← other writings

“The Big Picture” by Sean Carroll

Part One: Cosmos

Carroll begins the book by proposing a philosophical framework he calls poetic naturalism.


  1. There is only one world, the natural world.
  2. The world evolves according to unbroken patterns, the laws of nature.
  3. The only reliable way of learning about the world is by observing it.

The poetic aspect comes to the fore when we start talking about the world:

  1. There are many ways of talking about the world.
  2. All good ways of talking must be consistent with one another and with the world.
  3. Our purposes in the moment determine the best way of talking.

The poetic component is a recognition that many human concepts, such as person, cause, choice, reason, justice, morality, and meaning, while not fundamental are still real and are worthy of our consideration. There is also a warning that we should not apply these higher-level concepts in inappropriate ways.

Carrol’s discussion of determinism and free-will is a good example of how poetic naturalism combines deep ontological truths with everyday language:

[Laplace] realized that there was a simple answer to the question “What determines what will happen next?” And the answer is “The state of the universe right now.” …

We know that the quantum state of a system, left alone, evolves in a perfectly deterministic fashion, free even of the rare but annoying examples of non-determinism that we can find in classical mechanics. But when we observe a system, it seems to behave randomly, rather than deterministically. …

The momentary or Laplacian nature of physical evolution doesn’t have much relevance for the choices we face in our everyday lives. For poetic naturalism, the situation is clear. There is one way of talking about the universe that describes it as elementary particles or quantum states, in which Laplace holds sway and what happens next depends only on the state of the system right now. There is also another way of talking about it, where we zoom out a bit and introduce categories like “people” and “choices.” Unlike our best theory of planets or pendulums, our best theories of human behavior are not deterministic. We don’t know any way to predict what a person will do based on what we can readily observe about their current state. Whether we think of human behavior as determined depends on what we know. (chpt. 4)

Thus, if you ask a poetic naturalist “do humans have free-will,” they would reply “fundamentally, no, but the words free-will, choice, and person are useful everyday descriptions of reality.”

Carroll’s presentation of determinism in Chapter 4 is oversimplified, although he does discuss the topic in more depth later in the book. I believe his claim that the world is deterministic is only true within one of several interpretations of quantum mechanics. There is no scientific consensus as to whether the universe is deterministic or not. A great deal of debate still surrounds the so-called “collapse of the wave function” and the “measurement problem.”

The Copenhagen interpretation of quantum mechanics says that physical systems do not generally have definite properties prior to being “measured,” and only the act of measuring them causes them to taken on certain states. The Copenhagen interpretation has several problems.

The concept of the “system” and the “surroundings” are useful for modeling reality and writing equations for small parts of it, but they are not fundamental to our ontology. Thus, what does it mean to say the system is deterministic except when you observe it? We want to know whether the entire universe is deterministic or not. If there are parts of the universe that are not deterministic, even only occasionally when we “observer” them, then the entire universe is not deterministic. Determinism is an all-or-nothing phenomena.

Could we consider the entire universe as “the system,” such that there is no outside observer who could introduce randomness by “collapsing the wave function”? If this is possible then why does the act of dividing the universe into system and surroundings introduce randomness? It would appear that it can not.

One alternative to the Copenhagen interpretation is put forward by proponents of quantum decoherence. In this way of thinking, the “collapse of the wave function” and the “randomness” involved are simply useful ways of speaking about our incomplete knowledge of the environment’s very complicated wave function. Thus, there is no randomness involved. Quantum decoherence avoids making “measurements” and the “system” part of our ontology, which I find convincing. Quantum decoherence results in a deterministic universe.

Other alternatives to the Copenhagen interpretation exist, such as the objective-collapse theories, which may or may not result in a deterministic universe.

Thus, Carroll conveniently presents his minority-view as if it were the consensus of the scientific community. It wouldn’t surprise me if he does this elsewhere in his book.

I enjoyed Carroll’s discussion about the arrow of time:

For every way that a system can evolve forward in time in accordance with the laws of physics, there is another allowed evolution that is just “running the system backward in time.” There is nothing in the underlying laws that says things can evolve in one direction in time but not the other. Physical motions, to the best of our understanding, are reversible. Both directions of time are on an equal footing.

If there is a purported symmetry between past and future, why do so many everyday processes occur only forward and never backwards?

Boltzmann and his colleagues argued that we could understand entropy as a feature of how atoms are arranged inside different systems. Rather than thinking of heat and entropy as distinct kinds of things, obeying their own laws of nature, we can think of them as properties of systems made of atoms, and derive those rules from the Newtonian mechanics that applies to everything in the universe. Heat and entropy, in other words, are convenient ways of talking about atoms.

Given that, Boltzmann suggested that we could identify the entropy of a system with the [logarithm of the] number of different states that would be macroscopically indistinguishable from the state it is actually in. A low-entropy configuration is one where relatively few states would look that way, while a high-entropy one corresponds to many possible states. There are many ways to arrange molecules of cream and coffee so that they look all mixed together; there are far fewer arrangements where all of the cream is on the top and all of the coffee on the bottom.

