14 November 2014

Closing thoughts on model-dependent realism

Whenever I engage in debates with theists, I'm reminded of Tim Minchin's analogy of such discussions — that two people trying to win a debate while operating from completely different sets of assumptions are like two tennis players trying to score perfectly executed shots from the opposite ends of separate tennis courts. After pouring several hours of writing and research into my previous post, offered as a counter to Steven Jake's dismissal of Model-Dependent Realism (MDR), reading his response this week left me with a feeling like we were two ship captains in battle, and I'd just punched a massive hole in his ship's hull with a blast of broadside cannons — and yet, as his vessel sank and his crew abandoned ship, he stood on the deck shouting, "Haha, you missed!"

It's not that I feel Steven's response fails to adequately rebut my case so much as I feel that it doesn't address my case at all. I wrote my critique with the intention of clarifying what I felt were sorely misguided assumptions on his part about what model-dependent realism is; in his response, Steven has shown a frustrating unwillingness to critically examine those assumptions, repeated arguments I'd already addressed, and ultimately attacked a caricature of my actual position. He's said enough that I think it's worth replying to a few key points, but I'll let him have the last word should he choose — this post will be my closing thoughts on the matter.

With that, let's dive into his post, the entirety of which can be read on Steven's blog here. He begins:
I never claimed that MDR stated that models conform to reality—in fact this is my point of contention with MDR! I agree that MDR states that our models are only interpretive structures, so to speak.
Let me first clarify my point of contention here.This is what Steven wrote in his first post:
First, let it be understood that MDR does not claim that if two theories can both accurately describe or predict the same observations, then we cannot, at the moment, determine which theory actually conforms to objective reality. No, MDR is claiming that neither theory conforms to reality more than the other—that is, neither theory is more real than the other.
My issue here is that Steven is already off the mark with the whole concept of "conforming to reality". How do we know that reality exists? Presumably, through our sensory data. But it is not that simple, because our brains don't read sensory data as though we're watching a film or looking at a photograph — rather, our brains construct a model of reality based upon that data, and we navigate that phenomenological model with a process of reasoning heavily informed by subconscious metaphor. Unless one subscribes to naive realism, it's untenable to claim that we have direct, unfettered access to some singular, absolute reality — few would disagree that our conscious models of the world are often wrong, as illustrated by the rather striking number of cognitive biases to which we are susceptible.

Of course, as Steven points outs, we agree that objective reality exists. This is trivially true. But the important distinction that MDR makes is that our very conceptualization of reality is a conglomerate of distinct and overlapping models that work at different frames of reference. MDR is not comparing models against some singular "objective reality" — rather, it states that our very concept of what objective reality is is derived from our ability to construct models and test them against observation. 

Steven continues,
But for MDR to state that nothing at all can conform to reality, or that talk of models conforming to reality is meaningless, is to refute oneself, since this assertion itself is a claim about the nature of reality.
In addition to the error I've just pointed out (MDR does not state that "nothing at all can conform to reality"), Steven errs here in another way: when MDR states that our concept of reality is based upon distinct and overlapping models that successfully explain and predict data, it is not a statement about reality — it is a statement about the nature of models.

The claim that we do not have a model-independent description of reality is an inductively derived, provisional assumption that follows from two foundational assumptions: I exist, and my sensory data is sometimes correct. We can gleam the truth of this assumption from inconsistencies and errors in our phenomenological experience, our difficulty in communicating what we perceive as truths to others, and in the failed predictions or superfluous assumptions of misguided theories. If we had a model-independent concept of reality, we would never misperceive it or make cognitive errors. Indeed the entire enterprise of science would be unnecessary, since there would be no need to account for data or make predictions — our intuitions about reality would be 100% correct, 100% of the time. All human knowledge could be subsumed under a singular school of analytic philosophy, because our process of reasoning would always correspond correctly to reality.

Steven drops a sarcastic bombshell here that leaves me utterly guffawed:
The point is that MDR essentially claims the following: No model or theory is real, except, you know, the theory of MDR.
 MDR is neither a theory or a model. It is a provisional epistemological assumption about the nature of human cognition.

Steven appears equally hung up in misguided incredulity over the way MDR treats the concept of 'real', and trots out an example that I remember William Lane Craig using, and I remember feeling equally dismayed at how utterly badly he understood Hawking. Steven states,
On the contrary, I maintain that Mike has misunderstood MDR here. For when Hawking was discussing his theory that one model cannot be more real than another, he uses a very specific and revealing example—namely, that of creationism and the Big Bang theory. Examine this quote straight from the horse’s mouth: “this model—the big bang theory—is more useful than [creationism]. Still, neither model can be said to be more real than the other.” Did you catch that? The big bang cannot be said to be more real than creationism! Herein lies the absurdity of MDR. Again, remember that Hawking is not claiming that we can’t determine which theory accurately conforms to reality, rather neither theory conforms to reality at all, since conformity with reality is meaningless.
Like Craig, Steven is failing to distinguish between colloquial and philosophical usage of the concept of real (ironic for a couple of guys who confidently tout their expertise in philosophy, no?). MDR in no way prevents us from saying "the model presented by young-earth creationism is not real" in the ordinary, everyday sense we use the word 'real' — because, as Lakoff pointed out in the examples of cognitive neuroscience I highlighted in my previous post or as Sean Carroll points out in his post about the reality of free will, that use of the word conveys an ontology necessary to explain our experiences from a certain frame of reference (in this case, our everyday phenomenological frame of reference).

