Theoretical Physicist David Deutsch on the limitations of science

One of the most important limits of science is that it isn't philosophy. [...] Scientism is the purported application of science to problems that are really philosophical. Such as the question of whether animals really feel pain or not. We can tell whether animal's nerves are excited and whether their brains react to that. But whether an animal feels pain in the sense that humans do, or merely react in the sense that a robot does, that is ultimately a matter of philosophy, because it's only philosophy that can determine the criterion for science to use when trying to distinguish between those cases. So that's a limit of science--trying to reach into philosophy is scientism. [...] I would say to [those who claim that the only good explanations are scientific explanations] that that theory is not part of science, and therfore it rules itself out.

Taken from an interview with Robert Lawrence Kuhn.

How to (really) think about metaphysics

Today I was visiting some of the blogs I frequent and I stopped by the blog Atheism and the City, authored by a frequent commenter at my blog who goes by the name “The Thinker”.  He has just posted a review of Chapter 3 of Edward Feser’s book The Last Superstition. There is much to quibble with in Thinker’s post, but I’ll save an attempted refutation of his arguments for another day. My main focus here is regarding his comments on metaphysics and scientism.  I’ve gone on ad nauseum on this blog about the importance of metaphysics, and the falsity and self-refutation of scientism (see here),  but Thinker presents a different spin on this issue—a view he calls “weak scientism”—and therefore I felt the need to point out its falsity, thereby tossing it in the trash-bin with the other failed metaphysical frameworks.

Thinker begins to articulate his thoughts on weak scientism:

 I hold to what is sometimes called "weak scientism." Unlike strong scientism, which says that "the characteristic inductive methods of the natural sciences are the only source of genuine factual knowledge and, in particular, that they alone can yield true knowledge about man and society," weak scientism says that the natural sciences are given a privileged status over metaphysics and logic and all other methods of derived knowledge, but it stops short of saying that science and logic are the "only" ways of yielding true knowledge.

It’s hard to say exactly what Thinker means by the natural sciences entertaining a “privileged status” over metaphysics and logic. Perhaps he means that the natural sciences are more genuine and pure methods for attaining knowledge of reality. It is certainly true that the natural sciences have illumined the majority of beliefs and knowledge that we have about how the natural world behaves. But does this give one warrant to claim that science therefore entertains a privileged status for gaining knowledge about reality, over and above metaphysics? I don’t see how. For, as I’ve argued before, science tells us how the world behaves, but it does not, and cannot, tell us why—only metaphysics can hope to do this. Science might be able to tell us about the four forces of nature, for example, but it cannot tell us why those forces obtain at all as opposed to others. So I fail to see how science should be given a privileged status regarding methods of inquiry above methods like metaphysics.

The Thinker continues:

Furthermore, I apply this privileged status of science mostly when entertaining questions regarding ontology, such as the fundamental nature of reality—for which science is our most reliable epistemology, contrary to what Feser says.

So now Thinker says that science seems to be our most reliable guide for answering questions regarding the fundamental nature of reality. Again, this presents a couple problems. First, science by itself does not have the tools to comment on the fundamental nature of reality. This might seem like an arrogant statement against science, but it’s the truth. Scientific law only describes what already exists and how it behaves. But simply describing what already exists is purely abstract, in that it doesn’t actually enlighten us to the intrinsic nature of what it’s describing. So science simply doesn’t cut deep enough to penetrate the fundamental nature of reality.

Second, the natural sciences necessarily describe only the natural world. Therefore, if there were other aspects of reality, science could not enlighten us one way or another regarding their behavior. In fact, even if there exists no supernatural aspects of reality, science could likewise not comment on its non-behavior. The point is that science is completely silent on the question of what comprises the set of reality and being. To construct my point in another fashion, the description of a set of elements is not sufficient to conclude that only the set exists. Something else would be needed to ground this proposition, and it couldn’t be a mere description of a set’s behavior.

