This idea is an outgrowth of the "Skepticism vs. Cynicism" thread, in which we attempted to clarify the difference between the two. I first floated this idea in this post, and it seemed to resonate with some readers. This also follows my previous efforts to treat Conspiracy Theories in the abstract, such as my Inflationary Theory and its limiting case.
Introduction
In evaluating alternate theories, and conspiracy theories in particular, one often desires to distinguish whether a theory is plausible, if unlikely, or whether a theory is absurd. This distinction is important to a skeptic, because a skeptic should always respond logically and rationally to a plausible theory, even if it contradicts previous theories and especially if it contradicts poorly supported belief and conjecture. An absurd theory, on the other hand, may and in many cases should be simply dismissed. No amount of logical analysis can salvage an inherently illogical proposition.
Unfortunately, the question of how to identify an absurd theory remains open. The problem of evaluation is in many respects parallel to determining the burden of proof, in the sense that the more plausible theories have a lower burden of proof, or conversely that an argument partially or wholly satisfying its burden of proof is inherently more plausible. This consideration, however, does not give us any actionable formula, since burden of proof is also subjective. While some attempts to quantify burden of proof exist, primarily in the field of trial law, this standard is not directly applicable to philosophy or science.
In the absence of a clear standard, perceptions of bias are inevitable, as effort is spent debunking some theories that are entirely without merit, whereas other competing ideas may be judged and rejected prematurely. We will attempt to rectify this situation through a general process.
Description
Sorting the plausible from the absurd is inherently a subjective process, and no truly abstract quantification is possible. However, we can look to methods of estimation present in the engineering sciences for inspiration. In particular, I will draw upon the process of Probabilistic Risk Assessment, or PRA.
Risk management through PRA is an attempt to quantify individual risks in a given process, project, or product according to likelihood, potential impact, and possible mitigation. In a PRA exercise, these quantities are often gathered from the strictly subjective opinions of experts, supported with more focused studies or historical statistics where possible and needed, and propagates these quantities in order to arrive at an aggregate risk estimate. This process is similar to ours in that the inputs are inherently subjective, a large number of individual factors affect an overall conclusion, and we desire a quantfied result with enough fidelity to make a decision.
Using this as a guide, we will model a skeptic evaluating a Conspiracy Theory as a simple engineering process, with characteristics of the Theory and the ensuing discussion as contributors or mitigations to overall risk. The risk itself in this case is the risk that a Conspiracy Theory is in fact absurd, leading to an ultimately fruitless discussion with no grounding in reality.
Method
All quantities in the following discussion are subjective. Users of this method may adjust the numerical parameters as they see fit.
Baseline
We begin our approach by assessing the type of argument offered. This is broken into three categories. We assign a numerical score according to the best classification of the argument, hereafter referred to as the plausibility score, as follows:
The three categories are separated by a factor of 0.5 -- one step from Complete Theory to Partial Theory, and another step from Partial Theory to Anomaly or Artifact. This multiplicative approach and valuation is consistent with the following steps, as we adjust the plausibility score according to specifics of the argument.
Adjustments
We will model adjustments to the plausibility score as either positive or negative, and either major or minor, as follows:
Each incidence of an adjustment should be noted and classified. One by one, each adjustment has a multiplicative effect on the cumulative plausibility score as follows:
Following the method above, the higher quality an argument has, the higher its plausibility score will be. The score will change over time, as discussion either reveals or corrects flaws in the argument, and new evidence permits a more accurate assessment. This may cause a previously plausible argument to be revealed as absurd, and in rare cases, an absurd theory may be salvaged. In like fashion, a Partial Theory without flaws has an equivalent score to a Complete Theory with one major flaw, etc.
Decision Threshold
Much as individual project managers have different tolerances for risk, individual posters may have different tolerances for a hopeless argument. The author recommends a threshold value of 0.05, signifying an approximate confidence level of 95%. Any argument with a plausibility score below this value may be said to be "absurd," and unworthy of further contemplation, until it is significantly improved. This process also captures why the argument fails the test, and further provides an estimate as to how much reworking is needed for it to become a useful point of discussion.