Hawking is saying that fundamentally, the concept of real is only analogous and emergent — a construct of the mind. We call things 'real' when they contain a theoretical ontology necessary to account for data and make predictions. The field of computational neuroscience, for example, is based on a semantic framework that is entirely metaphorical: the Brain As Computer metaphor. There are not literally numbers being crunched the brain, but the metaphor is absolutely essential to our understanding of how the brain works. This is what is meant in saying that because the computational metaphor contains a necessary ontology for explaining and predicting phenomena, we ascribe to it the quality of 'real'.

But I want to go a step further in my critique of Steven here. I think he's attempting to sustain his criticism of Hawking and his incredulous rejection of MDR by adopting an interpretation of Hawking's work that is so unapologetically uncharitable that it borders on complete absurdity. If Steven's view of Hawking's work were correct, Hawking would be openly — not even tacitly! — claiming that reality is whatever we prefer it to be, and that we can adopt whatever scientific explanations we choose to describe the world. Steven is taking Hawking's use of the word real as a colloquial synonym for words like true and valid, when Hawking is using it in a highly specific philosophical context. For Hawking to adopt the view Steven has ascribed to him would mean for Hawking to be disregarding his entire career of theoretical work for a naive refrain of "Hey man, reality is whatever you want it to be". Steven is entitled to think (mistakenly, in my estimation) that Hawking lacks the philosophical sophistication of his Scholastic heroes. But does he really take Hawking to be that stupid?

Steven then returns to solipsism and realism:
[On] MDR neither realism nor solipsism is true, or false. In fact, such talk is, on MDR, superfluous. But this, again, is where the absurdity lies. For either a reality exists apart from subjective observers, or it doesn’t. This pure logic: either A or not-A. What we cannot say is “neither.”
This dichotomy is true, and indeed MDR does affirm the existence of an external reality — though it is an inductive and provisional assumption, not a claim of a priori truth as in the case of metaphysical realism (Steven's repeated oversight of this fundamental distinction again betrays a lack of nuance in his critiques) . But while we can say that ontologically one or the other must be the case, we can also rightly point out that epistemologically, it is a useless and superfluous discussion — which is precisely what follows from MDR.

Steven then shoots one so far over the bow that I can't even feel it breeze by:
For the main point of Mike’s post is an attempt to demonstrate that I have misunderstood MDR, and that my attacks against it are invalid. But wait. How, on MDR can Mike say that any “model” one espouses, whether mine or anyone else’s, is incorrect? He can’t. Remember Hawking: “it is pointless to ask whether a model is real[.]” Talk of real, unreal, true, or false, is meaningless here.
Again Steven is hastily and erronously conflating a colloquial usage of the word 'real' as a synonym for 'true' or 'valid' and uncharitably butchering Hawking's highly contextual intended meaning.  On MDR, models that have lesser (or zero) ability to account for data and make predictions are discarded as more reliable models form a coherent picture from convergent evidence in multiple frames of reference. Hawking is clear on this issue even in the full quote of the section Steven quotes above (emphasis mine): "According to model-dependent realism, it is pointless to ask whether a model is real, only whether it agrees with observation." Nothing stops us from saying that unreliable models of reality are 'not real' in our everyday parlance — creation 'science', such as it is, clearly does not agree with observation. Hawking's point is that the concept of real is ascribed to successful models rather than being a fundamental part of their ontology.

Steven closes his argument:
Mike, again, seems to miss the blatant contradictory nature of MDR here. He, and Hawking, claim that we do not have a model-independent concept of reality, and that we can only find utility in models, not truth or reality in them. But this itself is a claim about the nature of reality. Just ask the question “Is it really the case that we have no model-independent picture of reality?” The answer will commit one to make an objective claim about reality.

MDR does not claim we cannot find "truth or reality" in models — it claims (I'm repeating myself here) that our very concept of reality is derived from models. This is not a claim about reality; it is a claim about our epistemological relationship to the world, and it is the only such claim that is consistent with our empirical knowledge of the human mind while making a minimum of assumptions.

Steven closes by asserting that MDR "remains an incoherent, absurd, and self-refuting philosophic position". But Steven has only been able to sustain his criticism by constructing a caricature of the concept and displaying a wanton disregard for charity in reading Hawking's work, butchering Hawking's words into something barely recognizable from the nuanced and contextual concepts he put forth in The Grand Design.

The supreme irony here is that what Steven accuses MDR of doing is precisely what he hopes to do in undermining it. MDR tells us how we frame ideas and test them, slowly constructing a layered, composite conceptualization of reality through overlapping theories from a variety of reference frames. In rejecting MDR, Steven can claim to have a model-independent, privileged access to deep ontological truths that are free from the burden of testing and falsification. He becomes free to believe in whatever he wants — after all, if you want to shoehorn a deity into your picture of reality, a parsimonious and empirically responsible philosophy like model-dependent realism (and with it, Lakoff's embodied realism) will quickly disabuse you of such a fantasy. Given his theistic bent, is Steven's incredulity really surprising?

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.