Let’s move on:

No logician could ever derive the physics of quantum mechanics from the laws of logic, or from metaphysics. Only empirical evidence could enlighten us to such phenomena, and the universe is ultimately quantum mechanical in nature.

Thinker continues to make the similar mistakes here. First, I agree that science, and only science, enlightens us to the physics of quantum mechanics and general relativity etc. But nobody ever said that metaphysics was the method of inquiry that should have done this. You see, science uncovers the physical contingencies of the universe that can be repeated and predicted using abstract mathematical equations—equations that could have been different. But, metaphysics enlightens us to the necessities of the universe, and anything that could or would exist. It tells us the ontology of causality, identity, first principles, time and free will etc. So, to once again engage in chest-thumping on behalf of science because of its discoveries of the physical contingencies of the world is simply misplaced here, since metaphysics wasn’t attempting to do this in the first place. It’s like a basketball player praising his dunking ability, while playing soccer. Metaphysics tells us the fundamental nature of reality, and science tells us how this reality happens to behave. The latter is not equivalent to the former.

Second, Thinker shoots himself in the foot here with his last statement about the universe being quantum mechanical in nature. How does he know the universe is entirely quantum mechanical in nature? Well quantum mechanics cannot tell him this, since, to reiterate, QM is only a description of sub-atomically existing matter. How does he know this description constitutes the fundamental nature of reality? Even if he has an answer to this question, it will not be given to him by QM. He will have to have some other fulcrum to lay this proposition upon. And this goes for any set of natural sciences you want to erect as the end-all be-all description of reality. A description of a set simply will not be sufficient for concluding that only what’s in the set is what exists, or that the set does not have an aspect of its nature not captured by the original description. The Thinker’s attempt, then, to ground weak scientism seems to have failed.

Embodied Realism Part I: Correspondence Theory of Truth

After much prodding from Mike D over at The A-unicornist—which he has now stopped writing for, unfortunately—to read his Holy Bible on embodied cognition, namely Philosophy in the Flesh, I have begun to do just that. And rarely does a page goes by where I don’t find myself scribbling in the margins regarding the things I find fallacious. With that in mind I felt the need to write up a series of posts regarding the parts I vehemently disagree with as I read along. So stay tuned for many posts to come regarding Embodied Realism.

___________________________________________

Embodied Realism (or Embodied Cognition), as articulated by Lakoff and Johnson, is predicated on a few themes: 1) The mind is embodied, 2) reason is mostly unconscious and 3) is structured by neural connections that conflate conceptual domains which lead to metaphor—which means that most of our concepts and cognition are metaphorical. Embodied Realism does not simply say that we need a body to reason, but, rather, that our reason itself is shaped and structured in unconscious ways by our embodiment.

My first bone to pick with Embodied Realism is its jettison of the correspondence theory of truth. The correspondence theory of truth basically says that a statement—what is called a truth-bearer—is true if it corresponds to an actual state of affairs—what is called a truth-maker--or obtains, in reality. As an example, the statement “There is a computer in front of me” is true if, in reality, there is actually a computer in front of me, and false if there is not.

It is this very theory that is called into question because of the supposed embodiment of cognition promulgated by embodied realism. To argue against the correspondence theory the authors attempt to demonstrate that concepts are embodied at the neural level, the phenomenological level, and the cognitive unconscious level. Let me illuminate these as best as I can, so we can articulate where the asserted difficulty arises.

The neural level basically regards the physical circuitry that characterizes, grounds, and structures all cognition and conception. The phenomenological level is the level that we are consciously aware of, and the level where we experience all of our qualia. And the cognitive unconscious level is the level of cognitive operation that evades our conscious subjective experience; that is, it is what structures our conscious experience, but is something that we have no internal access to—hence the term unconscious.

