(The following section soon moving to HistoryOfFreeWill)


I hope that “FreeWillAndDeterminism”, this introduction to free will and determinism, can be refined into something that will help many people come to a better understanding of the current state of human knowlege.

FreeWillAndDeterminism contains discussion summarized from an older discussion that degenerated into a flame war.

I decided to re-create that page, nominating myself as PageMaintainer, summarizing the ideas that were on the “free will and determinism” page before it was deleted.

Helmut and Lion made some great thought-provoking comments – I hope I’ve salvaged most of them. (But leaving the silly flame war deleted and forgotten – good riddance!).

I summarized the conversation into DocumentMode, removing the indications of who said what. I hope the original authors won’t mind. I hope this reduces the chance of another flame war.

I’m sure I’ve introduced a bunch of errors into this essay. When you find them, remember this is a wiki – you know what to do. Well, perhaps you don’t, so let me explain: rather than smugly think to yourself “What a moron”, or edit this page to add a vague one-liner at the bottom saying “Somewhere on this page is a glaring error”, please give some correct explanation, and some evidence as to why it is true.

And you may know that some evidence or experiment or explanation conclusively shows (or at least suggests) that one or more or all of the opinions I mention are completely wrong. Please tell me about them.

PageMaintainer DavidCary

(The following section soon moving to HistoryOfFreeWill)

why bother discussing determinism on CommunityWiki?

We expect to continue discussing AI, reason, decision making, etc. Perhaps a clearer understanding of determinism will help clarify those discussions. One central place also lets us talk about some of the more subtle details. When mentioning determinism on some other page, a brief summary glosses over many details and a full account would seem out-of-place.

Centralizing all the talk here is a form of InternetConcentration. It’s better to keep all these little ideas about determinism here in one place (rather than repeated over and over in pages discussing AI, decision making, etc.) to make it easier for me to point out “wrong, wrong, wrong, wrong, not even wrong” :-).

Currently this page is verging on Wiki:TooBigToEdit. Perhaps some ideas have grown large enough to bud out into their own page (as described in BigBucketsFirst).

(The following section soon moving to HistoryOfFreeWill)


why such strong emotions

Perhaps if we know a little history of this discussion, we can avoid repeating past mistakes, and instead make new, more educational mistakes :-).

Why does this complicated subject stir up such strong emotions?

This discussion touches some questions of quantum mechanics, cosmology, causality, predictability, and randomness. I think we all agree that coming to a correct understanding of the universe we all live would be really nice, but apparently not essential to the survival of the human race.

This is intended to give information and (hopefully) give insight. Shouting down and insulting people who express “wrong understanding” is not the best way to reach a correct understanding. We’re not trying to “win an argument” here.

Many determinists – people who believe that the universe is deterministic – suspect that it cannot be proven to always be deterministic.

There are brilliant social activists who do great works and who talk at great lengths about creativity and the freedoms of peoples the world over (quite properly, justly, creatively, beautifully, ethically, intelligently, and so on,) who are within a BoundedRationality?.

It isn’t their job or their interest to think much about these questions of machines and physics and brain processes and so on. There is an easy and sensible picture of the workings of the world that does nothing to destroy creativity, intellect, justice, beauty, and so on, but that people find none-the-less threatening.

Perhaps one reason this discussion stirs up such strong emotions is that we are inspired by stories where someone does something that no human has ever done before. So we are appalled when we hear one person tell another, “you’ll never be able to do that”, especially when we suspect it is not true.

Both sides sling this accusation (based on misunderstanding) at each other.

Another, less emotional accusation both sides sling at each other: “That’s not scientific”.

… put Feynman quote about “the expected failure of science” from QED here …

Perhaps another reason this discussion stirs up strong emotions: Arguments that conclude that artificial intelligences will never have “true” intelligence or “free will” make some people very uncomfortable. To them, those arguments sound so very similar to arguments justifying slavery of “inferior” races, and similar arguments justifying denying suffrage to women.

(The following section soon moving to HistoryOfFreeWill)

voting on philosophical issues

Perhaps we should stick in something about “Yes, but computers don’t have free will.”

Rather than going into long, philosophical debates on the subject of “free will”, trying to understand each other, I’ve found a much simpler approach: Let people vote: Determinism or free will?.

Once we get that silly issue all squared away, we can turn to the real stumper: Which are tastier, apples or oranges?.

Perhaps I should move this bit of all-too-dry humor over to the VotingIsEvil page.

(The following section soon moving to HistoryOfFreeWill)

voting vs experiment

I’m a little sad about what happened with the “free will and determinism” question. I admit I was hoping for a little bit of a reaction when I mentioned it. (I guess that means I was trolling. Sorry.)

