Thoughts on definitions
April 4th 2007 07:41
You don't need definitions to use words.
-- Can a child define "the" or "is"?
-- Is a definition of "carpentry" a requirement of the job? Are you forbidden, otherwise, from calling yourself a carpenter? -- Or is a definition of "politics" necessary for politicians? Or even -- God forbid -- "philosophy" for philosophers?
-- Assume you do need definitions to use words. Well, aren't words only ever defined in terms of other words, and aren't those words defined in terms of still other words? -- and isn't the system then self-enclosed, with no entry point?
What is the defining task? What do lexicographers do? Can the defining task be completed?
-- You can point to things, and this has its own problems, among which is that you mightn't know what is being pointed to, and mightn't know what to generalize.
-- Presumably lexicographers gather raw data, and look for common denominators and patterns.
-- But, as in science, aren't there infinite hypotheses to fit any single set of facts? And even if you choose some arbitrary methodological principle, like Occam's razor, to select between theories, can't two theories be equally parsimonious, and aren't there a variety of ways in which something can be parsimonious?
-- A dictionary might give a one-sentence definition. -- Well, is one sentence sensitive enough for all the shades of meaning, and all the contexts, and all the effects in those contexts? -- All the interactions with etymology, codes of formality and politeness, group identity, metaphors and metonyms, rhetoric, sound and visual symbolism, cultural reference...?
-- And isn't usage constantly changing?
-- In what space do you differentiate? What counts as a common denominator? What allows you to recognize one? Why are these two homonyms treated as separate words, and why are these two heteronyms synonymous? Why is this single word divided into multiple meanings? -- Aren't you appealing to some system of categories, in terms of which you differentiate? And isn't this system something extra that you bring to the raw data, something you don't find within it?
-- Isn't there vagueness, always, at the edge of the word? Isn't there uncertainty as to whether, in particular places, the word may or may not apply?
-- I've rarely learnt how to use a word from the dictionary entry alone, as with the example of "gamin". But this doesn't mean the entry is useless.
Language is a tool to interact with other people's behaviour and psychology, to affect other people, and via them the world. Stripped down to its essential, it's no different from birdsong.
-- You create effects, achieve results, though language doesn't serve this purpose alone -- for instance, there are effects on yourself from the sheer act of speaking.
-- Language is also a tool to interact with your future self.
-- It's tempting to assert that languages must be growable, because you interact with different people using words in different ways, and because individual and collective experience grows. -- Tools must be simple enough to use readily, and flexible enough to cope with change: -- vagueness is essential.
-- Nouns don't stand for finite sets. When you learn "dogs", you're not told of all dog members. At best, "dogs" refers to an infinite set -- but even this picture isn't quite right, because you're not given a series like "1, 2, 3..." -- you're not given a clear rule to generalize.
-- Does anyone use nouns with single rules? In the course of your life, don't you modify the rules, or switch to different rules?
-- The older you get, the more rules and complications you learn.
-- And, to revisit an old idea, aren't a lot of nouns more a matter of family resemblance than necessary criteria?
-- And if you did have a clear rule, why would you have to stick to it? It's your prerogative to improvise, to employ a screwdriver as a hammer or as a paper-weight. Everyone must develop their own toolkit.
-- The suggestion, therefore, that language use is creative, that the sum of necessary conditions, if there are any, is not sufficient. When you apply a word, do you know that it'll stick? Perhaps you always have a choice. Perhaps every application of old word to new object is a leap in the dark. Perhaps every dog you meet expands the meaning of "dog", and expands the vagueness.
-- The world in its wit presents us with cases that don't fit neatly into existing category systems. And so we must invent new words, and extend old words. But even if there were no new cases -- even if we'd experienced everything there was to experience -- could we then compartmentalize neatly? Is reality itself clear? Could we find clear rules? Would it be practical to want to?
The concept of Pi or Fermat's theorem is elucidated in mathematical space, and is then named. But if you pick any ordinary concept (courage, dog, the colour orange), the concept exists in behaviour space, yes, but how can it be spoken of outside of words? How can I write to you now and answer the questions "What is courage?", "What is a dog?" "What is orange?" outside of words? Where is the mathematics for this?
-- Isn't a "What is" task almost always a lexicographical one, a defining one, with all of defining's problems? Isn't the definition of a concept the definition of a word, and a job for dictionary-writers?
-- Isn't the task of "conceptual clarification" more a matter of conceptual replacement, and rule replacement, gauged by precision with respect to particular practical purposes?
-- So few of our existing words are neat -- there is little starting point on which to build a complete set of neat boxes. We would have to invent a new language, like biology, and hope that that new language was adequate.
-- When someone inquires "What is love?", they might think they're asking for a definition, but need they be satisfied with one?
-- Aren't they asking, rather, after understanding, and aren't there other paths to understanding than definitional?
-- Can a child define "the" or "is"?
-- Is a definition of "carpentry" a requirement of the job? Are you forbidden, otherwise, from calling yourself a carpenter? -- Or is a definition of "politics" necessary for politicians? Or even -- God forbid -- "philosophy" for philosophers?
-- Assume you do need definitions to use words. Well, aren't words only ever defined in terms of other words, and aren't those words defined in terms of still other words? -- and isn't the system then self-enclosed, with no entry point?
