When I was growing up, Bertrand Russell’s essay, Why I am not a Christian, created a lot of buzz. It summarizes Russell’s understanding of Christianity and then proceeds to refute, unsuccessfully in my view, its basic tenets. In a similar spirit, I am moved to tackle one of the sacred orthodoxies of our own time, Determinism.
Throughout history, there has been a persistent intuition that all of the events that make up what we call Universe might somehow be ‘pre-determined’. While we lump all such notions under a single label (Determinism), deterministic theories actually come in many flavors, flavors which are often incompatible with one another and which have radically different implications for ontology and cosmology.
Determinisms may be distinguished from one another in four principal ways:
(1) What is the mechanism of determination?
(2) What is the sequence of determination?
(3) What is the domain of determination?
(4) What is the scope of determination?
Specifically, are events in Universe determined by something outside of Universe or is Universe the source of its own determination? Are current events determined by past events, or by future outcomes, or by simultaneously occurring events? Does Determinism apply to absolutely every phenomenon in Universe or are some categories exempt? Are events determined down to the minutest detail or does the determination only apply to the overall pattern of events?
This essay is not intended to provide a complete catalogue of all possible deterministic theories. The initial goal is simply to point out some of the main cross currents and to give a sense of the breadth and variety of viewpoint possible within the ‘Determinist School’.
At one end of one spectrum, for example, we find Determinism that relies on an outside agent or agency for its mechanism. Chief among such agents is ‘God’. According to Divine Determinism, whatever happens happens according to a divine plan. If God is providential, this form of Determinism closely resembles Leibniz’ idea that we live in ‘the best of all possible worlds’.
Divine Determinism, in turn, comes in flavors of its own. One flavor holds that God’s plan encompasses all things, even our own personal choices and actions. This version of Determinism is related to the religious doctrine of Predestination.
Other flavors make room for ‘free will’; God’s power to determine events ends at the tips of our fingers. Does that make us co-creators of the future along with God? Maybe, maybe not!
It has often been said, “Man proposes, God disposes!” Free will may operate outside of God’s plan but that ‘plan’ need not be rigid or static. Like the GPS system in your car, when an actor ‘makes a wrong turn’, God may gently readjust the route. Eventually, we may get where God wants us to go no matter how many wrong turns we make along the way.
This version of Divine Determinism is also a version of Teleological Determinism. According to teleological models, the end is determined, at least down to some level of detail. That end exerts an influence on present events, perhaps just as a lure, perhaps as the sort of ‘guidance system’ suggested above. Alfred North Whitehead, for example, advanced a version of Teleological Determinism in which his ‘Consequent Nature of God’ functions a lure, not a law.
Biblical references to teleological determinism (Je 31:3, Ho 11:4, Jn 12:32) use the Hebrew and Greek words for “to draw”. God draws the future out of the present, he draws Universe to him.
Couldn’t Teleological Determinism be absolute; couldn’t it determine every prior event down to the minutest detail? Certainly one can imagine such a model. However, there would be no way to distinguish that model from efficient Divine Determinism (discussed above) or Mechanical Determinism (discussed below). If every event is perfectly determined, there is no way to tell whether that determination is flowing from future to past or past to future. More to come on this problem!
Nor do Determinisms that rely on an outside agency necessarily have to be theistic. Nietzsche, for example, believed that events, at least in broad outline, were determined by fate, but he did not believe in God. Such Fatalistic Determinism forms the foundation of much ancient Greek literature and it is the ontological basis for the possibility of an oracle (at Delphi or elsewhere). Consider, for example, Sophocles’ dramatic story of Oedipus Rex.
It is difficult to prove or disprove a version of Determinism that relies on an outside agent. That agent typically lies outside the universe of discourse. In that case, there is no way to distinguish a determined sequence of events from an undetermined sequence. Ultimately, Determinisms that rely on an outside agent are more a matter of faith than reason.
That said, Divine Determinism has always been plagued by ‘the problem of evil’. It is very difficult to believe that this is actually “the best of all possible worlds”. I would coax out this argument a little bit further and talk about ‘the problem of noise’? There is so much duplication and interference in the real world that it is almost impossible to believe that it is ‘the most efficient of all possible worlds’. When events actually do occur, they appear not so much to be ‘determined’ as ‘over-determined’. Yet we might expect that a course of events following a pre-determined plan would be perfectly efficient. Otherwise, what do we mean by ‘determinism’?
Interestingly, this criticism need not apply to Teleological Determinisms. We would expect the path to a pre-determined end to be noisy. Think of a maze. There’s only one correct solution, only one way out, and yet people may wander for hours, or years, until they happen upon it. So it may be with our world as it gropes its way toward the Kingdom of Heaven.
