Eristics as a formal scientific theory

For those of you who asked "is this really science?", here is my answer

Symbol of the eristics personality test. ©2024 Rex Riepe

Author’s Note: from this point forward, I will be elaborating on eristics theory independently of its founder, Rex Riepe, although we are in communication. Rex is in no way responsible for any mistakes I make in describing his theory.

I may also propose theoretical innovations that Rex disagrees with. While I do not intend to “fork” eristic theory, it may sometimes be convenient to distinguish between our visions by talking about “normie” eristics versus “Rex”/standard eristics.

I would like to thank everyone who engaged with my first post on eristic theory and how I believe it has changed my life. The strongest theme in the critical responses was the claim that eristic theory is not really science. In some sense I agree — Rex hasn’t published any formal experimental results, and I myself only began engaging with the theory meaningfully in the past couple weeks. That said, I think eristics is on a track to becoming real science (or being falsified!) and in this post I’m going to lay out the roadmap.

What is a scientific theory?

A scientific theory should have three components:

1. Observation: a true observation about the real world.

2. Hypothesis: a theorized root cause of the observation, with predictions that follow from the root cause. Ideally, these predictions would be independent of the initial observation, so that the theory isn’t merely explaining the data it was developed with (evoking the train-test split in machine learning).

3. Experiment: an act of collection of data from the real world, with a method to evaluate alignment of the predictions to the experimental data.

Now I’m going to explain how eristic theory can be fit into this template.

Observation

Humans systematically vary in their dispositions to emotion, thought, and action. Some people have a disposition to be relatively more guilty, proud, fearful, daring, angry, peaceful, etc.

Hypothesis

Eristic theory

1. Emotions are evolutionary adaptations which motivate us to adapt our behavior to our context. Variations in human emotions are thus parametrized by the two dimensions of context and motivation.¹

2. There are three general contexts: self, (non-human) world, and society/tribe.

3. There are two general motivations: internalize and externalize.

   1. How to think about motivations is a complex topic in eristic theory. It is definitely context-dependent, and there is probably not a single best way. Important internalize/externalize dualities include add/remove (for the self), model/modify (for the world), and contribute/drive (for tribe).

4. Context-motivation space can be partitioned by crossing the three contexts with the two motivations, resulting in six general emotional “arenas” where contextualized motivation to behavior occurs. Rex applies traditional emotional labels to the six arenas to guide intuition about how they show up in our feelings.

   1. Self (includes “extended self”, the family)

      1. Love: add to the self.

      2. Disgust: remove from the self.

   2. (Non-Human) World

      1. Fear: model the world.

      2. Anger: modify the world.

   3. Society (or Tribe)

      1. Guilt: contribute to the tribe.

      2. Pride: drive good behavior within the tribe.

5. The six arenas can also be thought of as eristic arguments — imperative sentences that our evolved psyches say to us to persuade us to do things. Just as basic arguments can be chained together to reach complex conclusions, basic eristics can be chained together to reach a complex motivation.

6. The simplest chaining schema, the beat, involves two halfbeat arguments, and has the form “do A by means of B”. For example, making a beat from Guilt and Disgust might yield “contribute to society by means of removing pollution and unworthy elements from yourself,” which Rex labels as Envy.

7. People are too complex to be usefully bucketed by basic eristic arguments, but they do tend to adopt a single beat more than any other, and this beat characterizes their personality.

8. The basic eristics are ordered according to their “energy” as Love < Fear < Guilt < Disgust < Anger < Pride, yielding the acronym LFGDAP. The bottom three are considered to have broadly “low” energy expenditure, while the top three are considered “high” energy expenditure. (To emphasize the low-high grouping, Rex writes the acronym as lfgDAP in his code behind the web-based eristics test.)

9. There are 6 * 6 = 36 possible beats, but certain beats cannot be archetypal because they are impractical as bases for a person's general personality orientation:

   1. Reverse-order and double beats: for example, the Pride » Love and Disgust » Disgust beats can occur, but they’re too high-energy to be a sustainable basis for a personality. This condition removes 21 beats and leaves 15.