If I remember right, I intended to focus on voting. When I mentioned “voting on free will vs. determinism”, the “free will vs. determinism” was a bit of a red herring. I was hoping to provoke a bit of an argument – but not between the “free will” supports vs. the “determinism” supporters. Instead, between the “voting” vs “experiment” supporters.

I expected people, such as myself, who think voting is a wonderful, to weigh in on how voting is a much better way to resolve this issue than beheading those who disagree with you. (Such as Johan van Oldenbarnevelt, beheaded in 1619 because he supported the “wrong side” of the free will vs. determinism issue. Well, at the time it was actually a free will vs. predestination issue).

In opposition, I expected people, such as myself, who think that voting is unnecessary when a single individual can verify, all by himself, that something is true. And even when he doesn’t know exactly how to verify some fact, someone can tell him some steps to take to confirm that fact. Then he can verify, all by himself, whether or not the proposed experiment will in fact discriminate between whether the claimed fact is true or false (or if it is a Wiki:NonSequitur). After that, he can, all by himself, conduct the proposed experiment, verifying that fact.

But the conversation turned from “voting” to “free will vs. determinism”.

(The following section soon moving to HistoryOfFreeWill)


In 1619, Johan van Oldenbarnevelt was beheaded because he supported the “wrong side” of the free will vs. determinism issue. Well, at the time it was actually a free will vs. predestination issue.

Around 1700, Sir Isaac Newton described universal gravitation. Newton said, “Gravity explains the motions of the planets, but it cannot explain who set the planets in motion. God governs all things and knows all that is or can be done.”

This leads to the idea of a “clockwork universe”, constructed and set into motion by a creator, perhaps with rare and infrequent miraculous interventions.

In 1862, Louis Pasteur showed that living things do not spontaneously arise from non-living matter.

This led many people to believe that all non-living systems would become easily predictable, but living creatures (with vitalism’s “spark of the divine”) exert their will.

In 1898, Jacques Hadamard pointed out that in some situations, if you measure the state of affairs to some given precision, the range of possibility rapidly diverges, so that your predictions become less and less accurate further into the future. This led to chaos theory.

In 1905, Albert Einstein showed that the “random walk” of Brownian motion can be entirely explained by the collision of individual molecules of the liquid. Some general features of this walk can be correctly predicted if you know a few bits of information about the kind of molecules in the liquid, and the temperature. But the exact path is difficult to predict.

More and more of the universe – objects that previously we had assumed had intelligence, life, or at least choice – were being completely explained using a “clockwork universe” theory. Even with the minor setback of deterministic chaos, it appeared that soon we would have a Theory Of Everything that explained everything in the universe – even living things and humans – in terms of simple atoms relentlessly following a few simple, “mechanical” rules of motion.

In classical physics, it was believed that if one knew the initial state of a closed system with infinite precision, one could predict the behavior of the system infinitely far into the future.

“Causality” is the idea that events can only be directly influenced by events in the past. (We often observe certain events that seem to be influenced by events in the future – for example, people buy frozen pizzas “because” in the future, they will want to eat those pizzas – but so far we’ve almost always found a way to explain those events in terms of past events).

A “local theory” is a restricted form of causality – one where events can only be directly influenced by nearby events. (We observe certain events that appear to be influenced by distant events – such as photographs of distant supernova – but in most cases, we’ve found a way to explain these events in terms of particles that were once local to those stars, then traveled at or below the speed of light, then impacted the camera).

Perhaps the last great triumph of classical physics is the general theory of relativity developed by Albert Einstein. Its predictions of events rely only on events in the past light cone of that event.

In 1905, Albert Einstein proposed a theory that light always consists of individual photons.

In 1925 Werner Heisenberg and Erwin Schrödinger invented quantum mechanics.

This idea that light is “quantized” in discrete amounts (“quanta”) led to quantum physics.

In 1927, Werner Heisenberg (based on work by Louis de Broglie) published his “Heisenberg uncertainty principle”, which (unfortunately for us) seems to imply that if you don’t already know the initial state, quantum physics makes it impossible to make the measurements with enough precision to make such predictions infinitely far in the future.

In 1927, Niels Bohr and Werner Heisenberg developed the “Copenhagen interpretation” of quantum physics. Unfortunately for us, this idea implies that even if we do know the initial state exactly, we cannot know which one of several possible outcomes will happen, but we can very precisely predict the probability of each outcome. This is similar to the way that we do not know which one of the 11 possible outcomes (2 … 12) will happen on a particular roll of a pair of fair 6-sided dice, but we can predict that (over large numbers of rolls) it will come up “snake-eyes” (1 and 1) precisely 1/36 of the time.

Many people don’t like the implications of the “Copenhagen interpretation”.

