Science Building, Room 227
For further information contact Paul Grobstein.
Summary by Paul Grobstein
Steven intiated a discussion of Information: The New Language of Science by Hans Christian von Baeyer (Harvard University Press, 2004). The conversation was based on a set of Steven's notes about the book. It got through the first four chapters, and will be continued in future meetings to be scheduled (with Eric talking next week about issues in linguistics).
The book is an effort to bring to general attention the need for a better understanding of information, rather than a description of such an understanding (though it points in what von Baeyer feels are promising directions). The first chapter suggests that such an understanding ought to satisfy both "scientific" and "humanistic" intuitions/needs, but the book is clearly primarily written largely from the perspective of physics. A distinction between "natural" and "artificial" information provoked some interesting discussion. Is this a distinction between perception and creation? between permanent and creatable/destroyable? between non-human and human?
The second chapter contrasts "atoms and the void" with the "perception of a material world" and suggests that "information" is what connects the "real world" with "mind/consciousness". This triggered some considerations of the degree to which the "material world" and "mind/conciousness" were or were not separable from one another and the likelihood that causal relationships were bidrectional. The case of ambiguous visual images was noticed as an example. If the brain (including "mind/consciousness") is a part of the material world, then that world is indeed changed when one sees things differently. And one can of course further alter the material world by virtue of how one sees images (PG ... what is the origin of the skepticism about whether "mind" can alter "matter"?). A possible reading of Chapter 2, akin to last week's discussion, is that it represents von Baeyer's route to the conclusion that information is the "organized" aspect of matter/energy. Also interesting in this chapter is the notion that a "theory" contains very large amounts of information, in the sense that it can be unpacked to make lots of effective predictions (PG ... how do either/both of these things relate to the earlier "connecting" idea or to the information-as-that-which-is-decoded idea from last week?).
Chapter 3 of von Baeyer starts from the proposition that historically "in-formation" involved a belief in the imposition of form onto something. This fit the notion of information as related to "organization", the existence of relationships between things. A key point of discussion here was whether an "outsider" was needed to "impose" form or whether instead von Baeyer was simply referring to the origins of the term. Issues were also raised about whether the use of terms like "sign" and "symbol", and "form" and "pattern" were or were not interchangeable. von Baeyer seems not to make a distinction between "potential" information and effective information of a sort that seemed important in our previous discussions. The question of whether an entity that took in signals and generated outputs independent of them was a decoder was raised, and it was noted that the decoder notion of information wasn't (yet?) sufficiently well developed to deal with concepts like "accuracy" or "usefulness".
Chapter 4 talked about Shannon and "bits", adding to it Gell-Mann's "Information Gathering and Utilizing System" (IGUS) as a way to make the point that it is not enough to simply "transmit" information; it needs to have an effect. In this context the "decoder" notion was criticized as sounding too much like "regurgitation". Must a decoder have "intention"? (PG ... or is it enough that it generates an output based on its internal organization? does that internal organization have to be changed?). The issue of whether a "bit" is a general and fundamental information notion or not was addressed (at least such a discussion was begun) in terms of whether "bits" are an appropriate way to think about language (PG ... clearly the "information" and even the "information content" is different for a "meaningful" utterance of a given length than for a random one of the same length). One issue was whether there was one kind of "bit" or whether there could be several kinds. A still broader issue was whether the formalism of a finite one dimensional string of elements chosen from a finite population was adequate for thinking about "information" in general terms. Would it make a difference, and in what ways, if one allowed an infinite element population? infinite strings? multidimensional relationships among the elements? These and the possibly related question of opting for digital versus analog characterizations were put on the table for future discussion, along with the possible future topic of "hypercomputation".
| Calendar | About
| Getting Involved
| Groups | Initiatives | Bryn Mawr Home | Serendip Home
Director: Liz McCormack - email@example.com | Faculty Steering Committee | Secretary: Lisa Kolonay
© 1994- , by Center for Science in Society, Bryn Mawr College and Serendip
Last Modified: Wednesday, 09-Jun-2004 08:14:10 EDT