As a lifelong agnostic, I never thought that I would find myself arguing for the existence of “souls”, much less claiming that I’ve found proof through Philosophical reasoning. And yet, as I consider the alternative, I find myself unable to accept any other conclusion. If we were to assume that human consciousness arises purely out of physical phenomena, the logical conclusion that it leads to is simply far too absurd.
In order to understand why, let’s first assume the contrarian proposition, and see where that leads us.
Assumption 1: Human consciousness arises from purely physical phenomena.
This is something I myself have long assumed on the basis of Occam’s Razor. There is no proof of anything supernatural, so there is no reason to believe that it exists. Hence, human consciousness must necessarily arise out of purely physical phenomena, such as electromagnetism, quantum mechanics, gravitation, and such. Physical phenomena which give rise to the behavior of our neurons, which in turn give rise to our consciousness. Physical phenomena which we may not fully understand as yet, but are physical phenomena nonetheless, with no “supernatural” “mystical” or “other-worldly” properties.
Assumption 2: Any physical phenomena can be functionally simulated in a computer.
For example, the behavior of neurons, and the interactions of groups of neurons. The simulation may be completely hypothetical with no real-world analogue (eg, a solar system with only the Earth and the Sun). The simulation may not exactly match what actually occurs, or will occur, in reality. However, we can still simulate any physical phenomena at a functional level. The same way that a simulated dice-roll is functionally equivalent to a physical dice-roll, even though they differ in outcome. The same way that NASA is able to functionally simulate the trajectories of its rockets, even though it ignores the precise behavior of each individual atom.
Where physical phenomena is concerned, we can in theory simulate something that is functionally identical, for all relevant purposes, to something that occurs in reality. Philosophers allude to this when they talk about Substrate Independence, but more on this in the following section.
Assumption 3: A computer simulation can be just as conscious as a human being.
This assumption naturally follows from assumptions 1 + 2, and it’s one that a large number of philosophers subscribe to, as a natural conclusion to the principle of Substrate Independence. A computer simulation can be just as capable of experiencing pain and suffering, similar to what we experience. Perhaps Humanity will never be able to accomplish this due to practical limitations, but in theory, it is possible. Even if it’s impossible to precisely simulate a specific person’s brain, we can still mimic all the physical phenomena that give rise to our consciousness, in order to produce a brand new computer-simulated consciousness.
Assumption 4: It is immoral to subject conscious beings to torture and suffering, without justifiable cause.
The race, gender, and physical makeup of the conscious being is irrelevant. Even if we assume that some conscious beings possess greater moral worth than others, depending on their level of consciousness and past/future actions, it is still morally reprehensible for someone to subject an innocent helpful conscious being to torture and suffering, purely for personal amusement. If someone is engaging in such behavior, it is appropriate to condemn and imprison them, just like how we condemn rapists, murderers and terrorists.
Assumption 5: The fact that a computer simulation can be conscious, is independent of the CPU speed of the computer.
A computer that is running on a very slow CPU, is just as capable of generating consciousness as a computer running on the latest and greatest CPU. The former might take exponentially longer to achieve the same results, but it is just as capable of running any simulation that isn’t dependent on external run-time-inputs and interactions. Any and all interactions that are required for consciousness can be baked into the simulation itself, in order to produce an executable that runs identically on extremely slow CPUs.
Assumption 6: Anything a computer does can be functionally replicated by a human being writing 1s and 0s in a massive ledger.
As a software engineer who’s helped design Intel processors in the past, I can personally vouch for this “assumption” myself. Every single computer software can be broken down into “assembly instructions”, each of which is so simple and primitive that it can be executed by a human being writing 1s and 0s on a ledger. The precise timing behavior of computers-vs-humans might differ, but the functional outcome is identical. Anything a computer can do, so too can a human being with a lot of time and space.
Conclusion: A person writing 1s and 0s in a massive book, in a “wrong sequence” over an extended period of time, should be found guilty of atrocities such as torture and murder.