***
-- You can point to things, and this has its own problems, among which is that you mightn't know what is being pointed to, and mightn't know what to generalize.
-- Presumably lexicographers gather raw data, and look for common denominators and patterns.
-- But, as in science, aren't there infinite hypotheses to fit any single set of facts? And even if you choose some arbitrary methodological principle, like Occam's razor, to select between theories, can't two theories be equally parsimonious, and aren't there a variety of ways in which something can be parsimonious?
-- A dictionary might give a one-sentence definition. -- Well, is one sentence sensitive enough for all the shades of meaning, and all the contexts, and all the effects in those contexts? -- All the interactions with etymology, codes of formality and politeness, group identity, metaphors and metonyms, rhetoric, sound and visual symbolism, cultural reference...?
-- And isn't usage constantly changing?
-- In what space do you differentiate? What counts as a common denominator? What allows you to recognize one? Why are these two homonyms treated as separate words, and why are these two heteronyms synonymous? Why is this single word divided into multiple meanings? -- Aren't you appealing to some system of categories, in terms of which you differentiate? And isn't this system something extra that you bring to the raw data, something you don't find within it?
-- Isn't there vagueness, always, at the edge of the word? Isn't there uncertainty as to whether, in particular places, the word may or may not apply?
-- I've rarely learnt how to use a word from the dictionary entry alone, as with the example of "gamin". But this doesn't mean the entry is useless.
***
Language is a tool to interact with other people's behaviour and psychology, to affect other people, and via them the world. Stripped down to its essential, it's no different from birdsong.
-- You create effects, achieve results, though language doesn't serve this purpose alone -- for instance, there are effects on yourself from the sheer act of speaking.
-- Language is also a tool to interact with your future self.
-- It's tempting to assert that languages must be growable, because you interact with different people using words in different ways, and because individual and collective experience grows. -- Tools must be simple enough to use readily, and flexible enough to cope with change: -- vagueness is essential.
-- Nouns don't stand for finite sets. When you learn "dogs", you're not told of all dog members. At best, "dogs" refers to an infinite set -- but even this picture isn't quite right, because you're not given a series like "1, 2, 3..." -- you're not given a clear rule to generalize.
-- Does anyone use nouns with single rules? In the course of your life, don't you modify the rules, or switch to different rules?
-- The older you get, the more rules and complications you learn.
-- And, to revisit an old idea, aren't a lot of nouns more a matter of family resemblance than necessary criteria?
-- And if you did have a clear rule, why would you have to stick to it? It's your prerogative to improvise, to employ a screwdriver as a hammer or as a paper-weight. Everyone must develop their own toolkit.
-- The suggestion, therefore, that language use is creative, that the sum of necessary conditions, if there are any, is not sufficient. When you apply a word, do you know that it'll stick? Perhaps you always have a choice. Perhaps every application of old word to new object is a leap in the dark. Perhaps every dog you meet expands the meaning of "dog", and expands the vagueness.
-- The world in its wit presents us with cases that don't fit neatly into existing category systems. And so we must invent new words, and extend old words. But even if there were no new cases -- even if we'd experienced everything there was to experience -- could we then compartmentalize neatly? Is reality itself clear? Could we find clear rules? Would it be practical to want to?
***
The concept of Pi or Fermat's theorem is elucidated in mathematical space, and is then named. But if you pick any ordinary concept (courage, dog, the colour orange), the concept exists in behaviour space, yes, but how can it be spoken of outside of words? How can I write to you now and answer the questions "What is courage?", "What is a dog?" "What is orange?" outside of words? Where is the mathematics for this?
-- Isn't a "What is" task almost always a lexicographical one, a defining one, with all of defining's problems? Isn't the definition of a concept the definition of a word, and a job for dictionary-writers?
-- Isn't the task of "conceptual clarification" more a matter of conceptual replacement, and rule replacement, gauged by precision with respect to particular practical purposes?
-- So few of our existing words are neat -- there is little starting point on which to build a complete set of neat boxes. We would have to invent a new language, like biology, and hope that that new language was adequate.
-- When someone inquires "What is love?", they might think they're asking for a definition, but need they be satisfied with one?
-- Aren't they asking, rather, after understanding, and aren't there other paths to understanding than definitional?
| 98 |
| Vote |
Add To: 
subscribe to this blog
Comment by katyzzz
Photography Tips
MS Paint Art
I see your propensity for hard work and elucidation of the unelucidateable has not lessened, in fact I think it has increased, but that is just my opinion, of course; I was ritually pleased to see mathematics get a mention, that pleases me no end.
Nice to see you, pop by again and as the 'old expression goes' - come up and see me some time, and even that needs interpretation, although I think it was quite clear to Mae West what she meant, then again, she was a blonde, wasn't she?
katyzzz
Comment by Damo
I think you will need to change your name to the Phantom of Orble.
Language is a fascinating bunch of vague expressions we use to communicate. I was in some pretty tough meeting some years ago and the biggest problem we had was agree upon the definition of word and terms. Same company but seperate sub-cultures had hijacked the meaning of key words.
Even the Oxford has forsaken me from time to time. The fact that people can communicate with words is nothing short of amazing.