With Boltzmann’s definition in hand, it makes complete sense that entropy tends to increase over time. The reason is simple: there are far more states with high entropy than states with low entropy. If you start in a low-entropy configuration and simply evolve in almost any direction, your entropy is extraordinary likely to increase. When the entropy of a system is as high as it can get, we say that the system is in equilibrium. In equilibrium, time has no arrow.

What Boltzmann successfully explained is why, given the entropy of the universe today, it’s very likely to be higher-entropy tomorrow. The problem is that, because the underlying rules of Newtonian mechanics don’t distinguish between past and future, precisely the same analysis should predict that the entropy was higher yesterday, as well. Nobody thinks the entropy was higher in the past, so we have to add something to our picture.

The thing we need to add is an assumption about the initial condition of the observable universe, namely, that it was in a very low-entropy state.

What we know is that this initially low entropy is responsible for the “thermodynamic” arrow of time, the one that says entropy was lower towards the past and higher toward the future. Amazingly, it seems that this property of entropy is responsible for all of the differences between past and future that we know about. Memory, aging, cause and effect—all can be traced to the second law of thermodynamics and in particular to the fact that entropy used to be low in the past. (chpt. 7)

It is interesting to consider that “cause and effect” are not fundamental entities—they are just useful ways of speaking about reality, like entropy and heat are. Carroll mentions a useful way of defining a “cause”:

We can make sense of statements like “A causes B” by thinking of different possible worlds: in particular, worlds that were essentially the same except for whether the event A actually occurred. Then, if we see that B occurs in all the worlds where A occurred, and B does not occur when A does not occur, it’s safe to say “A causes B.” (chpt. 8)

Part Two: Understanding

Part two of The Big Picture is a summary of Carroll’s epistemological system, beginning with Bayesian reasoning.

Bayes’s theory can be understood as a generalization of contraposition. In propositional logic, where propositions are either true or false, we have:

(~H -> ~E) -> (E -> H)

Bayes’s theorem generalizes contraposition for the situation where propositions have a “probability” or “degree of belief”. Here is Bayes’s theorem:

P(H|E) = P(H)*P(E|H)/P(E)
       = P(H)*P(E|H)/[P(H)*P(E|H) + P(~H)*P(E|~H)]

Bayesian reasoning, while interesting, must be a smaller part of one’s epistemological system then Carroll presents it to be. There are many short-comings with Bayesian reasoning that Carroll does not consider, but which I believe are important.

One of the most intriguing and useful ideas presented is:

Evidence that favors one alternative automatically disfavors others. … The converse is also true: if some observation would have favored one theory, but we obtained the opposite of that observation, that result necessarily decreases our credence for the theory.

This idea is useful, because, if we are considering a hypothesis, H, we can imagine evidence, E, which would make this hypothesis more likely to be true. If this evidence doesn’t exist, we can take this as evidence that the hypothesis is not true. Performing thought experiments like this is not particularly difficult. E.g., imagine evidence that would support the hypothesis that “An all-powerful loving god exists?” Now, consider which of this evidence does exist, and what evidence doesn’t?

I like his analogy of “planets of belief.” This analogy helps one visualize the recursive nature of belief and the unavoidable interconnection between epistemology and metaphysics.

I also like (and agree with) Carroll’s distinction between faith and reason:

You will sometimes hear the claim that even science is based on a kind of “faith,” for example, in the reliability of our experimental data or in the existence of unbreakable physical laws. That is wrong. As part of the practice of science, we certainly make assumptions—our sense data is giving us roughly reliable information about the world, simple explanations are preferable to complex ones, we are not brains in vats, and so forth. But we don’t have “faith” in those assumptions; they are components of our planets of belief, but they are always subject to revision and improvement and even, if necessary, outright rejection. By its nature, science needs to be completely open to the actual operation of the world, and that means that we stand ready to discard any idea that is no longer useful, no matter how cherished and central it may once have seemed. (chpt. 15)

Part Three: Essence

Carroll claims, in Chapter 23, that:

The laws of physics underlying everyday life are completely known.

The logic behind our audacious claim is simple:

  1. Everything we know says that quantum field theory is the correct framework for describing physics underlying everyday life.
  2. The rules of quantum field theory imply that there can’t be any new particles, forces, or interactions that could be relevant to our everyday lives. We’ve found them all.

Note he is not claiming that we know all there is to know about physics, just that we know everything there is to know that could affect our everyday life.

Science will not be able to explain why the universe exists.

Did the universe always exist, or did it have a beginning? Science also can not tell us about this, since we can not see past the big bang. But poetic naturalism does have some useful things to say about this topic. In particular, it provides a refutation of the arguments from first cause:

Talking about “causes” is not the right vocabulary to use when thinking about how the universe works at a deep level. We need to be asking ourselves not whether the universe had a cause but whether having a first moment in time is compatible with the laws of nature.

