[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
--- In jboske@yahoogroups.com, "Jorge Llambías" <jjllambias@...> wrote: > > On 10/4/06, John E. Clifford <clifford-j@...> wrote: > > > > Indeed, I noticed that the notion of levels does not work directly at > > all well. The level below a pluality node is all the pluralitis > > lacking one memner of the head, but it is also all those members > > individually, so that successors are already level mixing, if levels > > count. > > The notion of levels works in a kind of relativistic fashion. Same-level > is akin to present in relativity: there is an absolute future and an absolute > past, but what counts as present depends on the frame of reference. > Similarly there are absolute upper levels and lower levels of the lattice > for a given node, but what counts as same-level will depend on the > particular model. So Spot and Fido minus one hair can be in the same > level but Spot and Spot minus one hair (or Spot and the Golden Retriever) > cannot. I take all this as saying something like (though I'm not sure just what) that, from an absolute lattice, say Mr. Dog and all among it, a particular conversation constructs a sublattice of relevant items, maintaining their relative positions, but dropping intermediate nodes in various ways, to suit the circumstances. Thus, for example, talk about dog breeds would create a piece of the dog lattice in whihc all the breeds were on "the same level" regardless of how they fell outin the original. Similarly, just talking about dogs might put all dogs (even various defective ones or both the part and the whole of one) on the same level. And this level becomes the focus of remarks, the other levels available as framework but not directly as topic -- or so only with some extra effort. > > > In tthe > > model I presented, "among" is exactly the relation between a node and > > any node above it on a constantly upward path. In particular, each > > dog is among Mr. Dog. I hope you can explain what you thought I meant > > or, more exactly, what you did mean (and still do, I suppose). > > I thought you meant the relationship that exists for example between > a word token and a word type (a token can be among several tokens > but not among its one type) or between a particular event occurrence and > the event that is repeated. The occurrence can be one among several > occurrences, but not one among the one event that is repeated. > > Maybe this is just a matter of terminology, but I think it's useful to > distinguish the metalinguistic subsumption relationship between things > that are normally not in the same domain of discourse, and the "among" > object language relationship that holds between things in the same > domain of discourse. Well, I don't understand two of the distinctions here: among v subsumption (especially the latter) and token v type as different again from amongness. On the latter, I would think that the main place (surely not the only one) for a token to be would be among its type: if not there, then in what sense is it really a token of that type? They may be proper to different discourse, but that doesn't affect how they are in fact (metalinguistically if you will) arranged. And indeed it seems that token-type pairs tend to turn up a lot in discourse -- as does breed-dog and other such conventional genus-species pairs.