Systematic Ellipses in Our Knowledge

Friday 11 August 2023

One of the more flattering portraits of Immanuel Kant.

In last week’s newsletter I discussed the possibility of a Kantian ethic for space exploration. I noted in that newsletter that Kant’s formulations in ethics concern themselves with any rational being, so he has already made the initial step toward transcending a purely anthropocentric formulation of ethical problems. I also noted in that newsletter that Kant’s work in cosmology meant that his reference to any rational being should be understood as referring to rational life on other worlds. However, I found this in Kant’s Metaphysics of Morals (Introduction, I):

“The very concept of duty is already the concept of a necessitation (constraint) of free choice through the law. This constraint may be an external constraint or a self-constraint. The moral imperative makes this constraint known through the categorical nature of its pronouncement (the unconditional ought). Such constraint, therefore, does not apply to rational beings as such (there could also be holy ones) but rather to men, rational natural beings, who are unholy enough that pleasure can induce them to break the moral law, even though they recognize its authority; and even when they do obey the law, they do it reluctantly (in the face of opposition from their inclinations), and it is in this that such constraint properly consists.”

Thus while Kant may have had in mind other rational beings as beings on other worlds — essentially, peers beings to human beings — Kant may also have had in mind the possibility of holy beings, i.e., angels, as other examples of rational beings. Aquinas and other Scholastic philosophers spilled a significant amount of ink on the analysis of angels (and demons) as other forms of intelligence, and we have much to learn from the Scholastic tradition. (I wrote about this in Angelology and Artificial Intelligence, and a number of interestingly related threads are to be found in God, Human, Animal, Machine: Technology, Metaphor, and the Search for Meaning by Meghan O’Gieblyn.)

So the idea of the community of rational beings (Kant’s Kingdom of Ends, i.e., all beings that are ends in themselves because they are moral agents) is larger than human beings is an old idea, and had been studied by the Scholastics with their usual exhaustive thoroughness. We can also expanded the Kingdom of Ends by including other rational beings on other worlds, or by including other beings on Earth, perhaps not as intelligent as we are, but still possibly intelligent enough to belong to the Kingdom of Ends. If we include all these categories, and throw machine consciousness into mix too, the Kingdom of Ends starts to get crowded, whereas most traditional conceptions of ethics confined moral considerability to human beings alone. This is how Kantian formulations in terms of any rational being serve to expand our ethical conceptions, an effort which, if iterated, could take us in some interesting directions. I have often suggested that we need to press our thought experiments to the limits of our imagination, elaborating them until we can eventually transcend our anthropocentric conceptual framework, and this is one of those opportunities: pushing our moral conceptions as far beyond the familiar as we can get, and Kant opens one of these avenues for us by shifting from human beings to any rational being.

Last week I also posed the question of what the threshold of intelligence ought to be for a being to be counted among rational beings, hence to be properly included in the Kingdom of Ends. Ought we to administer IQ tests to determine membership in the Kingdom of Ends? And what do we do with those who flunk the IQ test? Another twist on this has occurred to me: it’s not just about quantity of intelligence (what I sometimes informally refer to as “pure processing power”) but also about quality of intelligence. Intelligences may different qualitatively in interesting ways, especially when we start stretching the possibilities for rational beings to include aliens, angels, animals, and AGI.

In another project I am working on I have attempted to be more specific about the nature of human rationality (or intelligence), which is rationality that is embodied, evolved, biological, primarily visual, diurnal, mammalian, sexually dimorphic, and so on. This is just an extemporaneous list of some of the properties that may be relevant to human intelligence. Since the Enlightenment, intelligence (or, if you prefer, rationality) has been construed in universalist terms, which implies that, as human reason develops (hopefully toward greater perfection) the development has the character of leaving aside any remaining parochialism and striving toward a pure, universal reason. I do not believe this to be the case, but an adequate analysis would be a big job in itself.

If we hold that reason is one and universal, then all beings striving toward bettering their intelligence will be converging upon the same rational ideal. However, if we hold, on the contrary, that rationality is not one but many, and may take a plurality of forms, then a number of rational beings striving to better their intelligence may not converge on a single, universal rational ideal. They may, in fact, diverge, and the further they push their development, the more they will diverge. This is mere assertion; I do not consider it to be proved, but I state it here so that the reader will understand my presuppositions in coming to this problem. As I said, the full exposition of this would be a significant undertaking, and I will not attempt it here.

Is reason one or many?

It’s more than just the ideal of reason involved. The possibilities of mind (or, if you prefer, consciousness) in the universe aren’t yet known to us, either poorly or vaguely — they aren’t known at all. And this is compounded by the problem that mind-as-we-know-it supervenes upon biology, and we also do not know the possibilities for life in the universe. This later claim is truism that has been repeated any number of times, but it is worth trying to clarify what I mean here. A distinction can be made between the things we don’t yet know, and the conceptual tools that would both be developed in order to discover the unknown, as well as being developed in response to the unknown so as to comprehend it as part of the same body of knowledge as our existing conceptual framework (i.e., so as to integrate it into an expanded science). If I wanted to be pretentious I could call this latter second-order knowing (I would prefer some other, more imaginative label), which is when we know the category to which something belongs, and placing it in a category systematically relates it to other things in the universe, which either fall into the same category or into some other category.

