Eternalism, Nihilism, and Nebulosity

Eternalism and nihilsm are two poles of a tendency in relationship to either domains of meaning or systems of meaning. Eternalism seeks to stabilize and reify a model or way of seeing, whereas nihilism seeks to destabilize and dereify. I get these concepts from David Chapman; they come originally from at least a specific lineage of Tibetan Buddhism, though I suspect his interpretation is a bit modern and idiosyncratic.

Eternalist views see some domain of meaning or reality as stable, fixed, undeniable, or singular. Some examples:

• Justice, as determined by eg. divine command, or "natural law".
• Positivism tries to construct logic as an undeniable and flawless basis for all knowledge
• Soulmates as a feature of many folk fantasies or perhaps even "theories" of romantic love
• Purpose, as a story for the goal and value of a particular human life, eg. "I was put on this earth for a purpose"
• Souls as the basis and substance of human personality and identify

Unfortunately I sound pretty unsympathetic here, but it's hard to treat these sympathetically because (from my perspective) they're so obviously confused, or just empirically or even logically false.

Nihilist views construct a domain of meaning as unreal, insubstantial, or irrelevant:

• Atheism, generically, both denies cosmologically the existence of God, but also usually regards divinity, ritual, and sacredness in general, as meaningless or harmful.
• Views such as that one's legacy is meaningless, because in the future one will be dead, so therefore one's legacy is "not real" or "not meaningful" to oneself.
• Many formulations of moral relativism maintain that moral beliefs are of no consequence with respect to one another, and often that morality itself is basically meaningless or inconsequential.
• Similarly, some forms of epistemic relativism effectively deny that "knowledge" can exist whatsoever.
• Views to the effect that human meaning and value are actually meaningless because they are believed to be based in biological processes

While all of these examples are given propositionally because this is a written medium, the more interesting aspect is something like perceptual or existential. These kinds of views, most of them more or less ways of seeing, have intuitive, phenomenological effects, besides whatever arguments or ideologies they cause us to deploy or ally with.

Also, to clarify, 'eternalism' doesn't necessarily refer to stability in time, rather it means something more like "utterly fixed in meaning." Similarly, 'nihilism' doesn't here just refer to the view that everything is meaningless and pointless, eg. "existential nihilism," but more to views that specific things are.

Nihilist views say something like, "I shall not care about this," whereas eternalist views say "I shall care unceasingly about this." They both try to admit no wavering in the perception of meaning. Eternalist views are often about constructing stable, unambiguous justifications, whereas nihilist views often say, "there can be constructed no stable justification for this, therefore you get nothing at all."

Nebulosity

Eternalism and Nihilism, in this model, are ways of creating stability in the face of nebulosity. Chapman gives two standard examples to illustrate what nebulosity means:

• "Nebulous," literally "cloud-like": clouds are a real phenomenon, but blend indistinguishably into each other and into the background sky. Up close, or inside a cloud, like in an airplane, it's unclear if the cloud even exists, and it has no apparent boundary at all.¹
• How many pebbles are in a bucket? Well it depends on what counts as a pebble, since there's probably some grit and sand in there as well, and of course grit and pebbles are really the "same" thing, just of different sizes. There isn't quite a "correct answer" except for a specific purpose.²

Those are excellent examples but I want to stress that nebulosity is much more general. Often we're talking about ontological (and eventually axiological) nebulosity. For example, even physical quantities which seem quite certain and definite are not necessarily so simple, and sometimes were unknown until recently in human history:

• Mass is generally understood to be definite and clear, up to the accuracy of the measurement. This is correct, but if you put two objects in someone's hand that are similar in mass but different in density, and ask which is heavier, most people will pretty consistently "feel" the denser object to be heavier. If I recall correctly, this is true up to substantial differences in mass, holding the difference in density steady.
  • This is also ignoring that units of "weight" are defined as a force exerted on an object by a particular gravituational field, rather than an intrinsic property (mass). People say, everything is "lighter" in the freefall of an orbiting satellite—wait what? I thought weight was definite? etc.
  • Mass is a fine concept, and often more useful than what we now regard as either derived or compound units like weight and density, but it's not obvious that human-scale perceived weight "needs" to be the same as mass.
• Temperature, up to around the 18th century, was taken to be too confusing and variable to be reduced to a consistent measurable quantity. Perceived temperature is also a function of wind chill, humidity, air pressure, and the person's physiological state, even though we can now measure temperature quite accurately.³
  • Indeed, "how hot is it out" has much more to do with perceived temperature than physical temperature.
  • Even things like "boiling point" and "freezing point" are actually fairly nebulous. Water will evaporate faster at temperatures well below boiling point at a given pressure, and very clean vessels which provide no nucleation points can avoid approaching a rolling boil well above boiling point. Similarly for the freezing point, highly purified water in a very clean vessel can stay liquid decently far "below freezing."

Concepts like the borders of a country, GDP, unemployment rates, and of course race, are much more nebulous than temperature and mass. This is of course letting alone those like "freedom," "love," or "democracy."

I would claim that nebulosity is much closer to the fact-of-the-matter of human concepts and meaning, namely that they are substantive, meaningful, and worthwhile, while also ambiguous, uncertain, and seldom sharply defined. I would claim that all of the examples above of eternalist and nihlist views are at least partly true or real, or true in some sense, but a. usually not in an exclusive sense, and b. in a sense which blends indistinguishably into other, apparently "contradictory" perspectives.