   2. Beats from only one context: a beat made of opposite motivations in the same context doesn’t make sense as a person’s typical thought pattern. It would be like, “do X by means of the opposite of X”. Perhaps not impossible, but pretty counterintuitive. Three same-context beats are removed by this condition.

   3. Excessively high-energy beats: Double-high pairs seem to be too energetically demanding to show up much as personalities in real people, so DA, DP, and AP are removed from the menu of possible beats underlying archetypes. (These last 3 pairs do exist as eristic beats, possibly Dark Triad-associated, but aren't really a sustainable basis for an entire personality.)

10. Like the six basic eristics, the nine beats can be labeled using imperfect but useful emotional mnemonics, and these are the characteristic motivations of the nine archetypes.

    1. Love » Fear = Attachment (Artist)

    2. Love » Guilt = Devotion (Giver)

    3. Love » Anger = Frustration (Hero)

    4. Love » Pride = Satisfaction (Defender)

    5. Fear » Guilt = Anxiety (Scientist)

    6. Fear » Disgust = Duress (Dragon Slayer)

    7. Fear » Pride = Revelation (Architect)

    8. Guilt » Disgust = Envy (Fixer)

    9. Guilt » Anger = Remorse (Wizard)

    Measuring archetypes in humans

1. Each context contains a motivation duality, and in each context, a given person tends to choose one side more than the other side. We can measure the inclination using eristic dilemmas — plausible real-world scenarios in which the context is fixed and we must choose one side of the associated duality.

2. If set up a test containing dozens of eristic dilemmas with binary² response options aligned to the presented duality, invite participants to select a response to each question, and tabulate the results, we get a histogram with six bins. We rank the bins by frequency and construct a beat from the first and second-ranked result. This is the person’s archetype.

3. More generally, a person’s relative alignment to each archetype can be calculated as f(pA, pB), where pA and pB are the relative frequencies for the halfbeats A and B composing the archetype’s beat, and f is monotonically increasing in each argument. Currently, I suspect the product, f(x, y) = xy, is the “best” simple alignment measure, as it comports with a specific probabilistic model (random IID model of beats). One could even develop a distribution of archetypes for a given person by taking the softmax of the calculated alignments to each archetype.

4. I also suspect that more advanced tests will pose more complex dilemmas that directly measure beat alignment, and perhaps even allow the fitting of Markov models describing sequential dependencies between a person’s halfbeats or beats.

Prediction

1. Self-evaluated alignment with assigned archetype: After taking the eristics test, a person will “see themselves” more in the archetype assigned by eristic theory than a random archetype, as measured by a 5 or 7-point Likert scale.

2. Prompted autobiography alignment with assigned archetype: Suppose a person writes a prompted autobiography and also takes the eristics test. I predict that third-party human raters (or perhaps even LLMs) will tend to find the assigned archetype the best fit to the person’s life story.

Experiment

1. Self-evaluated alignment: have, say, 100 people take the eristics test. Split them into Treatment and Control groups of 50 each. Treatment receives the real archetype, Control receives a randomly selected fake archetype. Ask each participant how much they see themselves in the archetype they received on a 1-5 scale.

2. Prompted autobiography: have 100 people write a prompted autobiography, then take the eristics test, perhaps a couple weeks later to avoid contamination. I predict that third-party human raters (or perhaps even LLMs) will tend to find the assigned archetype the best fit to the person’s life story.

Potential applications

• Curing spiritual autoimmune disorder in Fixers by moderating the inner critic

• TBD — watch this space!

1. I currently speculate that there may be a need for an additional dimension which distinguishes between the “inner world” of the mental and the “outer world” of the physical. Thus, there might be an inner self (mental model of the self) and outer self (physical body), an inner world (mental model of the world) and an outer world (physical world), and an inner society (mental models of other people) and outer society (physical other people). This dimension seems like it is needed to properly formulate complex motivational patterns and archetypes, as we will see later.

2. In the real eristics test, not all questions have binary responses. I am not sure why this is. It may be that some questions are not restricted to one context, or perhaps they even measure beats directly rather than half-beats.