Wednesday, April 24, 2013

Review of Alex Rosenberg's The Atheist's Guide to Reality: Enjoying Life Without Illusions

Alex Rosenberg's 2011 book The Atheist's Guide to Reality: Enjoying Life Without Illusions attempts to boldly "tell it like it is" with respect to a rigorous metaphysical materialist's reality; what it accomplishes, however, is to boldly trip over its own feet.

Rosenberg defends what he calls "scientism": "This is the conviction that the methods of science are the only reliable ways to secure knowledge of anything; that science's description of the world is correct in its fundamentals; and that when 'complete,' what science tells us will not be surprisingly different from what it tells us today" (p.6).

As Rosenberg develops the implications of scientism throughout the book, it becomes evident that he builds into his definition of scientism the metaphysical commitment traditionally dubbed "eliminative materialism."  Eliminative materialism asserts a) that only physical objects and the deterministic laws pushing them around exist, and b) that whatever cannot be entirely reduced to physical objects must not exist (i.e. must be eliminated).   Rosenberg says flatly, "The basic things everything is made up of are fermions and bosons.  That's it....There is no third kind of subatomic particle.  And everything is made up of these two kinds of things....All the processes in the universe, from atomic to bodily to mental, are purely physical processes involving fermions and bosons interacting with one another" (p.21).

It is important to note that Rosenberg's project is not to defend scientism against competing worldviews--readers ought not open the book expecting systematic refutations of theism, deism, dualism, idealism, or any other non-materialist view:
[T]his book is written mainly for those of us who are already deniers [of theism], not just doubters and agnostics.  Although we will address the foibles and fallacies (as well as the wishful thinking) of theists, we won't treat theism as a serious alternative that stills [sic] needs to be refuted.  This book's intended readers have moved past that point.  We know the truth (p. xii).
Rosenberg instead aims to work out the implications of a world exclusively populated by fermions and bosons and only fermions and bosons.  Given that this is his project, it is odd that he used over 300 pages to work this out.  In a world populated by nothing but fermions and bosons, what more need be said other than, "the world is full of only fermions and bosons"?  Oddly, though, Rosenberg attempts to fill twelve chapters with discussion about all the ways we keep entertaining illusions about a world filled with more than fermions and bosons--for example, a world filled with meaning, purposes, desires, intentions, persons, choices, goodness, values, etc.  So, for twelve chapters, Rosenberg introduces these seemingly non-fermion/boson entities into the discussion only to repeat that each one is "just a configuration of fermions and bosons."  He even writes this of his own book, "This book isn't conveying statements.  It's rearranging neural circuits, removing inaccurate disinformation and replacing it with accurate information (p.193)."  Ignore for the moment that the sentence, "This book isn't conveying statements" is itself, of all things, a statement.  What he writes is still perhaps too friendly to non-boson/fermion thinking, as it introduces illusory concepts of "neural circuits," "accuracy," and "information" into the picture; I think it would be more accurate if Rosenberg were to say this instead: "This book isn't conveying statements.  It's a temporary configuration of fermions and bosons.  From moment to moment the fermions and bosons are moving around into different configurations of fermions and bosons."

The fact that Rosenberg attempts to say more than that the world is merely comprised of nothing but fermions and bosons is precisely the reason his project suffers from abject incoherency.  In writing anything, especially a 300 page book, one implicitly accepts a host of normative, meaningful, intentional features of consciousness--e.g. one accepts that a certain "person" exists who has certain thoughts "about" certain non-physical objects of "meaning" that "he" or "she" then translates into certain symbols that combine according to certain non-physical rules to form words that have certain non-physical "meanings" that combine with other words that have certain non-physical "meanings" into non-physical rules of grammar that convey further certain "meanings" to certain "persons" who decode those symbols back into certain non-physical "meanings" that those "persons" can then have thoughts "about."

Notice how many words in the foregoing sentence were neither "fermion" nor "boson."  Seriously, if there exists nothing other than fermions and bosons, how could any symbol ever refer to anything other than fermions and bosons?  What's more, all of the traditional referents of the words in quotes are specially targeted by Rosenberg as entities that do not fit the fermion/boson picture and thus must not exist:
"The physical facts fix all the facts.  It's because they do that thinking about stuff is impossible" (p.162).
"The fact that the mind is the brain guarantees that there is no free will.  It rules out any purposes or designs organizing our actions or our lives" (p.195).
"When it fixed the facts, physics ruled out the existence of selves, souls, persons, or nonphysical minds inhabiting our bodies" (p.220).
Any attempt to coherently put all the pieces of Rosenberg's position together simply smacks of contradiction.  He wants to assert that neither he nor the reader exists and that neither thoughts nor words are ever "about" anything, yet he also wants the nonexistent reader to entertain content-less thoughts derived from words written by a nonexistent author that could not possibly be about anything at all, much less about "scientism" in particular.  Frankly, this is just irritating.