“Die Theorie liefert viel, aber dem Geheimnis des Alten bringt sie uns kaum näher. Jedenfalls bin ich überzeugt, daß der nicht würfelt.” – Albert Einstein (“The theory yields much, but it hardly brings us nearer to the secret of the Old One. In any case I am convinced that God does not throw dice.”)

“I do not like it, and I am sorry I ever had anything to do with it.” – Erwin Schrödinger, speaking of quantum mechanics

“Anyone who is not shocked by quantum theory has not understood a single word.” – Niels Bohr

“I think it is safe to say that no one understands quantum mechanics.” – Richard Feynman

“If that turns out to be true, I’ll quit physics.” – Max von Laue, Nobel Laureate 1914, of de Broglie’s thesis on electrons having wave properties.

“…if there is one lesson we should all learn from the history of science, it is this: The fact that an observation is replicable does not mean that our interpretation of the observation is correct.” – Gary E. Schwartz

During the “Bohr-Einstein debates”, Einstein and others attempted to prove that the “Copenhagen interpretation” was incorrect. The most famous objection is the “EPR paradox”. (More on this later).

In 1957, Hugh Everett developed the “many-worlds interpretation” of quantum physics. Even more people don’t like the implications of the “many-worlds interpretation”. The implications of the “many-worlds interpretation” are widely regarded as absurd.

In 1974, Pier Giorgio Merli led researchers to perform the “one particle at a time” double-slit experiment for the first time. (More on this later).

(The following section soon moving to HistoryOfFreeWill)


What will the future bring?

New, improved “Bell test experiments” are expected Real Soon Now.

Some people are building “chatbot” software and “anthropomorphic robot” hardware that are obviously completely deterministic. Every year people improve these machines, making new machines that appear “smarter” and “more human” than previous machines.

Many people argue that humans are not deterministic, because they have free will.

Which of these possibilities will come to pass?

These three possibilities seems to be talking about machines and technology of the future, but it’s really more about humans today:

(The following section soon moving to HistoryOfFreeWill)

the present

If we really understood humans, we would be able to predict which of those three possibilities will happen. But unfortunately, as of 2007, no human completely understands the human condition. So the best we can do right now is describe what we do know and the hard evidence supporting it, and summarize the various opinions and the soft evidence supporting them.

Some people say that “free will” is something that is beyond the limits of what “science” is able to deal with. People have been wrestling with the idea long before the word “science” even existed. And “science” does have limits. People are certain that when trees fall, it causes air vibrations even when there is no one there to observe it. While “science” is difficult to apply to phenomenon that no one observes and has no indirect evidence.

(The following section soon moving to HistoryOfFreeWill - ?)


Wow, ..!

Talk about herculean effort!

However, there is an easy way of seeing if these causal explanations are real or just wishful thinking: “Can it predict?” If it can’t, then that explanation is useless.

The question of determinism vs. indeterminism is not one of utility, rather, it is one of speculation regarding actuality.

The quest “What is Truth?” can be independent “What is Useful?”

In my case, I am happy to entertain Determinism+, (which includes a random component,) but not Free Will, because of my inability to answer the challenges following, “What could Free Will even mean?” (ex: “If we went back in time, 30 seconds, and replayed everything- would a person choose something different?”, was one of the challenges.)

One positive point about indeterminism is, that it is more tolerant. If we assume that sometimes things happen that can’t be predicted, then we won’t be as surprised when, in fact, things don’t go as predicted.

Indeterminism is more tolerant than what? The notion that determinism is true, but we can’t predict what (some class of systems, including humans) will do? How is indeterminism more tolerant than unpredictable determinism? The base premise is identical: That sometimes things happen that can’t be predicted. Thus, the conclusions are identical.

When an enemy sniper randomly picks one of six soldiers to shoot, it is unfair to assume that the wounded soldier “must have done something”, and then praise the 5 other solders for being good at finding cover and blame the wounded soldier for being bad at finding cover.

If we assume that every situation is deterministic, then we may waste far too much time and effort obsessively trying to discover the real reason something happened – or worse, give up too soon with an (incorrect) reason something happened.

(The same response: Unpredictable determinism fares exactly as well as indeterminism.)

If we assume determinism, then we assume that the result is predictable – but some situations (especially in politics) are humanly unpredictable.

This is not true, because the condition is incomplete:

Determinism is the first condition. That you have complete state information is the second condition, and it was omitted.

Determinism can be true, and the outcome can be unpredictable, because we cannot gather the state information of the entire universe at a snapshot moment. (Even if we did, where would we store the information?)

Our assumption of a causal-mechanistic paradigm led to the conclusion that the type of push we selected is guaranteed to have the predicted effect. But this neglects the fact that the harder we push a human or a nation, the more likely that there will be an unwanted and unpredictable reaction.