Now at the other end of this same spectrum, we find determinisms that have no need for outside agents at all. No God, no fate! Here, Universe is its own agent and events can be explained solely by reference to Universe itself.
Chief among such determinisms is Causal Determinism. Proponents of this view often cite the work of Pierre-Simon Laplace (1749 – 1827), French astronomer and mathematician. Laplace wrote:
“We may regard the present state of the universe as the effect of its past and the cause of its future.”
Based on Newton’s laws of motion, Laplace reasoned that If we could know the position and the momentum of every entity in Universe at any one time (t) then we could calculate the position and momentum of every entity at any subsequent (or prior) time (t’). Such Determinism, derived from mechanics, might be called Mechanical Determinism.
Of course, it would be possible to base a Causal Determinism on something other than Newton’s laws but any such Determinism would need to have a theoretical basis that accounts for the fact that some events are determinative of other events.
According to Sir Thomas Frazer (The Golden Bough) ‘magic’ qualifies as one such alternative. Regarding Magical Determinism, Frazer wrote:
“Whenever sympathetic magic occurs in its pure unadulterated form, it assumes that in nature one event follows another necessarily and invariably without the intervention of any spiritual or personal agency. Thus its fundamental conception is identical with that of modern science…”
Notice how closely Frazer reads like Laplace. Just as the scientist has faith that he can reproduce results in his lab, provided he follows strict protocols, so the magician has faith that he can produce results in the world, provided he adheres to the formulas of his art. Whenever either fails, attention immediately focuses on the conjurer’s fidelity to his craft.
While the systematic basis of Magical Determinism is every bit as detailed and rigorous as that of Mechanical Determinism, most would argue that Magical Determinism does not in fact work. They would contend that the connections posited between events are coincidental or even imaginary.
Surprisingly perhaps, Mechanical Determinism needs to undergo just as rigorous a review. The strength of Laplace’s argument depends entirely on the validity and scope of Newton’s laws. Do these laws apply at all scales and to all phenomena? Is there any carve-out for chance or for consciousness or for free will? Is there an exemption for events that happen on a very small (or a very large) scale? Finally, are these laws descriptive of reality per se or are they just tools useful for calculating certain approximate values?
Quantum Mechanics suggests that it is not be possible to know the position and the momentum of every, or any, entity at every, or any, time (t). Worse, it suggests that entities do not even have well defined positions and momenta. This idea is known as Heisenberg’s Uncertainty Principle. To the extent that entities do not have well defined positions and momenta, Newton’s laws do not apply and causal determinism must fail.
The standard theory of Quantum Mechanics states that Schrödinger’s wave function evolves deterministically but collapses in an indeterminate and unpredictable way that mimics a probability distribution. This probability distribution in turn has suggested another form of Determinism that I’ll call Statistical Determinism.
Proponents of this view acknowledge that quantum events are not determined. However, they contend that since quantum events occur according to a pre-determined probability distribution, the result is ‘close enough’ to pure Determinism for everyday purposes. For this reason, this school has sometimes been called Adequate Determinism. This, I believe, is Stephen Hawking’s position.
But Mechanical Determinism’s problems with contemporary science are not limited to its confrontation with Quantum Mechanics. Ironically, this model, which in its current form owes its genesis and its vitality to modern physics, is actually the version of Determinism most at odds with contemporary science.
We have spoken, for example, of Causal Determinism (when current events are caused by past events) and Teleological Determinism (when current events are shaped by future outcomes), but what about Simultaneous Determinism (when current events are caused by other current events)? Is such a phenomenon even theoretically possible?
Einstein certainly would have said, “No!” In fact, he called the idea that two events outside each other’s light cones could have a determinative relationship, “Spooky action at a distance.” Well said, for it is certainly possible to view ‘spooky’ Magical Determinism as a species of Simultaneous Determinism, rather than as a Causal Determinism.
But in 1964, Irish mathematician John Bell proved Einstein wrong! Bell showed that ‘entangled quanta’, well beyond each other’s light cones, nevertheless retain a connection. If a scientist decides to measure a certain parameter on Quantum A, the result of that measurement immediately determines the value of that parameter on Quantum B, even if the two quanta are light years apart.
Unfortunately, the commitment of the intellectual elite to one form or another of determinism has led to a half century effort to disprove Bell’s Theorem, first theoretically, then experimentally. But just a few years after Bell’s theoretical proof of Simultaneous Determinism, Alan Aspect demonstrated the validity of Bell’s idea in a laboratory. Since then, Bell’s Theorem has been confirmed over and over again in subsequent experiments.
Because quantum entanglement is such a pervasive phenomenon in Universe, the demonstration by Bell and Aspect that Simultaneous Determinism is real suggests that this, not Causal Determinism, could be the dominant mode of determination in Universe.
