Dropped Arguments Are Not True Arguments by Jacob Nails

Debaters often claim that uncontested arguments should be treated as 100% true for the purposes of the debate round. On more occasions than I can count, I have heard the negative assert that the impact of their disadvantage is now 100% likely to happen because the 1AR “dropped” it (failed to respond). This paradigm for evaluating concessions is rarely justified in round, but it wins ballots largely because it appeals to an axiom many judges already accept.

While the notion has intuitive appeal, I believe that paradigmatically treating dropped arguments as certainly true is misguided. Rarely will an analytic or evidenced argument prove its claim with near certainty. What amount of evidence would it take to convince a well-informed person that living wage would collapse the economy for certain? Hopefully far more than a single study or expert prediction. Yet, even a single card on the matter – when uncontested by the opponent – is enough for most judges to treat the claim as guaranteed, even when no rational person would give the argument the same strength outside of a debate context. Judges frequently treat dropped arguments (even very bad ones) as true. This is a mistake.

The fact that an argument was dropped should not give it any special status.

For those that do grant that dropped arguments are true arguments, it is not immediately clear where to draw the line for which claims in a round can be treated as definitely true. The simplest paradigm would be to treat an argument as given only if there are no responses, i.e. it is totally dropped. This paradigm leads to painfully arbitrary conclusions.

Imagine a negative debater whose sole response to an affirmative case was to group it and assert that no AC arguments were warranted. The five words “group the case; no warrant” (or other extremely weak arguments) should hardly play a large role in any round. However, if judges should only suspend disbelief and grant the claims 100% truth if they are uncontested, then the whole case is now significantly weaker. Strictly speaking, none of the arguments has been dropped entirely, so none of the claims is conferred the status of definite truth that comes from being dropped.

Arguments don’t start at 100% likelihood.

One possible modification which would avoid the above conclusion would be to hold that dropped arguments aren’t true because they are dropped per se but because arguments in general, dropped or otherwise, start out at 100% truth until diminished by the opponent. On this account, if the negative debater only offered an incredibly weak response, then the argument is almost no weaker than if it were dropped and remains at essentially 100%.

Granting arguments a presumption of absolute legitimacy until responded to would arbitrarily favor the first person to take a stance on a given claim.

Consider a round where one debater forwards a weak argument that economic decline would cause nuclear war, e.g. “countries will get mad and nuke their rivals in anger” (not far less warranted than some impact cards). The other debater rejoins with an even weaker reason why nuclear war will not occur, such as “all WMDs might simultaneously fail to launch from their silos.” This argument only reduces the threat of nuclear by an infinitesimal fraction of a percent, so the claim remains functionally unrefuted and thus roughly 100%.

What if the roles were reversed? One debater preemptively gives an extremely weak reason to think nuclear war won’t happen, and the other responds with a relatively weak justification for why it will. “Countries will get mad and nuke their rivals” only offers very minor evidence of one possibility for nuclear war, so it would at most only reduce the starting proposition that nuclear war won’t happen to 97-98% likelihood.

The judge cannot grant both of two contradictory claims the presumption of 100% certainty, and, given identical arguments, the judge’s evaluation should not fluctuate between near certainty of the impact and negligible risk depending on which of the two claims is treated as the proposition to be tested. While I view this as a reductio ad absurdum, I have seen some debaters uncritically argue that dropped AC arguments outweigh dropped NC arguments by virtue of being conceded first. I sincerely hope that the debaters in question do not actually hold this viewpoint. Any legitimate paradigm should not determine the weight to empirical evidence based on who talked first.

The obvious solution is to directly compare the strength of the warrants. The stronger warrant should have more weight. Neither should start at 100%.

This logic should extend to dropped arguments as well. (Carving out an exception for dropped arguments runs afoul of the same arbitrariness problems as before.) When a claim is dropped, the comparison is between the initial warrant for the claim and no warrant to the contrary. Having some evidence for a proposition and none against it is not sufficient to confirm the proposition with certainty. If the affirmative offers one empirical example of a time that economic collapse led to conflict and the negative fails to provide a counter-example, a reasonable person would not conclude that the next economic collapse will definitely cause a war. Why shouldn’t the judge evaluate the round as a reasonable person?

The modern “tabula rasa” paradigm perversely favors bad arguments.

I suspect that the most common rationale for accepting dropped arguments without scrutiny is judges’ desire to be “tabula rasa” (blank slate), i.e. to avoid bringing biases and preconceptions into their decisions. Giving an argument less than “full weight” for any reason (other than the responses of an opponent) would strike many as blatant judge intervention.

