Genericity on the Standardized Tests Resolution

The views and opinions expressed in this article are those of the author and do not necessarily reflect the views of Victory Briefs.

Jake Nebel is an assistant professor of philosophy at the University of Southern California and executive director of Victory Briefs. You can email Jake here


The September–October resolution is “Resolved: In the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions.” As usual, some debaters will try to affirm this resolution by arguing that some particular colleges and universities ought not consider particular standardized tests in particular admissions decisions. As usual, I don’t think such advocacies affirm the resolution.

Let me be clear, once again: my argument is not that plans are bad. I do not hate plans; I really like them, and I wish that all resolutions, including this one, were worded to allow for a reasonable array of topical plans. For example, if the resolution said something like, “The United States ought to prohibit some or all public universities from considering standardized tests in undergraduate admissions decisions,” then it would allow the affirmative to propose a plan to prohibit particular public universities from considering standardized tests. But that’s not the resolution. And I think debaters should debate the actual resolution. Plans are good, but plans should be topical.

My argument is simple. First, on this topic, the affirmative should have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. This is the topicality rule. Second, even if some particular colleges and universities ought not consider particular tests in particular decisions, that doesn’t mean that colleges and universities ought not consider standardized tests in undergraduate admissions decisions. Therefore, showing that some particular colleges and universities ought not consider particular tests in particular decisions fails to meet the affirmative burden.

This two-premise argument strikes me as eminently reasonable, and I don’t think it should require much more explanation in order to win a debate. But many debaters and judges appear to think that the argument is not just unsound but obviously or embarrassingly so. This response puzzles me, given the intuitive plausibility of the premises. Many of those who reject the argument are reluctant to identify which premise of the argument they reject, and they should have to do that in order to reject the conclusion. But, once again, I will try to defend the premises of the argument in the context of the new topic.

Let’s start with the second premise.

1 Bare Plurals

“Colleges and universities,” “standardized tests,” and “undergraduate admissions decisions” are bare plural noun phrases. A bare plural is a noun phrase that lacks an overt determiner. Determiners include articles like the, possessives like my, demonstratives like these, and quantifiers like some. “Colleges and universities,” “standardized tests,” and “undergraduate admissions decisions” are plural, and they lack determiners, so they are bare plurals. (“Colleges” and “universities” are also bare plurals, but it doesn’t matter for our purposes whether we consider them separately or just consider the conjunctive noun phrase.)

Bare plurals are typically used to express generic generalizations. Generic generalizations include sentences like, “Dogs bark,” “Bees sting,” and “Birds fly.” It is helpful to understand generic generalizations by contrasting them with two other kinds of generalizations.

Existential statements say that there exist some things that satisfy a certain property. For example, “Some bees don’t sting” is an existential statement. It is true because there are indeed some bees that don’t sting. Existential statements can be affirmed by pointing to particular examples—e.g., mason bees.

Universal statements say that all things satisfy a certain property. For example, “All bees sting” is a universal statement. It is false because, as we just saw, some bees don’t sting—so it’s not the case that all of them do. Universal statements cannot be affirmed by pointing to particular examples, but they can be negated by pointing to particular counterexamples—again, e.g., mason bees.

Generic generalizations are neither existential nor universal.

Generics are distinct from existential statements because they cannot be affirmed by particular instances. For example, “Birds swim” is a generic. It’s false even though there are some birds that do swim: namely, penguins. You can’t affirm that birds swim by observing that penguins swim.

Generics are distinct from universal statements because they can tolerate exceptions. For example, “Birds fly” is a generic. It’s true even though there are some birds that don’t fly: namely, penguins. You can’t negate that birds fly by observing that penguins don’t.

Both distinctions are important. Generic resolutions can’t be affirmed by specifying particular instances. But, since generics tolerate exceptions, plan-inclusive counterplans (PICs) do not negate generic resolutions.

Bare plurals are typically used to express generic generalizations. But there are two important things to keep in mind. First, generic generalizations are also often expressed via other means (e.g., definite singulars, indefinite singulars, and bare singulars). Second, and more importantly for present purposes, bare plurals can also be used to express existential generalizations. For example, “Birds are singing outside my window” is true just in case there are some birds singing outside my window; it doesn’t require birds in general to be singing outside my window.

So, what about “colleges and universities,” “standardized tests,” and “undergraduate admissions decisions”? Are they generic or existential bare plurals? On other topics I have taken great pains to point out that their bare plurals are generic—because, well, they are. On this topic, though, I think the answer is a bit more nuanced. Let’s see why.

1.1 “Colleges and Universities”

“Colleges and universities” is a generic bare plural. I don’t think this claim should require any argument, when you think about it, but here are a few reasons.

First, ask yourself, honestly, whether the following speech sounds good to you: “Eight colleges and universities—namely, those in the Ivy League—ought not consider standardized tests in undergraduate admissions decisions. Maybe other colleges and universities ought to consider them, but not the Ivies. Therefore, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions.” That is obviously not a valid argument: the conclusion does not follow. Anyone who sincerely believes that it is valid argument is, to be charitable, deeply confused. But the inference above would be good if “colleges and universities” in the resolution were existential. By way of contrast: “Eight birds are singing outside my window. Maybe lots of birds aren’t singing outside my window, but eight birds are. Therefore, birds are singing outside my window.” Since the bare plural “birds” in the conclusion gets an existential reading, the conclusion follows from the premise that eight birds are singing outside my window: “eight” entails “some.” If the resolution were existential with respect to “colleges and universities,” then the Ivy League argument above would be a valid inference. Since it’s not a valid inference, “colleges and universities” must be a generic bare plural.

