some qualifications on agency
I was terse in my assertion that agency is a necessary component of morality, so that there are two fronts on which i think i could be clearer and more convincing (to myself, at any rate): (1) more precisely demarcate circumstances involving agency from circumstances not, and (2) addressing some plausible objections to the necessity of agency at all.
Within the terrain near the lower threshold of agency i positioned three examples: viruses, agent-based models, and deterministic games. The first of these may have been unnecessarily complex; the behavior of individual proteins and other molecules within a cell themselves exhibit agency, e.g. by traversing and (un)zipping nucleic acids and by ferrying other chemicals from one location to another. One might then object that these processes are themselves deterministically (in the colloquial sense) governed by regulatory processes that themselves may be described entirely in terms of the quantities and behaviors of substances in the surrounding environment. Thus, the whole cell, something rather more complicated than a virus, might be described analogously to a Rube Goldberg contraption, performing all sorts of tricks but with the underlying mechanisms entirely (or adequately) exposed. From this point of view, the perception of agency vanishes.
I want to argue that this is fine—that agency can depend upon point of view.
The principle is more clearly illustrated still by the agent-based model. Such a program may be deterministic or stochastic, but stochasticity itself is ultimately deterministic, depending on the string of bits governing a computer’s pseudorandom behavior and on the environmental factors that went into the choice of seed; so we may as well consider the program deterministic. If we pay attention to each signal produced by each procession along each bit of program, we do not see anything at all like agency; we see an automated process no different from the tumbling of dominoes (though a bit better-behaved). The agency we observe at the level of the visual representation of the simulated system is illusory.
The complication in both cases is that the scales at which we perceive the system exhibit agency versus openly mechanical behavior are separated by several orders of magnitude. We may watch the bits being processed one by one (in a sufficiently slowed reproduction), but we cannot on our own translate that process into anything resembling biological regulation. That must be done for us, by the acceleration of the process several hundredfold and its representation via colored clusters of pixels strategically arranged specifically for our own ease of understanding. Likewise, the interplay of osmosis and electromagnetism and other factors do not sensibly reproduce, by our own cognitive syntheses, a mental model of a living cell. We have to reorient ourselves and reconfigure our mental models at that scale.
One might characterize emergent phenomena by whether the property of emergence is a consequence of our limited mental faculties. Such is not the case for cubes of pyrite, or for cyclonic weather patterns, or for biological systems themselves, which all exhibit strict (for their size) boundaries and continuities; or even for Newtonian physics (thought of as a limiting case of quantum mechanics, a mathematical distinction as strict as any). Such is (based on the above) the case for agency, as it is for color (understood as distinguishable perceptions of different frequencies and intensities of light) and such evolving combinatorics as large networks and even jeu de taquin; such may even be the case for consciousness (in the sense of a brain having sufficient capacity to experience itself, which blurs the distinction by interpreting subjectivity itself as a physical phenomenon).
Such a characterization, however, exposes a vulnerability: If emergence is subjective (ignoring the previous parenthetical remark), how can any framework built upon it, e.g. (so i assert) morality, be objective? My answer is that there is no need to require that morality be any more objective than agency, so the objection is a non-starter. Indeed, without agency—again, regardless of choice or free will—there cannot be overt acts, which are the default template on which moral frameworks are founded, and (more importantly) what essentially everyone considers interpretable through whatever moral philosophy they happen to have adopted. This point is therefore not circular; it rests with generally accepted elements of morality, which as such must either be incorporated into any meaningful theory of morality or justified away by reason of (presumably generally unrecognized) conflict with some still more essential element(s), which i assert cannot be given (see next paragraph). On the other hand, if agency is granted, then (provided other essential elements are granted) morality might yet be objectively described.
So much for (1). Since i’ve already rambled quite a bit, i’ll save (2) for next time (which will still be asequentially before the “next time” alluded to in the previous post).