Of Minds and Language (57 page)

Now we ask the question, if you give them dissonant intervals vs. consonant intervals, do they show a preference for consonance as would be predicted from studies of human adults and infants? Results for tamarins and marmosets failed to reveal a statistically significant preference for consonance over dissonance either at the group or individual level. This shows that neither tamarins nor marmosets show a spontaneous preference for consonance. It is unlikely that this result is due to a psychophysical constraint as several prior studies, using both behavioral and neurophysiological preparations, have revealed clear evidence of discrimination. Rather, what our studies show is that despite a physiological capacity to discriminate consonance from dissonance, this is not a meaningful distinction for these animals in that it fails to generate spontaneous preferences for one stimulus over the other.

What about lullabies? We wondered whether there would be a preference for lullabies versus something else, so we started simply, contrasting a non-vocal, flute lullaby with a non-vocal piece of German techno. The younger members of my lab were voting strongly for the techno, and I was praying secretly (though I am not religious) for the lullabies. Consistently, both species preferred the lullabies to the techno. For some, this will either represent an exceedingly trivial result or one not worth discussing because the experiment is so poorly controlled. That is, there are dozens of differences between lullabies and techno, and so the crucial question is: what acoustic properties underlie the preference for what we are describing as a lullaby? As we planned all along,
the lullaby–techno contrast was simply an opening card, designed to see if we could find a systematic preference and if so, then attack the problem to determine what features are in play. I won't discuss all of the conditions, but will focus on one that is quite telling, specifically, the role of tempo. Recall that I mentioned the observation that lullabies tend to have slow tempos. As a result, perhaps the preference for lullabies merely reflects a preference for slow tempos. Thus, we presented a choice between short segments of sound played at either a fast (400 beats per minute) or slow (60 beats per minute) tempo. Consistently, subjects preferred the side playing the slow over the fast tempo. Thus, one crucial factor driving the preference for lullabies may be an evolutionarily ancient bias towards slow tempos. And one good reason for this preference is that if you look at a whole variety of species' alarm calls, they are typically associated with fast tempos. Fast tempoed sounds seem to be coupled with aversion or avoidance of what is going on in their natural vocalizations.

Following a discussion of this work, several colleagues challenged us with different versions of the following question: “Okay, your tamarins might prefer lullabies over techno, and slow over fast tempos, but do they prefer lullabies as we seem to do, and as children do, over peace and quiet?” The answer is a resounding “No!” Both species actually prefer silence to hearing a lullaby. So even though they have a preference for certain kinds of music, there seems to be a strong preference for silence over noise. Perhaps they are just ahead of their time, prescient animals who had to wait for minimalism and John Cage's “4'33”.”

Coming back to some of the themes of this conference, what might be the uniquely human aspect of the music faculty is the interface between evolutionarily shared systems of tempo and frequency discrimination together with the systems that are recruited for emotional processing. That is, we share most, if not all of our capacities for frequency and tempo discrimination with other animals, and a significant proportion of our abilities for emotional processing with animals, but it is the interface between these systems that perhaps uniquely constructed our music faculty.

Two final points to wrap up. What I have tried to argue in this paper is that one way of thinking about the nature of the human mind is to take the lead from much of what is happening in biology much more generally, what has been happening in linguistics more specifically, and running with the idea of universal minimalism, the idea being to look for basic rules and computations. I think this is consistent with some of the issues that Chris Cherniak brings up in his paper (see
Chapter 8
). If you look at some of the core computations that have been invoked in this conference and for the minimalist program more generally – notions like Copy, Move, Merge, Hierarchical Dominance, and so
forth – these are precisely the kinds of operations that are invoked by cellular and molecular biologists such as Mark Kirschner (Kirschner and Gerhart 2005). Secondly, once you invoke notions of modularity such as those that Gabby Dover brings up in his paper, you somehow need to create mechanisms that will enable interfaces between systems. The crucial question is: what is doing the translation, and how do the different representational formats “speak” to each other? In the case of language, for example, how does the representational format that codes for distinctive features in phonology interface with the representational format that codes for concepts within the system of semantics? Lastly, given the promiscuity of these systems to create the variation, ultimately what happens is that the environment is going to prune them back from the biologically given options, and this process will yield the distinctive signature observed in the local environment – thus, the move from I-language to E-language.

I hope this gives a reasonable sketch of the minimalist approach, and how it might open the door to new ways of thinking about our minds and how they evolved.

H
IGGINBOTHAM
: I wanted to raise a question on intended vs. foreseen. I think it is a bit tricky to make the distinction. You may remember that Kant famously said that you intend the consequences of everything you intend. So in the sense of Kant's dictum, in using the weight to save the five people, I also know that death is a consequence and therefore I intend the death of the skinny guy. Moreover, it is a bit of a trick if you ask how these things are conceptualized. Suppose I intend to take a drink of water. So then I stand up and I walk over and I pick up the bottle. Or I flip a switch because I want to find my eyeglasses. If you ask what I intended to do, one might view the situation in the following way. I intended to move my finger like this [stretching finger out] and the rest [moving arm forward] was foreseen consequences. So I think you have to frame it in some way that is rather careful, where you speak of the intended–foreseen distinction in some way that is categorized properly for the agent, and it is not so obvious how to do this just from a description.

Also a correlative question is, I have read a number of books, all of which I think are terrible, about moral permissibility, as if this were some kind of abstract stuff that you can sort of throw out. But is it possible to change any results, or have you considered asking the question in a more first-personal way? Would
you
pull the trigger, vs. should
he
pull the trigger – the dirty-hands thing?

