The Whole Person and Whole-Person Computing

Why begin with the whole person?

A person has physiological processes, health conditions, possible diseases, care histories, rehabilitation goals, movement demands, cognitive and behavioral patterns, personal objectives, and environmental constraints. These are not independent compartments. The same diagnosis can coexist with very different functional capacities in two people. The same capacity can lead to different performance under different loads, levels of fatigue, or real-world conditions. One organ, one test result, or one domain-specific model may therefore be insufficient to explain what this person can do now, why a task is difficult, or what evidence should be gathered next.

Whole-person computing asks how local evidence can retain its disciplinary meaning while contributing to a continuous understanding of the same individual. It does not imply that one model already contains everything about a human being, or that psychological and social experience can be reduced to a score.

Human function as a common computational coordinate

The established plain-language explanation of human function is: “The body's ability to be appropriately engaged and used.” This English rendering supports the fixed Chinese wording, “身体被正常调用的能力。” It does not replace that original definition.

As an organizing principle for computation, human function allows us to ask: What must the organism sustain internally? What does a real-world task demand? What capacities are available to this individual? How are those capacities actually engaged? What performance, cost, and feedback follow? Physiological mechanisms, disease information, movement observations, environmental conditions, and professional interventions can then be related through explicit functional questions rather than merely connected by similar words.

A central research question remains open: If human function is not used as the common computational coordinate, what alternative organizing principle can connect multiscale life processes, real-world tasks, personal capacity, actual engagement, medical intervention, and longitudinal change just as effectively? This is a claim to be compared and tested—not a declaration that alternatives are impossible.

Coordinating different domain models without replacing them

Within the HFWM research architecture, physiological models may characterize mechanisms and regulation; disease models may describe pathology, progression, clinical constraints, and risks; health models may represent broader health states and change; anatomical and movement models may describe structures and task performance; and cognitive, behavioral, and environmental models may characterize choice and context. These models need not share an algorithm or be collapsed into one monolithic simulator.

The proposed shared layer concerns how they refer to the same individual, time interval, need, and functional task; how each distinguishes observations, inferences, candidate explanations, and interventions; and how conflicts or missing evidence are preserved. Broad multiscale model integration remains a research and engineering direction, with actual integration subject to implementation and independent verification.

A co-evolving computational counterpart of the person

As a long-term vision, HFWM may be understood as a second computable self: an authorized, evolving computational counterpart that relates health and disease, bodily and cognitive capacities, real-world tasks, action, and feedback over time. This is a metaphor for a persistent model of an individual—not a copy of consciousness, a replacement for the person, or an already completed digital human.

A person changes over time

Sleep, illness, pain, treatment, training, behavior, and environment can alter the conditions under which capacity is engaged. A single assessment describes a moment, not a permanent truth. Longitudinal computing therefore needs events, measurements, professional decisions, actions, and feedback organized in time, with provenance, versions, corrections, and uncertainty.

This is not simply a larger medical record, nor a system that permanently treats every AI-generated guess as fact. The aim is a personal state model that can explain why its understanding changed as new evidence emerged.

A practical example

Two people report pain when climbing stairs. The symptom alone does not identify a disease, a particular anatomical structure, an insufficient capacity, an ineffective movement strategy, or a contextual constraint. A defensible assessment considers clinical history, the task and its demands, functional capacity, engagement during activity, observed consequences, environmental conditions, and appropriate tests. HFWM studies how to organize these observations, formulate competing explanations, and identify the next useful evidence. It must not infer an exact muscle, nerve, or joint diagnosis from pain or movement appearance alone.

The relationship between UOHF, HFWM, intelligence, and HFOS

UOHF provides shared terminology, semantic relations, and evidence constraints. HFWM continuously models the whole person and organizes information across domains and time. Human Function Intelligence investigates reasoning, judgment, memory, learning, and feedback under these constraints. The Human Function Operating System (HFOS) brings these ideas into professional workspaces, education, and service processes. A further research goal is a human-function interoperability protocol that different systems could implement and validate with appropriate authorization.

The objective is a scientifically testable way to understand one person more completely, continuously, and transparently.