Is there an alphabet gene?

I’ve been ruminating for the past couple of weeks about a little speculative article by Peter Frost, a Canadian evolutionary anthropologist whose primary work is on human sexual dimorphism.  In ‘The spread of alphabetical writing may have favored the latest variant of the ASPM gene’ (Frost 2008), Frost makes the remarkable assertion that there is a gene variant whose distribution is best explained by its use for cognitive tasks relating to alphabetic literacy (and specifically not non-alphabetic literacy).

This is a non-peer-reviewed paper in the journal Medical Hypotheses, which publishes papers that its editors (all medical scientists) decide are worthy of note, even when (or especially when) they challenge conventional wisdom.  I really like the publish-then-review model, but I do wonder whether in this case what is really needed is a journal (let’s call it, hypothetically, Social Hypotheses) to allow social scientists have a role in determining what is likely to be important or interesting.  Because, while the genetic evidence is fairly straightforward, the cognitive and more importantly the historical evidence are the truly controversial elements of Frost’s paper.  It rests initially on the following facts, which, not being an expert on human genetics, I’m just going to grant for the sake of argument:

– There is a gene, ASPM, that regulates brain growth, and that has evolved many variants, the latest of which emerged around 6000 years ago in the Middle East.

– It is much more common today in populations in Europe and the Middle East than in East Asia.

– While it relates to brain growth, it does not correlate with increased IQ, suggesting that its cognitive function is subtle.

Frost argues from this, quite plausibly, that this latest variant assists performance on some task relating to cognition and that expanded from a Middle Eastern origin starting around 6000 years ago.   He then moves on to the evidence I am more familiar with, to argue that that task was alphabetic writing.

– Writing developed in the Middle East around 3000 BCE, and phonetic alphabets around 2000-1000 BCE. This is sort of true; there is increased phoneticity in Near Eastern scripts over time, and purely phonetic scripts (like Ugaritic and Proto-Sinaitic) emerged as early as the 18th-15th centuries BCE (Lemaire 2008).    However, this is in the range of 3500-4000 years ago, not 6000, which makes the emergence of the ASPM variant at 6000 years ago rather early for his timeframe.

– Literacy levels in the ancient world range from 10% – 33% of adult males. False: this may have been true of Roman citizens (which is where Frost’s data come from), but was decidedly not true in the ancient Near East including both Egypt, Mesopotamia and the Levant.  Literacy rates certainly varied, but were probably in the 5-7% range at most in the period under consideration (van der Toom 2007: 10).

– East Asian writing is pictorial or ‘ideographic’. False; although there are logographic (word-writing) and ideographic (idea-writing) elements to the Shang Dynasty script (the earliest East Asian writing), just as with modern Chinese writing, there are phonetic complements, rebuses, and other linguistic aspects to the script from an early date.  John DeFrancis has essentially demolished the myth of Chinese picture-writing, showing it to be factually inaccurate and often marshalled for derogatory purposes (Defrancis 1989). It is true, however, that there was no ancient alphabetic writing in East Asia.  However, Japanese syllabic writing (kana) has been around since the 6th-7th centuries AD.

– Alphabetic writing has different cognitive advantages and demands than non-alphabetic writing. Yes, true, but not in simple or easily understandable ways (Olson 1995).  Frost is asserting on the basis of some experimental evidence that information is processed differently by Chinese literates reading in Chinese.  But Chinese is only one non-alphabetic script, and it has only been compared cognitively to the Latin alphabet we have no systematic comparative idea of how differently structured scripts affect cognition.  Frost’s reasoning certainly implies that Japanese writing should have similar effects to alphabetic writing, despite the near-absence of the ASPM variant there.

– Scribes were prestigious individuals who were highly valued, and those who excelled at scribal tasks would have been better-nourished and wealthier, thus more equipped to have greater reproductive success. Possibly true.  However, we have no evidence that they actually did have such success to a greater degree than non-literate elites, or indeed that they enjoyed such success at all.  Frost reasons from their social value to their reproductive success, which is plausible but unverified.

– Thus, if the ASPM variant did aid in the processing of alphabetic information, then it would be selected for among scribes and indeed among their offspring, some of whom would be literate but others of whom would not.

It will be clear, I should hope, that my difficulties with the points of evidence above render me highly skeptical of Frost’s conclusion.

As Frost notes, we would expect to find higher rates of the ASPM variant among societies that have long histories of alphabetic writing, if his theory is correct, and lower rates among societies that lack such histories. Good evidence for his position would be, as he notes, if African groups like Hausa and Fulani (long-term alphabetic scribal traditions) had high levels of the variant but other groups didn’t (similarly, if there were contrasts between Chinese and Japanese, or between Georgian and Chechen, this would tend to be confirmatory).

But he does not mention another important prediction that I think has been tested and refuted:  If the new ASPM variant plays the role he says it does, then East Asians should have difficulty learning alphabetic writing, even if raised using only alphabetic scripts.   There simply is no evidence that this is the case, and if it were true, would have incredible implications for public policy.  Moreover, the widespread use of purely phonetic scripts like the Japanese syllabaries, which really ought to have the same cognitive consequences as alphabets, should be problematic for East Asians.  It isn’t, and this is serious disconfirmatory evidence against the hypothesis.

Frost is right, though, that whatever factor selected for this ASPM variant must have been present / emerged around 6000 years in the Near East but should not have emerged or been significant in East Asia, and factors such as ‘food domestication’ and ‘urbanization’, which emerged in both regions, won’t suffice.

