Wednesday, October 8, 2008

Ancient reading and writing

The French journal L’Histoire has a special issue on reading and writing in ancient societies. One article, about Mesopotamia, makes several points that support an argument I have made: the invention of writing, especially alphabetical writing, created a strong selection pressure for people who had the rare ability to take dictation or copy written texts with a low error rate and over extended lengths of time (Frost, 2007).

1. In the ancient world, reading and writing required much stamina, concentration, and memorization, more than is the case today with current reader-friendly language. This may be seen in the long training needed to make a good scribe.

To learn cuneiform writing, the students followed a specific and very standardized curriculum that has been reconstituted thanks to the thousands of exercises that have been found. Training began with writing of simple signs and then writing of lists of syllables and names. Next came copying of long lexical lists that corresponded to all sorts of realities: names of trades, animals, plants, vases, wooden objects, fabrics, … Then came copying of complex Sumerian ideograms, even though Sumerian had become a dead language, with their pronunciation and their translation in Akkadian. Learning of Sumerian was completed by copying increasingly difficult texts: proverbs and contracts, and then hymns.

2. Scribes were not recruited from the general population. Their profession seems to have been largely family-transmitted, and was recognized as such.

Learning of cuneiform, in the early 2nd millennium, took place in a master’s home and not in an institutional “school”. The tradition was often passed down within families, with scribes training their children.

3. Although writing was generally done by scribes, many more people could read and, if need be, write.

It has long been believed that in ancient Mesopotamia only a very small part of the population knew how to read and write and that these skills were reserved for specialists, i.e., scribes. Several recent studies have called this idea into question and have shown that access to reading, and even writing, was not so uncommon. Some kings, and also the members of their entourage, family, ministers, or generals, as well as merchants, could do without a reader’s services, when necessary, and decipher on their own the letters sent to them. Sometimes, they were even able to take a quill—the sharpened end of a reed—and write their own tablets.

The last point may help us understand a chicken-and-egg question. If reading and writing are associated with specific genetic predispositions, how did people initially manage to read and write? (see previous posts: Decoding ASPM: Part I, Part II, Part III)

The answer is that these predispositions are not necessary for reading and writing. But they do help. Specifically, they help the brain process written characters faster. In this way, natural selection has genetically reinforced an ability that started as a purely cultural innovation.

This may be a recurring pattern in human evolution. Humans initially took on new tasks, like reading and writing, by pushing the envelope of mental plasticity. Then, once these tasks had become established and sufficiently widespread, natural selection favored those individuals who were genetically predisposed to do them better.

The term is ‘gene-culture co-evolution’ and it’s still a novel concept. Until recently, anthropologists thought that human cultural evolution had simply taken over from human genetic evolution, making the latter unnecessary and limiting it to superficial ‘skin-deep’ changes. But recent findings now paint a different picture. Genetic evolution has actually accelerated in our species over the past 40,000 years, and even more over the past 10,000-15,000. The advent of agriculture saw the rate increase a 100-fold. In all, natural selection has changed at least 7% of the genome during the existence of Homo sapiens. (Hawks et al., 2007; see previous post). And this is a minimal estimate that excludes much variation that may or may not be due to selection. The real figure could be higher. Much higher.


Frost, P. 2007. "The spread of alphabetical writing may have favored the latest variant of the ASPM gene", Medical Hypotheses, 70, 17-20.

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution. Proceedings of the National Academy of Sciences (USA), 104(52), 20753-20758

Lion, B. (2008). Les femmes scribes de Mésopotamie. L’Histoire, no. 334 (septembre), pp. 46-49.


Anonymous said...

One day, I tried to test recursive thought and sense of humour in one of my chinese collegue.

So I told him:
"chinese people call chinese food, food"

The joke only makes sense if we infer that chinese food is bad and is not really funny for a chinese unless he thinks of it as being though by a non-chinese.
(The amazing thing is that this joke comes from an americano-english man !)
Anyway, my chinese collegue was not amused and he didn't get it.

