Most systems that evaluate anything — health interventions, product quality, workplace performance, life satisfaction — use some form of weighted average. Take each factor, assign it a weight, multiply, sum. It's intuitive, simple, and wrong.
It's wrong because it assumes the factors are independent. That improving sleep has no effect on how much benefit you get from exercise. That nutritional adequacy doesn't change the impact of cognitive training. That attachment security is unrelated to stress regulation capacity.
The evidence says otherwise. Overwhelmingly.
What we measured
Across 11 evidence-grounded frameworks, we identified and characterized 63 pairwise interactions between dimensions. Each interaction has a mechanistic basis — a cited, biological or psychological reason why two dimensions are non-independent. Each has an interaction index between −1 (complete redundancy) and +1 (complete synergy).
| Interaction Type | Count | Mean Strength |
|---|---|---|
| Synergistic (both must be strong) | 46 (80%) | +0.30 |
| Redundant (one strong suffices) | 11 (20%) | −0.19 |
The ratio is 4:1. For every case where two dimensions partially substitute for each other, there are four cases where they amplify each other. Human needs aren't a checklist where you tick items independently. They're a network where getting one thing right makes getting other things right more valuable.
What synergy means, concretely
A synergistic interaction means: the combined effect of two dimensions exceeds the sum of their individual effects. This isn't metaphor. It's a measurable property of how human systems actually work.
The strongest synergies in the tree
| Interaction | Index | Mechanistic basis |
|---|---|---|
| Attachment security × Stress regulation Relationship Quality |
+0.50 | Secure attachment provides the neurobiological infrastructure for stress regulation — co-regulation with a trusted partner literally reduces amygdala activation (Coan's fMRI studies). Without attachment security, stress regulation relies entirely on self-regulation, which is more metabolically costly and less effective. Without stress regulation capacity, attachment security cannot maintain itself under pressure. Each enables the other. |
| Aerobic base × Neurobiological response Physical Activity |
+0.45 | The aerobic base IS the stimulus; the neurobiological response IS the adaptation. Erickson 2011 directly demonstrated the causal chain: aerobic training → BDNF increase → hippocampal volume increase → spatial memory improvement (r = 0.23–0.29). The exercise without the neurobiological response is movement without benefit. The neurobiological response without the exercise stimulus doesn't occur. |
| Sleep duration × Circadian alignment Sleep Optimization |
+0.45 | Leproult 2014: circadian misalignment DOUBLED metabolic disruption beyond sleep restriction alone. Eight hours of sleep at the wrong circadian phase produces measurably worse outcomes than eight hours at the right phase. Duration without alignment, or alignment without duration, both leave significant benefit on the table. |
| Vagal tone × Autonomic flexibility Autonomic Regulation |
+0.40 | High resting vagal tone without the flexibility to modulate it under challenge provides baseline calm but not adaptive capacity. Flexibility without vagal tone provides range without a healthy set point. The autonomic system requires both a strong parasympathetic baseline AND the capacity to shift dynamically. |
| Attention regulation × Emotional regulation Contemplative Practice |
+0.40 | You cannot regulate what you cannot attend to. Attention regulation (the capacity to notice and direct attention) is the prerequisite for emotional regulation (the capacity to work skillfully with what attention reveals). Contemplative traditions have known this for millennia — concentration (samādhi) precedes insight (paññā). |
What redundancy means, concretely
A redundant interaction means: two dimensions partially substitute for each other. If one is strong, the marginal value of the other decreases. This is less common (only 20% of interactions) but important for efficiency — it tells you where investing in both produces diminishing returns.
| Interaction | Index | Why they're partially redundant |
|---|---|---|
| Whole-diet pattern × Anti-inflammatory nutrition Nutritional Foundations |
−0.25 | Mediterranean and MIND diets already include the major anti-inflammatory components (omega-3 from fish, polyphenols from fruits and vegetables, fiber from whole grains). Targeted anti-inflammatory supplementation adds less value when whole-diet quality is already high. The whole-diet approach subsumes the targeted approach. |
| Breathwork × HRV biofeedback Autonomic Regulation |
−0.25 | Both target the same autonomic pathway — vagal tone enhancement through controlled respiratory patterns. If you're already practicing structured breathwork, HRV biofeedback provides measurement but not a fundamentally different stimulus. The physiological mechanism overlaps. |
| ACT therapy × Digital self-guided programs Psychological Flexibility |
−0.25 | Both deliver the same core intervention (Acceptance and Commitment Therapy principles). The therapist-delivered version is more effective but the content is substantially overlapping. Doing both provides some incremental benefit but less than the sum of their individual effects. |
Why weighted averages fail
A weighted average assumes every dimension contributes independently — that the value of improving sleep is the same regardless of your nutritional status, exercise level, or trauma history. The interaction data shows this is wrong in 80% of cases.
Consider a concrete example. A person who sleeps well but has chronic inflammation benefits less from sleep optimization than a person who sleeps well AND has managed their inflammation — because sleep and inflammation interact synergistically. Each 1-hour decrease in habitual sleep duration corresponds to an 8.1% increase in CRP. Sleeping well in the context of controlled inflammation means the sleep is doing more because the inflammatory cascade isn't degrading the very systems that sleep is trying to restore.
A weighted average gives this person the same "sleep score" regardless of their inflammatory status. The Choquet integral — the non-additive scoring function we use — captures the interaction. When both dimensions are strong, the combined score exceeds the weighted sum. When one is weak, it drags down the effective contribution of the other.
The 4:1 ratio is consistent across all branches
One might expect synergy patterns to cluster in certain domains — perhaps relationships are synergistic but nutrition is independent. The data shows otherwise:
| Branch | Synergies | Redundancies | Ratio |
|---|---|---|---|
| Sleep Optimization | 4 | 1 | 4:1 |
| Inflammatory Management | 4 | 1 | 4:1 |
| Autonomic Regulation | 4 | 1 | 4:1 |
| Cognitive Foundation | 4 | 1 | 4:1 |
| Trauma Recovery | 4 | 1 | 4:1 |
| Nutritional Foundations | 5 | 1 | 5:1 |
| Physical Activity | 5 | 1 | 5:1 |
| Relationship Quality | 4 | 1 | 4:1 |
| Psychological Flexibility | 4 | 1 | 4:1 |
| Contemplative Practice | 4 | 1 | 4:1 |
| Purpose Cultivation | 4 | 1 | 4:1 |
Every branch shows the same pattern: approximately 4 synergies for every 1 redundancy. The consistency is striking. It suggests this ratio isn't an artifact of any particular domain — it's a structural property of how human needs relate to each other. The dimensions of flourishing are overwhelmingly cooperative. They amplify each other far more often than they substitute for each other.
What this means for how we evaluate anything
If human needs are 80% synergistic, then any system that evaluates human wellbeing, recommends interventions, or scores products using independent weighted dimensions is systematically misrepresenting reality. It will undervalue interventions that address multiple synergistic dimensions (like exercise, which touches 6 branches) and overvalue interventions that score high on a single dimension while ignoring its interactions.
The practical implication: don't optimize dimensions in isolation. The evidence says that getting three related dimensions to "good" produces more flourishing than getting one to "excellent" and ignoring the others — because the synergistic interactions between "good" dimensions compound, while an excellent score on an isolated dimension has no interactions to amplify it.
This is why the knowledge tree uses the Choquet integral instead of a weighted average. Not because it's more sophisticated, but because it's more honest. It represents what the evidence actually shows about how human needs interact. And what the evidence shows is: synergy, not addition.