Research from beyond SpecGram has found some surprising associations between phoneme inventory size, population size, and distance from the likely East African origin of our species (Atkinson 2011). Larger populations tend to speak languages with more phonemes, and phoneme inventory size decreases with distance from East Africa. Subsequent work has confirmed these associations with larger datasets (Daland 2015).
Unfortunately, little attention has been paid to the most likely explanation of the data, which was given in a SpecGram podcast several years ago (Language Made Difficult, Vol. 5): phonemes have mass, and in order to survive the long migration out of Africa, people had to eat their phonemes. The starvation which necessitated phoneme consumption took lives, explaining why population sizes also diminish with distance from East Africa. An alternative theory for the population size data has it that smaller groups literally and physically cannot support large phoneme inventories. Eventually the phonemes become too massive to be carried around by a migrating speech community.
We wish to take up this neglected area of research in order to provide some vital information on the nutritional value of phonemes. This will allow readers to answer questions like the following: if I eat my back vowels the next time I want a snack, will I just be hungry again in five minutes? At this point we ask the editors to confirm that such questions are indeed frequently found in letters to the SpecGram editorial board.
Based on the figures in Daland (2015), we estimate that a community of 1,000 people need approximately 26 phonemes to sustain themselves. We assume that fears of radical phoneme loss lead to equal munching of all parts of the inventory, although further research should consider whether the vanishingly rare clicks are especially nutritious. With 2,000 kcal as the recommended daily intake, the average phoneme in an inventory of 26 would have supplied around 77 kcal per person per day. We will suppose that phonemes provided all of the nutrition in these trying times, with speakers unwilling to begin eating them until they had no other food. If we instead had a group of 5 speakers, a smaller inventory of 22 phonemes would be sufficient, each phoneme giving 90 kcal per person per day.
In larger inventories, individual phonemes thus provide less energy. This relationship can be understood intuitively: suppose that your favorite food is /t͡ɬʼ/, so that this phoneme satisfies a large part of your calorie needs. You can now see that you don’t need as much food from other phonemes, which consequently provide less of your energy. But if your diet is a healthy mix of a large inventory
We now call on you, dear reader, to continue this important work in future publications. Which phonemes are the most nutritious? And which taste best? Are /i, a, u/ common because they taste good or because they provide important nutrients? Is there a way for speakers of phoneme-
This bibliography is intentionally left blank