So, how does positing these different levels of embodiment call into question the theory of correspondence? Well, Lakoff and Johnson explain that “Truth claims at one level may be inconsistent with those at another.” In order to illustrate this problem, the authors discuss the embodiment of color concepts.  For example, consider the statement “Grass is green.” Surely we experience greenness inhering in grass, and thus this statement would seem to be true based on the correspondence theory of truth. However, the authors argue, this is only true at the phenomenological level of embodiment. On the neural level of embodiment “greenness” does not “inhere” in objects or things; rather, it is created by reflected light, our retinas, and our neural circuitry etc. The authors further articulate the alleged difficulty:

 At the neural level, green is a multiplace interactional property, while at the phenomenological level, green is a one-place predicate characterizing a property that inheres in an object. Here is the dilemma: A scientific truth claim based on knowledge about the neural level is contradicting a truth claim at the phenomenological level. The dilemma arises because the philosophical theory of truth as correspondence does not distinguish such levels and assumes that all truths can be stated at once from a neutral perspective.

 It is here where I believe the authors are mistaken, on multiple levels (see what I did there?).

Effects, formal and virtual

First, even though a Scholastic, like myself, would utter a proposition like “Grass is green” and indeed say that such a proposition is true, what they mean by such a statement is utterly foreign to what the authors attribute to any type of metaphysical realism—namely, that color is a single thing or property that inheres in substances. Thomist Peter Coffey illustrates:

 When, for instance, the normal perceiver apprehends snow as white, and spontaneously asserts that “snow is white,” he means not that the color-quality in question is wholly independent of the nature, structure, and conditions of his visual sense organs for its specific character as present to his consciousness.

 The point here is that when us Scholastics say “Grass is green” we don’t mean that there is a single property of greenness that inheres objectively in objects or things—that is, we do not say that color exists formally or actually. Rather, we would say that color exists virtually, or potentially. Therefore a Scholastic would agree with the authors when they say the following: “Colors are not objective; there is in the grass or the sky no greenness or blueness independent of retinas, color cones, neural circuitry, and brains.” Thus, to say that color exists virtually is to say that when all the necessary prerequisites are conjoined—the reflective properties of objects, our bodies and brains etc.—then and only then can we have the actuality of color. Upon taking this into account, proposing that color does not exist would only be true if effects could only obtain formally, which is something the Scholastic would not concede. For, to reiterate, effects can exist formally, but they can also exist virtually. And therefore to claim that only formal effects can obtain is to beg the question against the metaphysical realist.

Now, the significance here is that the statement “Grass is green” is still in fact objectively true, as long as what we mean by this statement is that grass is virtually, and not formally, green. However, the authors would still say that this cannot be an objective truth, because the property of greenness does not inhere objectively in the world. But to do so would be to conflate objective truth with truth obtaining objectively. That is to say, since color requires the existence of human embodiment for it to obtain, then color does obtain objectively, by definition. However, this doesn’t mean that the truth “Grass is green” is therefore not objective. Remember that all that’s required for the correspondence theory is for a truth-bearer to correspond to a truth-maker—again, either formally or virtually. And since grass is in fact (virtually) green, then the statement “Grass is green” is objectively true.

Levels of embodiment

Unfortunately (for the reader still awake at this point) this has all only constituted the first objection to the claims of Lakoff and Johnson. For the authors would still fire back that all my musings above assume a neutral perspective from which to promulgate my supposed truth. That is to say, my arguments above about the reality of color are all predicated on only one level of embodiment (or are they?) and to do so is to erect one level of embodiment as superior over another, thereby doing an injustice to the other levels. The authors articulate:

 Both the phenomenology-first and science-first strategies are inadequate in one way or other. If we take the phenomenology-first strategy, we miss what we know scientifically is true about color. We get the scientific metaphysics of color wrong. Our “truth conditions” do not reflect what we know to be true. If we take the science-first strategy, we do violence to the normal meaning of the word and to what ordinary people mean by “truth.”