This is a self-defeater, because it is making a causal statement about “what will happen if you push…”

That is, by it’s own (flawed) criteria for what “determinism” means, it itself is a deterministic statement! “If you push harder, then it will be more likely to produce an unwanted reaction.” This is a deterministic idea.

It would have been much more helpful to understand that a complex system like a state is in principle unpredictable, and that there is the risk of failure.

It could also be argued that, since systems like states are unpredictable, that it doesn’t matter whether you attack or not attack, because it’s all unpredictable. No use looking for lines of causality– just go on in and attack if you feel like it. No telling what will happen, no?

To me, “pessimistic hard determinism” and “compatiblism” are both true; The diagram is simply the mapping of a different meaning of the word “free will.”

I take determinism+ to be true.

Then you say, “Free will means that souls have a magical ability to affect the universe.” Suddenly, I am a hard determinist. (No, I don’t believe in free will.)

Then you say, “Wait! Free will means that vastly complex unpredictable systems like brains can resolve to a decision point.” Whoah! Suddenly, I appear to be a compatiblist!

Did anything change? I did not. The definitions of the symbols did. (The goal posts moved.)

David, thank you for tackling this refactoring task. Is there an archival copy of the original page? I’d like to check a few points. BTW it’s a pity that cw has even dropped the backlink function and leaves that to an unreliable Google search.

With respect to the current state of the page, I think it is slightly biased.

For example when you talk about Brownian motion and say ”…is difficult to predict…” then this might understood by a reader as “it’s possible, but actually too much work, so it isn’t done” while I would insist that “it is practically impossible to predict Brownian motion” (and I would assume that it has never been accomplished, not even in simlpe cases), because this would imply the knowledge of exact state and thermal motion of some 10^N molecules in the system (with N being reasonable large, maybe 10..20, so this means at least billions of billions molecules). The same is true when you talk about the roulette wheel. You do not take into account that the touching points of the jumping ball are actually vibrating atoms. So from the initial macroscopic state of the wheel and the ball, probably very little can be predicted. It is just assumed that deviations in predictions come from inaccurate knowledge of the initial state and not from the dynamics and random process after some initial moment.

Another example: when you say that “the problem of “free will” is an old one, and still not completely solved.” this might be understood by the reader as “the problem of free will is 90% or 95% solved, just some details remain to…” but actually the problem of free will is not solved at all. People have thought and written about free will and have formed a number of opinions, but that is pretty much all we have. The topic of free will is tightly bound to religious and worldview issues, having far-reaching consequences.

I also miss a more scientific view on the Bohm interpretation. As long as it is not adding to the predictive power of quantum mechanics, and is only repairing worldview issues by adding entities that are not observable, OccamsRazor should rule this out. It is valid metaphysics (Popper: not possible to falsify) but not valid science.

But apart from that, I congratulate you on the job, and I think that it may serve as a foundation to continue the discussion.

I agree with HelmutLeitner, that my speculation of determinism beneath the quantum mechanical fuzz can be cut from a scientific worldview, by OccamsRazor. That said, it does not mean that determinism is false, and it still falls within the realm of “determinism+” (determinism, with a statistical random component.)

It does not mean that there is “Free Will,” any more than a marionette being controlled by random die rolls is said to exhibit “Free Will.”

Helmut, did you see NestedSpheresOfCreativity, and my (requested) pushing on AlexandrianMethod?

Thank you, HelmutLeitner. Alas, I neglected to archive the original version of the page before I started hacking it up. After editing my local copy, and cutting-and-pasting this into communitywiki, all I have left in my local copy is a section on “infinity” that seemed to be a “red herring” to this discussion. Also there’s a few bits about atoms and electrons, but nothing that one couldn’t find presented even better in Wikipedia:electron_cloud and Wikipedia:atomic_orbital.


Alas, I spend too much time working with very deterministic digital logic circuits, and (when they do something unexpected) poking them with sharp pointy o’scope probes until I figure out what’s really going on. Perhaps if I spent more time with analog electronics – or in the opposite direction in the NestedSpheresOfCreativity, more time with complex software systems – I would be more comfortable around randomness and uncertainty.

I agree that this page is unfairly tilted towards the “compatibilist completely determinist” perspective. Given my current state of ignorance, there are many other perspectives that are completely rational. Please help add more information about those other perspectives to this article.

I agree that, in practice, it is impossible to predict Brownian motion.

However, "The Invention of the First Wearable Computer" by Wikipedia:Edward_O._Thorp claims that a few simple measurements were enough for the “shoe computer” to predict fairly accurately which quadrant the little roulette wheel would fall in. I suppose it is possible that his limited success is the best anyone could possibly do – but it seems just as likely to me that if one were to take more accurate measurements, and take more of them, and do some more sophisticated calculations with that information, that one could narrow down the range of pockets even more.

