6 ideas
19565 | How could the mind have a link to the necessary character of reality? [Devitt] |
Full Idea: What non-experiential link to reality could support insights into its necessary character? | |
From: Michael Devitt (There is No A Priori (and reply) [2005], 4) | |
A reaction: The key to it, I think, is your theory of mind. If you are a substance dualist, then connecting to such deep things looks fine, but if you are a reductive physicalist then it looks absurdly hopeful. |
19564 | Some knowledge must be empirical; naturalism implies that all knowledge is like that [Devitt] |
Full Idea: It is overwhelmingly plausible that some knowledge is empirical. The attractive thesis of naturalism is that all knowledge is; there is only one way of knowing. | |
From: Michael Devitt (There is No A Priori (and reply) [2005], 1) | |
A reaction: How many ways for us to know seems to depend on what faculties we have. We lump our senses together under a single heading. The arrival of data is not the same as the arrival of knowledge. I'm unconvinced that naturalists like me must accept this. |
2605 | If everything uses mentalese, ALL concepts must be innate! [Putnam] |
Full Idea: Fodor concludes that every predicate that a brain could learn to use must have a translation into the computer language of that brain. So no "new" concepts can be acquired: all concepts are innate! | |
From: Hilary Putnam (What is innate and why [1980], p.407) | |
A reaction: Some misunderstanding, surely? No one could be so daft as to think that everyone has an innate idea of an iPod. More basic innate building blocks for thought are quite plausible. |
2606 | No machine language can express generalisations [Putnam] |
Full Idea: Computers have a built-in language, but not a language that contains quantifiers (that is, the words "all" and "some"). …So generalizations (containing "all") cannot ever be stated in machine language. | |
From: Hilary Putnam (What is innate and why [1980], p.408) | |
A reaction: Computers are too sophisticated to need quantification (which is crude). Computers can work with very precise and complex specifications of the domain of a given variable. |
14894 | Indexicals have a 'character' (the standing meaning), and a 'content' (truth-conditions for one context) [Kaplan, by Macià/Garcia-Carpentiro] |
Full Idea: Kaplan distinguished two different semantic features of indexical expressions: a 'character' that captures the standing meaning of the expression, and a 'content' that consists of their truth-conditional contribution in particular contexts. | |
From: report of David Kaplan (Demonstratives [1989]) by Macià/Garcia-Carpentiro - Introduction to 'Two-Dimensional Semantics' 1 | |
A reaction: This seems so clearly right that there isn't much to dispute. You can't understand the word 'I' or 'now' if you don't understand both its general purpose, and what it is doing in a particular utterance. But will this generalise to other semantics? |
14700 | 'Content' gives the standard modal profile, and 'character' gives rules for a context [Kaplan, by Schroeter] |
Full Idea: Kaplan sees two aspects of meaning, the 'content', reflecting a thing's modal profile, which is modelled by standard possible worlds semantics, and 'character', giving rules for different contexts. Proper names have constant character; indexicals vary. | |
From: report of David Kaplan (Demonstratives [1989]) by Laura Schroeter - Two-Dimensional Semantics 1.1.1 | |
A reaction: This gives rise to 2-D matrices for representing meaning, and the possible worlds are used twice, for evaluating meaning and then for evaluating context of use. I've always been struck by the two-dimensional semantics of passwords. |