[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index

Infinite series, Zeno, simulations...



I've been doing a bit more research on the whole situation of sums and products in infinite series and I am struck by just how much information there is on the internet about all this stuff. Sheesh! My father had a knack for finding the hornet's nests, didn't he?!
 
Mainly, though, I have been forced to conclude that it's a waste of time to try and get inside the thought process involved... my entire problem with all mathematical logic can be easily illustrated by one example from what I've learned about convergence and infinite series: It's the convergent series: 1 + 1/2 + 1/4 + 1/8 + 1/16 +.......... = 2.  The mathematical "logic" involved in order for the field of mathematics to officially decide that arbitrarily close to 2 means that the series converges to 2; 2 is the limit of the series-- That's the kind of thing that my mind balks at. As if getting close to 2 is "good enough",  "might as well be 2" or "is the same thing as 2"... when mathematics ironically prides itself on exactness..... My conclusion is that what is referred to as formal logic is inconsistent, even within the realm of mathematics-- and the reason it is is because human minds are trying too hard to retain a limited notion of consistency in the face of complexity. That was pretty much Robert Rosen's diagnosis, as well, and while I admit to the overwhelming influence of a Rosennean point of view, both genetically and because of my life experience... my father was able to speak, understand, and navigate the language of mathematics as fluently as any pure mathematician. The fact that our two points of view wind up in the same place after we each do our analysis is what I have meant in the past when I said that Robert Rosen's scientific work passes my "common sense test".
 
For any who are interested, here's my (non-formal) logical assessment of the whole "infinite series" sums and products, thing:
 
If they're infinite, they're infinite. Period. If you make them into something finite, you're not working with the same system, anymore. Worse; the information gained by the exercise is not necessarily applicable-- so it is dangerous to use such information for answering questions which pertain to the infinite realm. Wouldn't it be more useful to study the various mathematical entailment relations of the finite and the infinite realms, via the behaviors of mathematical systems and their interactions in each, rather than attempting to make the infinite realm behave as if it's finite? This was partly where my little mathematical insight, posted last week, came from. We need to do the same exact thing in science.
 
The way I see what Cauchy was looking at,  there will never be convergence. It is not possible. An infinite series is, among other things, an infinite process-- of either addition or multiplication-- and it's the process that makes the series infinite, not the physical numbers themselves. Mathematics as a language is a complex system, in and of itself, yet mathematicians keep focusing on little pieces of it-- in this case, the numbers (as if the process or any other aspect doesn't really matter). In a finite world, that may make sense (commutativity, associativity, etc...) and when it does make sense, then the strategy may be useful. But this particular situation is about  processes taking place in the infinite realm. Humanity, it seems to me, has a tendency for always trying to apply finite rules to "the problems of dealing with" an infinite world-- even after we recognize that it's infinite (as is the case under discussion here).  One of the truths about mathematics as a whole that comes out of my father's work in complex systems theory is that any particular field of mathematics will have active relational information from the larger system of mathematics "encoded" into it. In this particular case, it's actually visible: Even though we are supposed to be dealing strictly with sums and products, both addition and multiplication are specified by a larger system of entailment which includes the opposites of both (adding fractions together is a case of adding processes of division; multiplying fractions is a case of multiplying processes of division; adding positive and negative numbers together is a process of adding processes of subtraction)...
 
The paradoxes arise precisely because a single realm of mathematics is treated as if it's not connected to the rest and does not carry the entailments of the larger system within it. People have labeled such relational aspects "semantics," defined semantics as an impediment, and tried to get rid of the relational aspects by replacing them with different entailment (syntactic/formal entailment). They oversimplified, which is practically guaranteed to cause side effects, and the paradoxes are the result. Paradoxes can be defined as a symptom of an imbalance somewhere in the inferential entailments of a formal system. They are a side effect of lost, or otherwise missing, information.  An analogy to this is the past experience in medical science of thalidomide damage to babies borne of mothers who were prescribed thalidomide at a critical point in their pregnancies. This was a classic case where a side effect was caused by tinkering in a complex system using medical answers based on information derived from simplistic models.
 
[Incidentally, the most important aspect those models failed to take into account in their experiments, in my opinion, were the relations of time. There are specific time relations involved in pregnancy that are related to but distinct from the time relations in "human female" organization, another distinct set of different time relations involved in human embryonic development, and then there are yet other time relations governing the interaction between the other two sets. They are all related, they all affect one another, but they are distinct from one another. When the researchers did their drug safety experiments for thalidomide, they assumed that all phases of embryonic development were equally vulnerable to any exposure of drug. This is not the case. The drug's characteristic side effect; a baby born with some variation on hands growing from shoulders and feet growing from hips, was generated because thalidomide only interfered with embryonic development during the phase of long-bone formation. This particular embryonic phase happens to occur during a pregnancy's first-trimester phase, so it coincides with the worst nausea for the pregnant woman. Thalidomide was being prescribed as a therapy for nausea.]
 