Such a person should be condemned to the same extent that we condemn rapists, murderers and terrorists. Not for planning a crime, but for actually committing one. Such a person should be imprisoned for life, and in extreme cases, subjected to the same contempt that we show towards people like Hitler and Stalin.
Clearly this conclusion should be rejected as being absurd. A man writing 0s and 1s in a private ledger should never be accused of committing atrocities on par with a murderer’s. A man writing 1s and 0s in a private journal should never be sentenced to life imprisonment for having committed moral atrocities. The idea that such a person should be found guilty of committing a crime, and should be subjected to punishments because he wrote the “wrong” sequence of 1s and 0s in a private journal, is simply absurd and should be rejected.
However, this conclusion flows naturally out of the earlier set of assumptions. The only way to reject this conclusion, is to reject one of the underlying assumptions. Out of all the assumptions presented, the weakest assumption is the very first one. All the others are supposed by solid logical reasoning, whereas the first assumption is merely supported by Occam’s Razor and the absence of evidence.
If we are to reject one of the assumptions presented, the only reasonable option is to accept that physical phenomena, such as the firing of neurons in our brains, is a necessary but not sufficient condition for consciousness. That in order for consciousness to occur, other “supernatural ingredients” are required as well. Supernatural ingredients that human beings can never simulate. Supernatural ingredients that for millennia, people around the world have referred to as as our “soul”.
If we are to reject the idea of people being condemned and punished for writing the “wrong” sequence of 1s and 0s on a private notebook, the only way out is to acknowledge the Human Soul.
Outlook Zen 
Brian, A computer has mass, therefore it does generate gravity (a very, very, very small amount, but there nonetheless); it’s existence creates gravity. I tend to accept your statement that “….it does not follow that all physical effects can arise”, but I’m not sure the gravity example is a good one. Perhaps something to do with physical or emotional pain would be better….? Just my 2 cents. Thanks, Bob
This is a fun thought experiment. First, it sounds like we’re assuming the malicious writer is aware of the sequence being a bad one (and therefore worthy of condemnation).
It seems there’s a big leap made from defining instructions to executing them. I suspect that this is where the argument breaks down. We don’t know exactly what it takes to bring consciousness into existence, but perhaps an instruction set itself is insufficient, and consciousness depends on the manner of execution of the instruction set (a certain type of processor).
Good points Brian. Yes, the assumption is indeed that the malicious writer is writing 1s and 0s in a ledger, with full knowledge of the type of consciousness and experience he is simulating.
And yes, we’re also assuming that the writer isn’t just defining the instructions (ie, writing software) but is also running the software (by simulating on paper how the register/memory contents change as the software is executed).
You bring up an interesting point about the material makeup of the computer itself. I had implicitly assumed this, but perhaps I should have listed it as an assumption. The way I see it, whether a computer can produce consciousness, is independent of the physical makeup of the computer. A computer made of vacuum tubes, or silicon, or carbon, or even dominos (haha, yes, they exist) are all functionally equivalent to one another, from the software layer’s perspective, and the consciousness simulation is really running at the software level. There’s nothing special about silicone cpus vs vacuum tube cpus, in terms of running a software application that would simulate consciousness. Hence why I put the “guy writing 0s and 1s on a book” cpu in the same category as well.
Ok, I’d like to take another crack at this:
A-1: Human consciousness arises from purely physical phenomena.
A-2: Any physical phenomena can be functionally simulated in a computer.
Even if all physical phenomena can be functionally simulated by a computer, it does not follow that all physical effects can arise. Take gravity, for instance. We simulate gravity in computers, but cannot (afawk) even in theory run an instruction set such that gravity is actually generated by the simulation. By the same lines, we can simulate consciousness, but it does not follow that consciousness is actually generated. We must distinguish between simulation of something and the generation of an actual physical phenomena.
So, to summarize, my objection is that the argument seems to hinge on an equivocation between simulation and generation of physical phenomena.
Really cool thought though!