Second, “colleges and universities” fails the upward-entailment test for existential uses of bare plurals. Consider the sentence, “Lima beans are on my plate.” This sentence expresses an existential statement that is true just in case there are some lima beans on my plate. One test of this is that it entails the more general sentence, “Beans are on my plate.” Now consider the sentence, “Colleges and universities ought not consider the SAT.” (To isolate “colleges and universities,” I’ve eliminated the other bare plurals in the resolution; it cannot plausibly be generic in the isolated case but existential in the resolution.) This sentence does not entail the more general statement that educational institutions ought not consider the SAT. This shows that “colleges and universities” is generic, because it fails the upward-entailment test for existential bare plurals.

Third, “colleges and universities” fails the adverb of quantification test for existential bare plurals. Consider the sentence, “Dogs are barking outside my window.” This sentence expresses an existential statement that is true just in case there are some dogs barking outside my window. One test of this appeals to the drastic change of meaning caused by inserting any adverb of quantification (e.g., always, sometimes, generally, often, seldom, never, ever). You cannot add any such adverb into the sentence without drastically changing its meaning. To apply this test to the resolution, let’s again isolate the bare plural subject: “Colleges and universities ought not consider the SAT.” Adding generally (“Colleges and universities generally ought not consider the SAT”) or ever (“Colleges and universities ought not ever consider the SAT”) result in comparatively minor changes of meaning. (Note that this test doesn’t require there to be no change of meaning and doesn’t have to work for every adverb of quantification.) This strongly suggests what we already know: that “colleges and universities” is generic rather than existential in the resolution.

Fourth, it is extremely unlikely that the topic committee would have written the resolution with the existential interpretation of “colleges and universities” in mind. If they intended the existential interpretation, they would have added explicit existential quantifiers like “some.” No such addition would be necessary or expected for the generic interpretation since generics lack explicit quantifiers by default. The topic committee’s likely intentions are not decisive, but they strongly suggest that the generic interpretation is correct, since it’s prima facie unlikely that a committee charged with writing a sentence to be debated would be so badly mistaken about what their sentence means (which they would be if they intended the existential interpretation). The committee, moreover, does not write resolutions for the 0.1 percent of debaters who debate on the national circuit; they write resolutions, at least in large part, to be debated by the vast majority of students on the vast majority of circuits, who would take the resolution to be (pretty obviously, I’d imagine) generic with respect to “colleges and universities,” given its face-value meaning and standard expectations about what LD resolutions tend to mean.

There are other reasons, too, which I’ll leave for debaters to piece together based on the literature. But, as I said, I don’t think it’s necessary to get into such reasons; when a claim is more obviously correct than the premises of any argument to the contrary, it is reasonable to take it for granted.

1.2 “Standardized Tests” and “Undergraduate Admissions Decisions”

Although “colleges and universities” in the resolution is certainly generic, I am inclined to think that “standardized tests” is not, and I’m not sure about “undergraduate admissions decisions.” To see why, let’s again isolate each bare plural. “Not” will distract us, so let’s also abstract away from the negative wording of the resolution—we’ll bring it back in soon.

Consider the sentence, “USC ought to consider standardized tests in undergraduate admissions decisions.” Suppose that USC ought to consider the SAT general and subject tests but no other standardized tests in undergraduate admissions decisions. This would, it seems to me, entail that USC ought to consider standardized tests in undergraduate admissions decisions. This suggests that “standardized tests” is an existential bare plural.

Consider next the sentence, “USC ought to consider the SAT in undergraduate admissions decisions.” Suppose that USC ought to consider the SAT in undergraduate admissions decisions for prospective humanities majors but for no one else. It’s not clear to me whether this would entail that USC ought to consider the SAT in undergraduate admissions decisions. So I’m not sure whether “undergraduate admissions decisions” is generic or existential; I’d be curious to know what others think.

At this point, you might be getting your hopes up that the affirmative can topically specify standardized tests and, possibly, undergraduate admissions decisions, so long as they specify more than one of each. Before disappointing you, let me build your hopes up even further.

Arguably, the “more than one” reading of the resolution’s possibly existential bare plurals is not even necessary. This is because not all existential bare plurals mean “more than one.” A classic example from Chomsky is “Unicycles have wheels.” Obviously this sentence doesn’t say that unicycles have more than one wheel. “Wheels” is dependent for its interpretation on “unicycles,” so it is called a dependent plural. Another example, involving an existential bare plural subject: “Dogs are wagging their tails outside my window.” This sentence is true only if there is more than one dog wagging its tail outside my window, but it doesn’t require any of the dogs to have multiple tails.

One might argue that “standardized tests” is a dependent plural with respect to “colleges and universities.” On this view, if the resolution were, “Colleges and universities ought to consider standardized tests in undergraduate admissions decisions,” the affirmative would not have to show that colleges and universities ought to consider multiple standardized tests in undergraduate admissions decisions. Now, this wouldn’t necessarily mean that a single-test affirmative would be topical. I hesitate because, in our other examples of dependent plurals, each unicycle has its own wheel, and each dog has its own tail. If all colleges and universities ought to consider the same test, would it be true that colleges and universities ought to consider standardized tests? It’s not obvious to me. Again, I’d be curious to know what others think. For now I’m content to challenge the idea that plurals necessarily mean “more than one” and to raise the question for further research.

For present purposes, however, these subtle distinctions don’t matter, because the resolution says “ought not.” Why does this matter? Consider again “Unicycles have wheels.” This sentence means, roughly, that each unicycle has at least one wheel (“roughly” because I’m glossing over the distinction between generic and universal for simplicity). By contrast, consider “Unicycles don’t have wheels.” This sentence means, roughly, that each unicycle has no wheels. It’s not just the logical negation of the original proposition, which would be the following: it’s not the case that, for every unicycle, there’s a wheel that it has—i.e., that some unicycle lacks a wheel.

