1. 1. Posted by S. Matthew Liao | November 20, 2007 2:07 am

   Thank you very much for this very interesting paper, Gil and Erica! I find the main claim that there could be a universal generative moral grammar very plausible. I just have a few questions regarding the extent to which the moral case is analogous to the linguistic case.

   1. It is has been claimed that in the moral case, one is often subject to framing effects. By this, it is meant that one continues to have a certain intuition about a certain case, even after it has been pointed out that the intuition may not be rational (e.g. Kahneman and Tversky’s Asian Flu Case). I was wondering whether there are any parallels in the linguistic case. That is, whether there are cases in which one continues to have a particular linguistic intuition even after it has been pointed out that the particular intuition may not be rational.

   2. In the moral case, some people believe that there are truly evil people. By this, they mean that these people are not disabled in some biological way; instead, these people do evil things freely and autonomously. Are there parallels of these kinds of people in the linguistic case? I accept that there could be people who intentionally use ‘wrong’ grammatical structures. But how ubiquitous are these cases in comparison to the number of evil people?

   3. Related, in the moral case, one can make sense of evil individuals’ flouting moral norms. For example, one can understand that Jack killed Bob in order to steal Bob’s money. Can one make sense of individuals’ flouting linguistic norms? What does one understand when one thinks that it is ungrammatical to say ‘Jack is have run,’ to borrow an example from you?

2. 2. Posted by Gil Harman | November 20, 2007 10:32 pm

   From Erica Roedder and Gil Harman:

   Dear Matthew,

   Thanks for the comments. What do you think about the
   following tentative answers to your questions 1-3?

   (1) Here are examples in which one may continue to have
   a particular linguistic intuition even after it has been
   pointed out that the particular intuition may not be
   correct:

   “The not unhappy king finally went to bed” seems OK
   even after noting the deviance of other cases of the form
   the + not + ADJ + noun, like “The not sad king …”

   “Jack built the house that the cheese that the rat that the
   cat that the dog chased bothered ate lay” seems deviant even
   though it is probably grammatical (but hard to parse).

   (2). This is a really interesting issue about possible
   parallels between those who do things they take to be
   morally wrong and those who say things that they take to be
   syntactically (or semantically) incorrect.

   First, we might think that some truly evil people are
   “biologically” disabled. For instance, if we take certain
   emotive responses to be part of normal moral processing
   (whether or not they are counted as part of the moral
   grammar, specifically), then psychopaths appear to be
   biologically disabled.

   But we take the more interesting question here to be about
   those who are not biologically disabled, but still evil.
   One way to handle these cases (and this might generalize to
   question 3), is to point to a distinction between moral
   action and moral judgment. Your moral grammar determines
   what you judge to be moral. Evil people, on this construal,
   continue to correctly judge what is right/wrong, but they do
   the wrong thing. There isn’t really an analog here in
   linguistics–i.e., people who have correct linguistic
   intuitions, but say the wrong thing (putting aside what we
   discuss in (3) below). This makes sense: we have, as
   you mention, reasons to act against our moral judgments
   (e.g., stealing in order to get a coveted item), but we
   rarely have reasons to act against our linguistic
   intuitions.

   This leaves, as a final case, those who are not suffering
   from a biological deficit, but who make incorrect moral
   judgments. There are a lot of reasons this might happen.
   For instance, someone may grow up in a racist society, and
   develop racist moral judgments. It’s unclear whether to
   include such judgments as part of that person’s moral
   grammar (e.g., depending on a parameter setting) or as
   “performance errors.” If such judgments reflect the
   person’s i-morality, this aspect of i-morality may or may not
   be part of the “core grammar,” which a child acquires
   naturally, or instead an artificial addition, like Esperanto
   or perhaps utilitarianism. We can hope that more empirical
   work on a moral grammar will sort these issues out.

   (3) We are not sure we understand the significance of the
   question about possible parallels between evil flouting of
   moral norms and individual’s flouting linguistic norms.

