Summary CAN .1 Primitive Syntax

331

5.42.4 Summary

Our corpus yielded the following syntactic combinations: CAN THINK CAN SAY CAN DO CAN BE GOOD The following combinations were missing from our data: CAN HAPPEN CAN BE BAD 332

Chapter 6 Analysis and Conclusion

In this chapter, we will be examining the raw results of our corpus search together with my own native intuitions to evaluate the viability of each of the syntactic valences in HCE. A gap in our data, that is, the absence of examples of a given syntactic valence, requires explanation. We need to explain why it is there and whether it should be interpreted as serious evidence against the existence of a particular valence. A gap might be incidental, the result of a defect in the NSM syntactic specification. Also, since the syntax of some primitives is unsettled, occasionally two alternative universal frames are proposed for the same meaning. For example, the frame YOU ARE A GOOD PERSON is a proposed alternative to the standard frame YOU ARE GOOD. Presumably, finding either one of these frames will establish that HCE has access to that meaning. A gap may be accidental, that is, we have so few examples of a given primitive that the probability of finding an example in the appropriate syntactic frame is relatively small. On the other hand, a gap may be systematic. If the gap manifests itself systematically for a particular class of syntactic constructions, this would indeed be persuasive evidence against the existence of a particular syntactic valence. For example, the most recent version of NSM which has not yet been as throughly investigated contains the primitive FAR. Our corpus contains examples of YOU ARE FAR, but no examples of I AM FAR. Should this be regarded as an accidental gap in the data? Or is there a real problem with the proposed valence? Both native intuition and common sense reject the latter possibility. First of all, FAR is relatively rare in our corpus. Secondly and more importantly, there is no principled reason why, in a given language, one would be able to say I AM FAR but not be able to say YOU ARE FAR. It simply doesn’t make any sense. Suppose, however, FAR was never predicated of people in HCE, but always of places. This kind of gap would be far more serious.

6.1 YOU, I, SOMEONE, and PEOPLE

There were no problems with YOU, I, and SOMEONE as subjects to the mental predicates THINK, KNOW, WANT, SAY, and FEEL. There is, however, a possible problem with the primitive PEOPLE. Earlier, I argued that contrary to the NSM proposal, the primitive PEOPLE did not combine with the NSM determiners and quantifiers. While determiners and quantifiers allow us to refer to individuals, the primitive PEOPLE does not. In the same way, is it possible that FEEL is a property of individuals and not of groups? I don’t think so. I think rather the absence of the combination PEOPLE FEEL is an accidental gap. Both FEEL and PEOPLE are relatively infrequent lexical items. More importantly, sentences such as the following sound perfectly fine to me: 128 pipo fio baed if dei no kaen heop m aut ‘People feel bad if they can’t help someone out’ 129 pipo fio baed if no mo nating fo du ‘People feel bad if there is nothing to do’