Alas, I’ve been told that casinos now ban such precision measuring instruments :-).

David, I do not doubt that one can increase the winning probability by observations or measurements. I doubt that even a precision measurement can predict single results like “17” reliably, however hard you try. Macroscopic measurements do not completely describe the reality. There is no way around the uncertainty principle, for example. The preciseness of artifical symbolic systems, like logic or mathematics, can not be transferred to nature.

Helmut and I agree on this.

It would be valuable, for the discussion, to have a diagram and a language of “predictability,” saying, “Here are the sorts of things that are predictable, here are the sorts of things that are less predictable, here are the sorts of things that are utterly unpredictable.” Perhaps add in an axis showing how much effort it would require to predict, or something. (“Given a computer a trillion times the size of this universe, modeling every single component in this one, and perfect state information, assuming such state information could be (A) accessed, and then (B) represented, …” being a far mark on that axis, …)

Lion, one could try to create such lists.

These lists are assumptions, each entry hypothetical in nature, open for negotiation and scientific investigation. The separation in predictable and unpredictable (the existance of the unpredictable) is not.



I repeat this for the occaional reader: it is important to understand, that “unpredictable” does not mean that we have not yet learned to predict this, but that these things are in principle unpredictable. To put it extreme: not even God would be able to predict them, they are not determined by the present state of the universe. More exactly: We have not even a concept of a defined state of the universe. This means that the future is open, depending on responsible autonomous individuals. No-one can fully control complex systems (e. g. there is no chance for tyranny to persist, there is no chance to build a completely fail-save technical system).

Well, that’s Helmut’s idea, at least. With the exception of the last couple sentences, I’m not so sure, …

My understanding is that Brownian motion does appear to be deterministic.

If we build a computer simulation of Brownian motion – a rigid container containing many water molecules and one larger rigid object, and let them move according to a few simple deterministic rules (“kinetic theory”) – then the motion of the large object appears to move “randomly”. However, if we run the simulation again, with the same initial positions and velocities, then the large object will move along exactly the same path. That path is completely “determined” by the initial state – the positions, sizes, and velocities of all the objects.

The exact path of the large object is highly sensitive to the exact starting positions and velocities of every water molecule in the container. It is very difficult to measure all that information in the real world. Fortunately, "some characteristics" of that path depend only on the temperature, the size of the molecules, Avogadro’s number and a few other easy-to-measure items.

Even though in the real world, the exact path of an object in Brownian motion is never the same twice, those characteristics of the path are repeatable. The theory that most closely matches the experimental reality of Brownian motion – “kinetic theory” – is entirely deterministic.

Although Brownian motion is completely described by a deterministic theory, there are many other things that, as far as I can tell, are not deterministic. But it is still too soon to tell whether those things are truly non-deterministic, or whether – like what happened with Brownian motion – the apparent randomness in those other things will someday be explained by a deterministic theory.

Isn’t the fact that they are repeatable just an assumption required to build the theory? I mean, you can create deterministic mathematical models and obviously they will always yield the same results for the same input data (this is a feature of deterministic mathematical models). We don’t have, however, any idea on how accurate those models are, as we are unable to set up an experiment with enough precision. Or am I just missing something and the models have been already verified to be “correct” within reasonable error margin?

Yes, any deterministic mathematical model will give the same results for the same input data.

The point is that not all physical theories are deterministic.

  • Kinetic theory is deterministic. A kinetic theory simulation, given a particular input state, run for a given amount of (simulated) time, will always end up in some other output state.
  • Quantum theory is not deterministic. A quantum theory simulation, given a particular input state, run for a given amount of (simulated) time, will give slightly different results from one run to another – unless we are very careful to use the same random number generator and reset it to the same seed value on every run.

Ideally we would set up a simulation exactly the same as some physical experiment, with each water molecule precisely placed, and the simulated mote would wander around in exactly the same path as the real mote. You are right that such an experiment is far beyond our current technology.

We are, however, able to set up an experiment with enough to precision to “verify to be correct” (“fail to falsify” in the Karl Popper terminology) the kinetic theory of Brownian motion.

One watches the little motes wander about, and every 10 seconds writes down the new location. Then one measures the distance from each location to the previous location 10 seconds earlier. This “distance traveled in 10 seconds” is the “some characteristics” mentioned earlier. Even though the exact path during those 10 seconds is very erratic and (so far) not humanly predictable, the distance measured is fairly constant. And when we adjust the temperature, or drop a bigger or smaller particle into the water, the theory predicts a new distance that is easy to verify (or falsify).

