Fig 1.
Languages with ejective consonants in the PHOIBLE database.
Fig 2.
Languages with uvular consonants in the PHOIBLE database.
Table 1.
The eleven macroareas used in this study to control for large-scale contact-induced areality.
Table 2.
Mean and median of altitudes for languages with and without ejectives and uvulars.
Table 3.
Mean altitude of languages with and without ejectives and uvulars for the eleven macroareas defined for this study and the number of languages in each group.
Fig 3.
Number of uvular consonants (left) and ejective consonants (right) in the PHOIBLE database as modified for this study.
Fig 4.
Plots of intra-area and intra-family means for large language families with uvulars against altitude means; lines represent Lowess scatterplot smoothers.
Fig 5.
Plots of intra-area and intra-family means for large language families with ejectives against altitude means; lines represent Lowess scatterplot smoothers.
Fig 6.
Presence or absence of ejectives and uvulars in extant Indo-European languages for the PHOIBLE data alone.
Fig 7.
Presence or absence of ejectives and uvulars in extant Indo-European languages for the amended data by means of grammar mining.
Fig 8.
Presence or absence of ejectives and uvulars in extant Sino-Tibetan languages for the PHOIBLE data alone.
Fig 9.
Presence or absence of ejectives and uvulars in extant Sino-Tibetan languages for the amended data by means of grammar mining.
Fig 10.
Estimated probability of ejectives as inferred by stochastic character mapping in Indo-European for the PHOIBLE data alone.
Fig 11.
Estimated probability of ejectives as inferred by stochastic character mapping in Indo-European for the amended data by means of grammar mining.
Fig 12.
Estimated probability of uvulars as inferred by stochastic character mapping in Indo-European for the PHOIBLE data alone.
Fig 13.
Estimated probability of uvulars as inferred by stochastic character mapping in Indo-European for the amended data by means of grammar mining.
Fig 14.
Estimated probability of ejectives as inferred by stochastic character mapping in Sino-Tibetan for the PHOIBLE data alone.
Fig 15.
Estimated probability of ejectives as inferred by stochastic character mapping in Sino-Tibetan for the amended data by means of grammar mining.
Fig 16.
Estimated probability of uvulars as inferred by stochastic character mapping in Sino-Tibetan for the PHOIBLE data alone.
Fig 17.
Estimated probability of uvulars as inferred by stochastic character mapping in Sino-Tibetan for the amended data by means of grammar mining.