Friday, October 21, 2005
Geese, Elephants, and Blind Men
Jeff Pollock has a 3-part article, titled "Semantic Elephant" 1,2,3, at AlwaysOn about the predicted semantic web. He compares our present understanding of what it will entail to the classic story of the blind men and the elephant. In recent entries here, I made several posts that can be related to this metaphor.
But instead of the blind men and the elephant, I used the metaphor of a flock of geese. Meaning (semantics) is like the V-shape formed by a flock of geese in flight. In Meaning is Real But Beyond Our Grasp I speculate that a real machine semantics will be never be in the possession of a single machine, not in the form of an ontology nor any other form. The reason for this, as I describe in another post, Meaning as the Coordination Between Agents About a Signal is that the meaning of a signal is a coordination equilibrium between two communicating agents. And just like in the geese (or the blind men), it doesn't require any sort of agreement, nor does each goose need a global model of the V-shape (or the elephant) for it to form. In spite of this, I am not at all pessimistic about the prospects for a semantic web. I just believe we need to work out the mechanisms by which swarms of machines can interact socially in a way that lets true ontologies emerge, perhaps as described by the work cited in Emergence - A Missing Piece of the W3C's Semantic Web.
The work done so far seems to be to be too much of a 'private language'. As Wittgenstein says of it, "Why can't my right hand give my left hand money? --My right hand can put it into my left hand. My right hand can write a deed of gift and my left and a receipt. --But the further practical consequences would not be those of a gift. When the left hand has taken the money from the right, etc. We shall ask: "Well, and what of it?" And the same could be asked if a person had given himself a private definition of a word; I mean, if he has said the word to himself and at the same time has directed his attention to a sensation" - Wittgenstein, Philosophical Investigations, 268. So to a business that has described its data with an ontology, without participation in a social network that is coordinated about the common knowledge of that ontology, you have to ask, "And what of it?" There is more to language than thinking. Language is a sort of shared thinking.
But instead of the blind men and the elephant, I used the metaphor of a flock of geese. Meaning (semantics) is like the V-shape formed by a flock of geese in flight. In Meaning is Real But Beyond Our Grasp I speculate that a real machine semantics will be never be in the possession of a single machine, not in the form of an ontology nor any other form. The reason for this, as I describe in another post, Meaning as the Coordination Between Agents About a Signal is that the meaning of a signal is a coordination equilibrium between two communicating agents. And just like in the geese (or the blind men), it doesn't require any sort of agreement, nor does each goose need a global model of the V-shape (or the elephant) for it to form. In spite of this, I am not at all pessimistic about the prospects for a semantic web. I just believe we need to work out the mechanisms by which swarms of machines can interact socially in a way that lets true ontologies emerge, perhaps as described by the work cited in Emergence - A Missing Piece of the W3C's Semantic Web.
The work done so far seems to be to be too much of a 'private language'. As Wittgenstein says of it, "Why can't my right hand give my left hand money? --My right hand can put it into my left hand. My right hand can write a deed of gift and my left and a receipt. --But the further practical consequences would not be those of a gift. When the left hand has taken the money from the right, etc. We shall ask: "Well, and what of it?" And the same could be asked if a person had given himself a private definition of a word; I mean, if he has said the word to himself and at the same time has directed his attention to a sensation" - Wittgenstein, Philosophical Investigations, 268. So to a business that has described its data with an ontology, without participation in a social network that is coordinated about the common knowledge of that ontology, you have to ask, "And what of it?" There is more to language than thinking. Language is a sort of shared thinking.
