I was asked by a reporter from the journal Nature to comment on an article entitled “Mozambican Grass Seed Consumption During the Middle Stone Age” by Julio Mercader in the journal Science, citing evidence that humans consumed grain much earlier than was previously thought. In addition, one I was asked to comment on an article in Scientific American entitled “Humans feasting on grains for at least 100,000 years,” by Katherine Harmon.
“Dear Dr. Cordain,
I am a huge fan of your books and have been eating a Paleo Diet for years. I’ve had your graph of the land/water meat/fish, and fruit, nut, seed % breakdown taped to my fridge for years, although I’m so familiar with it that I no longer need to look at it for reference.
The Paleo Diet is predicated upon the fact that humans did not have grain cultivation and consumption until 5,000-8,000 years ago, which coincides with the advent of ‘modern’ civilization diseases. Up until now, this hypothesis have not been challenged as the archival evidence of grain agriculture matches it.
However, the current issue of Scientific American has an article, “Humans Feasting on Grains for at least 100,000 Years” stating archaeological evidence that humans were eating grains 100,000 years ago.
I am most curious about your opinion about this and how it effects the Paleo dietary theory.
Dr. Cordain’s Response
“Mozambican grass seed consumption during the middle stone age,” is an interesting paper as it may push probable (but clearly not definite) cereal grain consumption by hominins back to at least 105,000 years ago. Prior to this evidence, the earliest exploitation of wild cereal grains was reported by Piperno and colleagues at Ohalo II in Israel and dating to ~23,500 years ago (Nature 2004;430:670-73). As opposed to the Ohalo II data in which a large saddle stone was discovered with obvious repetitive grinding marks and embedded starch granules attributed to a variety of grains and seeds that were concurrently present with the artifact, the data from Ngalue is less convincing for the use of cereal grains as seasonal food. No associated intact grass seeds have been discovered in the cave at Ngalue, nor were anvil stones with repetitive grinding marks found. Hence, at best, the data suggests sporadic use (and not necessarily consumption) of grains at this early date. Clearly, large scale processing of sorghum for consumption for extended periods seems unlikely.
Further, it should be pointed out that consumption of wild grass seeds of any kind requires extensive technology and processing to yield a digestible and edible food that likely did not exist 105,000 years ago. Harvesting of wild grass seeds without some kind of technology (e.g. sickles and scythes which are not present at this time) is tedious and difficult at best. Additionally, containers of some sort (baskets, also not present at this time), pottery (not present) or animal skin containers are needed to collect the tiny grains. Many grain species require flailing to separate the seed from the chaff and then further winnowing (baskets (not present), or animal skins) to separate the seeds from the chaff. Intact grains are not digestible by humans unless they are first ground into a flour (which breaks down the cell walls), and then cooked (typically in water – e.g. boiling (technology not present)) or parched in a fire which gelatinizes the starch granules, and thereby makes them; available for digestion and absorption. Because each and every one of these processing steps requires additional energy on the part of the gatherer, most contemporary hunter gatherers did not exploit grains except as starvation foods because they yielded such little energy relative to the energy obtained (optimal foraging theory).
If indeed the grinder/core axes with telltale starch granules were used to make flour from sorghum seeds, then the flour still had to be cooked to gelatinize the starch granules to make it digestible. In Neolithic peoples, grass seed flour most typically is mixed with water to make a paste (dough) that is then cooked into flat breads. It is highly unlikely that the technology or the behavioral sophistication existed 105,000 years ago to make flat breads. Whole grains can be parched intact in fires, but this process is less effective than making flour into a paste and cooking it to gelatinize the starch granules. Hence, it is difficult to reconcile the chain of events proposed by the authors (appearance of sorghum starch granules on cobbles or grinders = pounding or grinding of sorghum grains = consumption of sorghum). I wouldn’t hang my hat on this evidence indicating grains were necessarily consumed by hominins at this early date. To my mind, the Ohalo II data still represents the best earliest evidence for grain consumption by hominins.
Loren Cordain, Ph.D., Professor Emeritus