This threshold is referred to as the "Gravy Line." In sporting terms, this is reminiscent of the "Mendoza Line," a threshold of futility below which performance can be considered not up to professional standards.
Why the "Gravy Line?" The plausibility threshold resembles a "water mark" in that it attempts to classify a wide range of arguments against a single limit. Gravy, like water, seeks its own level, but unlike water is opaque, and also viscous, with arguments falling below the Line unlikely to ever resurface or to be seen again. Arguments that fail this test can be said to have been debunked by Gravy.
Examples
We close with two brief case studies to illustrate the process.
1. With Pakistani Intelligence serving as a link, the CIA supported the September 11th attacks. (thread)
2. Analysis of television news footage proves that, instead of jetliners hitting the World Trade Center as was widely reported, the event was hoaxed using digital video editing techniques and special effects. (thread)
Disclaimers
All effort in this whitepaper was conducted using my own materials on my own time. All opinions are strictly my own. I do not represent any agency, public or private. The method outlined above is only the barest approximation of a true PRA approach. Use at your own risk. The discussion above contains no export-controlled technology or schematics. Enjoy.
Introduction
In evaluating alternate theories, and conspiracy theories in particular, one often desires to distinguish whether a theory is plausible, if unlikely, or whether a theory is absurd. This distinction is important to a skeptic, because a skeptic should always respond logically and rationally to a plausible theory, even if it contradicts previous theories and especially if it contradicts poorly supported belief and conjecture. An absurd theory, on the other hand, may and in many cases should be simply dismissed. No amount of logical analysis can salvage an inherently illogical proposition.
Unfortunately, the question of how to identify an absurd theory remains open. The problem of evaluation is in many respects parallel to determining the burden of proof, in the sense that the more plausible theories have a lower burden of proof, or conversely that an argument partially or wholly satisfying its burden of proof is inherently more plausible. This consideration, however, does not give us any actionable formula, since burden of proof is also subjective. While some attempts to quantify burden of proof exist, primarily in the field of trial law, this standard is not directly applicable to philosophy or science.
In the absence of a clear standard, perceptions of bias are inevitable, as effort is spent debunking some theories that are entirely without merit, whereas other competing ideas may be judged and rejected prematurely. We will attempt to rectify this situation through a general process.
Description
Sorting the plausible from the absurd is inherently a subjective process, and no truly abstract quantification is possible. However, we can look to methods of estimation present in the engineering sciences for inspiration. In particular, I will draw upon the process of Probabilistic Risk Assessment, or PRA.
Risk management through PRA is an attempt to quantify individual risks in a given process, project, or product according to likelihood, potential impact, and possible mitigation. In a PRA exercise, these quantities are often gathered from the strictly subjective opinions of experts, supported with more focused studies or historical statistics where possible and needed, and propagates these quantities in order to arrive at an aggregate risk estimate. This process is similar to ours in that the inputs are inherently subjective, a large number of individual factors affect an overall conclusion, and we desire a quantfied result with enough fidelity to make a decision.
Using this as a guide, we will model a skeptic evaluating a Conspiracy Theory as a simple engineering process, with characteristics of the Theory and the ensuing discussion as contributors or mitigations to overall risk. The risk itself in this case is the risk that a Conspiracy Theory is in fact absurd, leading to an ultimately fruitless discussion with no grounding in reality.
Method
All quantities in the following discussion are subjective. Users of this method may adjust the numerical parameters as they see fit.
Baseline
We begin our approach by assessing the type of argument offered. This is broken into three categories. We assign a numerical score according to the best classification of the argument, hereafter referred to as the plausibility score, as follows:
1. Complete Theory (Score = 1.0): An argument that is complete and self-consistent, allowing examination of a single coherent hypothesis.
2. Partial Theory (Score = 0.5): An argument that is generally complete and self-consistent, but vague in at least one major detail, requiring examination of an entire family of related hypotheses.
3. Anomaly or Artifact (Score = 0.25): An argument that criticizes an existing theory while proposing no competing hypothesis of its own, such that if the argument is accepted, uncertainty actually increases.