David, again I have to insist that

“The exact path of the large object is highly sensitive to the exact starting positions and velocities of every water molecule in the container. It is very difficult to measure all that information in the real world.”

shows a terrible bias. It is not “very difficult” but simply “impossible” to even measure the state of a single water molecule. The basic assumption, that a water molecule state is fully characterized by a position and a velocity, is totally absurd. There is not even a concept to describe a realistic state of a water molecule, including the rotations of the structure, the vibrations of the atomic nuclei and the dynamic charge densitity fluctuations coming from internal dynamics and the interactions with the neighbouring molecules. The description of single water molecule in a second might be as data intensive as the complete description of a hurricane with all its clouds and wind velocity, temperature and air pressure distributions over weeks or months.

BTW I actually spent about 3 years doing computer simulations of aquous systems during my former life at university. So I tried all this, I did all the proper and improper simplifications and know how far the current physical descriptions are from reality.

Even famous quantum mechanics is unable to numerically simulate a single water molecule - even without motion, vibrations and external interactions - with a reasonable accuracy (let’s say a few percent). So a molecule is far from being modelled with the same quality as a falling stone is modelled as a point of mass following the law of gravitation.

This is not only a question of incremental development. One will never be able to measure the charge density distribution in a single atom or molecule, because (1) there are no measuring tools (rays or particles) available and if one thinks this could change magically (2) because of the uncertainty principle.

Of course all complexity doesn’t mean that such systems can’t be deterministic, but … if one knows this overall situation and … that physics has to build on indeterministic processes in the world of the smallest phenomena … then I think the idea of an overall deterministic world becomes an illusion. Again:

“But it is still too soon to tell whether those things are truly non-deterministic, or whether – like what happened with Brownian motion – the apparent randomness in those other things will someday be explained by a deterministic theory.”

is a statement that is not state-of-the-art physics. It is true, that not each and every expert scientist and philosopher have accepted the idea of an indeterministic universe … but current physics leaves no choice and most expert scientists have.

The world as a whole is not only (1) unmeasurable and (2) unpredictable, but (3) really indeterministic. If one accepts (1) and (2) but rejects (3), this is a belief decision without consequences. We enter the world of psychology and have to ask “Why do people cling to the idea of a purely causal world?”. Immanuel Kant stated that the idea of causality is not a concept that people learn, but something that exists a priori in our mind, like the ideas of space and time. Evolutionary epistomology explains this by genetic developments during evolution. So maybe it’s an statistical advantage to look for causal relationships and believe in them. So people believed that behind the thunder there was a “god of thunder” in a situation with no better information to create a better causal explanation. So “causal nonsense” was more satisfying than a “it just happens without a cause that we can identify”.

I guess my argument for determinism at all levels goes something like this:

  • Where the world is (1) measurable and (2) predictable, it is pretty clearly deterministic.
  • It is easy to imagine deterministic phenomenon give rise to deterministic phenomenon,…
  • It is hard to imagine that indeterministic phenomenon give rise to deterministic phenomenon.

So, given the inability to measure and predict something itty bitty, I just guess, “Well, it’s probably deterministic down there, even though we can’t see it.” Otherwise, I have to mentally explain for myself, “Why is it deterministic here, but not there?”

It’s not a question of deterministic places and indeterministic places. Indeterministic phenomena can become predictable when they become macroscopic or subject to some statistical simplification.

We know that unpredictable phenomena (single throw of dice) may lead to predictable results (a known random distribution). A stone falls predictable although each single atom of it is in unpredictable vibration and movement.

It seem unscientific to have phenomena that are unpredictable by principle and assume that they are causal nevertheless.

It seem unscientific to have phenomena that are unpredictable by principle and assume that they are causal nevertheless.

I don’t think so; I think that assumption is exactly how science has played out.

Science is built on several assumptions.

Can you clarify “unpredictable by principle” ..?

stepping into this with serious reservations, but…

  • I’ve spent a lot of time discussing various definitions of “the Scientific Method”. The resulting opinions I currently hold include…
    • There is no single, accepted definition, but rather there are several slightly different ones.
    • A couple of common characteristics that are generally accepted are that “Results are repeatable” (predictable Cause and Effect) and that independent, 3rd parties can replicate the “procedure”.
  • I find it interesting that various theoretical “models” can be developed to explain a set of Observations. Then, as our “tools” improve, newer Observations of previously indiscernible discrepanci lead to the formulation of new theories. My favorite example that illustrates this is the shift from the theories that placed the earth at the centre (of the universe) to our current Observations of the universe and the various currently competing theories that try to explain them.
  • Personally, I very much doubt that our knowledge of science is by any means complete. Hence, models of the universe that are based on assumptions such as “the speed of light is a constant” are likely to change as our tools continue to improve and we discover new micro and macro forces that we currently cannot discern.