This means that, if “standardized tests” is a dependent plural with respect to “colleges and universities,” the resolution means that colleges and universities not consider any standardized tests. Compare: if the resolution were “Unicycles don’t have wheels,” they would have to argue that unicycles don’t have any wheels, not just that there are some wheels unicycles don’t have (e.g., the wheels on my car). This is because the negation of an existential statement (“it’s not the case that some do”) is a universal statement (“all of them don’t”). This is the observation about quantifier scope I made about the Jan–Feb 2019 resolution, and it applies straightforwardly to the standardized tests topic because of the “ought not” wording. So, if “Colleges and universities ought to consider standardized tests” means roughly that colleges and universities ought to consider at least one standardized test, then the sentence “Colleges and universities ought not consider standardized tests” would mean roughly that colleges and universities ought to consider no standardized tests.

Crucially, this argument is not specific to dependent plurals. It applies to all existential bare plurals. Consider “Mary saw zebras.” That’s true just in case Mary saw more than one zebra. It means the same thing as “Mary saw some zebras.” By contrast, “Mary didn’t see zebras” is true just in case Mary saw no zebras; it cannot be interpreted to mean merely that there are some zebras Mary didn’t see. It does not mean the same thing as “Mary didn’t see some zebras,” which at least has a reading on which it’s true as long as some of the zebras were unseen by Mary. So, if “Colleges and universities ought to consider standardized tests for undergraduate admissions decisions” means roughly that colleges and universities ought to consider more than one standardized test for undergraduate admissions decisions (or: ought to consider standardized tests for more than one undergraduate admissions decision), then “Colleges and universities ought not consider standardized tests for undergraduate admissions decisions” would mean that colleges and universities ought to consider no standardized tests (or: ought to consider standardized tests for no undergraduate admissions decisions.)

Here is another example, using a bare plural subject (for those who continue to insist that “colleges and universities” is existential). Suppose I say, “Lima beans are not in the pantry.” You and I are looking in the pantry, and there they are: lima beans. I then point to the floor, and there they are: more lima beans! I say, “See, some lima beans are not in the pantry. Therefore, lima beans are not in the pantry.” The conclusion obviously does not follow. The sentence “Lima beans are not in the pantry” cannot be interpreted to express the true claim that there are lima beans that are not in the pantry (“wide scope”); it means that it’s not the case that there are lima beans—i.e., that there are no lima beans—in the pantry (“narrow scope”). So if, contrary to fact, “Colleges and universities ought to consider standardized tests” means that some colleges and universities ought not consider such tests, then “Colleges and universities ought not consider standardized tests” would, contrary to fact, mean that no colleges or universities ought to consider such tests—making it harder to affirm than the correct (generic) interpretation.

Summing up: if “colleges and universities” were existential (which it isn’t), the resolution would mean that no colleges or universities ought to consider standardized tests; if “standardized tests” is existential (which I think it probably is), the resolution means that colleges and universities ought to consider no standardized tests; and if “undergraduate admissions decisions” is existential, the resolution means that colleges and universities ought to consider standardized tests for no such decisions.

Here is a way to see this point in action. Consider the sentence “In the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions, and colleges and universities ought to consider standardized tests in undergraduate admissions decisions.” This sentence expresses a contradiction. But if the resolution meant what our hypothetical affirmative needs it to mean, it should sound perfectly consistent. After all, “In the United States, some colleges and universities ought not consider some standardized tests in some undergraduate admissions decisions, and some (i.e., possibly different) colleges and universities ought to consider some (i.e., possibly different) standardized tests in some (i.e., possibly different) undergraduate admissions decisions” is perfectly consistent. This shows that the resolution could not be interpreted in a way that allows for the specification of colleges, universities, tests, or decisions, even if all of the bare plurals in the resolution were, contrary to fact, existential. At the very least, the onus is on our hypothetical affirmative to explain why the sentence above sounds contradictory if their desired interpretation of the resolution is even possible.

1.3 Summary

We have seen that “colleges and universities” is generic in the context of the resolution. So, even if some colleges and universities ought not consider standardized tests in undergraduate admissions decisions, that doesn’t mean that colleges and universities ought not consider standardized tests in undergraduate admissions decisions. (And, since generics tolerate exceptions, PICs that exclude particular colleges or universities do not negate.)

I have also suggested that “standardized tests” is likely an existential (possibly dependent) bare plural. If it is, then, since the resolution says “ought not,” it requires that colleges and universities not consider any such tests. I am unsure whether “undergraduate admissions decisions” is existential or not, but the same argument would apply if it is.

So, on any reading of the bare plurals in the resolution, specifying particular colleges and universities, particular tests, or particular decisions could not show that colleges and universities ought not use standardized tests in undergraduate admissions decisions.

2 Some Objections

I’ve just defended the second premise of my argument. Let’s now consider how affirmatives tend to reply to this premise.

2.1 Pseudo-Counterinterpretations

Affirmatives tend to respond with an interpretation along the lines of, “I can specify colleges/universities/tests/decisions.” That’s not a topicality interpretation, because it’s not an interpretation of the resolution. On the face of it, it allows the affirmative to specify parameters of the resolution even when doing so would not to be topical.

Consider an example. A recent policy topic read, “Resolved: The United States federal government should substantially increase its economic engagement toward Cuba, Mexico or Venezuela.” Suppose the affirmative increases economic engagement toward Mexico by one dollar. If I read an interpretation of “substantially” on which a one dollar increase is not substantial, you can’t just reply, “Counterinterp: I can increase by one dollar.” That so-called counterinterpretation doesn’t tell us what “substantially” could possibly mean such that a one-dollar increase affirms the resolution. If “substantially” in fact means something greater than one dollar, the pseudo-counterinterpretation implies that the affirmative can defend a nonsubstantial increase—i.e., doesn’t have to affirm the resolution. It’s fine to reject the requirement of topicality, but then just say so.

