On Nov 16, 2006, at 2:22 AM, Mark Jones wrote: > ******* > Subject: Re: Diachronic conlanging; linguistics forums > > On Tue, 14 Nov 2006 18:24:33 -0600, Eric Christopherson > <[log in to unmask]> > wrote: >> Actually, I think it'd be really cool to "reverse engineer" a natlang >> (or natural protolang) to create a conlang which could be its >> parent. Has anyone attempted that? > > Many times by many famous people. For instance, what else do you > think "Proto-Indo-European" is? > ******** > > Eldin is quite right that Proto-Indo-European (and any other Proto > Language, for that matter) is essentially a bit of reverse > engineering, via the techniques of comparative and internal > reconstruction, though with a Conlang you would have complete > control over the extent and quality of the data, so the challenges > would be similiar in kind, but not in degree. And you don't have > esteemed colleagues telling you you're wrong for perfectly good > (but differently weighted) reasons. Yes, that is correct. I think that what I asked was (accidentally) only part of what I intended to ask. I was actually wondering if anyone had attempted to construct a protolang for two languages which are not known to be related, e.g. PIE and Proto-Semitic. I think that'd be really tough, but the result could be very cool.