Call me old fashioned, but I think debates over the next few months should be about whether or not just governments ought to require that employers pay a living wage. Some debates will inevitably focus on whether particular countries should have particular kinds of living wage laws. I don’t think that such debates are, in themselves, better or worse than debates that focus on abstract questions about justice and the living wage. More generally, I don’t really care for clash-of-civilizations debates between policymakers and philosophers: I think it’s a mistake to privilege, without much empirical evidence, either abstract philosophical debate or concrete public policy debate as more fair or educational than the other. Some approaches are more germane to certain resolutions than they are to others. Most resolutions call for elements of both, but the balance depends on precise aspects of wording.
I think the best way of motivating this view is by example. So I’m going to look at the wording of the living wage resolution with an eye to what kinds of things, if any, the affirmative may legitimately specify. I believe that the affirmative may not specify a particular government or set of governments, but may perhaps specify a particular way of requiring employers to pay a living wage. In this post, I’ll focus on “just governments.”
My primary aim is not really to argue against agent specification on this topic. It is, rather, to motivate a general approach to questions about specification and topicality that takes the precise wording of the resolution as central. The approach is to figure out what the resolution means, not what you want it to mean. Another aim is to expose the reader to certain distinctions and techniques that may be helpful beyond this particular case.
I believe that debaters shouldn’t specify a government on the living wage topic. The standard argument for this is simple: “just governments” is a plural noun phrase, so it refers to more than one just government. Most debaters will stop there. But there is much more to say. (Some seem not to care about the plural construction. I plan to address this view in a later article about the parametric conception of topicality.)
Some noun phrases include articles like “the,” demonstratives like “these,” possessives like “my,” or quantifiers like “some” or “all.” These words are called determiners. Bare plurals, including “just governments,” lack determiners. There’s no article, demonstrative, possessive, or quantifier in front of the noun to tell you how many or which governments are being discussed.
We use bare plurals for two main purposes. Consider some examples:
- Debaters are here.
- Debaters are smart.
In (1), “debaters” seems equivalent to “some debaters.” It is true just in case there is more than one debater around. If I enter a restaurant and utter (1), I speak truly if there are a couple of debaters at a table. This is an existential use of the bare plural, because it just says that there exist things of the relevant class (debaters) that meet the relevant description (being here). In (2), though, “debaters” seems to refer to debaters in general. This use of the bare plural is generic. Some say that generics refer to kinds of things, rather than particular members of their kinds, or that they refer to typical cases. There is a large literature on understanding generics. Here my aim is not to figure out the truth conditions for the generic reading of the resolution; I shall simply work with our pre-theoretical grip on the contrast between sentences like (1) and (2).
This distinction bears importantly on the resolution. If “just governments” is a generic bare plural, then the debate is about whether just governments in general ought to require that employers pay a living wage. If it is an existential bare plural, then the debate is about whether some just governments—i.e., more than one—ought to require that employers pay a living wage. Only the second interpretation allows one to affirm by specifying a few governments.
To my ear, the generic reading is correct. I think the best evidence for this is simply the undistorted judgments of ordinary speakers. No competent speaker of English would, without distorting influence or additional evidence of generalizability, endorse an inference from a plan involving two just governments to the resolution. Suppose Sally, an American citizen, believes that the U.S. and Canada should require employers to pay a living wage, but that no other government (just or unjust, actual or possible) should. She would not represent her view by asserting, “Just governments ought to require that employers pay a living wage.” She would deny this claim and hold that the U.S. and Canada are exceptions.
One might object that Sally would endorse this assertion if she believed that the U.S. and Canada are the only just governments. Maybe she would, but that is explained by the generic reading, because she would then be making a generalization about (what she believes to be) just governments. And the onus would be on the affirmative, when specifying particular governments, to add such a premise. Moreover, many linguists would add that Sally could not regard it is as mere accident that these governments are just and that they ought to require employers to pay a living wage: the resolution requires there to be some explanatory connection between the justness of governments and the living wage requirement (see Carlson 2005).
This is good evidence because ordinary speakers have an implicit (but not infallible) mastery over the language in which the resolution is stated. The resolution is stated in English, not in some special debate-specific dialect of English. Facts of usage constrain interpretation. The existential interpretation is not even, as I see it, eligible. So its pragmatic benefits are irrelevant. Compare: I think it would be better if the resolution were, “It is not the case that just governments ought to …” But that’s not the resolution, so it’s not even an eligible interpretation in a T debate. (Here I assume a controversial view about whether pragmatic benefits can justify a semantically inadequate interpretation of the resolution. I cannot defend this view here, but I welcome questions and objections in the comments to be addressed in a later article.)