In order to show that an affirmative is topical, we need an interpretation of the resolution: what do they think the resolution means, or what do they even want it to mean, such that it can be affirmed by specifying colleges, universities, tests, or decisions? Without an answer to this question, we cannot tell whether the affirmative actually affirms the resolution; being told that it does doesn’t tell us why.

In order to defeat the argument while granting the topicality rule, the affirmative would need to tell us what they take the resolution to mean such that they are affirming it. The answer is presumably something like this: “‘Colleges and universities ought not consider standardized tests in undergraduate admissions decisions’ means that there are some colleges and universities, some standardized tests, and some undergraduate admissions decisions such that those colleges and universities ought not consider those tests in those decisions.” It sounds absurd when you make it explicit, which is presumably why it’s seldom made explicit. As we have already seen, the resolution could not possibly have that meaning, any more than “Pterodactyls didn’t have wings” could possibly mean merely that that there are some wings that some pterodactyls didn’t have. The affirmative would need to give some reason to believe that the resolution could possibly have the meaning they need it to have—much like, when defeating any other topicality argument, the affirmative needs some evidence that the words in the resolution at least might mean what they need them to mean. I am willing to bet that there is no such evidence, and that any purported evidence for it is misrepresented or misinterpreted. Without any such evidence, we don’t need to consider the counterinterpration (whatever it is) any more than we need to consider an interpretation on which the resolution means that states ought to eliminate their nuclear arsenals.

A card that says “Bare plurals can be existential” is not enough. Obviously I agree that bare plurals can be existential; I think at least one of the bare plurals in the resolution is existential. But even if all of the bare plurals in the resolution were existential, affirmatives that specify particular colleges, universities, tests, or decisions would not show that colleges and universities ought not consider standardized tests in undergraduate admissions decisions, for the reasons we have already seen.1

Some people read a card about “ought” from this excellent paper by philosopher Fabrizio Cariani. The card points out that to say I ought to do something does not imply that every way of doing it is permissible. His example is that you ought to give food to your pets, but ought not give them poisonous food. That is surely true. Debaters somehow warp the card into saying that “‘ought’ implies specification.” That doesn’t mean anything. But if it’s supposed to mean that any ought statement can be affirmed by specifying any of its constituents, it’s clearly false, and in no way supported by the card. “Everyone ought always to tell lies” cannot be affirmed by specifying that, in some situation, someone ought to tell a lie, and it’s absurd to suppose that the author thinks otherwise. That there are impermissible ways in which we might do things we ought to do in no way supports the idea that any ought sentence can be affirmed by particular instances no matter what those sentences say. It’s a complete non sequitur. To use a negative version of Cariani’s example: that you ought not give poisonous food to your pets does not entail that you ought not give food to your pets. Isn’t that obvious? Besides, even without knowing anything about the card, it’s easy to see that it could not possibly refute the generic interpretation of the resolution, because our arguments for the generic interpretation already took the “ought” in the resolution into account.2

I’ve elsewhere commented on the Leslie cards that some affirmatives warp to support an existential counterinterpretation. I’ve seen a few other cards, too, that simply cannot be parsed as making an argument for the affirmative’s interpretation. And these are just the cards readers have sent me. But affirmative misrepresentation of evidence on these issues is extremely widespread. This is unsurprising because no linguist would defend the affirmative’s interpretation of the resolution, since it’s transparently incorrect. Of course, I don’t think that people are deliberately misrepresenting evidence here. The more plausible explanation is that people very much want the affirmative interpretation to be credible, so—to their credit—they search the literature. But the literature is extremely complicated. So, when they find passages that contain strings of words like “existential,” “bare plural,” “singular,” “specification,” or whatever, they interpret those strings of words into arguments that support the hypothesis they want to be true—even though, of course, no such argument is there. The words look like tea leaves, so why not divine the future from them? That is confirmation bias and motivated reasoning in action (more on that later). This process makes it easy to see how debaters could unwittingly misrepresent evidence as supporting the affirmative interpretation on this issue. Still, debaters have a responsibility to understand their own evidence, so lack of understanding doesn’t really excuse misrepresentation.3

Instead of misrepresenting evidence to say that the resolution means something it clearly doesn’t mean, the affirmative should just admit that the resolution doesn’t mean what they want it to mean, and that they aren’t affirming the resolution. They should reject the first premise of the argument, and grant the second. That seems to me a more principled and intellectually honest strategy that would permit real clash on the heart of the issue: whether or not the affirmative should have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. We’ll turn to that issue soon.

2.2 “Pragmatics First”?

Affirmatives tend to defend their (pseudo-)counterinterpretations on this issue by arguing that topicality should primarily be a matter of “pragmatic” rather than “semantic” considerations. (See my “Priority of Resolutional Semantics” for this distinction. Please note that these terms are being used in a debate-specific sense; the “semantic/pragmatic” distinction in linguistics and philosophy of language is completely different.) Semantic arguments are arguments about what the resolution means; pragmatic arguments are about what it would be better or worse to interpret the resolution as meaning. My view is that semantics function as a side constraint on resolutional interpretation, and that pragmatics can help us choose between different interpretations that are semantically eligible. If the resolution is genuinely ambiguous between two or more propositions—a claim that requires evidence of some form (e.g., explicit definitions or linguistic tests)—we can ask which of those propositions would be better to debate. But there are many propositions that the resolution just doesn’t mean; one, in the context of this resolution, is that some colleges and some universities ought not consider some tests in some admissions decisions. The alternative “pragmatics-first” view seems to hold that the pragmatic benefits of debating something can be sufficient to interpret the resolution as meaning that thing, even if the resolution doesn’t mean that thing.