One of the lines of thought that cause me to write this is one line of thought that that emerges from…

It seem unscientific to have phenomena that are unpredictable by principle and assume that they are causal nevertheless.

There is a considerable body of evidence that our species will always strive to see a “cause and effect” relationship. If this can’t be done using “the Scientific Method”, folks will do it with “superstition or “religion”. Given a choice, I personally prefer to use the current scientific “state of the art” as opposed to accepting dogma on “faith”. But that’s likely just the result of being trained as an engineer. :-)

[duplicated to WhatIsScience]

I think it was assuming that there are cause-and-effect relationships that led to science.

When other people were saying, “Oh, don’t fret about it; It just goes this way some times, and that way other times,” there were people saying, “Well, if we could control this environment just so, I’ll be we can figure it out.” And figure it out they did (assuming determinism.)

Now, with quantum mechanics, scientists discovered: “We will never be able to control this environment just so.”

Does that mean that we should think it’s really random underneath?

I don’t think so!

Honestly, though, I don’t see the relevance of any of this. I’m perfectly at home in a deterministic universe. People still make choices, regardless of whether the process is deterministic or otherwise.

We should probably isolate “RelevanceOfDeterminism?,” “QuantumMechanicsAndDeterminism?,” and so on.

I might be pouring oil into the fire, but I really like this:

  • Hence, models of the universe that are based on assumptions such as “the speed of light is a constant” are likely to change as our tools continue to improve and we discover new micro and macro forces that we currently cannot discern.

The laws of physics might be changing. They may be changing according to some overarching meta-laws or they might do it randomly, or adapting to us, or even in ways completely beyond our understanding. That’s not it. The thing is, if we allow the assumption that they might be changing, we suddenly find ourselves much less knowledgeable than we were before.

For example, we have various estimations of the life of the universe, or the life of our sun, of the life of our planet. These estimates are based on observing what happens now (or appropriate amount of time before now, if light-years distances are involved) and assuming it was always like that. This might be untrue: in particular, the singularity called Big Bang might be just an artifact of how the laws of physics changed. And there is no way to tell, because it affects our perception. It’s like Heron’s paradox – from his point of view the arrow never hits the turtle, simple because the way he reasons creates a singularity at this point, that prevents him to see anything that happens after it.

Now, assumption of determinism is a similar thing. We do it not because it is “right” or “true” in some global, universal way, but because it helps us do something. The theories we build might be completely wrong because of that assumption – but they are better than no theories without it. I believe that one has to be careful to not extend this assumption – this tool – to areas where it is no longer useful. If it doesn’t help us get better results, or if we can get the results without it – why should we carry it as a dead weight?

I wonder if we could agree on the areas where determinism is definitely useful and needed, and ones where it gets in the way?

I agree with Radomir’s statements, regarding relevance. I don’t think it matters much whether determinism is true or not, regarding physics or sociology, since we have the fact of unpredictability, and the fact of the inability to construct tools to make necessary measurements towards predictability.

There are interesting philosophical questions, but the contribution of science to those questions is iffy: The universe gives us no clear case (my present understanding of qm) for or against determinism.

Hans is right: If we find reason to believe c is different elsewhere, then we’ll adjust our models. I share Hans’ skepticism that science is complete. In fact, I may go further: I suspect that we know little more than nothing. My picture of the universe is that the known universe we live in is an itty bitty island of stability. People who study the universe note that the ObservableUniverse? is almost certainly a tiny fraction of the whole of space, and many speculate that even all the connected space may well just be one of many, and there is no way of knowing.

Assumptions are the vehicle towards progress (SelectivelyOpenMinded.)

I think we would agree on the areas where determinism is definitely useful and needed, and ones where it gets in the way.

Maybe we would need a number of pages

“I actually spent about 3 years doing computer simulations of aqueous systems…”

If you’re still interested in that field, you might find the SklogWiki useful. But please keep writing here as well.

David, thank you for the hint. Yes, I’m still interested, although I stopped active academic work in that field.

"Free Will and Scientificalism" by Peter Unger

I like his gently mocking term “Scientificalism”.

… says that “Determinism”, “Fatalism” and “Inevitabilism” all mean the same thing.

Presents this argument:

  • First Premise: If Determinism holds, then, as everything we do is inevitable from long before we existed, nothing we do is anything we choose from available alternatives for our activity.
  • Second Premise: If Determinism doesn’t hold, then, [while some things we do may be inevitable from long before our existence and, as such, it’s never within our power to choose for ourselves] it may be that some aren’t inevitable - but, as regards any of these others, it will be a matter of chance whether we do them or not, and, as nothing of that sort is something we choose to do - nothing we do is anything we choose from available alternatives for our activity.
  • Third Premise: Either Determinism holds or it doesn’t.


