Causal Determinism (was "God's Warriors")

This started out as a tangent from "the God's Warrior's" main Topic thread. Perhaps all has been said in the many cross-Comments there. However (at Joel's suggestion), here is my new Topic on determinism. Although Spinoza and Einstein cannot speak for themselves, having passed away, I believe I have inherited some measure of their "spirit" and they would agree with most of my views. (Other Spinoza/Einstein fans might disagree :^)


The best and most readable scholarly treatment I've found on the topic may be found in the online Stanford Encyclopedia of Philosophy at: http://plato.stanford.edu/entries/determinism-causal/



DEFINITION (from Stanford):

Causal determinism is, roughly speaking, the idea that every event is necessitated by antecedent events and conditions together with the laws of nature. The idea is ancient, but first became subject to clarification and mathematical analysis in the eighteenth century. Determinism is deeply connected with our understanding of the physical sciences and their explanatory ambitions, on the one hand, and with our views about human free action on the other.

Given the "initial conditions" of the world, and the laws of nature, everything that has happened and everything that will happen, have been and are destined to occur. However, that does not imply either perfect predictability of the future (at least not by humans) nor what is ususally meant by "fate".



"Fate" is a mystical concept and sometimes the subject of fiction: Person A is "fated" to die in a certain way or on a certain date, he or she learns of that "fate" and tries to avoid it, but it happens anyway. For example, Macbeth thinks he is safe from being killed by Macduff because his "fate" is that he "cannot be killed by woman borne". Macduff tells him he was born by Cesarian section, and kills Macbeth! Or, in a modern context, a person learns he is to die on a certain day and it turns out he has airplane tickets for that day. So, he cancels his flight and stays in his apartment, only to die of an explosion due to a gas leak! The airplane on which he was scheduled to fly lands safely.



Perfect predictability depends not only upon the universe being causally deterministic (which I happen to believe it is), but also on perfect knowledge of all the laws and forces of nature and positions and velocities of all objects (which, other than in certain limited domains, is impossible for humans, even with the aid of computers, to achieve.) The Stanford link provides an 1820 quotation from Laplace that "the perfection [of] the human mind ... affords but a feeble outline of such an intelligence."



When Joel speaks of the "predictability" at the heart of the scientific method, I believe he is refering to highly controlled experiments where extra care is given to isolate the experimental chamber from the rest of the universe. If an earthquake or lightning or a clumsy assistant causes the beaker to shatter, the results are discarded and the experiment must be repeated.

Something is a scientific fact if and only if any competent scientist can replicate the experiment under the same or similar controlled conditions.



An Actual Example of Determinism With (and Without) Perfect Predictability

I mentioned my HexLife computer simulation of evolution of life forms in the previous thread. HexLife is a very well-controlled situation and the model for my concept of causal determinism. It is run within a PC and is therefore both finite and discrete.

I make use of the PC "random number generator" to determine when and where genetic mutations and crossovers occur, how much of the "sunlight energy" falls on any given location, and so on. However, I allow the user to select one of a number of pre-set "initial random seeds" (or make up his or her own). Thus, if something interesting happens during a given run, that run can be repeated and will proceed bit-for-bit in the exact sequence to allow the user to analyze and exactly recreate that interesting event. In that sense, HexLife is perfectly predictable, so long as the PC does not malfunction.


In any case, I was quite pleased when my HexLife program turned out to exhibit a logarithmic "power spectrum" of extinctions (many small extinctions, a lesser frequency of moderate extinctions, and few large extinctions) just as Stephen J. Gould ("punctuated equilibrium") says characterizes the evolutionary history of life on earth!

I did not specifically program HexLife to have extinctions. Each simulated organism lives its own life in accordance with the behaviors and asexuality or sexuality set in its genetic code and its interactions with other organisms. It absorbs energy and matter (which are conserved in the total simulation), and reproduces asexually (if it gains enough mass and energy) or sexually (if it gains enough mass and energy and happens to meet a suitable sized mate). It dies when it gets eaten by another organism, or has an accident, or of old age.



Although not specifically programmed to do so, the total living biomass increases as new organisms evolve and become more effective at exploiting the environment and other organisms. The biomass stabilizes as a balance is reached between resources (dead matter and energy available in each location). Then, as yet other organisms evolve, there are extinctions, which are noted as significant reductions in the living biomass. Extinctions are characterized on a logarithmic scale and, amazingly, each increment of the scale eventually gets about the same number of extinctions. That result indicated to me that I had captured, on a very small scale and under controlled conditions, some "true" aspect of evolutionary life.


"God" and His/Her Possible Powers

With respect to HexLife, I am "God" -- having made up all the rules for DNA codes and reproduction and energy and mass distribution and so on, and having written every line of Pascal code. (Therefore, *I* am the closest thing to "God" you may ever meet on earth, so pay attention! :^)

When I start with a different set of initial conditions and/or a different initial random seed, I cannot predict in detail what will occur. Based on prior runs with different conditions, I can predict the general results: if the simulation is started with randomly-generated genetic codes, and if it does not die out early in the run, genetic diversity (number of different genomes in live organisms at any given time) will decrease and morphologic complexity (size and variability of genetic code string) will increase. After a while, these measures will stabilize. Then, there will be an extinction during which genetic diversity will increase a bit and morphologic complexity will decrease a bit. This will go on and on in sequence and there will be a power spectrum of extinctions.

Therefore, for a new run, only the general results are more or less perfectly predictable by me, but the details are not. However, if I have run the same initial conditions before, and kept the data, I know exactly, bit-for-bit, what the detailed results will be. Assuming there is a God of the Universe separate from His/Her Creation (which, as a Pantheist I do not accept), if the Earth has not been "run" before but other biological runs have been done, God only knows the general results. On the other hand, if the Universe has been run before with the same initial conditions, God can perfectly predict each and every event that will happen in full detail.