But what about the possibility that the ASPM variant helps promote encephalization with respect to particular plant  domesticates – e.g., wheat / barley, the classic Near Eastern domesticates, as opposed to rice and millet (the ancient East Asian domesticates)?  Here, the explanation is that individuals with the ASPM variant in the Near East had greater reproductive success because they were better able to use local domesticates to promote encephalization.  Individuals who lacked the variant still derived nutrition from these foods, but not in a way that contributed to brain growth. Obviously, I’m not a nutritional anthropologist or a dietician or even an expert on human evolution.  I’m not proposing this simply as a plausible alternative, given the complete insufficiency of the alphabetic hypothesis.  It is also testable – one would expect that grain-eating societies would have higher levels of the variant than non-grain-eaters in the same general geographic areas, all other things being equal.  But now we are back in the realm of Medical Hypotheses and outside anything to which I could claim to be anything more than an interested nonspecialist.

Works cited

DeFrancis, John. 1984. The Chinese language: fact and fantasy. Honolulu: University of Hawai’i Press.

Frost, Peter. 2008. The spread of alphabetical writing may have favored the latest variant of the ASPM gene.  Medical Hypotheses 70(1): 17-20.

Lemaire, Andre. 2008. The spread of alphabetic scripts (c. 1700 – 500 BCE). Diogenes 55(2): 45-58.

Olson, David R. 1995. Towards a psychology of literacy: on the relations between speech and writing. Cognition 60(1): 83-104.

van der Toom, Karel. 2007. Scribal culture and the making of the Hebrew Bible. Cambridge, MA: Harvard University Press.



  1. How are literacy rates calculated for antiquity? It was only recently brought to my attention that the ability to read does not necessarily correlate to an ability to write; and that historically, far more people would have been able to read than to write. The argument as I encountered it was for the Victorian period, but it’s led me to reassess what little I know about the history of literacy and how it’s deduced.

  2. The answer to that question is roughly ‘With large margins of error and a great deal of guesswork’. There are enormous definitional issues regarding literacy even in modern contexts, and for antiquity, these problems are compounded by a lack of accurate recordkeeping and the perils of survival of materials. There are also important issues in antiquity related to genre, because many people may have had limited-scope literacy restricted to particular types of writings (esp., e.g., in non-alphabetic scripts with hundreds of glyphs of which any person may only know the ones of relevance to a particular task).

    If you are interested, I can strongly recommend John Baines, ‘Literacy and ancient Egyptian society’, Man, 1983, 18(3): 572-599 (which is in JSTOR). Baines takes the most conservative approach and argues that Egyptian literacy was at its height no more than 1% of the populace.

  3. Pingback: A clarification « Glossographia

  4. I always appreciate comments, and yours are no exception. I’ll try to answer your points one by one.

    1. “The new ASPM variant emerged prior to fully phonetic scripts (by about 2,000 years).” If this variant facilitated mental processing of alphabetical characters, its selective advantage should have become manifest with the earliest shift to phonetic characters in the pictorial systems of the Middle East. I should point out that this variant is also present in East Asians, but at lower levels, so even a mixed ideographic/pictorial system would have generated some selection for it.

    2. “Literacy levels did not range from 10% to 33%.” We don’t really know. The problem is not simply a lack of good data. It’s also the concept of literacy itself. Many people in the ancient world could read street signs and signs on storefronts (which were written in block letters and not lengthy). Few could read books written in cursive text. In any case, the selection pressure was on the small minority who could read and write lengthy texts. This ability was exceptional and was seen so at the time.

    3. “East Asian writing systems are not fully ideographic” – This is a straw man. East Asian writing systems are more ideographic than writing systems in South Asia, the Middle East, and Europe (with the exception of Korean, which is a recent exception).

    4. “we have no systematic comparative idea of how differently structured scripts affect cognition.” MRI scans show that Chinese characters are processed in the brain differently than alphabetical characters, see:
    Liu, C., Zhang, W-T., Tang, Y-Y., Mai, X-Q., Chen, H-C., Tardif, T., & Luo, Y-J. (2008). The visual word form area: evidence from an fMRI study of implicit processing of Chinese characters. NeuroImage, 40, 1350-1361.

    5.”Frost’s reasoning certainly implies that Japanese writing should have similar effects to alphabetic writing, despite the near-absence of the ASPM variant there”. Japanese has two writing systems, an ideographic one and an alphabetical one. As I understand it, the former was traditionally favored by scribes.

    6. “However, we have no evidence that they [scribes] actually did have such success to a greater degree than non-literate elites, or indeed that they enjoyed such success at all.” The first proposition is true, but beside the point. Scribes had greater reproductive success than the bulk of the population (who were not elite). Their profession was highly regarded, as seen in the ‘Praise of a Scribe’ in the Book of Sirach (chap. 39):

    “He shall serve among great men, and appear before princes: he will travel through strange countries; for he hath tried the good and the evil among men. …
    Many shall commend his understanding; and so long as the world endureth, it shall not be blotted out; his memorial shall not depart away, and his name shall live from generation to generation.
    Nations shall shew forth his wisdom, and the congregation shall declare his praise.
    If he die, he shall leave a greater name than a thousand: and if he live, he shall increase it.”

    The last verse is the one most relevant to this point. In the ancient world, “leaving a great name” did not simply mean being written about in history books. It meant having many progeny who would perpetuate one’s name long after one’s death. The concept of ‘name’ had a strong hereditarian meaning back then that is much less significant today.

    7. “East Asians should have difficulty learning alphabetic writing, even if raised using only alphabetic scripts” – Are we talking about modern alphabetical writing or ancient alphabetical writing? Remember, ancient texts were written in run-on style with no punctuation and no word or paragraph breaks. So, yes, I would make that prediction — just as I would predict that many Europeans would also have trouble. I’m talking here about an ability that was considered exceptional in the ancient world, so it would probably be exceptional even today.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s