A Cultural Law of Gravity

Alfred Russel Wallace was the co-discoverer of natural selection. His paper provides a framework for thinking about how culture and environment constrain varieties to stay true to type and also how changes enable varieties to stray indefinitely from the original type.

It is now well established that language cannot follow just any old rules. Linguists a few decades back thought there was no limit to the variety of language, but research has since identified a number of formal constraints that mark boundaries. Language can work within those borders, but not cross them. The trouble with those borders is understanding what these constraints mean psychologically and neurologically. There must be some reason beyond the formal rules for why these constraints existed. We hardly know how to think about these matters, let alone explain them. A letter in the most recent issue of Nature reminds me, however, that clues are coming in from, of all places, songbirds.

The letter (abstract here) expands on research reported last year at the Evolang conference in Barcelona (see: Reality Blogging: The Survivors). The work caught this blog’s notice then because it paralleled work being done by Simon Kirby. In Kirby’s experiments computers generate a language which is taught to one person. That person teach the language to another, who teaches the language to yet another. As the language moves from generation to generation it becomes more structured, more like a normal language.

Feher and her team did something similar with zebra finches. They raised one in isolation so that it did not hear what a normal zebra finch sounds like and could not imitate it. The result was a finch that sang a song very unlike the song of zebra finches in the wild. Then a young finch learned to sing the song of the isolated finch and another young finch learned to sing that song. Over the generations the song becomes more and more like a typical finch.

The letter draws explicit comparison with real world examples:

Our findings resemble the well-known case of deaf children in Managua, Nicaragua, spontaneously developing sign language as well as linguistic phenomena such as creolization.

All of this material is well-known on this blog, but it is important enough to reconsider and the Feher team’s discussion of just how distinguishable are cultural and genetic traits  Conceptually, they are quite distinct. Genetic traits derive from an individual’s biological processes and are fixed. Cultural traits are learned and can take many forms. But as a practical matter the differences are thickly entangled. As the Feher team’s letter opens:

Culture is typically viewed as consisting of traits inherited epigenetically, through social learning. However, cultural diversity has species-typical constraints, presumably of genetic origin.

There is a rub, at least some of the constraints on culture are of genetic origin and so the two concepts become intertwined. This entanglement immediately introduces contradictions between genetic fixity and cultural freedom.

• Genetic trait: as a concept, the trait is fixed and controlled, similar to a program controlling a machine, but in some species the genetic controls are free enough to permit social learning.
• Cultural trait: as a concept, the trait is free; anything might turn up, but the genes that enable are the social learning are too fixed to actually permit everything.

The formal constraints on syntax that linguists have identified reflect the biological limits to social learning in this geno-cultural entanglement.

A critical factor in all of these examples—the Feher team’s zebra finch experiments, Simon Kirby’s experiments, the deaf children of Nicaragua, and the creolization of pidgins—is that it takes several generations for the entanglement to show its effects. It is as though there was a kind of geno-cultural law of gravity stating that cultural discontinuities will be pulled backwithin genetically-constrained norms over n generations.

Feher’s team proposes a simple equation: Pupil’s output = Genetic contribution + Environmental contribution. (P=G+E) They break down the environmental part into the tutor-dependent environment (E_t) and the other parts of the environment (E_o). Run this equation, P=G+E_t+E_o, for each generation until the pupil’s output (P) has been normalized.

A striking feature of this equation is that the constraints are both genetic and environmental. If the genetic and non-tutor-dependent environment is stable, then their strength will determine how quickly the tutor’s input is overcome. But if the environment begins to change, the return to normalization will depend on the strength of the genetic part alone.

 This notion of a geno-environmental pull over generations might prove useful in thinking about the constraints on language. Particularly notable is the role of generations, which offers a way of roughly quantifying the force of a constraint. Each time another generation is required for normalization, the weaker the constraints. Something that returns to normal without even requiring another generation—such as a slip of the tongue—is evidently very powerfully constrainted. In more dramatic cases, such as the process of creolization, the subject + verb structure that appears in the first generation is powerfully constrained while a trait that appears in the third generation is less forcefully constrained.

But attractive as this thought might seem, there is a problem. Human culture seems to have gone way beyond what such a law of gravity might allow.

 The same problem bedeviled early thinkers about biological evolution and Alfred Wallace’s great paper setting forth natural selection began by considering this very issue. In fact he even titled his paper, On the Tendency of Varieties to Depart Indefinitely from the Original Type and opened it:

One of the strongest arguments which have been adduced to prove the original and permanent distinctness of species is, that varieties produced in a state of domesticity are more or less unstable, and often have a tendency, if left to themselves, to return to the normal form of the parent species [reprinted in Galileo’s Commandment, p. 389].

Wallace went on to present the reason “successive variations” could depart “further and further from the original type.” Even granting a tendency to return to the original type, says Wallace, the conditions of the environment can push in the opposite direction with strong effect. In other words, changing the environment can overcome any genetic constraints supporting the original type. Wallace then noted the divergent type can itself produce varieties that will diverge even more seriously from the original type.

An argument Wallace did not make, but with our own greater knowledge can now be made, is that over time variety becomes more common. As the number of varieties increase, the environment changes, altering the constraints and perhaps weakening the pressure to return to the old ‘normal’ variety.

This acceleration of changing constraints suggests something about the origins of speech. Originally, human culture was likely to be subject to strong geno-cultural gravity simply because the setting was so stable. Genetic structure changes very slowly and the environment contained, as yet, little cultural variety. Thus, even if a few members of a generation moved language in a new direction, the strength of geno-cultural gravity would have quickly brought language and culture back to what it had been. This process could explain the fact that Homo erectus culture was so stable despite its expanded brain. It is tempting to sneer at them as morons, but there may have been a powerful “gravitational pull” that kept speech and tools very stable.

 Living as we do in such an unstable culture where anybody over thirty has to be astonished by how deeply and permanently the world has changed,  it is hard for us to divorce fixity from questions of intelligence and imagination.  But experiments and history suggest that a powerful stabilizing force exists and slowed the whole  process of speech origins.

Comments

[mariana]

very intresting. txs.
It sound reasonable that language variation stoped, I do not see why all things that evolve have to do it either at the same speed, or at the same acceleration rate.
The epigenetic and the genetic evolution in a species Seems to be analogous to the evolution in one human regarding the neural circuits and the environmental factors.

[JoseAngel]

Death, especially early death, is a powerful stabilizing force. A culture that can simply be transmitted but not significantly increased, because it depends exclusively on individual acquisition, can only reach a given level. Technologies of cultural transmission take long to develop in such a society. But once they are established they serve as a foothold for progress to a further level. Take for instance navigation. Or agriculture. Or architecture. Or, perhaps especially, writing, which expands cultural memory beyond the individual's life.

[Charles H. Smith]

Only one comment: Your portrait is not of Alfred Russel Wallace, it is of Robert Owen. I know because I have this image posted at my Wallace site. Oh--one other thing--Wallace actually wrote a fair amount on the origin of language, specifically.
---------------------
BLOGGER: Oops. Maybe I can fix that. Meanwhile, the commenter's Wallace site can be found at http://www.wku.edu/~smithch/index1.htm