Now, even on the “pragmatics-first” view, it’s not at all clear that it would be better to interpret the resolution to mean that some colleges and some universities ought not consider some standardized tests in some undergraduate admissions decisions. After all, there are over five thousand colleges and universities in the United States. On the affirmative’s interpretation, there are vastly more topical affirmatives than there are atoms in the observable universe. (There at most 1082 atoms in the observable universe. The number of non-singleton subsets of two or more items out of a set of five thousand is 25000 − 5001, or approximately 1.4 × 101505. For simplicity I’ve ignored the distinction between colleges and universities, since I don’t know the breakdown between them. It’s still an underestimate, though, because it doesn’t count affirmatives that specify different tests or decisions.)

Against a generic interpretation of the topic, some complain that there is “only one aff.” That may be true, if you strangely insist on individuating affs only by their advocacy rather than their arguments or advantages (e.g., they think your Kant aff and my util aff are one and the same aff, which I find bizarre).4 But, however we count affs, it is much better for there to be only one—the generic generalization expressed by the resolution—than for there to be more than a googol. It’s impossible to adequately prepare specific answers to such a vast number of affs. That preparation burden leads to worse debates, since it leads negatives to have lower-quality answers to any particular aff; everyone knows this, so debaters are incentivized to quickly mine for surface-level arguments for obscure affs, so they can use the rest of their time finding the minimal viable quantity of barely responsive evidence against other people’s affs, or beefing up maximally generic arguments that are recycled from topic to topic. That is a worse outcome than everyone researching the very same proposition, knowing that everyone else will research that proposition, thereby creating an incentive for in-depth research, expertise, and innovation. Limits are not obstacles to creativity; they are essential to it.

But even if our hypothetical affirmative’s interpretation would be better for debate, we should reject the “pragmatics-first” view that the pragmatic benefits of debating something can be sufficient to interpret the resolution as meaning something it doesn’t mean. This is because the pragmatically best proposition to debate could be something that has nothing to do with the topic or is straightforwardly inconsistent with the topic. So all the reasons to debate the resolution are reasons to reject the pragmatics-first view. Moreover, no one knows or could ever know which of all possible propositions would be best to debate, since there are infinitely many propositions that could be debated and no way of figuring out which of them would be best. So, on the “pragmatics-first” view, no one knows how the topic should be interpreted. That is an embarrassing result, because the purpose of having a resolution is to establish a commonly known basis for preparation. What’s more, it is incredibly unlikely that the proposition affirmed by the affirmative is the best possible proposition to debate, so the affirmative is incredibly unlikely to be topical even in its own terms.

One reply to this argument is to say that we don’t have to consider all of the possible propositions that could be debated, but all and only those that are advanced via interpretations in any given debate. This doesn’t solve the problem. Suppose the negative interprets the current LD resolution to mean that the United States federal government should substantially reduce Direct Commercial Sales and/or Foreign Military Sales of arms from the United States. Since the affirmative defends no such reduction, they aren’t topical. This interpretation would be much better for clash, affirmative flexibility, limits, advocacy skills, and so on. Obviously it’s not what the resolution means, but that’s not what topicality is about—right? On the view we’re considering, pragmatic benefits can justify interpreting the resolution to mean something it doesn’t mean. As long as the negative can identify at least one proposition much better for debate than what the resolution actually means, the affirmative is not topical.

Perhaps the view is the pragmatic benefits of an interpretation can justify it so long as it’s semantically good enough. I have no idea what it is for an interpretation to be semantically good enough, if it’s not just to be something the resolution might mean. And, as we have shown, the requisite interpretation for the affirmative is just not something the resolution could mean. There is no dialect of any language in which even one meaning of the resolution is that some colleges and some universities ought not consider some standardized tests in some undergraduate admissions decisions. If the resolution were genuinely ambiguous between the affirmative’s interpretation and others, or if we just couldn’t tell whether the resolution might mean what the affirmative wants it to mean, then perhaps its pragmatic benefits could justify it. But the resolution is not ambiguous along that dimension, for all the reasons we have already seen.

It seems to me that, despite appearances encouraged by the format of topicality arguments, the “pragmatics-first” person isn’t really rejecting the second premise of my argument: that, even if some particular colleges and universities ought not consider particular tests in particular decisions, that doesn’t mean that that colleges and universities ought not consider standardized tests in undergraduate admissions decisions. They are instead rejecting the first premise of the argument: that, on this topic, the affirmative should have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. They are, in other worse, rejecting the topicality rule, as I define it. So let’s now turn to that premise.

3 Topicality, Bobicality, and Focal Points

The first premise of my argument is more general: on this topic, the affirmative should have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. It seems to me that the most principled response to my argument is to reject this premise, the topicality rule. And many debaters do independently claim that the affirmative should not have to be topical.

Unfortunately, it is outside the scope of this article to defend the topicality rule in depth. But I take it that almost every debater and judge would accept that there is at least some reason for the affirmative to be topical and that there should therefore be at least a defeasible presumption in favor of topicality. I have tried to say a bit about that elsewhere. And, to the extent that there is such reason, we have reason to expect the affirmative to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. Here I want to take a somewhat different route.

It seems to me that one of the strongest reasons to accept the topicality rule is the badness of its alternatives. These alternatives are seldom proposed or defended explicitly. Even when the affirmative explicitly rejects the topicality rule, they sometimes fail to explain precisely what the affirmative should have to do instead of affirming the resolution, in more general terms than what they did. This seems to me the most important question in debate theory. What is the affirmative burden, if not to show that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions?

Consider the most radical possibility. Suppose that the affirmative burden is to identify some proposition—anything whatsoever—and prove that proposition. The negative burden is to refute whatever proposition is defended by the affirmative. Call this division of burdens “the unconstrained parli rule” (in some parliamentary formats, the affirmative proposes its own proposition in each debate, within certain constraints). This rule would be pretty bad. Unconstrained parli would not be fair, fun, or particularly edifying in any way. For example, the affirmative could choose to defend the proposition that 2 + 2 = 4, leaving the negative with the meager ground of mathematical nihilism (“There are no numbers!”) and no recourse, given the negative burden of refuting whatever proposition the affirmative chooses to defend. I leave other objections to the reader.