Stepping back from the particulars for a moment, let us recall why one ever proposes or entertains a given theory in the first place.  In general, we are interested in explaining, describing, or understanding the world we happen to live in.  To satisfy these intentions, a theory must both be internally consistent (that is, not self-contradictory) and externally consistent (that is, it must faithfully "map-onto" or "match" events in the external world).  Presumably, there is only one way for a theory to be externally consistent (since we live in a single universe), but there are potentially infinite ways for a theory to be internally consistent (since there are an infinite number of possible worlds).  Thus, our ultimate aim when assessing theories is to find one that satisfies both internal and external tests of consistency.  The internal test may be dubbed the "rational test," whereas the external test may be dubbed the "empirical test."

At the risk of oversimplifying too much, let me suggest that the field of mathematics is a paradigm example of rational analysis, whereas the field of science is a paradigm example of empirical analysis.  The axioms, postulates, and theorems of mathematics beautifully embody internal consistency, but they don't, by themselves, tell us anything necessarily true about the external world.  For example, if geometrically perfect cubes existed in the world, then mathematical formulas would describe the features of that cube without error.  But, of course, that conditional does nothing to show that there ever did, or ever will, exist a geometrically perfect cube in this universe.  In order to know whether a geometrically perfect cube has ever, or will ever, exist in this world, one needs to step into the realm of empirical analysis--i.e., one needs to look at our world and see what's in it.  A great example of this is the story of Neptune's discovery.  Astronomers had used Newton's laws of gravitation to compute Uranus's orbital path on the assumption that no further celestial body existed beyond Uranus's orbit.  Observations of Uranus's orbit over time, however, revealed deviations from the path outlined by Newton's equations.  Thus, there was a conflict between theory (only seven planets) and observation (deviations from predicted orbit).  In these situations, one can either revise the theory or deny the observations.  Orthodoxy generally dictates that one do the former, and that indeed is what happened with Uranus--the observations were ultimately reconciled with the existence of an eighth planet pulling Uranus off its predicted path.

Now let's return to Rosenberg and his scientism thesis.  He claims to strictly be following the methods and findings of empirical science--indeed, he derives scientism from the word science itself.  The last thing Rosenberg should want to do, then, is allow detached theorizing to dictate what one actually finds in the world.  In other words, Rosenberg should never let what exists in the world be contingent on, derivative from, or beholden to what this or that theory says.  The direction should always be in the reverse: the theory should be contingent on, derivative from, or beholden to what is observed to exist in the empirical world.  Rosenberg quite explicitly maintains that scientism predicts the non-existence of intentional states (i.e. aboutness), subjective points of view, and enduring selves.  Furthermore, Rosenberg's scientism predicts that all anyone should ever find when looking at the world is fermions and bosons (afterall, that's all that supposedly exists).  The natural next step for one committed to the empirical nature of science would be, of course, to look at the world to see whether this theory checks out.  Depending on what one finds, the theory is either affirmed or denied.

So, when Rosenberg looks, what does he see?  He says, "Without scientism, we look at life from the inside, from the first-person POV [point of view].  The first person is the subject, the audience, the viewer of subjective experience, the self in the mind" (p.194).  This seems to suggest that when Rosenberg looks, he finds, just as his readers reliably find, a first person point of view, a subject, a self--that is, he finds what his theory says he shouldn't find (selves, etc.), and he doesn't find what his theory says he should find (only fermions and bosons).  Despite these observations, Rosenberg asserts, "Scientism shows that the first-person POV is an illusion" (p.194). Now, on first blush, this appears to be an instance of retroactively modifying an observation in order to fit a given theory.  This would be analogous to claiming Uranus's observed irregularities were merely "illusory" in order to preserve the seven-planet solar system model.  Such reasoning is the reverse of empirical honesty.  If Rosenberg's scientism were the correct model of our world, there should not even be "illusions" of first-person points of view, much less real points of view.  Again, all there should be are fermions and bosons--no selves, no meaning, no points of view, and certainly no illusions.  There shouldn't be any "experiences" of any kind.  But, remarkably, Rosenberg goes on to say this:
Since scientism admits that experiences exist, they will have to be physical for scientism to be true.  If the facts about experience can't be fixed by physics, scientism will be false.  More specifically, if the facts about my experiences can't be fixed, explained, accounted for by neuroscience, then scientism will be false (p.227).
Now, why in a world of just fermions and bosons would there be "experiences"?  Does physics tell us that fermions or bosons have "experience" properties in addition to mass, spin, or whatever?  Furthermore, why in a world barring first-person points of view would experiences nevertheless be permitted?  How is it even coherent to speak of "experiences" without points of view "having" the experiences?  To Rosenberg's credit, he does appear to recognize, to some extent, the gravity of this problem.  He admits that these "experiences" need to somehow be absorbed into the fermion/boson world if scientism is going to live.  In chapter ten, he (sort of) attempts to answer objections to the-mind-is-the-brain conjectures in the philosophy of mind.  He brings up three famous arguments (one each from Rene Descartes, Thomas Nagel, and Godfried Von Leibniz), each of which illustrating conceptual problems inherent in turning the mind into a physical thing.  I'll simply sketch Descartes' argument here as that will suffice.