I think the main problem of the infinite series "solutions" in mathematics is also partly the "forgetting" or formal elimination of the relations of time that are an integral aspect, built into the process. Any infinite process can only be infinite if time is part of the organization. Taking time out of the series and dealing with the result as a finite process is analogous to trying to scientifically analyze a complex system via "states".
 
Unfortunately, the finite system that convergence relations create (no offense to Cauchy) to deal with the symptoms (paradoxes) is based on the same effort and the same thinking that caused the symptoms in the first place. As often happens in such a situation: the therapy seems to eliminate the symptom... but it does nothing about the generator of it; the symptom was merely an observable consequence of entailments. Rather than addressing whatever created an imbalance in the system entailments, this approach gets rid of a symptom by applying more of the same therapy that caused the imbalance in the first place. The result is guaranteed to generate MORE SYMPTOMS; mixed up together because we now have two generators. In a relational world, we can pretty much expect that the two generators will begin interacting with each other and with the rest of the system, over time. What is the next solution? To fractionate further and create a whole family of convergence relations, each designed to deal with specific categories of symptoms.... and in effect, they have created a new infinite process of their own; what RR called "an infinite regress". There are currently many different sets of convergence relations for addressing all manner of different kinds of infinite series-- it's just nuts.
 
The properties of finite mathematics which were so starkly inapplicable in  infinite series of addition or multiplication are commutativity and associativity. Commutativity [ 5 + 6 = 6 + 5 ] and associativity [ (1 + 2) + 3 = 1 + (2 + 3) ] are properties that hold in the (finite) process of multiplication as well. Therefore, these are considered to be mathematical laws in multiplication and addition and the fact that "you can depend on this" has been built into a whole bunch of other operational rules and mathematical procedures. I guess that's why the realization that they don't apply in infinite cases caused such distress. But there's a lot more than just those examples in the set of things you can do in a finite world that don't transfer to the infinite. For example, in a finite world all of the following are absolutely equal to each other and to the final answer, which is exactly 1: [6-5], [-5+6], [-5- -6], and so on. Furthermore, in a finite world, the following are all equivalent to each other, because their end results are all equal: [3+3-5], [6-6+1], [1006-1005], etc. Furthermore because the first group and the second group all have the same final answer, that means the first and second group are equivalent to each other. My immediate reaction to such statements during my research was: Just because the final answer is the same doesn't mean that these examples are the same-- in fact they are all different. It's just that the ways they are different from one another don't impact the conclusion: the final result of each of these different processes is the same. On the other hand, that is only because they are happening in a finite realm, where time stands still and where relational aspects are minimized.  
 
There is another analogy here between mathematics and science: to reduce an entire set of mathematical processes down to a single, numerical result and define each process in the set entirely on the basis of the result is just like calling a simulation "the same as" the system it simulates. That's pure nonsense if the entailment underneath is entirely different. However, mathematics and science have both been really lucky: When dealing with simple systems, the negative impact is "equivalent" to zero!!! (sorry... couldn't resist!) In a finite realm, if we create a simulation of a finite system, like a car-- a new system that has all the same outward appearances and all observable properties and behaviors of a car, it will perhaps actually BE A CAR. So, if we simulate it accurately enough, we may actually create another of the system being simulated. But even if there is no negative impact from doing so, it's a bad habit to get into.  The rules that appear to work in a finite realm are not necessarily applicable at all; it's just not showing up as negative impact. This is exactly the situation with the machine metaphor in physics.
 
In the course of this past week or so, I researched the three most famous of the "Zeno paradoxes" (one of which is "Achilles and the Tortoise"). I did so because the Zeno paradoxes were often invoked in any description of the ideas surrounding convergence and infinite sums, etc. After reading these paradoxes and the stories of the trouble they caused in the history of both science and mathematics, I confess to a strong feeling of bemusement. Frankly, I think the Zeno paradoxes are just as illogical as the "problem" and "the solutions" when dealing with infinite sums and products in a mathematical series. In a strange way, they are sort of inverses of each other. The Zeno paradoxes were created by trying to make the finite, infinite.  In contrast, the mathematical convergence rules (regardless of which "type" of convergence it is) all try to make the infinite finite. In both cases, the system entailments are completely warped after being processed in that way and therefore, the results are less than useful.
 
I find it a little mind-blowing that those paradoxes drove people nuts for centuries. Even more mind-blowing is the fact that they never came up with the analysis that it is something they were doing to the system prior to the generation of the paradox that needs to be looked into-- rather than concentrating solely on the paradox itself and labeling IT as the problem to be solved.
 
Judith