The issue with this view is the assumption that the full weight of every argument is 100%. Naively accepting every assertion with the semblance of a warrant to back it is not objectivity; it’s lazy judging. A claim with a number of strong examples supporting it should carry more weight than a claim backed by one weak example. A judging paradigm that treats either one as enough to establish complete certainty does a disservice to debaters making stronger arguments. It flips the incentives, so short arguments with extreme conclusions have more strategic value than developed arguments with reasoned conclusions. A truly objective tabula rasa paradigm would give the latter more sway in the decision, but most disadvantage impacts, theory standards, and philosophical frameworks in LD tend to reflect the former. Current debate norms do not favor objectively better arguments.

The problem is particularly stark in the case of statistical arguments. Suppose that one side asserts that economic collapses lead to war and supports the claim by citing a study showing that collapse is linked to military aggression 50% of the time. Suppose further that the other side drops this argument. Surely, the judge should not treat the prospect of war as 100%. To do so would be to go against the very argument being conceded (that the change of the next economic collapse causing war is something on the order of 50%). And if judges shouldn’t evaluate a statistical argument as 100% certain, then why should weaker arguments with more outlandish claims get a pass?

The baseline for evaluation should be the warrant, not the claim.

One last middle ground position that someone still subscribing to status quo norms of non-intervention might take would be to argue that dropped arguments should be taken at face value. Rather than deciding whether the warrants of a card justify the conclusion the author reaches, judges should accept the conclusion as sound. This view wouldn’t entail outlandish conclusions of 100% certainty because most authors make far more modest predictions.

If taken seriously, even this paradigm would substantially diverge from common practice. Take an average four card disadvantage. More than likely, each of those authors hedges their rhetoric with terms like “might” and “could” (often in multiple places per card), which don’t always make it into debaters’ underlining of the evidence. If such “could” claims appear four times in the disadvantage, and each “could” means something like 20-40% risk, then the combined risk of the disadvantage would start at somewhere between 2% and .2%. That’s only after pricing in the uncertainties admitted by the authors, not the opponent’s responses.

While that modification alone is already a far cry from the “100% risk” claims that currently beset advantage/disadvantage debates, I don’t believe it goes far enough. Evaluating an argument by the strength of the claim leaves in place the same perverse incentives for debaters to rush to bad arguments with hyperbolic conclusions. If an argument is as strong as the claim it makes, then the most strategic sources are the authors on the fringe of the literature base who make the most sweeping and radical claims. By contrast, the authors who acknowledge counter-arguments and draw restrained conclusions would have less strategic utility.

Minimizing bias does not mean treating all arguments as equally legitimate. A fair paradigm should give stronger weight to stronger arguments, so if judges striving to be “tabula rasa” find themselves consistently voting for bad arguments, this should be evidence that the dominant paradigm needs to be recalibrated.

In any case, I am skeptical that simply accepting dropped claims on face obviates the need for the judge to do “work” after the round. Authors rarely quantify their predictions precisely, and there is usually room for interpretation whether a journalist saying that a minimum wage hike “might” swing the next election intended to convey a 30% risk or a 3% risk. With analytic claims made by the debaters themselves, the task is even tougher. Each side will of course claim that their arguments prove near certainty in their conclusion, but how much is genuine conviction, and how much is strategic posturing?

In my experience, theory debates in particular are plagued by bad analytics which are falsely legitimized by “tabula rasa” judging paradigms. Debaters blitz out one-liners like “education comes first because schools won’t fund debate otherwise,” or “time skew outweighs ground skew because you need time to make arguments” in hopes that the opponent drops them. These are barely coherent thoughts, let alone well-warranted arguments. Even conceded, they should count only slightly (at best) in favor of their respective conclusions.

Most judges are reticent, and with good reason, to insert themselves and their opinions into the round, but reluctance to intervene has slid too far in the direction of outright laziness in judging, and many judges have abdicated all responsibility for evaluating arguments. This article leaves open-ended how to weigh argument quality fairly, but it definitely does not mean treating every warrant as 100% proof of its conclusion, and it does not mean taking every dropped claim at face value.


Jacob Nails debated for 4 years for Starr’s Mill high school in Georgia, graduating in 2012. As a debater, he won the Saint James School tournament and the Georgia state tournament, cleared at NFL nationals, and qualified to the TOC. As a first year policy debater in college, he won the JV Southeast district championship and cleared at every regular season tournament he attended with GSU. In his first year as a coach, he coached a sophomore to quarterfinals of the TOC.

  • Josh Roberts

    This entire discussion IMO points to the huge importance of judge paradigms for both students and coaches. The judge paradigm is the best way for judges to articulate how they evaluate arguments/the round and provide debaters with insights they can use to adapt to that judge. I agree in principle with the idea of truth > tech, especially when it comes to more critical arguments that question validity of representations or methodologies. But other judges obviously feel differently.