In computer language, two powerful algorithms are often used: iteration and recursivity. Recursivity is more elegant and goes deeper into the essence of things but is very expensive in memory ressources. Similarly, a conditional iteration (a while loop) can give exactly the same result than a repetitive "for" loop, but is much more computer intensive and therefore costs more energy.
It is likely that the mental tasks that we perform in our brains could be done similarly by different algorythms at different costs.
But the "best" algorythm is only the best if we assume infinite energy ressources, because the brain is an hungry organ.

A smaller brain, able to perform a task for a longer time because it spend less energy per neurone to compute this task, could very well outperform a bigger brain with more hungry neurones and more enery available.

A new physiological feature, acquired throuh mutation, that would allow more enery (sugar) to feed a particular group of neurones necessary for a task could possibly lead to improvment in that task, such as writing and reading, but a decreased in energy cost per/neurone via a better algorythm and a more compact wiring, could have produce the same effect as well.
That means that mutations with apparent opposite effects (more energy available versus better utilisation of less or same energy available) are to be looked for.

Surely, brain imagery can partially respond to that, but I would take a guess and say that, based on the trend in computer hardware compaction and algorythm design, the energy saving route is more promising than the energy producing route.
So maybe the smartest brains in long, tenuous intellectual tasks, are those who look like the coolest and the less active in brain scan imagery. Does it make sense ?


Anonymous said...

Any comment on this?

Anonymous said...

Spelke, Hirst, and Neisser (1976) studied the reading of short stories for comprehension while writing down words to dictation. Might this be an example of the kind of activity that requires effortful pushing of mental plasticity at first. There are other cases of apparently successful performance of two complex tasks discussed in Cognitive Psychology: A Students Handbook by Michael Eysenck(google book search preview).

"When the Jewish refugee professors of Classical Philology turned up in England in the 1930s what astonished them most was the linguistic standard of public-school boys who not only translated unseen pasages which(as on of them put to me 'I should have been lynched for setting to doctoral candidates' but- what is far more to the point- translated superb examples of their own English prose tongue into passible Latin and Greek prose and verse. For that no substitute will be found." (J. Enoch Powell)

Having a few languages gives first hand experience in real time translation (a great perspective on the ASPM theory), gives access to French journals, and of course it's a very marketable skill. The benefits are endless!

The National Geographic Neanderthal doesn't look like a big ape but the creators are artists so it could have been worse. The actual thing has blue eyes, digitally changed to hazel because of Hans Eiberg's estimate of blue eyes first showing up as a single mutation 6000 to 10000 years ago. They were determined to give it light eyes. If the Neanderthal women looked like that Darryl Hannah should have been playing one in Clan Of The Cave Bear.

According to NG "Neanderthals appear to share with modern humans a version of a gene called FOXP2 that contributes to speech and language ability."

Anonymous said...

To me, what makes the difference in reading and writing is the ability to focuse mentally for a long period of time. Playing chess is similar and you can feel yor brain warming. But why do you say that these mutations have to appear in a specific caste of scribes in middle orient, namely jeuish scolars ?

Such tasks are certainly different than playing tennis or hunting, where we have to think smart and quick, but not for very long. It is also different from memorizing.

But I see at least one activity for a prehistoric man to stay focused and think hard for a long time, it is the story telling of a hunter IN ABSCENCE of an elaborate langage.
Try to explain to a group of other hunters the new strategy for the next hunt (or the next war) if you have only few words available. You need to describe the places according to synbolic features (red river, round stone etc), you need to name the people etc. If you've been hunting you know It's quite complex actually and also can be quite long. I could not imagine the trouble of doing it without a complex symbolic language. THe chief hunter needs to think hard for more than just a couple of minutes and in the process he might have to invent new symbols and words from scratch to describe places and assign people to these places and forecast consequences of doing this or that, like in playing chess basically.
Mutations that would have boosted mental focus could have been as usefull in these societies of hunters as in the farmers and scribes of the middle east.

Another example is more recent, it is the "astronomers and mathematicians castes" of predicting solar and lunar cycles and creating an algebra to predict events.
Chineses astronomers have been reporting celestial events 5000 years ago. Creating a calendar is mentally quite demanding, and here again, mutation that help mental focus could have been as usefull in china than in the middle east.


Anonymous said...


You've raised a point I hadn't thought of: a hardwired algorithm takes up less mental space than a softwired one and thus consumes fewer metabolic resources. A hardwired algorithm is stripped down to those steps and decision rules that are needed to produce a specific range of outcomes.

