> Claudio Gnoli, many years ago, both questioned and reconsidered this
> and proposed what has long been known to some of us as the "Gnoli
> triangle." The point he was making was that most (all?) conlangs fall
> somewhere in the triangle, rather than right on the apexes. I put a
> slightly modified version of the Gnoli triangle on my webpage:
>
http://www.carolandray.plus.com/Glosso/Glossopoeia.html
>
> I retained the Claudio's triangle image, tho I did at one time wonder
> if three intersecting circles might be more appropriate.
>