    Validity of an argument is very much a part of the argument itself. I agree with Jake’s point about dropped arguments. If someone makes an argument like “no warrant” and it’s “dropped” in the sense that nothing along the lines of “I have a warrant because…,” but a warrant is extended, that denies the validity of the no warrant argument. I am not sure how you can ignore the extension of a warrant and believe an argument that’s blatantly false.

    However, I feel like there’s a difference between analytic claims that make statements easily proven false and Dave’s example about vaccinations where, although opinion/science leans in one direction, it’s still a debatable question. Perhaps this is a nonstarter, but to me there seems to be a distinction. I suppose that would be considered and example of where I “draw the line” on truth over tech.

  • jnebel


    I agree that these are problems with the way judges treat dropped arguments. But before abandoning the principle, let’s see whether we can at least partially solve these problems by being really clear about what we mean by “dropped,” “argument,” and “true.”

    1. “Dropped.” Let’s say that an argument is dropped if and only if the opponent says nothing that contradicts any part of it. This includes the claim/conclusion of the argument. This definition explains why embedded clash is clash. On this definition, “group the case: no warrant” will go dropped only if the aff never presents her arguments as having warrants. It also means that the judge would never be in a position in which she has to treat two contradictory claims as true. In the unfortunate situation where debaters assert contradictory claims and make arguments without addressing their opponent’s, the judge simply has to make a judgment about which argument is stronger.

    2. “Argument.” I think most agree that an argument requires a warrant, so dropped claims (without warrants) needn’t be treated as true. But there is disagreement about what constitutes a sufficient warrant. One partial solution is to return a bit more explicitly to the Toulmin model, which is the basis for the CWI construction. The original Toulmin model requires a claim to be supported by a reason (also called “data,” “evidence,” or “grounds”), and the warrant is the explanation of why the reason supports the claim; the warrant may also require further “backing.” The warrant is often implicit in informal contexts, but let’s require an explicit warrant (in Toulmin’s original sense) for purposes of treating dropped arguments as true. On this definition, “Presume aff because neg time skew makes it harder for me to win” is not a sufficiently developed argument. One would need to explain why the time skew’s making it harder to win justifies aff presumption, and back up this explanation.

    We can make much the same point without the Toulmin model. The point is that many blippy arguments are enthymemes: they leave some premises implicit. In ordinary contexts, that’s fine, because the suppressed premises are usually shared assumptions, and it would be a waste of breath to repeat what everyone already thinks. But in debate, the primary way in which an assumption becomes shared is by being made explicit and then dropped. So we should require premises to be stated explicitly in order to count something as conceded if dropped.

    3. “True.” Suppose one drops an argument whose claims is that “X might happen.” As you rightly point out in the article, treating this claim as true doesn’t mean that the probability of X goes up to 100%. It just means that the probability of X is not 0%. Similarly, treating “X will probably happen” as true doesn’t mean assuming that X will happen; treating “X is the best explanation of Y” as true doesn’t mean assuming that X is true or even that if Y is true then X must be true.

    This might solve some but not all of the issues you mention. Here is another partial solution, which I’m less sure about. Suppose the aff offers some evidence to justify some claim, and that the evidence is conceded. (And suppose it is not power tagged, because then the judge needn’t treat the dropped card as justifying the claim.) One possibility is to treat the claim as true. But we could instead just treat the evidence as true, and treat the evidence as supporting the claim, but leaving the extent to which the evidence supports the claim unspecified. It might be very weak evidence for the claim, or it might be very strong evidence for the claim. You might leave this up to the judge. I’m not necessarily endorsing it, but just mentioning it as an intermediate option between two extremes.

    • Jacob Nails

      I’ll admit that I’m not as confident in the last point I make in the article about warrants vs claims — that a strongly worded but weakly warranted card should count for little — but I think that even accepting your #3 point would depart substantially from the way most rounds are judged now. Most disad evidence is of the “X might happen” variety, so even when dropped the full weight of a disadvantage that relies on a long string of “might”s should be relatively low (the point I make in the article about .2% risk).

      I don’t know if that’s a conclusion you want your model to draw, but I view it as an advantage. The main issue that started me thinking about dropped arguments was the prevalence of existential risk advantages and DAs in policy debate (and, to a lesser extent, in LD). When most rounds are decided on indefensibly specific nuclear war link chains, it suggests to me that the current dominant paradigm is severely miscalibrated. If these aren’t rationally the best arguments, why do they win most rounds between equally skilled debaters in front of competent judges?

