Axioms
Preface
This dialogue is an analysis of Models1, by Pranav Balakrishnan. Therein, Balakrishnan offers a critique of model-based adjudication of the competitive norm. This critique is correct in its argumentation and power, but incorrect in its scope. Balakrishnan provides an account of the impossibility of models that only truly applies to three systems of adjudicating competitive norms:
Debatability
Contingent Predictability
Weak Categorical Predictability
Here, we import the same assumptions as Debatalist No. 12, but one additional argument will be imported later from Notes No. 1 on Debatalist No. 13.
Propositions
The same axioms as Debatalist No. 1 are imported here.
The Text
Analyzing Models
Balakrishnan begins by introducing an example of his delineation of ‘models-based’ and ‘in-round’ impacts.
The two framework impacts articulated most often in debate are clash and fairness. Clash operates in a model-based framework: if we debated under a single interpretation, which would produce better education? While fairness can operate under this same models-based approach, it’s more often articulated as in-round offense: something unfair was done in this round, and only the ballot can remedy it.
Balakrishnan neglects an opportunity to mention that this in-round approach can be analogized to pure utilitarianism—that is, ‘something bad was done that can be fixed,’ whereas the model approach can be analogized to rule utilitarianism—that is, ‘something bad violates a rule that on net produces more good than bad.’
There are two forms of this model argument:
‘Voting against violations of our rule is intrinsically good since the rule would be a good thing.’
There may be a more compelling articulation of this viewpoint, but proposition-exchangers seldom find it. This viewpoint is attractive because it attends to a near-universal intuition—even if the behavior in this round has no concrete impact, it would be bad if everybody acted that way. However, accounting for an intuition isn’t sufficient to constitute a non-arbitrary system.
‘Voting against violations of our rule is instrumentally good since it deters bad actions and brings about the model.’
Balakrishnan identifies the correct critique of this form:
Regardless of how a judge votes in a single debate round, it is impossible for them to actually enforce or guarantee that critical affirmatives are not read in the future. Even a hundred rounds would barely shift the needle.
While the general instrumental argument may be loosely true, it is easily outweighed by other ambient behaviors or impacts.
Let us return to the point about remedying a loss of fairness:
Something unfair was done in this round, and only the ballot can remedy it.
It may be said that the only thing the ballot can remedy is a loss of fairness. The judge voting negative to reject settler-colonial epistemologies does not remedy the fact that those epistemologies were propagated, but voting affirmative to correct an unfair burden being placed upon the affirmative in that round does remedy the terminal to unfairness—the ballot being submitted one way or the other.
However, it is not solvency for fairness that is the problem; it is the impact itself. This will be discussed later.
For now, Balakrishnan identifies three forms of models:
What is a model? Even though the term is thrown around in many debates, it lacks a clear definition that describes it. Here are a few, representing a “model” where affirmatives must be topical:
Type one:
Judges universally vote against untopical affirmatives, but they can be read.
Type two:
No untopical affirmative is read, and teams somehow know that they cannot read untopical affirmatives.
Type three:
No untopical affirmative is read, and teams don’t know that they cannot read untopical affirmatives, but are somehow “fiated” to read untopical affirmatives even though they don’t know if they are forced to.
It should be noted that The Debatalist Papers defends a Type One understanding of the competitive norm (and believes that the only possible competitive norm is Type One).
Balakrishnan begins this section with a critique of Type One modeling:
“Type one” models (where judges universally vote against untopical affirmatives) cannot solve the clash impact to framework. The impact is solely about whether the arguments read in debate are refutable. If a “model” for debate allows teams to read critical affirmatives at all, then it cannot access the clash impact.
This is true. Type One models intercede at the ballot’s level, and can reap no benefits of imagining a ‘better round’ (unless the Type One model defends an instrumentalist understanding of modeling, which was dispelled above).
He continues:
These types of models also mitigate the framework team’s predictability impact in the long run. In the world of the counter-interpretation (still a type-one model, where judges vote against teams that do not read an affirmative that fits within the counter-interpretation), policy teams would suddenly start losing every debate. Eventually, they would realize that they must read counterinterp-adjacent affirmatives in order to win, which would eventually result in debate being exclusively over the counter-interpretation.
There are two errors. The first is the general argument from categorical predictability:
The long-run generation of ‘predictability’ in the instantiation of the counterinterpretation is actually the generation of contingent predictability, which is not a preferable metric. This is true because the shift of judge preferences to the counterinterpretation is not categorically predictable (it is not derivable from a proposition presented by the administrator). No. 1 has already condemned the possibility of contingent predictability.