We could potentially learn quite a bit about any given object of knowledge without being able to place it within a taxonomy that divides all the individuals of the world into some taxon or another, but when we not only know the individual but also know the taxon to which it belongs, and we have an adequate taxonomy, then we are able to place the individual object of knowledge in its proper context. So what is an adequate taxonomy? We begin with local (i.e., parochial) taxonomies and then extend our taxonomies until they converge on epistemic totality. However, the very idea of epistemic totality is a philosophical concept that the sciences cannot define, and many scientists would prefer not to define.

We can, nevertheless, posit epistemic totality as the end point of the development of knowledge at which everything is known. The expansion of conceptual frameworks toward greater adequacy points to this end, but we can deny this possibility as easily as I can deny that universality and unity of reason. I think I can show that these are two different things, but I will not attempt to make the argument now. Again, an adequate discussion of this would lead far afield and cannot be settled here.

But grant me, for the purposes of argument, that our conceptual frameworks expand over time, becoming more comprehensive, and, as the conceptual framework of science expands, it places our knowledge in a larger context. The limit of our scientific knowledge is bounded by the possibility of observation, meaning that scientific knowledge is bounded by the limits of the observable universe. New techniques and technologies of observation may expand the scope of the observable universe, but at any one moment in history we are forced to accept the limits of the universe observable in our time. Knowledge that transcends empirical observation may also transcend the observable universe, but that is not science. Perhaps the expansions to science that I have discussed in previous newsletters will eventually lead us to an integration of science and philosophy that will also allow for the integration of empirical knowledge and non-empirical knowledge, but at present that is little more than an optimistic hope for the future.

We have a taxonomy for life on Earth, but not yet an astrobiological taxonomy for life.

Sticking more-or-less to what science means today (science-as-we-know-it), scientific knowledge still has ample room for expansion. I have many times discussed the ways in which science might be extended, and among these ways are the expansion of the conceptual framework within which any scientific knowledge must be located. Accepting as a given the limits of the observable universe, we can still be reasonable specific about the conceptual frameworks that we lack, and the development of which is either within our grasp at present, or will be within our grasp soon, if only we will explicitly recognize the epistemic ellipses and make the effort to fill them. For example, we lack:

1. A system of planetary classification
2. A metric for planetary complexity
3. A system of planetary system classification
4. A metric for the complexity of planetary systems
5. A taxonomy of biospheric development
6. An astrobiological taxonomy of life
7. A system of classification for intelligence
8. A system of classification for technology
9. A theory of science (a science of science)
10. A theory of civilization (worthy of the name)
11. A theory of history (worthy of the name)

This list is not intended to be exhaustive; no doubt there are other — many other — conceptual frameworks that need to be developed. I attempted (1) in my talks at St. Andrews and Edinburgh last year. In a talk to the University of the Third Age last year I posited a scale for (2), and many papers implicitly approach the problem, but there is no explicit recognition of this need. There are some papers on (3), but nothing really systematic, but it is exciting that we are beginning to gather enough data on exoplanets and exoplanetary systems that we can realistically contemplate a taxonomy of planetary systems. I don’t know of any work on (4), but it would be an interesting question to work on. I also attempted (5), which was the focus of my talks last year. I have many times quoted Carl Sagan on the need to deparochialize biology, which is the point of (6), but we cannot realistically do this until we have other forms of life to study. (7) is what I have been discussing in the above paragraphs. Daniel Dennett wrote a book, Kinds of Minds: Toward an Understanding of Consciousness, that explicitly recognizes that there are different kinds of minds, but he doesn’t attempt a taxonomy. Many philosophers of technology have worked on (8), so there is at least a literature on the problem, but no consensus such as could found a scientific research program. And I have many times discussed (9), (10), and (11): both scientists and philosophers have seen the need, and while some have attempted, unsuccessfully, to supply the need, most are unwilling to take up these largest questions as the chance of success (or even the chance of making incremental progress) is slim, and the likelihood of ridicule for making the attempt is high.

There is a sense in which we can say of all the parochial forms of complexity, i.e., forms of complexity at this time only known from terrestrial instances, that there are many instances — many kinds of life on Earth (biodiversity), many kinds of mind on Earth (neurodiversity), and many kinds of civilization on Earth. (The taxonomies of terrestrial biology and terrestrial minds are the limited and local taxonomies I mentioned above, that still need to be placed into a comprehensive context, in that way that astrobiology is attempting to place terrestrial biology in a cosmological context.) On the other hand, we can also say that there is only one kind of life on Earth, and we will not begin to understand the diversity of life in the universe until terrestrial life can be compared to life elsewhere. Similarly, we know only terrestrial minds, and we know only terrestrial civilizations. In the case of life, we can identify all life on Earth as using the same DNA replication mechanism, and in the case of minds we can identify all terrestrial minds as supervening upon the central nervous systems of organisms that all employ the same DNA replication mechanism. While the diversity of civilizations on Earth cannot, in the absence of a theory of civilization, be traced to a common root in quite the same way, still, there is a commonality: all terrestrial civilizations are human civilizations. Elsewhere I have quoted Carl Sagan to the effect that, “Civilization is a product of the cerebral cortex.” (Cosmos, Chapter XI) In this way, all terrestrial civilizations are unified by supervening on the human mind, as all terrestrial minds are unified by supervening upon terrestrial biology.

Taking science as we know it in the context of the above survey what it is missing, we cannot maintain that there are only mere occasional ellipses in science, but rather that our scientific knowledge is still an overwhelmingly local and parochial affair. Science is nowhere close to completion and stagnation (though it may stagnate for extraneous reasons), but rather it is an enterprise that has only barely begun.