X-Message-Number: 20890 From: Date: Thu, 16 Jan 2003 14:30:22 EST Subject: information conservation again --part1_18c.14a06745.2b58624e_boundary Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Michael Price says there is no evidence for conservation of information. Without trying to define "evidence," I can think of several lines of thought indicating the possibility or likelihood of information conservation. 1. On a practical and approximate level, there is much more information potentially available than most people think. For example, to determine the genome of a particular individual to a high degree of accuracy, you don't need any part of him--only a few small specimens from some blood relatives. Likewise, a tremendous amount of information about a cryopatient's brain can be inferred from outside records, including the memories of friends and relatives. 2. Cryptology tells us that, while it is easy to degrade information, it is exceedingly hard to conceal it--even when you are really trying, which Nature is not. 3. The universe may be deterministic, as many first-rate people have thought. In that case--at least in a finite universe--information in principle is conserved. (It is irrelevant that, in a local region of spacetime, one cannot infer the past with complete confidence, since sometimes B would follow from A and also from A1 and A2 etc. In any sizable system the possibilities quickly dwindle to one. Any guess about the past or present must conform to known records, such as written records, fossil records, cloud chamber tracks, etc.) (David Deutsch thinks the "multiverse" is deterministic, although a particular "universe" is not from the point of view of its inhabitants; but he also thinks that, in any universe, the past is fixed. If he is wrong, and there are "many histories" instead of many worlds, that's a different kettle of fish.) (The current majority interpretation of the current version of quantum theory is not only on shaky ground, but from an historical perspective is almost certain to be wrong in fundamental ways. There are wildly divergent views. Bart Kosko thinks the current quantum physics is almost certainly wrong simply because it is linear, and experience tells us that nature abhors a straight line.) 4. Although now more or less outmoded, Einstein's 4 dimensional block universe suggests a "map" of eternity already coexisting--past, present, and future equally definite and fixed, the "flow" of time an illusion. In that case, everything is conserved--except perhaps human error, which always finds more room to grow. 5. It should be noted that "information" means different things to different people and in different situations. Shannon's classic work on theory of communication made this distinction carefully. A "lot" of information in an engineering or communication sense does not necessarily mean a lot of information semantically. Suppose a zillion monkeys work at a zillion keyboards for a zillion years, and finally one of the print-outs is a verbatim copy of the U.S. Declaration of Independence. How much information is there in this printout? "None" would be a good guess, but it depends on your frame of reference. Key questions include the "intent" of the sender (even if the sender is Ma Nature) and the "interpretation" of the receiver. The point is that definitions of "order" or "disorder" are not so simple and any one is applicable only in its own appropriate niche. As previously noted, we might ask whether being dealt a royal flush in spades at Bridge is a common event or an unusual one. Most people at first thought would say "unusual," but they have already had their biases built in. After more thought, we might ask, "Compared to what?" If you compare a royal flush in spades to any other distribution, previously exactly specified, even with honor points zero, the probability is precisely the same for any hand. Furthermore--and this is frequently lost sight of--the effective or semantic content of a message or observation need not depend entirely or even primarily on its complexity. It also may depend heavily on CONTEXT. For example, a Yes or No response to a previous question (e.g. "Is Natasha Nogoodnik a spy?) might convey (in one sense) a tremendous amount of information, answering many implied questions. Robert Ettinger --part1_18c.14a06745.2b58624e_boundary Content-Type: text/html; charset="US-ASCII" [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=20890