To elaborate a bit on some of these, at least specifically from my own perspective:

• I believe there are, relatively speaking, justice and injustice, and that they are meaningful and important, but I don't believe that they are "mind-external" in the way that physics is, and I don't believe that they were given or defined by an intelligent creator. I also think justice is substantially culturally bound, even though I still feel strongly about notions of justice which other cultures, present and past, would disagree with.
• I believe there is profound intimacy and "real," sometimes earth-shattering, romantic love, but that it's contingent on circumstances and personal history, and that it's limited in time, at least to natural lifetimes but perhaps much shorter, ie. that perhaps romantic love is both "real," and "not meant to last" past the length of time needed to raise children, or even less.
• I don't believe in a cosmological God, but I believe that divinity and holiness are "real" in a sense such that throwing them out is something like "wrong" or "mistaken," and that it's often "correct" to construe something like God, at least in some culturally specific context.

The point here is robust agnostic of these specific cases. I'm trying to describe structures which might we might want to, or do often, describe as though fixed and definite, and which I think are in fact unstable and nebulous, while still "real" or substantial.

Nebulosity and comprehensiveness

I want to pick on a specific confusion I see all over the place, in philosophy in general, and especially among rationalists.

Models are lossy, or as people often (mostly wrongly) repeat, "wrong." Many models are presented by their authors, explicitly, as comprising some useful heuristics, or some set of generative assumptions; these can however be mistakenly taken as comprehensive and authoritative by someone who isn't paying enough attention. (See inappropriate reification.)

That's a boring case but still a common failure mode. Examples that come to mind are eg. psychiatric diagnosis again (where, specifically as described by the DSM, the diagnostic categories are not regarded as necessarily forming natural kinds), perhaps also speciation in both biology and linguistics. Even concepts like covalent vs. ionic bonds are actually continuous, on the same dimension! In these cases the model is presented as lossy but is still predictive in a range of contexts.

More interesting are cases where it's ambiguous or uncertain whether a model is comprehensive. My understanding is that the modern posture taken in physics is something like, "(because of the failure to yet synthesize QM and relativity) indeed, something about these models is wrong and we're pretty confused, nonetheless they're extremely predictive in the domains in which they're regarded as predictive, and so you should absolutely not expect to be able to achieve physical effects contrary to the existing models." (Perhaps with some exotic theorized exceptions like vacuum collapse, etc.)

What's remarkable metaphysically and culturally about physics is that it's referred to outside of academia as singularly authoritative and comprehensive, which is not even the posture taken by actual physicists. My point here is not to take a jab at physics, or to claim that some metaphysical gymnastics will let you perform miracles—I have no such expectation. My main gripe here will be with respect to theories of value, but I think the point applies for all of metaphysics:

If you have a model or a theory, at what point can you say, "indeed, this is a satisfactory reductionist account of this domain, and in fact not lossy, up to the resolution specified by the model," as opposed to "this is predictive but known to be lossy"? What reason do you have to believe the ontology you chose accounts for all of the behavior observed in the actual domain?

I don't think this point is marginal. I see this mistake being made constantly, even by fairly sophisticated people, namely of too readily taking an ontology or a broader model as comprehensive before that has been demonstrated. Probably, considering that even physics, the reductionist domain par excellence, is not quite "done", we should be pretty skeptical of any claims to have completed a reductionist account of any other empirical domain!

Two examples which I'm going to harp on throughout this book:

• Everything that intersects with the study of the mind. Again, often researchers at the actual leading edge are much more parsimonious in their claims, but the point is well enough made anyhow.
  • It seems like quite a lot of people take the computational theory of consciousness as a fait accompli, even still to this day in the absence of a materialist account of consciousness! (Note that I mostly expect CTC to win.)
• I often see utilitarian and decision-theoretic models being deployed as a fundamental theories of value. It seems to me that decision theorists have largely skipped the work of justifying that you can even reduce human value to a consistent utility function. Worse, in many cases they recognize that human value is not consistent (in this narrow sense) and go on merrily with their philosophizing.

In these cases I want very strongly to say "hold up there, who ever said that was comprehensive?" To be clear, however: sometimes one has a research program, and indeed it's very useful for a person or group to pursue some set of assumptions while they still remain provisional. That's not what I'm objecting to.

Often I think one should be asking something like, "is this the kind of domain where we're going to get a nice comprehensive account, or is our intuitive ontology made out of so many heuristics and kludges, and if you try to hammer it into something flat and consistent you'll just be throwing out important functional aspects of what's normally regarded as the domain?"

I'll also say, even in domains in which I expect a comprehensive model is possible, it will be at a much higher resolution than that at which people are yet trying. Think of the riotous complexity of infectious diseases, which (my impression is) we having fairly straightforward models of at this point. People have of course been trying to understand disease for the entire history of our species, and we were working with almost complete nonsense for a large category of diseases until less than 200 years ago, and the picture continues to get messier and more complex as we learn more.

Footnotes

1. David Chapman. Meaningness ↩

2. David Chapman. Metarationality ↩

3. Hasok Chang, 2004. Inventing Temperature, pp. 8-11 ↩