   However, we would point out that real speech is
   full of all sorts of grammatical problems. For instance,
   people change their minds in the midst of their sentences.
   So, humans are robustly able to understand sentences that
   flout linguistic norms. Presumably this is because human
   language is highly redundant and we can use contextual cues
   to determine what the person probably meant to say.

   Even out of context, it is possible to understand a
   sentences like “The baby seems sleeping,” which is
   ungrammatical but its meaning is clear.

   And someone might with stylistic purpose deliberately
   exploit this ability people have to interpret ungrammatical
   remarks. (Perhaps certain uses of metaphor are of this
   sort?)

   Also there are cases of doubletalk in which a speaker
   deliberately produces nonsense in order to confuse or amuse
   an audience.

   What do you think?

   Erica and Gil

3. 3. Posted by S. Matthew Liao | November 21, 2007 12:49 am

   Thanks very much for the responses, Erica and Gil, which are very helpful. Here are some further thoughts:

   You said:

   “The not unhappy king finally went to bed” seems OK even after noting the deviance of other cases of the form the + not + ADJ + noun, like “The not sad king …”

   I’m not sure about this example, because first, ‘unhappy’ already involves a negation of a noun, that is, un-happy, and secondly, there are two types of negations here, namely, a ‘not’-type negation and an ‘un’-type negation. Perhaps it is ok to have double negations especially in cases that involve more than one type of negation (not being a linguist, I’m not sure about this).

   You said:

   “Jack built the house that the cheese that the rat that the cat that the dog chased bothered ate lay” seems deviant even though it is probably grammatical (but hard to parse).

   Perhaps you can explain how this sentence is grammatical, and I’ll see if I continue to think that it is ‘ungrammatical.’ Recall that in the Asian Flu Case, people continue to have the intuitions they have even after the loss/no gain distinction has been explained to them.

   You said,

   (3) We are not sure we understand the significance of the question about possible parallels between evil flouting of moral norms and individual’s flouting linguistic norms.

   Sorry I wasn’t being very clear in (3). By ‘understanding’ and ‘making sense,’ I just mean (as you have already discussed in (2)) that we can often understand the motivation of people who would act in an evil way, whereas in contrast, it seems hard to understand the motivation of people who would intentionally flout linguistic norms.

   In other words, in (2), I’m asking whether you think there could exist people who would intentionally flout linguistic norms, and if so, how ubiquitous such cases would be; and in (3), supposing that there could be people who would intentionally flout linguistic norms, I’m wondering whether we could make sense of their motivation to do so in a way that is comparable to the way we can make sense of the motivation of people who would act in an evil way. As a possible response to my point, you might just deny that we can make sense of the motivation of truly evil people. I’m not sure if such a response is adequate, but some people sometimes say ‘I don’t understand how he can do something like this…or, it is just unimaginable that she did this…’

   I hope these comments are somewhat useful.

4. 4. Posted by Erica Roedder | November 21, 2007 4:32 pm

   Dear Matthew,

   Thanks for your further comments. These are quite interesting and useful to us. We hope the following response makes things clearer.

   First, about “the not unhappy king finally went to bed” — we cannot demonstrate that the sentence is ungrammatical, but it may be. (Langendoen and Bever 1973 argued for this onclusion.)

   One argument is, as we mentioned, that given a phrase of the form “the ADJ NOUN,” the phrase “the not ADJ NOUN” is normally deviant. Exceptions are cases where the ADJ begins with a negative prefix like “un” and perhaps “in”.

   Not every such case is clearly an exception. Consider “the not unafraid king finally went to battle,” which strikes us as deviant.

   So, one view in linguistics is that a phrase like “the not unhappy king” is ungrammatical even though it seems OK, because to introduce a special grammatical rule to cover this cases does not fit with other aspects of the grammar. This would mean that the apparent nondeviance of “the not unhappy king” is due to performance factors rather than grammar alone.

   We offer this as a possible example without being able to provide an appropriate article length defense of this possibility.

   Second, about “Jack built the house that the cheese that the rat that the cat that the dog chased bothered ate lay in” — consider the following examples:

   Jack built the house that the cheese lay in.
   Jack built the house that the cheese (that the rat ate) lay in.