Here the argument gets tricky. Not all particles in Universe are entangled. Therefore, Simultaneous Determinism, pervasive as it is, cannot exhaustively account for all the phenomena of Universe. Something else must be going on. Accepting Bell’s Theorem does not make one a Determinist.
Mechanical Determinism, on the other hand does purport to account for all the phenomena of Universe; remember Laplace? Therefore, if there are events in Universe that are not determined mechanically (e.g. events determined simultaneously), then Mechanical Determinism fails.
There is, however, a loop hole in Bell’s Theorem. Bell assumes that the scientist measuring the quantum is free to choose which parameter she will measure. If the scientist’s choice is also pre-determined, then Bell’s reasoning does not hold.
A Causal Determinism that posits complete determination down to the level of human thoughts and choices is called Super Determinism. But Super Determinism has problems of its own. All events in all domains and at all scales must be rigorously pre-determined. All future states are entirely implicate in all past states. Therefore nothing actually happens. There are no true events. Universe is like a multi-dimensional screen on which a single all-encompassing image is projected. This is Laplace on steroids.
Such a Universe is called a Block Universe: there is no evolution, no unfolding of events, everything just is. But that’s a problem. If everything just is, then it is just as true to say “nothing is determined by anything” as it is to say “everything is determined something”. In a Block Universe there would be no way to distinguish among Causal Determinism, Teleological Determinism, Simultaneous Determinism and Indeterminacy; after all, everything just is!
In a related problem, Causal Determinism also seems to run afoul of Information Theory. Information can be defined as the difference between potentiality and actuality. According to Causal Determinism, all events at t´ are entirely determined by events at t. Therefore, in any domain where Causal Determinism holds, there is no potentiality. There is no evolution from t to t´: t´ is encoded in t. Therefore, between t and t´ no information manifests. Taken to its logical conclusion, between α and Ω, no information manifests. But if Universe contains no information, is there a universe at all? “It from bit,” wrote renowned physicist John Wheeler. To extend his thought, “No bit, no it!”
The concept of time itself poses a challenge for Causal Determinism. According to Super Determinism, we cannot distinguish the impact of the past on the present from the influence of the future on the present. Time then becomes a superfluous term, and by Occam’s Razor, it must drop out. But is a model of Universe without time compatible with any version of Determinism?
On the other hand, any Mechanical Determinism that stops short of Super Determinism must rest on the classical laws of motion which allow time (and therefore events) to flow bi-directionally. But if time is bi-directional, then given enough time, every event, every chain of events (e.g. an individual’s life) will reverse itself and ultimately annihilate itself.
Later we will see that Causal Determinism seems to require an Aristotelean model of Universe; in that model time is infinite. Therefore, no event can ever be a settled matter of fact; no event can ever manifest information. At best, events will be like virtual particles, endlessly forming and annihilating. Such a deterministic Universe will ultimately be an empty Universe; whatever happens must ‘unhappen’, retroactively. Mechanical Determinism is nihilism!
Big Bang Theory poses yet another problem for Causal Determinism. According to this familiar theory, Universe began as a point-like singularity. The basic assumption of Causal Determinism is that all events are caused by prior events (and not by an outside agency). This begs the question: Was Big Bang an event? If so, then what caused it? If not, doesn’t that make Big Bang an outside agent, like God or Fate?
Ironically, this is not a problem for Divine Determinism which presupposes the agency of a God who is causa sui (his own cause). In his previously mentioned essay, Why I am not a Christian, Bertrand Russell cited the alleged absurdity of an ‘uncaused cause’ as one of his main arguments against Christianity. Of course at that time, there was no such thing as Big Bang Theory. I wonder what Russell would say today, seeing that this same argument can be used to combat Causal Determinism.
Whether one believes in Big Bang or the Hand of God or Bootstrapping or some other model of creation, it seems there must be a mechanism by which a completely novel event can occur. Mechanical Determinism would seem to disallow any such mechanism.
Of course, it’s possible to offer a version of Determinism that allows Universe to spring into being in an entirely undetermined way but requires it to evolve deterministically thereafter. This looks a lot like cheating; but it also begs the question, “Why couldn’t other entirely undermined events occur at various other times in the history of this Universe?” This version of Determinism lays the groundwork for a ‘science of miracles’, the uncaused in-breaking of novelty, and this is certainly not something most causal determinists would embrace.
According to the Big Bang model of cosmogenesis, the initial state of Universe was one of pure energy (no particles) which turned quickly into a Quark-Gluon plasma (‘Quark Soup’). This plasma had a temperature of more than one trillion degrees Kelvin, many orders of magnitude hotter than the center of our Sun (or any other star). It defies credulity to imagine that Universe was so highly structured in the first second following Big Bang that all future events (including this essay) were implicit in that structure.