H
AUSER
: Yes, great questions, and I am sympathetic particularly to the first. I mean I think I was trying to foreshadow your question in the sense of saying that I think the notion of principles that I am picking up is really crude, and I think it is crude in precisely the way that you pointed out. For example, try pushing the analogy to language even further. Let's say that the notion of an action, a representation of an action, or what we might call an “acteme,” is like a phoneme – completely meaningless in isolation, and only gaining in meaning as a function of particular sequences, underpinned by intentional states, and generating particular consequences. In this sense, bending your finger may be either meaningful or meaningless. It depends on how it is strung together with intentional states and other surrounding actemes or actions. John Mikhail, who has written a very nice thesis
⁶
and is really my co-collaborator in much of this, intellectually at least, has tried to make much more subtle kinds of distinctions appealing back to some of the philosophy of action, and especially Goldmanesque decision trees. I think the problem is that these trees are not at the right level of grain. And I think all the complications that have been raised in the philosophy of mind and language about the notion of intentionality are not cashed out. Frances Kamm, I think, is one of the few people actually engaged in moving these ideas forward.
⁷
She has created extraordinarily complex dilemmas that largely target the same scenario, the famous trolley problems. For Kamm, however, the issue is not one about empirical research or deep questions about the mind, but about probing our intuitions to decide what is prescriptively or normatively permissible. That said, I am convinced that the kind of work she has put into play will make significant contributions to the empirical studies that we are engaged with.

On the second point, we have approached this question from several other directions. Let me tell you about two of these. If everyone carries around some version of the categorical imperative, then they should answer these scenarios in the same way if they are judging (a) their own actions as the bystander, (b) a third party as the bystander, and (c) themselves or another as one of the possible victims on the track. This would be exquisite evidence of a folk theory of the categorical imperative! Now the problem is, how to dissociate what will clearly be a very strong emotional response to being on the track and saving your derrière. And this is precisely where studies of patient populations enter, and in particular, patients with damage to brain areas associated with emotional processing. This isn't going to directly answer the You/I suggestion, but let me give you a flavor of the move. Consider Antonio Damasio's classic
studies of patients with damage to the ventromedial prefrontal cortex.
⁸
Much of the work on these patients suggests that there is a problem with the connection between the emotional areas in the amygdala and decision areas (this is crudely described) in the frontal lobes. Due to such deficits, these patients appear to have severe problems in the socioemotional domain, including moral decision making. In collaboration with Damasio and several other colleagues, including my two students Liane Young and Fiery Cushman, we have refined our understanding of this deficit by systematically exploring a broader part of the space of moral judgment.
⁹
Cutting a long story short, these patients seem to have a very selective deficit: they only show differences with normals on a certain class of dilemmas which are other-serving personal cases. They are true moral dilemmas in the sense that there is no adjudicating norm that clearly arbitrates between the options, and where one option is to engage an action that is aversive, but where the consequence is to maximize aggregate welfare in terms of saving more lives. Under these conditions, the frontal patients go with the utilitarian outcome, as if the aversiveness of the act was irrelevant.

H
IGGINBOTHAM
: You said the categorical imperative, but actually you meant the utilitarian, I think. It is the utilitarian for whom it doesn't matter who carries out the action.

H
AUSER
: I meant the categorical imperative in the sense of, I think this is a permissible action, I just think it is permissible in the sense of being permissible for any…

H
IGGINBOTHAM
: An imperative is an imperative about maxims, it is not about individual actions, it is about reasons for doing them. It is the utilitarian who has the problem here.

H
AUSER
: Right. More questions?

P
ARTICIPANT
: If I didn't misunderstand you, you related the lack of lexicon in primates to an inability for vocal imitation. Is that right?

H
AUSER
: That is not the sole reason. What I was saying is, that no matter how rich the conceptual system, there are at least two problems, one of which is that even if they could externalize, they can't pass the information on. So there are two problems. Thinking about FLN again, there is both the problem of the mapping between sound and meaning, but there is the additional problem of being able to pass it on.

P
ARTICIPANT
: Yes, but then what I am questioning is your saying it just in terms of vocal imitation, because certainly they can pass on gestures, but they cannot pass on signs, or sign language.

H
AUSER
: Even gestural imitation is extremely impoverished in primates. “Monkey see, monkey do,” just for the record, is a myth. No evidence. The best gestural imitation is weak, very very weak, relative to humans. It has taken literally thirty years to show even the most slight evidence of it. So it is absent vocally, it is weak at best visually.

C
HOMSKY
: On this same point, there is another form of transmission, namely by inheritance. So suppose that you get a smart ape, one that comes up with a combinatorial system. That ape has advantages. It can think, it can plan, it can interpret and so on, and its descendants will have the same advantages even without vocalization. If those advantages are sufficient, they could take over the whole breeding group. They'd all have these capacities and then vocalization could come along later because it is useful to interact. So I think there is a crucial (I'd like to expand the difference between Steve Pinker and me) – I think a possibility is that that is the way the transmission took place.

H
AUSER
: Yes. My quick answer to that is that that would not affect the story I told. Indeed, it would add to it, which I think is perfectly reasonable. In fact, it does enlarge the gap because it says that much of primate thought could have been really moving in quite extraordinary ways by genetic transmission, and then it may have even been a more simple trick of something about the auditory–production loop that got fused, and that could have been a trivial step.