However, if we grant that the above is somehow resolved, Balakrishnan’s argument may appear threatening for The Debatalist Papers, which purports to defend categorical predictability. However, this is only a threat to Weak Categorical Predictability. There is also Strong Categorical Predictability (that, perhaps, can hardly be called ‘predictability’). The difference is as follows:
Weak Categorical Predictability: There is a certain type of predictability value brought about by in-round conduct called categorical predictability, and that categorical predictability value ought to be maximized.
Strong Categorical Predictability: It is functionally irrational for debaters to not adhere to competitive norms that are categorically ‘predictable.’
Both ideas share the name ‘categorical predictability’ because they analyze debaters as a category and what competitive norms can be rationally derived from what those debaters concede.
However, where Weak Categorical Predictability cares about the categorical derivability of competitive norms because they are universally predictable, Strong Categorical Predictability cares about them because they are universally binding.
This will be significant soon. However, Balakrishnan has more arguments to make:
“Type two” models clearly mitigate the predictability impact. Teams know (maybe by NSDA decree, tournament rules, or something similar) that they must read an untopical affirmative. That makes the reading of critical affirmatives predictable for both teams, because they know that every team at the tournament must be reading one.
Here, he identifies the flaw of default debatability as The Debatalist Papers did in No. 1. It errantly assumes that teams are aware of the competitive norm, which isn’t correct.
Then Balakrishnan continues to critique the Type Three Model:
“Type three” models are strange and difficult to explain. They almost feel like an abuse of fiat, creating a model in which teams universally agree to an interpretation without knowing that they have. It’s akin to “mindset fiat” or “object fiat” as described in other debates.
However, this seems like the only way for the framework team to access the full scope of the predictability impact. Under the counter-interpretation, teams would be just as surprised to see a critical affirmative as they are under the “affirmatives must be topical” interpretation, even though every single team would read a critical affirmative.
There are also clear disadvantages to thinking about debate in this manner. It could never be actualized, warps reality, and feels as though it has been exclusively manufactured to preserve the impact to predictability.
This is exactly correct.
Balakrishnan concludes with a final comment:
There is also usually little discussion of what a model should include, either leaving a judge making guesses or simply overlooking this seemingly very relevant issue.
This is a correct analysis of debate’s state.
A Path Forward
Balakrishnan’s critique is powerful evidence for categorical predictability.
Condemning models and suggesting that the ballot can only remedy in-round unfairness is untenable. It fails No. 1’s critique of contingent predictability; here is how:
The only impact the ballot can resolve is unfairness occurring in this round.
Predictability matters insofar as an argument not being predicted hurts the fairness of its introduction.
One can email the team their argument a week before the tournament and make it predictable, which resolves that unfairness offense. Now it becomes a matter of pure limits.
However, Strong Categorical Predictability insulates itself against this approach in a stronger fashion. Instead, Balakrishnan’s critique is a reason that debatability cannot account for the arguments of categorical predictability. To review:
The administrator lays down certain propositions with implied or explicit imperatives that define its debate format.
Debaters elect to debate that format.
Intentionally violating those imperatives would ‘bastardize’ the debate, making it not an example of that format.
Thus, debaters who intentionally violate the imperative of a format at once intend to debate that format and prevent their debate from being an example of that format. This is a contradiction in intentions.
A contradiction in intentions is a formal error; it means one’s conduct violates ethical obligation in either intent.
It is irrational to come to debate and refuse the terms presented for the tournament. This is the true source of the competitive norm—it denies the ballot to proposition-exchangers who violate a term of the format, because insofar as we are debaters (and concede that we intend to engage with debate), that violation can never be good.
The ballot does not remedy in-round fairness loss; it safeguards the integrity of this particular round as a member of the format. Thus, ‘reject racist violence’ is a valid procedural even if it does not harm fairness, because the format explicitly excludes racist violence. The question of a bright line is entirely separate.
These terms are ‘models’ because they purport to concern more than just this round’s abuse, and are not remedied by particularities (say, Berkeley Preparatory RV’s experience with kritikal affirmative skewing the debate in their favor when they are read). Even if one’s opponent was prepared and could fairly debate, the rules are the rules.
Balakrishnan critiques rule-utilitarian models, but fails to consider deontic models like this. The deontic model for competitive norms survives.
(Pranav Balakrishnan, 9-24-2024, "Models," Bolt from the Blue, https://boltfromtheblue.substack.com/p/models, accessed 9-24-2024)
(Chinmay Khaladkar, Adam Humphrey, & Adhi Thirumala, 8-21-2024, “Debatalist No. 1 - General Constructive,” The Debatalist Papers, https://substack.com/home/post/p-147730271, accessed 9-24-2024)
(Adam Humphrey, Chinmay Khaladkar, & Adhi Thirumala, 8-24-2024, “Notes No. 1 on Debatalist No. 1 - Typology of Propositions,” The Debatalist Papers, https://debatalistpapers.substack.com/p/notes-no-1-on-debatalist-no-1-typology, accessed 9-24-2024)