It may be, then, as we noted above, that Causal Determinism is only compatible with a Universe that follows Aristotle’s idea that the sequence of events has no beginning and no end. Superficially at least, this view seems to contradict the evidence of astronomy and cosmology, but beyond that, it begs the question: Why then is there something rather nothing?
Even if every event in Universe has a prior cause ad infinitum, what causes Universe in the first place? An outside agency? Or is Universe itself causa sui, contradicting the fundamental axiom of Causal Determinism?
Again, Russell is dragged into the discussion. His ‘Theory of Types’ provides an escape hatch. Everything in Universe may require a cause but that does not mean that Universe itself requires a cause. The set is not necessarily one of its own members (though it can be in some circumstances).
Even if you find Russell’s distinction convincing, there’s another problem: does the notion of an infinite causal regress (inside the set) even make sense? One is reminded of Stephen Hawking’s story of a woman who claimed that the World rested on the shell of a tortoise. When Hawking asked the woman what the tortoise rested on, she became indignant: “Another tortoise of course! It’s tortoises all the way down.”
I’m not sure how many causal determinists would be comfortable aligning themselves with that woman in her debate with Stephen Hawking.
On top of all of this, Causal Determinism has perhaps an even greater failing: it lacks any heuristic power.
“A butterfly flaps its wings in Borneo and hail stones fall in Manhattan.” In the universe we actually live in, there is a virtually no connection between these two events but in the chaotic world of Causal Determinism, the two events must be hard wired. An infinitesimal perturbation (wings flapping) on a micro scale can cause a catastrophic event (hail falling) on a macro scale.
But that raises its own set of problems. In an entirely random universe where there is no causal connection between events, the exact same two things are just as likely to occur: butterflies flap, hailstones fall, who can say why?
Therefore, it is impossible to distinguish observationally or experimentally between a totally determined universe and a totally random universe. Determinism was supposed to introduce an ordering principle into our analysis of experience; instead it eliminates order from the picture entirely.
The world of everyday experience is neither random nor determined; it is above all ‘ordered’. Events are connected by a causal web but the efficacy of that causality is significantly dampened by interference, indeterminacy (chance), spacetime extension and, God forbid, “free will” (i.e. the phenomenon of unconditioned choice exhibited in varying degrees by all events at all ontological levels but especially evident in the course of human events).
The idea of ‘compression’ (fractal or holographic) seems to fit the data of experience much better than ‘determination’. The past is immanent in the future but in a ‘compressed’ form. Not all the data is perfectly reproduced and there are holes in the image. Result: there is a broadly consistent evolution of Universe from t to t’ but there is room for minor variations and occasional black swans. The real world is ordered enough to allow planning but novel enough to be interesting.
Whitehead’s cosmology accounts for this co-existence of novelty and order. His ‘ontological principle’ states that every actual entity is the product of the other actual entities it ‘prehends’ (cause) and its own subjective aim (motive). Earlier we mentioned Whitehead in the context of ‘teleological determinism’; how does that fit in? Whitehead’s ‘end’ (Consequent Nature of God) is an actual entity that novel actual entities prehend. Therefore, ‘final cause‘ (teleos) functions along side ‘efficient cause’ in Whitehead’s system.
In my view, these arguments successfully undermine the coherence of any strict version of Causal Determinism. But what about less strict models?
Suppose, for example, that Determinism governs the ‘outside world’ but does not determine our personal choices/actions. According to such a model I (and others?) am an island of freedom in a sea of determination (the physical world). Then I am like a ‘ghost in a machine’ and philosophically we are right back where we were before Gilbert Ryle wrote his seminal work, The Concept of Mind in 1969. Gilbert Ryle took the Ghost out of the machine; Physical Determinism puts it back in.
Suppose, again, that Determinism governs the world on the classical scale but does not govern it on the quantum scale. In my view, any coherent cosmology must start with the assumption that there is one Universe. That universe may behave is very different ways in different domains and at different scales but a final Theory of Everything must account for that variety. We cannot throw up our hands and say, “so far but no farther”.
Then Determinism per se must be an all or nothing proposition. Models that restrict the domain or the scope of determination are either disguised versions of mere ‘causality’ or ‘teleology’ or undisguised ‘dualisms’.
We still have much to discover about how the world works. How exactly are the past and future related? How are we to understand the phenomenon of Presence (the elusive ‘now’)? What mechanism underlies Quantum Entanglement? What is consciousness? Can we unify Quantum Field Theory and Relativity into a single Theory of Everything? Before we can begin serious work on these questions, we must lay to rest the appealingly simple but fatally flawed notion that all events are pre-determined.
This is why I am not a Determinist!