There is a spectrum of possibilities in between the unconstrained parli rule and the topicality rule. Some of those possibilities assign no privileged role to the resolution. They can be seen as more constrained variations of the parli rule we just considered. I won’t consider such constrained parli rules here. Let’s instead consider those rules that assign some privileged role to the resolution, other than the topicality rule as I’ve defined it.

3.1 Bobicality

Bob Overing has provided a helpful list of such alternative rules, which I will call Bobicality to distinguish them from topicality as I understand it. Let’s go through his list in order.

Bobicality1 says that the affirmative “plan and case ought to justify the conclusion the resolution asks to be drawn” (based on Allen & Burrell 1985). This seems pretty close to, though not exactly, what I think. The conclusion this resolution asks to be drawn is that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. Even if some particular colleges and universities ought not consider particular tests in particular decisions, that doesn’t justify the conclusion that colleges and universities ought not consider standardized tests in undergraduate admissions decisions. Generics cannot be justified by particular instances. So the affirmatives under consideration are not Bobical1.

Bobicality2 is that the affirmative must “provide evidence in favor of the resolution.” This requirement is much too weak. Showing that colleges and universities have some obligations would provide evidence in favor of the resolution, but it’s obviously not sufficient to affirm. So would showing that colleges and universities exist, that there are some things some agents ought to do, that there are prospective undergraduates, or anything else that slightly increases the probability of the resolution. But, also, affirmatives that specify particular colleges, universities, tests, or decisions may not even be Bobical2; they may fail to provide evidence for the resolution. For example, suppose that two colleges and two universities ought not consider the SAT main or subject tests for prospective humanities majors because instead all colleges and universities ought to consider the ACT and other standardized tests for all students. Then the fact that some colleges and universities ought not consider some tests (namely, the SAT tests) would be strong evidence against the resolution.

Bobicality3 is that the affirmative advocacy “must be an example or instance of the resolution.” This is a common alternative to the topicality rule, as I’ve defined it. (It seems to be used pretty much interchangeably with the formulation that the plan must be a “subset” of the resolution.) But the resolution is a sentence, and I do not know what it is for something to be an example (or a subset) of a sentence. I suppose it must be something like this. If I say, “Dogs have four legs,” I might add, “For example, Fido has four legs.” The Fido statement is an example used to illustrate the generic statement. If it is felicitous to follow up a sentence with, “For example, p,” then we’ll count p as an “example” of the sentence. This heuristic works for many policy resolutions. For example, consider last year’s high school policy resolution: “The United States federal government should substantially reduce its restrictions on legal immigration to the United States.” Assuming that opening borders would count as substantially reducing restrictions on legal immigration, it would be felicitous to follow up the resolution with, “For example, the United States federal government should open its borders.” This suggests that the open borders plan is topical, which (given our assumption) it is: supposing that the United States federal government should open its borders (and that opening borders would count as substantially reducing restrictions on legal immigration), it would follow that the United States federal government should, indeed, substantially reduce its restriction on legal immigration to the United States—namely, by opening its borders. For this resolution, Bobicality3 is coextensive with topicality.

Although the example heuristic works for certain resolutions, it can’t be right in general. If I say, “All dogs have four legs,” I might add, “For example, Fido has four legs.” The Fido statement would be considered an example even if the original statement were explicitly universal. This makes Bobicality3 too permissive. Suppose the resolution were, “In the United States, no colleges or universities ought to consider any standardized tests in any admissions decisions.” It would be felicitous to follow this up with, “For example, Princeton ought not consider my ACT score when I apply next year.” So it counts as an example, and is therefore Bobical3. But it would be absurd to suppose that Princeton not considering one person’s SAT score is sufficient to affirm this explicitly categorical resolution. More generally, Bobicality3 seems to imply, strangely, that the resolution’s division of ground would remain the same even if the bare plural “colleges and universities” were replaced with “some/all/most/many/seven colleges and universities.” It would be bizarre to think that such drastic changes to the resolution’s wording could not affect what the affirmative would have to defend, on the grounds that the affirmative only has to defend a single example (or “subset”) no matter what. Now, maybe there is some way of refining or reinterpreting Bobicality3 to avoid its insensitivity to different quantifiers. But then it is hard to see how it could make single examples Bobical3 on a generic resolution, since single examples affirm generics no more than they affirm universals.

Bobicality4 requires the affirmative advocacy to “fall within the bounds of the resolution.” I don’t know what exactly that means, but I think it falls prey to the same objection above. Either particular instances fall within the bounds of universal generalizations (in which case Bobicality4 is too weak and is objectionably insensitive to massive differences in resolutional wording), or they do not fall within the bounds of generic generalizations (in which case the affirmatives under consideration are not Bobical4 on the present LD topic).

Bobicality5 says that, “All plan provisions must bear a rational relationship to a topical scheme of action” (Unger 1979 cited by Hingstman 1985). This statement uses the word “topical” in what Bob wants to be a definition of “topical.” That is not helpful. And the idea of a “rational relationship” is insufficiently precise to be helpful.

