Conclusions and Directions for Future Research. Throughout the paper, we have hinted at a possible trade-o between ex ante co-operation and ex post e ciency. Unfortunately, since we do not model the frequency with which cooperation is needed, nor the magnitude of the bene ts arising from it, our model does not o er predictions on how this trade-o is resolved. This could be a fruitful way the paper can be extended in future research. is related to Xxxxx (1997), who argues that xxxxx, which are better at ghting in court, should hold senior debt, because this will minimize the resources wasted in legal battles. There are two main di erences, however. First, we analyze the allocation of control rights, while Xxxxx (1997) is focused on property rights. Second, in Xxxxx the allocation of property rights itself is su cient to rule out dissipative activities, while in our paper the allocation of control rights a ects the direction of transfers, which then a ects dissipative activities.
Conclusions and Directions for Future Research. In this paper, I have attempted to shed some light on the historical use of any as an NPI. Corpus searches revealed that any was used with high frequency under negation in OE, lower frequency in ME, and high frequency in EMoE. This mysterious U-shaped curve was the primary motivation for this study. We do not see similar U-shaped trends in the historical distribution of any in the other NPI licensing contexts (i.e. conditionals, questions, and comparatives). Previous studies (Xxxxxx 1995; Xxxxxx 1997; Xxxxxx 2002; Kallel 2011) argued that any emerged as an NPI as a product of the loss of negative concord in Middle English. One notable flaw with this hypothesis, which I called the ME Negative Concord Loss Hypothesis, is that it cannot explain why any occurs with such high frequency in conditionals, questions, and comparatives in OE and ME (before they hold that it became an NPI). As these are NPI licensing contexts, it is odd that any would be licensed in these contexts but not negation, since negation is the primary licenser of NPIs. In §5.5, I proposed a four-staged model based on syntactic and semantic features of any and na/no to explain why we see this U-shaped trend. In OE, any was an emphatic negative meaning „not a single‟ or „none at all‟. This emphatic feature was lost in EME. This meant that any and na/no meant the same thing and were competing for the same structural position. This resulted in a drop in the frequency of any under negation. After negative concord was lost between LME and EMoE, any started to fill these same structural positions, resulting in the increased frequency of any under negation in EMoE. This project was a very broad overview of a very narrow phenomenon. Polarity sensitivity is a seminal topic in discussions of the architecture of the grammar of human language (Israel 2011). I began this research because I was disappointed at the few instances of historical accounts of NPIs available. After undertaking it, I found that, like many things in historical linguistics, it is incredibly difficult to tease cause and effect apart. For example, did the loss of negative concord lead any to develop into an NPI, or did the increase in any-words lead to the loss of negative concord? Future research in the historical behavior of any as an NPI is needed to make more sense of what I have found in the present study. Further, this issue should be examined on more dimensions, including historical dialectal variation in the use of any as an N...
Conclusions and Directions for Future Research