      • jnebel

        Agreed that the more honest perspective of risk is an advantage and that X-risk scenarios are not the best arguments. Not sure that this fully explains the prevalence of X-risk scenarios. Don’t people claim that even the smallest finite probability of extinction comes first? If people are winning that (contentious) claim, then pointing out the much lower probability justified by their arguments would have little effect.

        • Jacob Nails

          I agree. I think that argument would make dueling nuke war debates look a lot more sensible, but I don’t think it explains the prevalence of extinction arguments, mostly because I see that argument far more often in LD, but extinction impacts are more common in policy.

          It also doesn’t explain why huge but non-existential risks, e.g. regional nuclear war, win rounds, and I would expect to see a lot more importance put on billions of deaths vs total extinction if that were the rationale being applied.

  • mcgin029

    One more thing – on the “debate context” point – I think you’re badly wrong on this. Other contexts are *not* the debate context. When an opponent concedes a point (say, for instance, an empirical point) the judge’s evaluation is not “I now believe this to be literally true,” but rather, “the best evidence in the debate round suggests the truth of this argument, so the argument is true, according to the best evidence.”

    Take the vaccines topic for a moment. When we debated this issue, relatively few teams carded “anti-vaxx” “studies,” mostly because those studies were of poor quality and it wasn’t hard to debunk them from the aff position.

    However, imagine a round where the aff, with a framework of “preserving health,” presents no empirical evidence to support vaccine efficacy, but rather assumes that vaccines work. And now, say the neg presents a study (however poor) that says that vaccines cause autism, irritable bowel syndrome, and spontaneous combustion. And finally say that the aff concedes this evidence.

    In that round, vaccines are (or should be) deadly. If a judge inserts themselves into the debate and rejects the neg evidence because they personally know that the evidence is bunk, then that judge is debating, not judging.

    And if you think that sounds ridiculous, imagine if the situation were reversed: say the judge were an anti-vaxxer, and the aff presented “vaccines good” studies which were conceded by the neg; would you accept a world where the judge treated the conceded evidence as false based on their personal beliefs?

    The norm that debaters are 100% in charge of the outcomes of rounds, with as little involvement as possible on the part of the judge, is a good norm. It is certainly flawed, but it is vastly less flawed than the alternatives.

    • Jacob Nails

      I don’t think you need to be an anti-vaxxer (nobody should be an anti-vaxxer) to recognize the fact that, in a debate with zero evidence supporting vaccines and some studies indicting them, the best evidence would favor the anti-vaxx thesis. I’d vote neg there.

      But suppose you replace “vaccines might cause autism according to a study” with “vaccines definitely cause autism; I heard it from a friend.” Does the latter have more weight?

      If you’re 100% hands-off as you say at the end, then it seems like the answer is Yes. The debater has made a stronger uncontested claim in the second case. But if you value “the best evidence,” and I do think you should, it seems obvious that the study should carry more weight than the heard-it-through-the-grape-vine argument; it’s better evidence.

      Or maybe you would give both claims 100% truth. I don’t see how that’s justified, seeing as in neither case does the best evidence in the round rise to the level of 100% certainty.

  • mcgin029

    I don’t mean to be flippant, but this article reads like the author has never debated or coached in a world where tab judging and mutual preferencing were not the norm.

    There are certainly problems with tab judging, but please, for the love of heaven, take my word for it — or consult any coach over 30 — the world you’re describing is a nightmare of arbitrariness.

    It’s potentially troubling that the community cannot, given a general norm of tabula rasa judging, identify any upper limits on argument expectations (although this certainly varies widely, even among highly-preferenced judges), but the alternative is a world where the vast majority of the judge plumbs the lower limits on a regular basis. There was a time when RFDs in bubble rounds routinely involved repeated use of the phrase “I just didn’t buy…” — if the judges even *gave* RFDs.

    The degree to which judges insert themselves into the round should be circumscribed as greatly as possible. It is not our role to make decisions based on our beliefs about the relative quality of arguments on their own merits, but rather to evaluate the interaction of arguments that the debaters make. There are certainly good paradigms that allow for judge intervention on a variety of issues — intervening to weigh competing evidence or claims where the students do not weigh; intervening to reject sexist, racist, or otherwise oppressive speech acts; even, for some, intervening to reject objectively false claims (ie, I once had a novice debater lose a round because their core argument depended upon the assertion that there was still a Soviet Union; their opponent conceded this point but the judge rejected it as factually inaccurate.) There are grey areas and dealing with those through experience, preferencing, and adaptation is fine.

    But we should not accept the bad old norm that judges, in general, should evaluate the relative quality of arguments in a way generally divorced from the arguments of the debaters. Thither lies madness.