With HexLife I can intervene during a run, increasing or decreasing the strength of the "Sun". I can also stop the run or revert it to a previous data point and add or subtract organisms and/or change the genetic codes of any organisms. During the development cycle, I rewrote the Pascal code to change the "laws of nature" within the simulation. However, when I did these things with a situation that had been run before, any specific data I had about future events became more or less moot. Thus, if there is a God separate from Creation, and He/She performs a "miracle", God cannot predict the long-term consequences perfectly. Of course, if God has access to multiple Universes, He/She could put our Universe on hold while He/She tests out a "miracle" on a doppleganger Universe and then come back and start our Universe up again with perfect predictability!


Random Number Generators and Enumerability


The random number generator that comes with the PC is pretty good and would pass most statistical tests of "randomness". However, even the best computer random number generators are not "truly random". Digital computers are "finite automata" and have a finite number of possible states. Thus, the random number sequence will, at some point, repeat itself. How often? Well, depending upon the design, it might not repeat for days or years or decades or longer -- but it *will* repeat.

Therefore, any digital computer random number generator has a sequence that is FULLY DETERMINED and ENUMERABLE. Given any initial state, it is possible to determine the exact sequence of "random" numbers that will follow and you could store each and every "random" number in that sequence on a CD or other storage medium.

Do you ever get the impression that TV programs repeat? Well, it is true. Even if you have the latest digital HDTV "1080p" (1920 x 1080 pixels, each with 12 bits for color), that is "only" about 8 billion different FULLY PRE-DETERMINED and ENUMERABLE instantaneous pictures, most of which would look like confetti and are therefore unlikely to be broadcast. Thus, there is a much smaller number of instantaneous pictures actually broadcast. Of course, there are 50 individual pictures each second and they are presented in different orders, so the number of TV programs is enormously large, but still FULLY PRE-DETERMINED and ENUMERABLE.

It is possible, at least in principle, to enumerate all 30-minute TV programs that could be broadcast in HDTV. That enumeration would include all videos broadcast so far by Osama bin Laden and all future Osama diatribes that will be presented in the future. Of course, it would also include many Osama videos that will never occur in actuality, including the one in which he appologizes for 9/11 and converts to Judiasm :^)

This reminds me of the old thought experiment of the "million monkeys randomly typing on a million typewriters for a million years". They will eventually "randomly" type out all the great literature of the world.

Of course, you would have to get a million language professors to paw through all this stuff to find the gems. Imagine one professor, after years of reading gobbledegook, picking up a sheet that starts: "To be or not to be ..." (But then, it continues: "that is the gribnick? Michan&83( jkjhs nnnneo ...")


Is the Universe Both Finite and Discrete?

Although I am not sure if philosophers have made this point, I believe causal determinism depends upon the Universe being both finite and discrete (and I believe it is).

Here is my reasoning: According to chaos theory, a tiny difference in initial conditions can result in a major difference over the long-term. The example usually given is the "butterfly effect". Imagine an experiment in which the Universe is run twice with absolutely identical conditions, except, in the first run one particular butterfly flaps its wings slightly to the right and in the second slightly to the left. Well, in the first run, six months after that butterfly flap, hurricane Katrina strikes New Orleans, while in the second run, it strikes Miami.

If the Universe is continuous, that tiny difference could be 1/infinity, which, for all practical purposes, is zero. Therefore, with a difference approaching zero, all long-term determinism goes out the window. On the other hand, if the Universe is discrete, the difference must at least one quanta of energy or matter or space or time.

There is an old thought experiment where a donkey is placed exactly midway between some food and some water and it is exactly as hungry as it is thirsty. The donkey will be exactly equally attracted in each direction and will therefore die of thirst and hunger. Of course, that experiment depends upon a level of exactness and constancy that is unlikely in the animal world. The donkey would probably sense the water before the food (or vice-versa) or, after standing motionless, would become thirsty before becoming hungry, etc.

Let's do that thought experiment a bit differently. Take a common plastic straw and stand it upright on a flat table. If there are no strong breezes or gusts, it will remain stationary, falling neither to the left or right, front or back. Now try it with a toothpick! Unless the toothpick has a flattened end, or is poked into a hole in the table, it will fall over no matter how carefully you place it. The slightest waft of air or vibration of the table will knock it over. The difference between a straw and a toothpick? One has an end with a discrete diameter, the other an end with a near-zero diameter.

As I mentioned in the previous thread, we know that energy comes in discrete quanta. Since we also know energy and matter are interchangeable, that would mean matter is also quantized. That leaves only space/time, and I believe they may also come in quanta. Thus, I think it is a good guess that the Universe is discrete.

The second requirement, a finite Universe, is, I believe fairly obvious. I think the current scientific consensus is that the Universe began with a "big bang". (The alternative -- watch your daily newspaper -- is that the consensus will change to "continuous creation" of matter which I believe also entails continuous recycling of matter to keep the balance.) In either case, there is only a limited amount of energy/matter in the Universe.

Furthermore, the Universe is curved in some or all dimensions, which would entail space/time being finite as well. (The surface of a sphere, for 2D "flatlanders" appears infinite in that they can go forever in a given direction without reaching the "end". However, we 3D-ers recognize the sphere as finite, and we explain the delusion of the 2D-ers as being the curvature of their 2D space.) Thus, to us 3D-ers, what appears to be infinite time and space may actually be finite (and would appear so to a 4D-er or higher).

Ira Glickstein