Continuing with the themes from my previous essay discussing motivation and functional autonomy, I will start by expounding upon my claim that this principle is a factor in addiction. Next, my goal is to provide a selective overview of pertinent topics related to reinforcement learning and perception before concluding with neuroscience and psychiatry topics that I will likely continue to investigate in subsequent writings.
To state simply, the principle of functional autonomy espoused by Gordan Allport (1937) is significant in that it demonstrates that not only does a complete set of instincts or inherent drives not exist but, in fact, initial motives for behaviors can be severed and replaced with novel ones that function to maintain motivation to conduct the same behavior. Therefore, there exists an expansive, diverse, and emergent set of potential motives that drive a person’s behavior at a specific point in time.
Since I first was introduced to the concept of functional autonomy years ago while completing my undergraduate degree, the connection between this principle and addiction has remained in my memory primarily due to an illustration my professor provided to explain functional autonomy. While I do not remember the specific class (or professor to give them credit), I recall the general gist of the example because of how directly it corresponded with functional autonomy.
Imagine there is a young person in high school or college who has never smoked a cigarette and is detested by the smell, in addition to being frightened of the health risks. However, they desperately want to socialize with a group of peers and realize that smoking cigarettes is a means to achieve this end. Let us assume this plan is successful, as this hypothetical peer group has only smoking as their shared interest.
At this point, this fictional person likely feels a sense of reward from attaining their goal of a sense of belonging; however, as time passes by, this person’s initial motive of smoking for social inclusion is replaced by the rewarding pleasure of smoking itself. It is here, at this point, that their initial means (smoking cigarettes) to the end (social inclusion) has been transformed into an end in itself. Consequently, this new end goal of feeling satisfaction from smoking is capable of functioning independently of the initial end goal of social inclusion—in fact, it may even be that smoking cigarettes is hindering social inclusion with desired groups, but this person now is driven by a motive that is exerting a more significant influence relative to this other potential desire.
With this hypothetical case study in mind, let us broaden our view to examine how the brain’s general processes function to navigate daily life.
Our brains constantly receive and process information from our external environment via perception and our internal environment via interoception (or internal (Chen et al., 2021). For this paper, I will remain on the surface of these deep topics, but I hope to write about them in further detail in the future.
Throughout our days, these input data are being filtered, sorted, and processed for relevancy, particularly as it relates to the future. Our brains generate models based on past experiences and data from our current sensory inputs to create representational models that account for our current state of being and futuristic thinking. We conduct mental cost-benefit analyses for our future behaviors, taking into consideration the effort and energy required to perform some action or series of actions to obtain a given reward outcome; moreover, our brain seeks to minimize the energetic cost while maximizing the reward of our selected outcome (Peters, McEwen, & Friston, 2017).
These are topics rich in-depth that I hope to return to in the future. Still, this essay will focus on the more surface-level takeaways: Our brains process troves of information, and our values, prior experiences, future goals, and availability of environmental rewards all function to assist in sorting and filtering the stream of raw sensory data. This is by no means an exhaustive account of the cognitive process of perception, reinforcement learning, and decision-making.
However, embedded with this brief overview are core assumptions of human motivation that are applicable and well-documented for the general population; however, an additional inquiry is necessary for individuals who possess serious psychiatric disorders, particularly schizophrenia (though similarities are found in other psychiatric disorders).
For example, central to the mental processes I have highlighted for perception and predictive planning is the ability to create mental representations. Closely related (if not integral) to this process is reward valuation; that is, associating varying degrees of reward value with different reward stimuli, typically relying on past experiences and personal values to facilitate this determination (Der-Avakian et al., 2016).
The ability to formulate goals is indeed so significant that machine learning researchers Richard Sutton and Andrew Barto write in their 2018 book Reinforcement Learning: An Introduction:
“A learning agent must be able to sense the state of its environment to some extent and must be able to take actions that affect the state. The agent also must have a goal or goals relating to the state of the environment…Rewards are basically given directly by the environment, but values must be estimated and re-estimated from the sequences of observations an agent makes over its entire lifetime.” (pp. 5-6)
However, aberrant reward learning, dysregulated goal-directed behaviors as a result of inappropriate attribution, and the inability to accurately discriminate relevant stimuli from those irrelevant are precisely central areas of deficit identified by those researching the mechanisms of schizophrenia. In fact, researchers Der-Avakian et al. reported, “pleasure and valuation have been dissociated in schizophrenia, with patients showing intact capacity to experience pleasure, but deficits in properly representing the value of future rewards” (2016, p. 237).
Additional research findings further elucidate the implications of these findings and provide other intriguing points that extend the scope to incorporate other psychiatric disorders, such as mood and development disorders. The most intriguing aspect for me is discovering the common themes that exist across diagnoses. I believe this indicates overlaps in the underlying mechanisms responsible for the symptomatology of these different categories of disorders.
With this in mind, the theme that has become most apparent as a significant diagnostic challenge is further understanding how goals are constructed, what factors are involved in attributing and experiencing reward values as outcomes of goal-directed behaviors, and how does one’s time orientation affect their creation and implementation of futuristic goals. These are topics that I will seek to (or attempt to) unravel in my subsequent writings.
Allport, G. W. (1937). The Functional Autonomy of Motives. The American Journal of Psychology, 50, pp. 141-156.
Barto A.G., & Sutton R.S. (2018). Reinforcement Learning: An introduction (adaptive computation and machine learning series) (2nd ed.). The MIT Press: Cambridge.
Chen W., Schloesser D., Arensdorf A., Horowitz T., Vallejo Y., & Langevin H. (2021). The Emerging Science of Interoception: Sensing, Integrating, Interpreting, and Regulating Signals within the Self. Review Special Issue: The Neuroscience of Interoception. 44(1) 3-16. https://doi.org/10.1016/j.tins.2020.10.007
Der-Avakian A., Barnes S., Markou A., & Pizzagalli D. (2016) Translational assessment of reward and motivational deficits in psychiatric disorders. In: Robbins T.W., Sahakian B.J. (eds) Translational Neuropsychopharmacology. Current Topics in Behavioral Neurosciences. https://doi.org/10.1007/7854_2015_5004
Peters A., McEwen B.S., & Friston K. (2017). Uncertainty and stress: Why it causes diseases and how it is mastered by the brain. Progress in Neurobiology. Volume 156, pp.164-188. https://doi.org/10.1016/j.pneurobio.2017.05.004