There is an old saying, "Anything with science in the name, isn't." That applies here. Reading it I am reminded of the old linguistics argument that every other field is a branch of linguistics because they have to use language to express their discoveries. Fine, but is studying linguistics an effective way to make progress in learning other subjects?
Have you noticed how seldom physicists bother arguing about how fundamental they are? The value of what they do is so obvious to them that they don't need to defend it.
For a more detailed contrasting view read http://bentilly.blogspot.com/2009/09/what-makes-it-science.h.... FWIW in the list of great science discoveries there I nearly included several CS results (specifically the halting problem, the Church-Turing hypothesis and the identification of NP-complete problems) but none quite made the cut.
Replace linguistics with neuroscience (or more generally biology) and language with the brain (or more generally the body). Does your conclusion change? What does learning mean if we understand the biological processes that underlie it?
I am puzzled why you would think that my conclusion would change.
You've fallen into a classic mistake. Science focuses on tractable problems, not important ones. For instance no matter how wonderful world peace would be, as long as there is no clarity about what lines of research are likely to advance it, it simply can't be science. By comparison human happiness is unlikely to change if we find the last particle predicted by the Standard Model. However we know how to go about finding the Higgs boson, its discovery would confirm our understanding of how mass arises, and understanding it could help us towards reconciling GR and QM.
I fully grant the importance of understanding the human brain. I appreciate the effort that is being spent, and that lots of progress is being made. I've learned a lot by reading about what we have learned. However there is no coherent story yet, and no sense of which lines of research are likely to lead to fundamental improvements in understanding. As a result neuroscientists have less of a shared paradigm than the hard sciences.
Incidentally your comment below about learning is ironic. Researchers into memory learned a lot about how timely reminders improve long-term learning ages ago. They have been trying to shout the news from the rooftops for decades. There are computer programs that help people follow schedules that really improve how you learn. But has it had a practical impact on, say, classroom teaching?
"Science focuses on tractable problems, not important ones."
That doesn't make sense. Science isn't an entity for one thing, it can't focus on anything... for another, do you really think that mechanics, immunology, electricity, and ceramics, to pick a few examples, aren't important?
The reality is that scientists aim to solve tractable problems because they know that they can't solve the whole thing at once. It's part of a simple algorithm called "divide and conquer."
What does learning mean if we understand the biological processes that underlie it?
The same thing it has always meant? It's not like understanding addition grants us some brilliant revelation that changes the "meaning" of multiplication. I fail to understand why the meaning of learning would suddenly change because we understood how it worked.
(Side note: I'm not entirely certain about how you're using the word "meaning" here; are we talking about "meaning" as in "definition of", or "meaning" as in "metaphysical significance or purpose", or something else entirely? Eh, in either of the first two cases at least, my opinion's the same.)
I'd say that you can't have computation without physics. I am willing to bet that it'll eventually be proved that computation is inherently a physical process: information is just one form of energy, and computation is the process of converting energy and entropy into information. Computer Science is then best viewed as a subset of Physics.
If I could study ANYTHING without worrying about getting funding, I'd study computation as a physical phenomenon. I think the answer to P vs NP lies somewhere in the laws of physics; solving NP-Complete problems in polynomial time on the size of the problem input is either possible in our physical universe, or it is not.
Math is a language, not a science. To quote Einstein "Math deals with the relation between concepts. Physics deals with the relation between concepts and reality."
If you dig into that statement, it turns out to be a very nihilistic view. We should be careful not to fall into the trap of demoting math to a mere notation. It isn't. It's the truth.
Green's Theorem, for instance, is no more just part of a mere language for describing electromagnetic theory than the law of gravity is just part of a mere language for describing, say, spacecraft propulsion.
In other words, if math is just a language for physics, physics is just a language for engineering, and we don't have any "sciences" at all ... just languages for higher level disciplines.
Some math seems to be able to describe physics, but all math is implemented in physics (assuming thinking is material and not supernatural, something I do assume), so physics is primary. :)
Physics is nothing but modeling phenomena found in the world. Usually physicists use mathematical functions as their preferred class of models but they are free to use whatever they please. I don't believe there is a 'fundamental' science since it's all just modeling. Maybe the science of modeling is the most fundamental. Computer science, however, does seem to come closer to the 'study of models' then physics.
The answer is obvious (to me): Without yellow there would be no blue! Blue is built on the fundamental assumption that we can model the world using yellow. Yellow is the most green.
To a certain extent this really comes down to semantics. If someone is writing a complex algorithm to describe a physical phenomenon that has been observed in a lab, are they a physicist or a computer scientist?
I tend to disagree with the authors conclusion. I'll concede the theoretical point that anything can be computed, but this doesn't make computer science fundamental. By my definition, Physics is the study of the most fundamental components of the universe and algorithms are an indispensible tool in that study.
It is possible to write an algorithm that perfectly describes what happens when you get hit by a car. This is not the same thing as you actually getting hit by the car. In other words, algorithms are a description of what is happening, not the actual events themselves.
I think at any level level of reasoning where you compare computer science and physics they become pretty indistinguishable. At that point it's all about which words you use to describe it.
Some pretty important ones seem to be: entropy, energy, information, logic, and computability. Those seem to sit squarely between information theory, physics, and math.
I think what this debate really means is that we're not yet going to find some sort of nice hierarchy where every result in chemistry can be considered a practical simplification of something squarely in physics. Instead perhaps both physics and CS are attacking a problem so fundamental that eventually there will be a new body of study that considers both CS and physics both to be a practical simplification of its results.
Have you noticed how seldom physicists bother arguing about how fundamental they are? The value of what they do is so obvious to them that they don't need to defend it.
For a more detailed contrasting view read http://bentilly.blogspot.com/2009/09/what-makes-it-science.h.... FWIW in the list of great science discoveries there I nearly included several CS results (specifically the halting problem, the Church-Turing hypothesis and the identification of NP-complete problems) but none quite made the cut.