Descartes simply argued that he can doubt everything about the world except his own conscious awareness.  He can be wrong about the fact that there's an independently existing world "out there"; he can be wrong that he has a physical body interacting with such a world; he can even be wrong about logical truths like addition.  What he cannot be wrong about, though, is that he is a perceiving thing.  Even if every perception he ever has deceives him in some way, it still must be true that he is a perceiving thing.  Rosenberg's persistent appeal to "illusions" in his book very clearly meets Descartes' argument head-on.  The appeal to illusions requires that there be a perceiving thing of whom the illusion deceives.  Here's Rosenberg's response:
Does Scientism actually have to take Descartes's argument and others like it seriously?  Does it actually have to diagnose each of their mistakes, or any of them?  No.  Even before you hear them, science provides a compelling reason that they must all be wrong.  One has only to weigh the evidence for scientism--500 years of scientific progress--and the evidence against it--including those cute conundrums.  It's clear which side has the weightier evidence.
Scientism isn't required to figure out what is wrong with these proofs that experience can't be physical, so minds can't be brains.  That's the job of science--neuroscience in particular.  Scientism can ignore the conundrums, dilemmas, and paradoxes that the arguments generate.  Science cannot.  These problems are signposts in the research program of the science of the brain.  Unraveling them will be some of its marks of progress.  Meanwhile scientism can already be confident that they will be unraveled (p.227-8).
To call this response underwhelming would be an understatement.  This move does not bear the marks of one following empirical observations.  This is theory telling us what exists rather than what exists telling us what theory to hold.  His strategy is to downplay Descartes et. al.'s arguments as "cute," tell his readers to ignore them, and redirect attention to the 500 years of scientific progress in fields unrelated to the mind.  It is worth mentioning that Descartes (who lived from 1596 to 1650) is himself credited as one of the founders of modern science.  By appealing to the "500 years of scientific progress" while simultaneously mentioning Descartes, Rosenberg inadvertently draws attention to the fact that science has failed to resolve Descartes' mind-body problem for basically all 500 of those years.  How much "weight" does this add to scientism's evidence base?  Furthermore, it is entirely circular to appeal to the future findings of science to demonstrate the physical nature of experiences; the point of Descartes et. al.'s arguments is to show that experiences are not physical.  To assert that future science will show experiences to be physical is simply to assume that experiences are physical.  If one assumes what needs to be proved, then one is just arguing in a circle.  I suppose scientism's devotees can follow Rosenberg's suggestion and just ignore these problems, fortunately the rest of us can just as easily ignore scientism.

Allow me to close by repeating something I said earlier.  Rosenberg's project took a turn for the worse the moment he tried to say something above and beyond "the world is filled with only fermions and bosons."  If this claim were really true, there would simply be nothing more to say.  There would be no book to write, no audience to educate, no scientism to define.  There would just be the damn fermions and bosons and nothing else.  Such a world would be internally consistent.  The fact that Rosenberg did not stop with that one statement undermines his entire project.  Word after word, he just digs scientism a deeper and deeper grave by reminding the reader that there are all kinds of words that are "about" concepts other than fermions and bosons.  He cannot say anything without presupposing the very entities and concepts he wants to say do not exist.  In doing so, all 300 pages of his book decisively show that his scientism is not externally consistent--it may describe a world, but it does not describe our world.  Insofar as readers are interested in understanding this world, the best this book will do is provide some insight into what some people in this world are prepared to believe.