2. Partial Theory (Score = 0.5): An argument that is generally complete and self-consistent, but vague in at least one major detail, requiring examination of an entire family of related hypotheses.
3. Anomaly or Artifact (Score = 0.25): An argument that criticizes an existing theory while proposing no competing hypothesis of its own, such that if the argument is accepted, uncertainty actually increases.
The three categories are separated by a factor of 0.5 -- one step from Complete Theory to Partial Theory, and another step from Partial Theory to Anomaly or Artifact. This multiplicative approach and valuation is consistent with the following steps, as we adjust the plausibility score according to specifics of the argument.
Adjustments
We will model adjustments to the plausibility score as either positive or negative, and either major or minor, as follows:
1. Major positives: Attributes or events that significantly improve the quality, plausibility, and verifiability of the argument. Examples include presentation of repeatable and verifiable calculations, predictions that are made and confirmed, and adaptation of the argument to new evidence that strengthens the argument through discussion with non-advocates.
2. Minor positives: Attributes or events that have a secondary or circumstantial improvement on the plausibility of the argument. Examples include peer review or concurrence by relevant independent experts, testable (but so far untested) predictions, and verifiable anecdotes or sub-scale experiments that support the argument.
3. Minor negatives: Attributes or events that circumstantially cast doubt or add uncertainty to the argument. This includes dependence on assumptions that are plausible but unsupported, vagaries in the argument that leave possible alternative explanations, logical fallacies peripheral to the main argument, criticism by relevant independent experts, and verifiable counterexamples that are typical but not definitive.
4. Major negatives: Attributes or events that substantially diminish the argument's credibility. These include dependence on unsupported and exceptional assumptions, logical fallacies in the main argument, and a failure to correct the argument for serious errors found during discussion and analysis.
2. Minor positives: Attributes or events that have a secondary or circumstantial improvement on the plausibility of the argument. Examples include peer review or concurrence by relevant independent experts, testable (but so far untested) predictions, and verifiable anecdotes or sub-scale experiments that support the argument.
3. Minor negatives: Attributes or events that circumstantially cast doubt or add uncertainty to the argument. This includes dependence on assumptions that are plausible but unsupported, vagaries in the argument that leave possible alternative explanations, logical fallacies peripheral to the main argument, criticism by relevant independent experts, and verifiable counterexamples that are typical but not definitive.
4. Major negatives: Attributes or events that substantially diminish the argument's credibility. These include dependence on unsupported and exceptional assumptions, logical fallacies in the main argument, and a failure to correct the argument for serious errors found during discussion and analysis.
Each incidence of an adjustment should be noted and classified. One by one, each adjustment has a multiplicative effect on the cumulative plausibility score as follows:
Major postive: Multiply by 2 (i.e. divide by 0.5)
Minor positive: Multiply by 1.25 (i.e. divide by 0.8)
Minor negative: Multiply by 0.8
Major negative: Multiply by 0.5
Minor positive: Multiply by 1.25 (i.e. divide by 0.8)
Minor negative: Multiply by 0.8
Major negative: Multiply by 0.5
Following the method above, the higher quality an argument has, the higher its plausibility score will be. The score will change over time, as discussion either reveals or corrects flaws in the argument, and new evidence permits a more accurate assessment. This may cause a previously plausible argument to be revealed as absurd, and in rare cases, an absurd theory may be salvaged. In like fashion, a Partial Theory without flaws has an equivalent score to a Complete Theory with one major flaw, etc.
Decision Threshold
Much as individual project managers have different tolerances for risk, individual posters may have different tolerances for a hopeless argument. The author recommends a threshold value of 0.05, signifying an approximate confidence level of 95%. Any argument with a plausibility score below this value may be said to be "absurd," and unworthy of further contemplation, until it is significantly improved. This process also captures why the argument fails the test, and further provides an estimate as to how much reworking is needed for it to become a useful point of discussion.
