Working note. Laws, norms, and institutions are treated as mediated outputs of interacting welfare-bearing agents and as imperfect evidence about social evaluation—not as direct products of a social-welfare maximizer or as the source of welfare itself.
Society is not a single agent directly optimizing a Social Welfare Function. It consists of many VMS-mediated agents interacting within a shared State of the World.
For each person, that state already includes:
other agents
existing laws and norms
institutions and social structures
language, education, and media
welfare ideals, moral theories, and political principles
distributions of power and resources
These variables enter individual state-transitions and are altered by the outputs of those transitions. The basic dynamic is recursive:
State of the World
→ many VMS-mediated agents
→ action, expression, and non-expression
→ social interaction and discourse
→ institutional translation
→ norms, laws, and institutions
→ new State of the World
→ repeat
Other agents are part of each agent's world. Laws, norms, and institutions also become inputs into later VMS processes. They can change behavior, opportunity, manifestation, self-understanding, and sometimes person-level evaluative structure.
Discourse transmits welfare-relevant information.
Discourse connects individual evaluation with social outcomes:
individual evaluation
→ social expression
→ collective and institutional outcome
Voting, lobbying, protest, public discussion, academic research, journalism, legal argument, and policy proposals are different transmission routes. They allow welfare-relevant information to enter the social system, but none is a transparent channel.
At least two layers of mediation intervene.
The first is between welfare or individual evaluation and political or social expression:
welfare / evaluation
→ VMS mediation
→ expression or non-expression
Possible distortions include disability, cognitive limitation, learned behavior, socialization, fear, unequal confidence, unequal resources, and unequal ability to communicate, organize, or gain attention.
The second is between expression and institutional output:
expression
→ institutional mediation
→ law, norm, or collective outcome
Possible distortions include power asymmetry, political structure, representation failure, agenda-setting, media access, historical lock-in, and path dependence.
Equal welfare stakes therefore do not automatically become equal political force. Equal expression does not automatically receive equal institutional weight.
Law and norms are signals, not grounds.
Norms and laws can be read as accumulated outputs of earlier discourse, historical records of social evaluation, partial reflections of welfare concerns, products of conflict, and workable solutions retained through institutional learning.
They contain information. But the information is mediated and may preserve distortion as readily as discovery.
signal ≠ ground
manifestation ≠ source
stability ≠ correctness
A stable law shows a relatively stable configuration of operative social and political forces. It does not by itself show welfare maximization, utility maximization, or moral correctness.
Social development can be a discovery process.
Societies sometimes move toward better welfare-relevant outcomes because discourse reveals previously hidden harms, excluded agents become audible, institutions accumulate information, coordination improves, and systematic transmission failures become visible.
This is not guaranteed optimization. It is a possible approximation process. The same dynamics can also amplify error, stabilize unequal power, or make a distorted proxy appear authoritative.
The distinction is therefore:
social change may improve access to evaluative information
social change does not create correctness merely by occurring
as a hypothesized integrated evaluative structure.
I use for the human- or social-level integrated meta-ethical structure-point. and are proposed as the same geometric kind of object at different levels: each carries complex evaluative information, one at the person level and one at the integrated human level.
is not an agent. It is not a vote, a consensus, an average personality, a discourse equilibrium, a set of rules, or merely an aggregate count of current states. Its internal information may appear socially as welfare, dignity, inclusion, respect, rights, justice, or social utility. These are different manifestations or projections of an integrated evaluative structure rather than independent foundations.
This is a proposal, not a completed account of social aggregation. The unresolved question is what makes the structure integrated and how person-level evaluative information relates to it without reducing it to a simple sum, current preference, or political outcome.
enters human life through mediation.
does not act directly. It appears, approximately and sometimes distortively, through social variables:
norms and laws
education and language
institutional rules
social expectations
public discourse
precedents and policy categories
rights and harm language
These variables enter individual VMS processes. They may alter states, behavior, self-interpretation, and sometimes person-level evaluative structure. But they are not identical to .
M* → mediated social manifestations
manifestations → individual VMS transitions
observed output → partial evidence about M*
The distinction allows law and norms to be both causally powerful and epistemically imperfect.
Interaction does not imply homogenization.
Convergence is one possible directional effect under particular conditions, not a general property of social interaction. The same social variable can change different selves in different directions; alter behavior without altering person-level evaluation; change state or manifestation while leaving Self relatively stable; or have little effect.
is not a personality endpoint toward which every must move. A shared evaluative structure can include a region of permitted difference. If inclusiveness is internal to that structure, individual variation within that region is not a deviation demanding convergence.
Even outside such a region, convergence is not assured. Some selves may have low plasticity relative to a variable, atypical mediation routes, deep divergence, or histories that produce opposing transitions. Conflict, crime, and persistent difference therefore remain possible.
The claim is only that social states can be evaluated against an integrated welfare-relevant structure, not that human agents will become identical.
Policy as approximation under partial access.
If fully accessible, would generate a preference ordering among policy options. In practice, policymakers work through partial knowledge and mediated proxies:
law and precedent
empirical truth and scientific models
institutional practice
money and economic indicators
life expectancy and risk indexes
harm categories and rights language
surveys and stated preferences
expert judgment, philosophical reasoning, and public reason
These are not final grounds. A proxy is good when it captures the relevant variables and their relative weights well enough to approximate the ordering that adequate social evaluation would generate.
Some policy decisions appear easy because a high-weight variable is legible, differs clearly between options, and is already well represented by law, evidence, or precedent.
Hard cases arise when:
no known high-weight variable decides the case
the relevant variables do not differ clearly across options
the weights are uncertain
available indexes misrepresent the welfare structure
proxies are heavily mediated or illegible
law supplies constraints but not a final ordering
Hard cases require explicit investigation: which variables matter, what their relative weights might be, which proxies are misleading, and what ordering the underlying evaluative structure would generate. Philosophy, science, empirical inquiry, institutional reasoning, and criticism of proxies all contribute.
Law may constrain the permissible option set:
But law rarely supplies the complete welfare ordering in a hard case. Policy cannot always be mechanically derived from existing rules because the rules themselves are mediated structures whose justification depends on welfare-bearing selves.
Human-bounded objectivity and continuing limits.
The proposed meta-ethical structure is objective relative to human ontological structure, not necessarily universal across every possible being. Human welfare and moral significance are constrained by human biology, cognition, sociality, and selfhood. A radically different kind of being might have a different integrated evaluative structure.
This does not make current human discourse constitutive of correctness. Discourse is epistemic and mediating: it can help discover, represent, or distort the structure; it does not create it merely by reaching agreement.
The central limitations remain:
Self boundaries are uncertain
welfare is imperfectly observable
VMS mediation is unavoidable
institutional mediation is unavoidable
the integration represented by M* is not yet explained
Direct observation of true welfare is not expected. The practical project is increasingly refined approximation while keeping representation distinct from reality.