A serial founder effect model of phonemic diversity based on phonemic loss in low-density populations
Fig 2
Plot of the slope of the linear fit to the simulated number of phonemes versus distance from the most likely origin of the Out-of-Africa dispersal, versus the time of phonemic loss (e.g., if the phonemic loss time is 80 generations/phoneme, then 1 phoneme disappears every 80 generations).
In the main figure, results are shown for 40 to 120 generations/phoneme and four different initial populations. The insert shows, as an example, one initial population (speaking a mix of languages with 35–40 phonemes) over a wider range of phonemic loss times. Each error bar has been obtained from 100 numerical simulations, and represents the 95% confidence-level interval. At the start of the simulation only the central node of the grid is populated with a population of N = 10 tribes, speaking a randomly generated mix of synthetically generated languages within the given ranges of phonemes. The maximum number of tribes per node (Ns) is set to 10 tribes/node and the demographic threshold below which phonemes are lost is also set to 10 tribes/node. Net fecundity R0 = 1.4. Time = 2,280 generations or about 70 kyr, corresponding to the present time from the Out-of-Africa dispersal. The dash-dotted and dotted lines give the mean, upper and lower bounds to the slope of the observed cline, as computed in Ref. [18], Fig 1, by performing a linear fit of the number of phonemes of 366 present-day languages versus their distances from the most likely origin of the Out-of-Africa dispersal (UPSID database). The result was [18] slope = -(3.4–6.5)·10−4 phonemes/km (95% confidence-level interval), r = -0.317, p < 0.001 (from Fig 1 in Ref. [18]).