SUBJECT: Re : anaph Thanks for the work you did on the intrasententials . I 'm going the get the results from that and put them in the paper : when do you want to have a look at it ? I 'm leaving to go to &NAME on the 7th ( Monday ) late afternoon , but I have n't finished the paper yet ... I could give you a draft without any results tomorrow , but I 'll probably only be able to put results in Thursday-ish . Let me know . Thanks for the resolution corrections , most were OK according to my &NAME criteria : I only resolved third person pronouns , that 's why ' you ' was n't linked . Antecedents were only chosen from available noun phrases , so if the tagger got a part of speech wrong , a pronoun may have been assigned antecedent_does_not_exist . What 's the difference between xmods and xcomps ? I 've been trying to get my Re-implemented &NAME software to output basically all relations , but I 'm really struggling with the comps and mods . I do n't seem to be able to find a distinguishing pattern myself , so I ca n't get a computer to do it ! E.g. the xmods / xcomps with ' to ' : xcomp to wanted wait xmod to wait make in ' ... wanted to wait to make ... ' How can you have : xcomop to continue place but xmod to work pass ? Etc. Is there a reference I could have a look at that would explain it to me ? Thanks , &NAME . On &NAME , &NUM Mar &NUM , &NAME &NAME wrote : <QUOTE> I 've done as much as I can do now -- in ~&WEBSITE Next week I have to focus on a couple of grant proposals and am in &NAME &NAME / &NAME &NAME <END_QUOTE>