Entropy, Relevance and Meaning, Part 2
Goal conflict and introduction to relevance realization. How can we begin to solve the frame problem?
This is Part 2 of this series. You can read Part 1 here.
After having defined psychological entropy and identified the source of uncertainty-related anxiety, we made clear how certain explanatory frameworks (e.g. goals, narratives etc.) can reduce stress and provide meaning to an ever-changing landscape. In this part we are going to explore these ideas in more depth.
Goal conflict, goal hierarchies and combinatorial explosion
The concept of goal setting is as old as consciousness itself. Human beings recognize that the most efficient way to accomplish complex goals is by fragmenting them into smaller, more manageable subgoals. An accurate representation of the environment and adequate knowledge of the path forward is crucial in effectively minimizing entropy. Goal pursuit thus becomes a well-defined, masterfully refined process that reduces stress and furthers the individual's plans to reach a desired state.
Problems arise when goals are not effectively pursued or outside parameters undermine the individual's selected path towards accomplishing said goals. It is more common than not for people to set ill-defined goals with no clear outcome, or set a number of goals that lead to the desired state while simultaneously engaging in behaviour that sets the progress to zero.
Procrastination is a great example of how one prioritizes wasting time instead of pursuing high value goals. Although sometimes there are underlying issues that cause this behaviour, the overwhelming majority of procrastinators simply can't stop. This actively interferes with the pursuits of the individual and thus, when inevitably the future goals will not be reached, causes great amount of stress.
There is therefore a clear relation between the effectiveness of a goal in reducing entropy and the specificity and attainability of it. In general, conflicting goals can be understood in terms of the amount of energy a person consumes in order to pursue them. Since energy and entropy are tightly correlated, pursuing goal A (or more broadly pursuing a specific outcome that impedes the attainment of another) can sometimes be of detrimental effect on the possibility of accomplishing goal B (Hirsh et al., 2012):
For example, a person may wish to attain career success but also spend time with his or her family. In such a situation, the choice to work late to finish a project directly conflicts with arriving home in time for the family dinner. The highest levels of entropy and metabolic waste will exist when the goal itself is not well specified or in the case when a previously held goal is abandoned and has not yet been replaced by an alternative goal.
Neurophysiologically, Hirsh et al. (2012) and McNaughton and Gray (2024), point out that high entropy states activate the behavioural mechanisms that are responsible for dealing with equipotential affordances. This makes sense since there is no clear path forward and the emotions one experiences are that of fear and anxiety; similar to a fight-or-flight situation. This is the essence of the goal-conflict problem. Certain constraints have to be imposed - perhaps a rigid hierarchy of goals based on importance - since a complete lack thereof seemingly leads to existential angst.
The resolution of such conflict involves careful examination of the situation to determine the optimal response. What this suggests is that situations with the fewest constraints can be the most anxiety producing as a consequence of their inherent uncertainty (reflecting the large number of possible interpretive frames and response options).
Interestingly, the anxiety characterizing highly uncertain situations appears similar to the experience of angst described by existential philosophers, who argued that unconstrained behavioral freedom can lead to a state of despair and insecurity.
Goal hierarchies are thus a means by which we manage to sort out present and future uncertainties1 and chart a clear path forward. The low level of a goal hierarchy is inhabited by high-resolution, high-precision goals that are subordinate to more abstract major goals. For example a daily work schedule could be a subordinate goal to a promotion, a specified diet subordinate to a gain in muscle mass or weight loss, a daily writing schedule subordinate to writing a book, and the list goes on. Failure at a low level produces the kind of uncertainty that can be constructively manipulated so the failed goal can be changed to serve the higher one more efficiently. This also explains why we tolerate uncertainty at such a low level, even welcome it. It makes us more perceptive concerning the failings of our approach and facilitates the optimization of current habits or the adoption of new ones.
A great example is the identity crisis (Hirsh et al., 2012) that ensues when the individual feels dissatisfied at work and decides to quit (suddenly exposing himself or herself to a high level of uncertainty) in order to create a better future (eventually reducing overall long-term uncertainty). The uncertainty and stress generated at the low level is by no means unimportant. The person has to look for a new job which is a tiresome and nerve-racking procedure. But in the face of the overarching goal of being a loving spouse, a good parent and creating a bright future of oneself and one's family, it is practically insignificant. The dissatisfaction that surrounds an unpleasant daily life (in this case an unpleasant job) can gradually cause mood swings, changes in behaviour and psychotic breakdowns making life utterly miserable for the people closest to us.
