The modern physics (quantum observation) is galvanizing me to make the assertion that causality itself is not particularly useful to us– even if it is true and important– in a health intervention setting, but instead an emphasis should be on working with correlations and creating causality.
It matters less how smiling causes happiness than the fact that it does!
Consider a quantum information perspective: Physicists provide support for retrocausal quantum theory, in which the future influences the past
In my view, a coherent field description of activity and matter conceptualizes time as more of an ocean with waves and tides than a river - the whole system is available to itself always since it is itself, but perspective means localizing of possible observation and 'causality'. The wave seems to 'happen' but it returns to a coherence that it never left, and that happening is a pattern ongoing. Each wave is unique, but is never disconnected from the substrate of happening, the ocean. So the ocean is itself and is its state always, and its future even, but each wave must express time still and know only its unique state. This is an explanation that would allow 'intuition' to exist - a linear time with causes would mean reality is cut off from itself, which I have not found compelling evidence for.
Maxwell's demon doesn't work on its deepest level because discrete variables don't exist in natural systems, discrete variables are ideas natural systems (us) have about ourSelf yet these individual ideas are not the whole truth. Even in a tamed chaos box like a computer, still the system is beyond the full accounting of the design and designers.
That old adage about correlation ≠ causation is true, but without context of the role of observers this statement loses meaning.
If a correlation drives someone to do an action, has correlation not 'caused' an action?
The truth contains our illusions, and illusions have historically been powerful causes of 'truths' in a socio-technical setting at least, so bringing perspective into science and studies of causality seems our only hope to actually arrive at useful practices and methods.
Now of course, causality helps us mitigate problems and knowledge is power - finding causal links can also let us skip over extra processes to get at the root of effects. So it is essential to try and identify causality.
However I can't help but have a trickster twinkle in my heart about how our observations cannot be separated from our entire perspective.
We observers and researchers aren't neutral, we are coherent with the systems we study!
So the gentleman that thinks smiling is silly because the world sucks, will his smiling ever be as strong as someone who thinks the world is wonderful? His life and his studies will likely reflect his opinions, and so the only way to really 'control' for the full set of variables is to understand the perspective of oneself. Until this happens at scale, all studies will risk being reflections of local truth, rather than the full substrate of truth which contains all contradictions - waves on the ocean, rather than the ocean itself with its waves.