In contrast, a softwired algorithm includes a lot of unnecessary wiring and thus imposes unnecessary demands on the brain. Previously, I saw the advantage more in terms of learning costs: no time is lost learning an algorithm you will almost certainly have to use.

Re: Hammer et al's paper

The only surprise is the relatively high maternal to paternal genetic diversity among the Basques. The other groups are known to be polygynous. I'm not suprised that the San have the lowest ratio. The San are hunter-gatherers, so their incidence of polygyny is relatively low. Polygyny is high among sub-Saharan agricultural peoples because year-round agriculture makes women much less dependent on men for food provisioning. The costs of taking a second wife are thus correspondingly lower.


Is bi-tasking measured by any cognitive tests? How well does it correlate with performance on conventional IQ tests?

Anonymous said...

I think your questions were pointing out that learning to dual task involves automatic processing on one task and "automatiziation reflects a transition from algorithm based performance to memory based performance", ah well (Dual questioning is obviously a marker for strong grounding in the relevant literature).

Possibly relevant to Anonymous(RG)'s suggestion a Scientific American article called High Aptitude Minds says that "Neuroscientists have garnered support for this efficiency hypothesis using modern neuroimaging techniques. They found that the brains of brighter people used less energy to solve certain problems....In other cases scientists observed higher neuronal power consumption in those with superior mental capacities. Musical prodigies may also sport an unusually energetic brain. That flurry of activity may occur when a task is unusually challenging...whereas a gifted mind might be more efficient only when it is pondering a relatively painless puzzle".

According to CH Waddington "Many of the minor variants between the genes are in fact buffered out of expression and if one just looks at the normal animal in the normal circumstances, one would not realize they were there at all. However they are still present in the population...These unused potentialities do come through to expression in a very few people"
ASPM variant might have been widespread for reasons of drift before scribe work gave those few who expressed it a special utility reproductive advantage.

Could the most prestigious scribes and personal secretaries be said to be doing real time translation of a sort. They would not alter meanings but they would put the gist of what was being dictated down in a set format paraphasing the lord or king. This might involve a lot of mental rearranging of the text in "real time".

Anonymous said...

Expert pianists can repeat back or shadow heard speach while playing from sheet music. Expert typists can type and shadow at the same time. Presumably "expert" means most people can not do this no matter how they are practiced.
Spelke's subjects seem to have been Cornel undergradates, some at least were paid to take part. Session lasted for one hour a day 5 days a week The subjects who took part in Experiment 2, the most difficult stage, were Tom, Mary and Arlene. Tom dropped out after making very little progress.

Mary had a reading speed of 400 words per minute, the others in her group of 8 were under 300 words per minute.

Arlene was the most skilled subject by far and was was able to learn a task in a week that had taken other subjects in her group 6 to 10 weeks to learn. Arlene had previously worked as a secretary and reported being able to type copy while talking on the telephone. Arlene was a graduate student.
It seems safe to say that the two subjects capable of performing at dual tasking were well above average in IQ . Arlene's superiority at double tasking seems to have been greater than what might be expected if ability to double task was closely related to reading speed or IQ, she was a graduate student though.

Anonymous said...

Early telegraph messages were called out by an operator and written down to be transcribed into English. The skill of writing down the message in English or taking messages by ear was at one time very uncommon. Andrew Carnegie was said to be the third person in the US to be able to do it, which earned him a job as clerk and personal telegrapher to a railroad superintendent. Edison was also a telegrapher, it took him three months to learn the skill.

"At high speeds, over about 45-50 wpm, many experts agree that copying -- but not reading quickly becomes very exhausting, and can only be continued only for very short periods of time. For them as speeds go up, getting it from the ear to the paper demands the utmost of concentration, shutting everything else out of mind. Some have described it as almost being hypnotized"

Anonymous said...

All the contents you mentioned in post is too good and can be very useful. I will keep it in mind, thanks for sharing the information. Keep updating, looking forward for more posts. Thanks.seo services

Anonymous said...

gain height
Increase Height
You deserve the best and I know this will just add to your very proud accomplishments in your already beautiful and deserving blessed life. I wish you all the best and again. Thanks a lot..