There are cases, however, that the consequences are far greater. In the event of a disruption at the highest strata of the hierarchy, the individual is exposed into unprecedented levels of uncertainty and unmanageable anxiety (Hirsh et al., 2012).
A tragic example would be the death of a spouse. In the event of such a devastating happenstance most levels of the goal hierarchy are in danger of crumbling since minor goals have likely been revolving around the existence of a stable family structure.
Under circumstances of sufficiently severe and potentially traumatic uncertainty, as is likely to emerge when the highest levels of a goal hierarchy are destabilized, the individual can no longer clearly determine the significance of any given object, action, or experience; all of these must be understood and constrained in relation to a particular goal or reference point.
The multitude of existing pathways produces overwhelming confusion as the individual naturally lacks the mental capacity to consider all possible outcomes.
This known as combinatorial explosion and applies to most real world problems as they are usually ill-defined and require careful consideration and planning in order to be untangled and dealt with.
A practical example is the following (Vervaeke & Ferraro, 2013):
In a typical game of chess one can make thirty legal moves and one takes about sixty turns. So the number of alternative pathways to checkmate is thirty to the power of sixty or 4.239 x 10^88. Compare that to the number of neurons in the brain (estimated to be 1010), or even the number of synaptic connections (approximately 5 x 10^14). In fact, the number of atoms in the universe is 10^82. So even the massively parallel nature of the brain is not sufficient for searching the entire search space using a brute force, exhaustive strategy; the size of the search space is just too vast.
If this as an approximation of what the human brain tries to compute in the aftermath of a traumatic experience, how then can we learn to best navigate the world where such things will inevitably occur?
In order to provide a solution for this problem, we first need to take a look into the manner in which a cognitive agent (i.e. the human brain) zeros in on relevant information.
In simple terms, how do we decide what’s important to us?
The frame problem and how cognitive agents zero in on relevant information
The core of the frame problem is the question of how do we zero in on relevant information while effectively ignoring the rest of the world. The world is infinite in complexity, but we somehow manage to isolate specific information that suit our goals and facilitate our survivability.
There are two existing frameworks in cognitive science that explain how organisms achieve this: Predictive Processing and Relevance Realization. In a recent publication, Andersen et al. (2022), explain the differences and ultimate convergence of these two frameworks into solving the frame problem.
Predictive Processing can be explained as follows:
Predictive processing (PP) is an emerging framework in cognitive science which depicts the mind as being engaged in a proactive process of prediction-error minimization. This framework suggests that we separate signal from noise by forming expectations about the 'precision' (technically the inverse variance) of incoming prediction errors and assigning a higher 'weight' to prediction errors which are treated as highly precise.
Relevance Realization is described in the following manner:
Relevance realization (RR) is another emerging framework in cognitive science which takes a dynamical systems perspective to explain how the mind separates signal from noise. The RR framework suggests that the brain achieves this feat by attempting to balance the competing goals of remaining efficient in the current environment while also being resilient in the face of environmental perturbations. By self-organizing around a set of opponent processing relationships related to this efficiency-resiliency tradeoff, the brain is able to zero in on relevant information.
What the authors of this publication showed is that the precision-weighting which characterizes PP, is actually the means by which PP accounts for RR. This is an amazing discovery since two different frameworks ultimately point to the same underlying process.
More specifically:
Connecting these frameworks should serve to enrich the PP story by describing the tradeoffs that influence how an organism’s perceptual system assigns precision to select sensory inputs and top-down predictions. On the other hand, by connecting RR to precision-weighting, RR can be integrated into a large and growing framework which has been put forward as a way to unify the cognitive and psychological sciences.
Relevance realization is of great interest in our discussion of goals, hierarchies of goals and goal conflict. The world consists of an infinite amount of problems and, accordingly, every choice has an infinite amount of consequences. Nevertheless, through the process of realizing relevance, our cognition manages to isolate noise from signal, distractions from purpose.
In the next part we will focus more explicitly on this process and discuss the research published and the ideas presented by many experts on the field of cognitive science.
If you enjoy my work you can support me with a one-time donation by clicking below. Much appreciated!