Let me summarize the comments above into a more general problem that should afflict, for any n, Bobicalityn. The challenge is to identify a principled, precise, and predictable conception of the affirmative burden in relation to the resolution on which (1) particular instances meet that burden even when the resolution is generic, but (2) they do not meet that burden when the resolution is universal (or otherwise nonexistential but not generic). I know of no conception of topicality that avoids both horns of that dilemma. We should instead just accept that, insofar as the affirmative should affirm the resolution, they have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions—since that’s what everyone actually thinks anyway, if only in their less theory-laden moments. There is no need to grasp at ad hoc theoretical straws.5

My impression is that people appeal to views like Bobicality because they think they do the best job of explaining the practice of policy debate. What characterizes the relationship between a paradigmatically topical plan and the resolution is that they are Bobical: they are examples, or subsets, or fall within the bounds, or whatever. These ideas may be helpful heuristics that work in many contexts. But, like many helpful heuristics, we cannot always rely on them. As we have already seen, they break down—i.e., are too vague to be helpful or have absurd consequences—when we consider a wider array of resolutions. And we don’t need to appeal to Bobicality to characterize topical plans in policy debate. As I’ve explained elsewhere, paradigmatically topical plans entail the resolution. To reject this claim, one would need to identify a plan that is widely or should be considered topical and explain how it does not entail the resolution. Bob has asserted that my conception of topicality is “extremely difficult to meet with any plan, even on policy topics.” In the spirit of Bobicality3, he offered a single example. It was a quarter-century-old NDT topic written in the passive voice; I’ve answered it here.

3.2 Focal Points

Although most of this section has been about the problems for alternative views, they suggest an argument in favor of the topicality rule, according to which the affirmative has to show that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. The argument appeals to the game-theoretic concept of focal points (not to be confused with the buzzword “stasis point” derived from the rhetorical concept of stasis). A focal point (or Schelling point) is a solution that people can be expected to converge upon even when they cannot effectively communicate. A solution may be a focal point because it is especially salient or otherwise special. For example, if we agree to meet in Washington Square Park, but fail to set a more specific location, we can probably expect each other to show up at the arch or the fountain; those locations stand out and everyone can see them once they’re in the park, so most people tend to meet there, can expect that most others would tend to meet there, and so on.

The topicality rule is a focal point for debate. We all know that there is a resolution; that the resolution says that in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions; that one debater is designated “affirmative” and another “negative”; and we all know that we all know these things. In light of that common knowledge, the simplest, clearest, least arbitrary, and therefore most salient assignment of burdens is for the affirmative to show that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. That is a focal point. And there is no other comparably salient focal point. As we have already seen, it is not at all clear which alternative is accepted by those who reject the topicality rule. So we cannot reasonably expect others to converge on any such alternative, even if some such alternative does in fact avoid the problems we have considered. Thus, the topicality rule is the most salient focal point.

That is a strong reason to accept the topicality rule because debaters need a focal point in order to coordinate their preparation. Debaters face a problem: they have limited time in which to prepare for a tournament and do not know in advance who exactly they will debate, what their opponents will argue, what their opponents will expect them to argue, and what judges will expect them and their opponents to argue. The problem can be solved only if debaters can reasonably expect each other to converge on some basic assignment of burdens. Since the topicality rule is the most salient focal point, they can reasonably expect people to converge on that solution. And, since everyone can reasonably be expected to know this, it is reasonable for the judge to expect the debaters to converge on that solution. And, since there is no other comparably salient focal point, it would be unreasonable for the judge to expect debaters to converge on any other solution (e.g., the alternatives to the topicality rule we have considered).

This argument is not just a fancy version of a “predictable limits” argument, according to which the affirmative should have to argue that colleges and universities ought not consider standardized tests in undergraduate admissions decisions because, otherwise, debaters would have no predictable basis for deciding what to prepare for, which is bad for fairness and education. My argument goes further than that: even if there were some other adequate and predictable basis for preparation, that basis would be a less salient focal point than the resolution, so debaters could not be expected to converge on that solution to their coordination problem. Since everyone is in a position to know that, it would be unfair to use such an alternative basis and to expect others to use it.

Here is a simple example to illustrate the point. Suppose that I send a message to every debater in the country that I will debate last year’s Jan–Feb resolution in every round and will expect all of my opponents to debate that resolution as well (i.e., will run T against any aff that doesn’t affirm it). Imagine that I can somehow verify that everyone has received this message, so I know that it is now completely predictable for me to run cases on that topic, and to argue that others should do the same. Suppose also (as is plausible) that the Jan–Feb resolution is much better to debate. It would nonetheless be manifestly unfair for me to run Jan–Feb affirmatives against opponents who want to debate the Sept–Oct topic, or to argue that affirmatives affirming the Sept–Oct topic should lose for affirming the wrong resolution. This is because, even though my solution is predictable and perhaps would be better for debate if everyone adopted it, and even if some other people do adopt it, I cannot reasonably expect everyone else to converge on it, expect everyone else to expect everyone else to converge on it, etc., since it’s not designated by the league or tournament as the topic.

Let me summarize this section. The topicality rule says that, on this topic, the affirmative should have to argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions. This is the default, face-value affirmative burden that must be defeated if it’s not to be accepted. To defeat it, it’s not enough to show that it’s bad in certain ways; we would need to know what exactly the alternative is—what exactly any given affirmative has to do in order to win any given debate. But, first, no such alternative can reliably secure the precise and predictable limits needed for students to coordinate their preparation. And, second, even if there were some such alternative, it would not serve as a comparably salient focal point, since not everyone could reasonably expect that everyone could reasonably expect (and so on) everyone to converge on that assignment of burdens rather than the face-value one: the topicality rule.

4 Conclusion

Let’s put the argument back together. The affirmative should have to advocate that, in the United States, colleges and universities not consider standardized tests in undergrad admissions decisions. It’s not enough to advocate merely that some colleges and some universities not consider some standardized tests in some undergrad admissions decisions. For even if some colleges and some universities indeed ought not consider some such tests in some such decisions, that wouldn’t necessarily mean that colleges and universities ought not consider standardized tests in undergrad admissions decisions.

It’s a simple argument. Its simplicity is sometimes obscured by the format of topicality and by the complex vocabulary used to warrant the premises more deeply than they require, given their independent plausibility. Lots of people strongly dislike the conclusion. But if you don’t like the conclusion, which premise do you reject?