This threshold is referred to as the "Gravy Line." In sporting terms, this is reminiscent of the "Mendoza Line," a threshold of futility below which performance can be considered not up to professional standards.
Why the "Gravy Line?" The plausibility threshold resembles a "water mark" in that it attempts to classify a wide range of arguments against a single limit. Gravy, like water, seeks its own level, but unlike water is opaque, and also viscous, with arguments falling below the Line unlikely to ever resurface or to be seen again. Arguments that fail this test can be said to have been debunked by Gravy.
Examples
We close with two brief case studies to illustrate the process.
1. With Pakistani Intelligence serving as a link, the CIA supported the September 11th attacks. (thread)
Given that the premise accepts the other features of the attacks as reported, only adding this important detail, this can be viewed as a complete theory. Its initial plausibility score is 1.0.
From reading the thread, there is only one potential positive adjustment, that being the single unnamed source cited by CNN and others, claiming a $100,000 money transfer. We will choose to accept this as a minor positive adjustment.
There are, however, several negative adjustments:
While a low score, this is significantly above the Gravy Line, and therefore this topic is potentially worthy of further discussion. Even though the discussion so far indicates that it is false, this is not an "absurd" theory. This theory breaks no laws of physics, demands no miracles, and is definite enough to support an actual investigation.
From reading the thread, there is only one potential positive adjustment, that being the single unnamed source cited by CNN and others, claiming a $100,000 money transfer. We will choose to accept this as a minor positive adjustment.
There are, however, several negative adjustments:
- $100,000 is surely not significant in the eyes of the CIA, thus the unsupported but plausible assumption that there was more trading hands
- Revealed through discussion, confusion over the identities of the supposed actors, leading to a vagary in the argument
- The entirely credible alternate explanation that certain individuals in the ISI might have had a selfish interest in supporting the Taliban and Al-Qaeda, or even the attacks themselves, without requiring CIA involvement at all
- The assumption that the CIA is, in fact, directing the ISI -- this is an unsupported and exceptional assumption
While a low score, this is significantly above the Gravy Line, and therefore this topic is potentially worthy of further discussion. Even though the discussion so far indicates that it is false, this is not an "absurd" theory. This theory breaks no laws of physics, demands no miracles, and is definite enough to support an actual investigation.
2. Analysis of television news footage proves that, instead of jetliners hitting the World Trade Center as was widely reported, the event was hoaxed using digital video editing techniques and special effects. (thread)
The theory presented is not complete, because it does not describe how the attacks really were carried out (and leaves no readily apparent candidate), who did this, or why. We must treat this as an Anomaly or Artifact, and assign an initial plausibility score of 0.25.
The evidence presented in support of this theory is disputable on its face, and the theory makes no predictions, testable or otherwise. Methods of calculation are ad hoc and nonstandard, permitting no easy review. There are no positive adjustments. There are, however, several negative adjustments:
This is fitting, as the argument presented above violates the laws of physics, ignores current technological limits, gainsays witnesses, and rankles common sense. Major work is required before anything derived from this theory should be given the floor.
The evidence presented in support of this theory is disputable on its face, and the theory makes no predictions, testable or otherwise. Methods of calculation are ad hoc and nonstandard, permitting no easy review. There are no positive adjustments. There are, however, several negative adjustments:
- The unsupported and exceptional assumption that video editing capabilities of this sophistication exist
- The assumption that all major news organizations are complicit in this plot
- The repeated refusal to address those who witnessed the impacts, rather than finding out about them through news organizations
- Repeated arguments from personal incredulity regarding the phenomenology of the collapses that eventually followed
This is fitting, as the argument presented above violates the laws of physics, ignores current technological limits, gainsays witnesses, and rankles common sense. Major work is required before anything derived from this theory should be given the floor.
Disclaimers
All effort in this whitepaper was conducted using my own materials on my own time. All opinions are strictly my own. I do not represent any agency, public or private. The method outlined above is only the barest approximation of a true PRA approach. Use at your own risk. The discussion above contains no export-controlled technology or schematics. Enjoy.
But that "Mackey Mark" would actually have the names of both of you in it.