   Jack built the house that the cheese (that the rat (that the cat bothered) ate) lay in.

   Jack built the house that the cheese (that the rat (that the cat (that the dog chased) bothered) ate) lay in.

   At some point the center embedding becomes too deep to be processed, but this seems to be due to performance factors rather than grammar. But even if this claim about grammar is understood and granted, the final sentence (at least) will probably still have the appearance of deviance.

   Finally, you ask whether “there could exist people who would intentionally flout linguistic norms, and if so, how ubiquitous such cases would be; and … supposing that there could be people who would intentionally flout linguistic norms, … whether we could make sense of their motivation to do so in a way that is comparable to the way we can make sense of the motivation of people who would act in an evil way.”

   Here we are inclined to think the motivations of evil people and ungrammatical people are just plain different. There is an analogy here, but they are still separate spheres of human behavior. Morality and language have two very different purposes. So, we would expect the intentions of those who deviate from morality, as opposed to those who deviate from language, to be quite different.

   We would also repeat the point that deviant judgments are different from deviant actions. Deviant actions are, strictly speaking, beyond the scope of the moral grammar. The moral grammar generates moral judgments. So, we are not committed to the existence of an analogy between deviant actions and deviant linguistic utterances.

   Erica and Gil

5. 5. Posted by S. Matthew Liao | November 21, 2007 6:14 pm

   Erica and Gil, this is very helpful, thanks!

   I’ll have to think more about the distinction between judgment and action and what that means for your analogy.

   Just a quick comment on

   Jack built the house that the cheese that the rat that the cat that the dog chased bothered ate lay in

   Initially, ‘in’ was left out, so I couldn’t quite work out how this sentence could be grammatical.

   Now take,

   Jack built the house that the cheese lay in.

   Isn’t this already ungrammatical, as it ends a sentence in a preposition? Shouldn’t it be ‘Jack built the house in which the cheese lay’?

6. 6. Posted by Gilbert Harman | November 25, 2007 5:20 pm

   Dear Matthew,

   Sorry about the omission of the final preposition from our
   original example!

   To answer your question about the corrected example: For
   most people, it is grammatically OK to end an English
   sentence with a preposition, although some “prescriptive
   grammarians” advise against it. One response to such
   advice, often attributed to Winston Churchill, is “This is
   the sort of nonsense up with which I will not put.”

   How do you feel about a sentence like “This is the house
   that the cheese was in”?

   Anyway, it is easy to find examples that do not end in a
   preposition: “This is the cheese that the rat that the cat
   that the dog that the neighbor owned chased bothered ate” –
   “This is the cheese (that the rat (that the cat (that the
   dog (that the neighbor owned) bothered) chased) ate).”

   Erica and Gil

7. 7. Posted by S. Matthew Liao | November 29, 2007 7:19 pm

   Dear Erica and Gil,

   I wonder if the preposition case might in fact be a good example for you. It might be said that, from an a priori perspective, prepositions should not be at the end of a sentence, as they would then be postpositions. Yet, intuitively, many people think it is ok to put them at the end. This seems to mirror the Asian Flu Case in which many people report a difference between saving and killing, although, given the way the case is constructed, from an a priori perspective, this difference appears not to be relevant.

   On the

   “This is the cheese (that the rat (that the cat (that the dog (that the neighbor owned) bothered) chased) ate),”

   I have two thoughts regarding it. First, I have a hypothesis that the reason one thinks that this sentence is deviant is due to cultural considerations. In particular, I hypothesize that Germans might not be as inclined to think that this sentence is deviant, given that they put many verbs at the end of their sentences. Secondly, suppose the sentence were deviant. I wonder if the reason is because it improperly uses ‘that’ too many times. ‘That’ is a restrictive modifier, so perhaps it should only be used once. Using it more than once in a sentence creates ambiguities in the sentence. Indeed, in a simpler variation of the sentence, one would say

   This is the cheese that the rat, which the cat chased, ate.

   On Churchill, I know it’s suppose to be a joke, but isn’t ‘up’ an adverb rather than a preposition? :P

   Again, I find your main thesis to be very plausible. These are just tangential issues.

Post a comment