Let me conclude by explaining why I care about this issue. Why have I written so many articles and so many words about genericity, plans, and topicality in LD—especially given that I like plans and wish more of them were topical on more LD resolutions? Well, first, because I find the material interesting: I enjoy thinking about debate theory and especially about language and meaning (as is reflected by some of my more recent academic work), and I better understand new things when writing each article. But, second, I think it’s really important.

I am a pluralist about the value of debate. I think that debate is great for lots of different reasons, in different ways for different people. But the value that seems to me most foundational to all debate formats, all flavors of argument, and everything students learn in the activity, that is good in itself, for all students, and for the world, that is both durable beyond high school and portable beyond debate, and that debate is most uniquely positioned to promote is critical thinking: the “wide range of cognitive skills and intellectual dispositions needed to effectively identify, analyze, and evaluate arguments and truth claims; to discover and overcome personal preconceptions and biases; to formulate and present convincing reasons in support of conclusions; and to make reasonable, intelligent decisions about what to believe and what to do”—collectively, “disciplined thinking governed by clear intellectual standards.”6

One great obstacle to critical thinking is motivated reasoning: essentially, biased reasoning toward conclusions you want to believe. I would speculate that people feel some cognitive dissonance from (i) strongly wanting to run certain plans (e.g., ones that specify particular colleges, universities, tests, or decisions), (ii) believing that the affirmative should affirm the resolution (i.e., should argue that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions), and (iii) realizing, on some level, that the plans they want to run don’t affirm the resolution (i.e., do not show that, in the United States, colleges and universities ought not consider standardized tests in undergraduate admissions decisions). It would take some pretty extreme unlearning for (iii) not to be the case. This cognitive dissonance could explain a few data points: for example, why some people seem not just to disagree with but passionately despise the argument; why people have suggested such absurd strategies to avoid it; why debaters so commonly (and unwittingly) misinterpret and misrepresent evidence to try to refute it; and why the strategy against the argument tends not to be the principled one that the affirmative doesn’t need to affirm the resolution. To reduce cognitive dissonance, we often engage in motivated reasoning, which gets reinforced via confirmation bias (selectively seeking out and interpreting evidence to confirm your own beliefs).

The value of critical thinking and of mitigating cognitive bias is more important than whatever costs and benefits there are to any particular strategies on any particular debate topic. We should not encourage students to accept whatever they want to believe, but to believe whatever is best supported by the evidence. If the resolution does not mean what your affirmative needs it to mean, then just admit that you aren’t affirming the resolution, and don’t complain when you hit a nontopical aff. That seems to me the only principled response to the argument. If you don’t want to do that, because you believe that the affirmative should affirm the resolution, then maybe you should run a different aff—even if that makes it harder for you to win.


  1. A card that says “Bare plurals are/must be existential” (there is no such card) would be even worse, because it’s self-refuting. In order to have any substantial implications, it cannot itself be existential. For if it were existential, it would mean that some bare plurals are existential, which we already knew. But if it’s not existential, then “Bare plurals are/must be existential” is false.
  2. A similar point applies to an argument, which some affirmatives will inevitably make, that “in the United States” means somewhere in the United States, so they can specify colleges and universities even if “colleges and universities” is generic. This interpretation illicitly adds an existential quantifier to the resolution that we’ve already observed not to exist. One way to see this: “In the United States, mountains are not more than ten thousand feet tall” obviously doesn’t mean merely that somewhere in the United States are mountains that are no more than ten thousand feet tall. “In the United States” restricts the generic “mountains”; it doesn’t transform it into or add an existential quantifier. To make it even clearer, suppose the resolution read, “In the United States, most colleges and universities ought not consider standardized tests in undergraduate admissions decisions.” It would obviously be absurd to suppose that this means merely that somewhere in the United States (e.g., in California), most colleges and universities ought not….
  3. This is one reason I think debaters should disclose all rebuttal evidence and interpretations. Doing so would allow people to catch the misrepresentation of evidence more effectively, and thereby create a better deterrent to misrepresentation.
  4. Even then, I don’t think it’s true, because extratopical affs can be topical (and, I think, legitimate); there is clearly more than one extratopical aff. I offered some defense of extratopicality in my contribution to last year’s Sept–Oct briefs. But, as Marshall Thompson has pointed out to me, there’s especially strong reason to allow extratopicality on this resolution because of the negative wording. Suppose the resolution said, “In the United States, colleges and universities ought to consider standardized tests in undergraduate admissions decisions.” On that resolution, the negative could win by showing that colleges and universities ought to do something else—i.e., a counterplan—competitive with considering standardized tests in undergraduate admissions. Colleges and universities ought not do something if the opportunity cost outweighs the value of doing it. So, when the resolution is negatively worded, showing that colleges and universities ought to do something competitive with considering standardized tests (more specific than just not considering standardized tests) should affirm. Even if you don’t like extratopicality, though, at least you can agree that extratopicality is better than nontopicality.
  5. Another variation I have sometimes heard is that, to be topical, the affirmative only needs to “meet the words” of the resolution, not its actual meaning as a sentence. I do not know what it means to “meet the words” of a sentence. For example, what does it mean to meet the definition of “not”? It depends on the syntax of the sentence. You can’t arrive at a sensible or testable interpretation of “not” without at least considering its role in the sentence as a whole. More generally, the meanings of the resolution’s subsentential expressions (e.g., its individual words) are important because of how they contribute to the meaning of the resolution—i.e., the entire sentence. Otherwise we would just have topics be list of words and let the affirmative play Mr. Potato Head.
  6. G. Bassham (ed.), Critical thinking: a student’s introduction (New York: McGraw-Hill Higher Education, 2010), p. 1.

For helpful comments and discussion, I’m grateful to Jacob Nails, Nick Smith, Chris Theis, Marshall Thompson, and students in my topicality modules at VBI this summer.