Tomlin (1986) buruzki Millinson eta Balke (1981): "... a 100 language convenience sample..."

Tomlin (1986) dú jarraitzen bere revisioa aipátuz lana on Millinson eta Balke (1981), nortzuk aurkeztu zutén 100 hizkuntzako lagin bat, zeintan berriro naguzitzen dirén komenientzia irizpideak gain irizpideak on representazio orokorra, edo bestela esanda:

... a 100 language convenience sample..." [Tomlin, 1986:20]

Tomlin-ek adibidetzen ditú gabeziak zein aurkitzen dituén an lagina ganik Millinson eta Balke (1981):

The sample is biased genetically. For instance, Indo-European represents 10% of the Mallinson and Blake sample, but only 3% of the estimated total population. Nilo-Saharan represents 9% of the Mallinson and Blake sample, but only 2.5% of the world total. Large groups like Niger-Kordofanian or Austronesian are systematically undersampled, Niger-Kordofanian representing 9% rather than 20% of the estimated total, Austronesian 8% rather than 16%. In fact, only a few groups, Sino-Tibetan for example, approximate their estimated proportions in the languages of the world . The sample shows similar biases with respect to areal distributions. [Tomlin, 1986:20]

Esan nahi baita ze ezin atera daitezke konklusio ondo fundamentatuak respektu banaketa orokorra on munduko hitz-ordenak. [2755] [>>>]