Freud in Plato – The restraint of appetite as the beginning of rationality

These reflections are based on discussions in “Comparative Literature 227: Drive, Instinct, and Will in Philosophy and Literature” led by Dr. Lia Kleinbock.

Argument Reconstruction of Republic, Reeve ed., 122-129.

1. The ideal city has three components (rational, spirited, appetitive) responsible for distinct functions.
2. These components must derive from people within it, whose intentions are thus (probably) also made up of these components.
   1. This is not rigorous and vulnerable to at least two good counterarguments, and is better treated as a prediction.
3. We are interested in fundamental components. A component is fundamental if it cannot do one thing and its opposite at the same time. (Analogy of the top in 436d3)
   1. I believe this is definitional statement rather than an inference predicated on any prior statement, and is the premise on which the actual argument unfolds.
4. Sometimes a person wants to do an action and its opposite at the same time.
5. Hence, there must be at least two fundamental components within humans.
6. Apparently, one of these components is the rationally calculating one and the other is the appetitive one. (439c9)
   1. Plato provides no explanation as to why the two fundamental parts must be these specific parts, simply asserting this by channeling the power of Zeus.
7. Spirited impulses like anger do not have much to do with rationality, but also often come into conflict with appetite, as in the example of Leontius. (439e6)
8. Hence these spirited impulses constitute a third component to human motivations.
9. In fact, anger often partners reason to resist the appetite, exactly like how the auxiliary partners with the rulers to control the producers.
10. Hence, the components of the city is indeed reflected in the components of the human soul.

Does the restriction of appetite require rationality?

In Plato’s motivation for the existence of second part of a soul (beyond appetite) is an idea which some have argued encapsulates the central contention of Freud–that rationality begins with the resistance of appetite.

This definition of rationality is compelling to me because the ability to restrain one’s appetite seems to require the ability to make abstract generalizations, which is a crucial element of–if not the definition of–rationality.

This is because the appetite is a source of fulfillment that functions almost solely in the present and its particularities. Thus, from the perspective of the present moment, it is always optimal to pursue appetite. To not pursue appetite, then, implies that the agent is optimizing something greater than the utility of the present moment. One alternative objective it may instead be pursuing is the cumulative utility across a stretch of time. In order to do this, the agent most likely has the ability to hold some concept in its mind which abstracts and aggregates representations of utility across these moments in time. If this is indeed the path through which the agent spurns appetite, then abstractive capability would be a necessary condition of such resistance.

There are several cases where this may not be the path through which the agent restrains their appetitive urges. For instance, the agent may not be optimizing for any goal whatsoever. Instead, it may be functioning deontologically. Or, even if it is pursuing the cumulative utility across a stretch of time, perhaps it does not need to create an abstract representation which coheres the utility of each moment, and instead simply memorizes the exact, particularized representations of each particular moment, and attempts to optimize their sum.

Both of these seem quite unreasonable to me–the latter for compression reasons and the prior because the stochasticity of the types of agents I’m interested in (humans, AIs) seems to defy a rules-based account.

More importantly, in both cases, the agent still requires a cognitive or computational component that the pure-appetite agent does not: memory. A purely appetitive agent can be stateless–at each moment in time, it looks around, decides which direction best fulfills this appetite, and moves that way–but both the deontological agent and the one that memorizes representations of utility must persist something beyond the experience of the present moment.

This memory, which allows the agent to escape the appetite of the moment, mustn’t be the flavor of memory as in computer architecture, whereby a block of information is stored and retrieved deterministically and isolatedly. It could also be stored in a more fuzzy, compressed way, superposed and entangled with other pieces of information as in a neural net or a human mind.

The distinction between these is best demonstrated through an example. Suppose that I am faced with the choice of consuming foie gras (which I loved as a child) and decide to pass due to the cruelty of the farming practices associated with producing the food. My appetite draws me to it, but what Plato would call my ‘rational and calculating component’ resists. This, per previous analysis, relies on some sort of memory which extends beyond the present context. In this case, this ‘memory’ is the idea that the foie gras I see before me was extracted from some miserable duck or goose, and some broader understanding that my current refusal decreases the demand for this food item as well as a more general aversion towards being part of something that feels slightly evil even if I am powerless to change the institution as a whole.

Perhaps these memories of ideas derive from a hodgepodge of documentaries, articles, conversations, and childhood experiences. If my memory functioned in the flavor of computer architecture memory, I would have to retrieve each documentary, article, conversation, and experience, analyze them in the present moment, and come to a conclusion. Instead, it seems that these pieces of media form some messy constellation of cognitive patterns which activate in concert as I perceive the foie gras and decide not to buy it. This latter case can be quite well represented by a neural network, which is stateless/memoryless in the computer architecture sense, but still “remembers” in effect the content of its training data–although in this complex and entangled fashion.

Regardless of what manifestation of memory was the enabling factor behind my resistance of my appetite, the point remains that in order to resist it, I needed some understanding of the world which extended beyond the present moment. In a sense, this understanding–whether it is rote memory or a more compressed, abstracted representation of knowledge–is what we call rationality.

The statement that “the resistance of natural urges is where rationality begins” is morally loaded, because rationality is usually thought of as a good thing, so this statement seems to encourage self-repression, which Freud would say has led to a civilizational march towards misery. It is also intellectual confused, because at some point, the rational judgements I make become part of my natural urges–after some number of years, I simply stop wanting foie gras, and in this sense, what was only at first rational understanding has merged into my intuitive feeling. This create interesting questions about to what extent natural urges and rationality form a mutually exclusive dichotomy.

I will observe, however, that Plato’s tripartite framework itself can explain this change in natural changes while maintaining a clear split between appetite and rationality (the body and the mind?). Since “natural urges” are partitioned into the spirited and the appetitive, a change in urges can be explained by a change located uniquely in the prior. Plato may say that yes, our instinctive urges can change, but this is merely the triumph of a spirit-rationality alliance over appetite, not a fundamental transformation of the latter; the appetite, seething beneath, is never altered in direction, but in magnitude only.