Some speakers might balk at the generic reading of the resolution. How, they might think, could anyone assent to such a sweeping claim about what just governments ought to do? It seems to depend heavily on the details of each country. I can easily get into this frame of mind. But, equipped with this frame of mind, it’s not as if I would assent to, “Just governments ought to require that employers pay a living wage,” and expect my audience to pick up on the existential reading. I would instead either deny the resolution or suspend judgment about it. This means that the anti-generalization view is not evidence of an eligible existential interpretation; rather, it’s a reason not to affirm the resolution. One more argument for affirmatives to answer! Consider an analogy. Suppose I say, “Dogs are ugly.” You might think it’s silly to say of dogs in general that they are ugly: how could one support such a generalization about the aesthetics of dogs? So you’ll reject my statement. You won’t reinterpret it to mean that some dogs are ugly and agree with it.
I’m sure that many readers will be skeptical of directly appealing to how we ordinarily speak and think. Let me mention a more theoretical explanation of why the resolution is generic. Carlson (1977) suggests that the reading of bare plurals depends on the predicate of the sentence. He distinguishes between highly temporary stage-level predicates like “being here” or “being available,” and more intrinsic individual-level predicates like “having four legs” or “being altruistic.” He calls the former stage-level because they express properties of temporary stages of things: for example, sitting is a property of the present stage of Jake. One might argue that “ought to require” is an individual-level predicate: if just governments have an obligation to require that employers pay a living wage, that is not just a fleeting property of temporary government-stages. I mention this argument just as an illustration of how one might support the intuition with a theory, but I do not endorse the argument.
We can turn next to a less direct argument for the generic reading of “just governments.” But this argument may carry more weight in a T debate.
One of the most trolly observations to make in a debate on this topic is that just governments do not exist. It strikes me as plausible that no actually existing government is just. But most debaters will rightly trust their linguistic intuitions (in this case, but not in others!) and assume that this point is irrelevant to the resolution. The question is: why is it irrelevant?
If “just governments” gets an existential reading, then the point should be relevant. If there are no just governments, then it is not the case that there are some just governments that ought to require employers to pay a living wage. So the resolution is not true. Reading “just governments” as a generic bare plural, then, is key to avoiding the trolly observation as a knockdown negative argument (or a knockdown presumption trigger, if presupposition failure makes the resolution neither true nor false).
One might object that not even the generic reading can avoid the problem. After all, if there are no just governments, then how could it be true that just governments in general ought to do anything? Some linguists have held that generics never presuppose the existence of the kind of thing in question. But others, such as von Fintel (1996) and Greenberg (2003), have endorsed a more modest point, which is still enough for my purposes. This point starts with the fact that the resolution states a rule—namely, that just governments require employers to pay a living wage. Consider the rule,
- Trespassers ought to be prosecuted.
We can affirm this rule even if there are no trespassers. But consider next,
- Some trespassers ought to be prosecuted.
This statement is not true if there are never any trespassers. The lesson is that normative generics do not presuppose the existence of members of the relevant kind. Since the resolution is a normative statement, it does not presuppose that there actually are any just governments.
Note that my argument here is very different from the just-as-trolly response that many make against the trolly observation. This response is that if there are no just governments, then the resolution is vacuously true. This is defended on the grounds that “just governments” refers to all just governments, and that this universal generalization is vacuously true if there are no just governments. My argument differs from this trolly response because I claim that “just governments” is generic, rather than universal: generics, unlike universals, allow exceptions. The trolly response also rests on more controversial assumptions than my response: the word “all” in English, in many cases, presupposes the existence of the things in question. And although there is no existence presupposition when the universal generalization states a rule—e.g., that all trespassers be prosecuted—the absence of trespassers does not make the rule vacuously true: it would not imply, for example, that all trespassers should be shot. More generally, we can’t assume that the English “all” is governed by the same rules as the universal quantifier in an ideal logical language.
There are many other arguments one might make for or against the generic interpretation. Note that these arguments do not apply to all bare plurals, so one cannot assume that “employers” is generic for the same reasons. (I think it is generic, though, for what it’s worth.) Nor do they give any reason to think that the affirmative may not specify a way of requiring employers to pay a living wage.
What do you think?
Carlson, Greg N. 1977. “A Unified Analysis of the English Bare Plural.” Linguistics and Philosophy 1 (3): 413–57. http://link.springer.com/article/10.1007/BF00353456.
———. 2005. “Generics, Habituals, and Iteratives.” In The Encyclopedia of Language and Linguistics. Vol. 2. Elsevier.
Greenberg, Yael. 2003. Manifestations of Genericity. New York: Routledge.
von Fintel, Kai. 1996. “Specific Generics.” Talk Given at Rutgers Colloquium. http://web.mit.edu/fintel/fintel-1996-specific-generics.pdf.