  • Conclusion: Nothing we do is anything we choose from available alternatives for our activity.

Then spends the rest of the essay trying to weasel out of the conclusion he doesn’t like.

The most telling sentence to me is: “If we didn’t engage in full choice, at least from time to time, our lives would lack much of, even most of, the significance we commonly suppose our lives to have.”

…then followed by: “Here, I try to uncover what are, in this present day and age, the most forceful apparent threats to our belief that we have full choice, or free will.”

This does seem to be “trying to weasel out of the conclusion he doesn’t like.”

I did this myself, when I was younger. I was confident that Free Will must be right, because nothing would make sense in my personal meaning structure – my ideas about what make life meaningful and so on – unless it were true. I was thinking somewhat along the line of: “if we didn’t engage in full choice, …. our lives would lack … the significance we commonly suppose our lives to have.”

I knew (rightly) that my life had significance, so what remained was to perform the calculation-work to show that free will was something not due to (what the auther of this paper would call) “nonderivative choice.”

Over time though, bit by bit, I gathered the thoughts and observations that would seed the understanding that “derivative choice” (choices that are made as a consequence of history) are meaningful.

Part of this was a personal inquiry into the nature of meaning (“What is meaning? What is significance? Where does it come from?”), part of it was disillusionment with the search for the supernatural, part of it was being shown the work of prior thinkers and the EvolutionarySpirituality crowd, and part of it was just plain-old-fashioned thinking over time.

In particular, there was WhitMeyers? question for me: “What would Free Will mean? Would it mean that, if you relived the incident, with exactly what you knew, you might do something different?

Through that question, I realized that randomness is not a signifier of deeper choice, and I began thinking through the possibilities of “ethereal machinery,” or “how souls make decisions.”

And, hot damn, I just had an idea about what to do when these biiiig long unreworkable pages show up: “StartInTheMiddle?,” which is to say, just make pages for recurring ideas, while just letting the main baseline conversation get longer and longer (and perhaps thrust into archives occasionally.) Start using the new reworked pages titles, so as not to re-introduce the language over and over.

I’ll start with EtherealMachinery.

OK, I finished my account in LionsFlirtationWithFreeWill.

“Conclusions we don’t like” is, to me, the problem. The problem isn’t holding human value fixed – it is valuable. The problem is that we don’t always see the way through other frames. In our case, our entire thinking about determinism and the material world is strongly negative, but it is (I believe) just that thinking that is wrong. We can still have valuable humans, in a determinist universe.

I’m reading a Systems Theory book, and I am suprised to read

“… In the mechanistic world view, there is no place for goal-directedness or purpose. All mechanical processes are determined by their cause, which lies in the past. A goal, on the other hand, is something that determines a process, yet lies in the future.

The thesis that natural processes are determined by their future purpose is called teleology. It is closely associated with vitalism, the belief that life is animated by a vital force outside the material realm. Our mind is not a goalless mechanism; it is constantly planning ahead, solving problems, trying to achieve goals. How can we understand such goal-directedness without recourse to the doctrine of teleology?”

This text seems to be saying “The mechanistic world view cannot explain X … and yet X does occur in some situations … what can we do to patch up things to explain X without giving up our precious mechanistic world view?”

Yes, we have to patch things up, regularly, in science. For example, Newton observed F=Gm1m2/dd, and predicts accelerations of falling objects. But, oh-hoh, here I have a feather, which accelerates quite quickly; Clearly patching is needed. And patching is performed: The concept of wind and air resistance and so on.

This does not mean that science is a practice of pure fantasy, nor that F=Gm1m2/dd is wrong, nor that wind & air resistance are just imaginative patchwork to bamboozle the outsiders. They are all discoveries of the real, by looking deeply into apparant contradiction. This is penetration into paradox.

Natural processes have no teleology, they just follow natural laws. But minds, (and much less than minds,) have evolved in this environment; There is no doubt that protozoa act with the teleological “goal” of reproducing. It makes less sense to speak of goals for protozoa, (even though it is easy to project goals,) but when we get to the dog or the human, we define “goal” as that thing we do in our minds where we quest(ion) for something.

Put another way: Roses blossom, even though atoms do not, even though roses are made from atoms. The higher levels of complexity create different forms that require different language to describe them.

If I were a Christian, I would almost certainly be a determinist (predestination,) and then further, believe in universal salvation. That would not mean that choices made in life wouldn’t need to be thought over carefully.

Define external redirect: RelevanceOfDeterminism BoundedRationality WhitMeyers StartInTheMiddle QuantumMechanicsAndDeterminism ObservableUniverse

EditNearLinks: DocumentMode