (which could be implemented on a computer), and seriously interesting semantics about how language refers to and describes the world. |
| 21624 | It is known that there is a cognitive loss in identifying propositions with possible worlds [Williamson] |
| Full Idea: It is well known that when a proposition is identified with the set of possible worlds at which it is true, a region in the space of possible worlds, cognitively significant distinctions are lost. | |
| From: Timothy Williamson (Vagueness [1994], 7.6) | |
| A reaction: Alas, he doesn't specify which distinctions get lost, so this is just a pointer. It would seem likely that two propositions could have identical sets of possible worlds, while not actually saying the same thing. Equilateral/equiangular. |
| 6435 | You can believe the meaning of a sentence without thinking of the words [Russell] |
| Full Idea: If you have just heard a loud clap of thunder, you believe what is expressed by 'there has just been a loud clap of thunder' even if no words come into your mind. | |
| From: Bertrand Russell (My Philosophical Development [1959], Ch.13) | |
| A reaction: This seems to me important, and accurate. We should not be too mesmerised by language. Animals have beliefs, and this is a nice example of an undeniable non-linguistic human belief. |