 This is all to say that by claiming color exists virtually is to do damage to the phenomenological level of embodiment upon which color seems to exist formally. But my retort is this: so what? On the phenomenological view we can only talk about how we perceive sense qualities, and just because we perceive something does not mean it is there, or that it’s there in the fashion we perceive it to be, or that it inheres in the world. A perfect example to knock down the authors’ claims here is the experience of hallucinations. At the phenomenological level a hallucination is very real, in that we experience qualia with regards to said hallucination. But it is only at the neural level that we know that the hallucination is not actually real. And the crucial point is that everybody, including the authors, would take a "science-first strategy" here and claim that when somebody hallucinates a dead relative (for instance), the statement “my dead relative appeared to me” is unequivocally false. And the really unreasonable thing to do here would be to say that the existence of the dead relative is true at the phenomenological level, but just not at the neural level! Hence, it seems that, contrary to the authors, we must, and do, utilize certain levels of embodiment over and above others in different circumstances. Thus, there is no dilemma when one level of embodiment contradicts another.

So, the authors are simply mistaken to say that the correspondence theory is false for not distinguishing the different levels of embodiment. For the levels of embodiment do not at all call truth, as classically conceived, into question. In fact, we’ll see next post that this conception of truth is unavoidable, and that it is the truth of Embodied Realism that runs into various difficulties.

Science and explanation

Science works by uncovering certain conditional statements such as “If a occurs under certain ideal conditions, then b will occur” or “if a is present under certain ideal conditions, then b will be present.” For instance, every time water accumulates energy in the form of heat, once the water reaches a certain temperature (approximately 100 degrees Celsius) it will become gaseous. Thus, we can form the conditional statement that “if water reaches the temperature of 100 degrees Celsius, then it will become steam.”

Now this is generally how laws of nature come to be formulated. We run an experiment a few times under ideal conditions, and when a certain state of affairs reliably obtains, then we assume that this happens across the board and formulate a law of nature. This means that laws of nature are really just descriptions of the way nature tends to behave.

But what, then, is the explanatory use of laws of nature? For example, if one asks, per our example above, why when water reaches a certain temperature it subsequently tends to become gaseous, it won’t do any good to revert back to the law that states this very thing. For that would simply be to say that a tends to be followed by b because a law of nature says that b’s are reliably produced by a’s. Such an answer would thus amount to a tautology, and be explanatory vacuous.

Now, one could instead say that the boiling point of water is not explained by the law of nature, but rather that the boiling point of water is explained by molecular behavior, kinetic energy, and molecular forces. The problem, though, is that here a law of nature is simply being explained by more laws of nature—for molecular behavior, kinetic energy, and molecular forces are themselves simply physical descriptions. That is, this explanation is tantamount to saying that the reason a’s are followed by b’s is because c’s are followed by d’s, and e’s are followed by f’s etc. But, this only pushes the problem back a bit. For the original question was not how water becomes a gas, but, rather, why it does. And explaining laws of nature by other laws of nature only answers a “how” question, and not a “why” one. (Not to mention that if this method of explanation were valid, then eventually we would arrive at a law of nature which simply has no explanation whatsoever, which is absurd.)

This is to say that the ultimate question here pertains to asking why things tend to behave the way they do at all. And again, it will not do to answer this by reverting to any law of nature, because a law of nature is purely descriptive. But then what tool does science have for answering this “why” question, since all science can really do is formulate descriptive laws of nature useful for prediction and manipulation? The answer is that science simply does not have this tool, philosophy does. And this is why naturalist Bertrand Russell said the following:

[Physics] lays down certain fundamental equations which enable it to deal with the logical structure of events, while leaving it completely unknown what is the intrinsic character of the events that have the structure[…] All that physics gives us is certain equations giving abstract properties of their changes. But as to what it is that changes, and what it changes from and to—as to this, physics is silent.

The point here is that science can tell us about certain regularities in nature. But, it cannot tell us why these regularities happen without fail, or why the correlation between these regularities is necessary, or exactly why it is of the nature of substances to have certain dispositions to behave the way they do, or even why there is regularity at all.

Thus stated, we see that science does in fact have explanatory limits, and therefore we should not expect that science exhausts our knowledge of reality. Science is not the be-all end-all description of reality. It tells us nothing about the intrinsic nature of the things that it describes. It only tells us of the quantitative, and not the ontological.