Re: Double minimal pairs

J R wrote:

> Do you accept all of the random pairings, or will you reject ones that
> strike you as really off? Personally, I haven't used any random methods, and
> wouldn't, except perhaps in very small doses (Mark's catchy advert
> notwithstanding!). It kind of takes the con- out of conlanging for me. If I
> want randomness, I'll look at a natlang :-)
>
> Josh Roth
With Tirelat I've accepted only a few of the random pairings. Some of
the earlier languages, like Zharranh, kept a larger number of the random
assignments.