Popping into existence and “having a beginning” are not the same thing. Popping into existence implies there was a priori moment where there was nothing, but we are talking the entire universe—thus there could not have been a prior moment, rather, the first moment is the one without a prior moment.

Everyday objects don’t pop into existence because doing so would break the laws of physics. There is nothing about the laws of physics that precludes it having a first moment.

Some people argue “sure, you can say that physics doesn’t preclude the universe popping into existence, but there are metaphysical arguments which are more general than physics which preclude this!” But what are your arguments for the metaphysical principle that things (including the universe) can’t pop into existence? Isn’t this begging the question? While we don’t know why it did, we also don’t have good reasons to think that it couldn’t have.

Part Four: Complexity

Theodosius Dobzhansky: “Nothing in biology makes sense except in the light of evolution.”

Entropy, statistically speaking, always increases for a closed system. Thus, the entropy of the universe is increasing.

Complexity, in some systems with certain properties, can increase and then decrease over time.

Mixing milk and coffee together is an example of a system whose complexity increases and then decreases; at the start, the coffee and milk are well separated. There is low entropy and low complexity. Then, as the milk and coffee are mixed, the entropy increases, as does the complexity—there are beautiful and subtle swirls of milk with interesting structures. Finally, once the coffee is mixed enough, the milk and coffee are completely mixed and the entropy has increased further, yet the complexity has decreased.

Complexity is not as well defined as entropy is, although both are not fundamental concepts.

Perhaps humans are like the swirls of the coffee in the coffee cup of the universe?

The fine-tuning argument for the existence of a god is:

Carroll believes this is “probably the most respectable argument in favor of theism … but it’s still not a very good argument.” One of the reasons that Carroll doesn’t like it is because it uses “old evidence,” that life exists, to make the argument. And making a prediction and then gathering “new evidence” would be more convincing. I think this particular counter argument is weak. We shouldn’t expect to apply the same standards to deep questions about the existence of humans or a god as we do to typical science experiments.

There are a few issues with the fine-tuning argument which I think are more reasonable. One is that it may not be easy to determine whether a particular set of physical constants would result in life or not.

Another counter-argument is that some physical theories predict that there are many “universes” each with different physical constants—the so-called “multiverse.” If there is a multiverse, then we may simply be in the particular “universe” that has life-supporting physical constants.

Carroll makes another interesting counter-argument:

If the laws of physics were chosen so that life could exist, we would expect that each of the various features of those laws would play some important role in the unfolding of life. What we see, on the contrary, is something of a mess. All living beings are made out of the lightest generation of fermions—the electron and the up and down quarks, with occasional appearances from electron neutrinos. But there are two heavier families of particles, which don’t play any part in life. Why did God make the top and bottom quarks, for example, and why do they have the masses they do? Under naturalism we would expect a variety of particles, some of which are important to life and some of which are not. That’s exactly what we do observer. (chpt. 36)

Finally, if the universe was fine-tuned so that we could exist, why do we “live in a galaxy with more than 100 billion stars, in a universe with more than 100 billion galaxies”? The universe could have easily been much smaller; why make it so enormous if we are the point of it all?

Part Five: Thinking

Carroll begins his discussion of conciousness with an provocative evolutionary hypothesis: the development of conciousness began when creatures move from the sea to land. Vision is very limited underwater, thus, sea creatures don’t have much time to plan their actions and instead must react quickly to events. Once on land, where creatures can see further, there is evolutionary pressure to develop brains that can plan their actions.

There are some intelligent sea-creatures which seem to break this theory. In particular, octopuses.

If there is any one aspect of reality that causes people to doubt a purely physical and naturalist conception of the world, it’s the existence of conciousness. (chpt. 37)

This statement is true for me. But there are many reasons to believe that conciousness is physical:

Some philosophers speak about “the hard problem” of conciousness—explaining the existence of qualia, the subjective character of experience. Other philosophers view this as not being a problem, but rather just “conceptual confusion.”

The Zombie Problem: consider a human, now consider an identical copy of the human who does not experience qualia—a zombie. If zombies can exist then naturalism must be incorrect. Some philosophers argue that zombies are conceivable, thus they must be able to exist. But is this true? We can conceive of water without an oxygen atom, but this can not exist.

Part Six: Caring

Meaning, morality, and purpose are not fundamental parts of the universe but are emergent phenomena. This doesn’t mean they are not real. “Water doesn’t stop being wet when you learn it’s a compound of hydrogen and oxygen.”

Humans must be the starting place for any morality, purpose, or meaning.

Hume: You can not derive an “ought” from an “is.”

Carroll’s ten considerations:

  1. Life isn’t forever
  2. Desire is built into life
  3. What matters is what matters to people
  4. We can always do better
  5. It pays to listen
  6. There is no natural way to be
  7. It takes all kinds [of people]
  8. The universe is in our hands
  9. We can do better than happiness
  10. Reality guides us

I think this is an odd set of considerations to point out.