Complication Great and Small

The living world and human civilization are two great systems orchestrated by sequences and capable of open-ended evolutionary creativity. Each in its own way crossed von Neumann’s threshold of complication and, for each, crossing the threshold required a division of labor. The transition from preliterate human culture to our literate technological civilization paralleled the transition from the RNA world to modern DNA-RNA-protein biology.

While both precursor systems were governed by sequences, those sequences were unstable and impermanent, with poor replication fidelity, and they did not allow much accumulation of functional novelty. Impermanence limited the random access needed for self-referential regulation of sequences by other sequences. The precursor sequences were also entangled with dynamic processes, ribozyme folding and paralanguage, limiting their ability to function as rate-independent abstractions. Finally, their ability to interact with their environments was limited by the poor binding and catalytic power of ribozymes and by the time bound and space bound perceptual and motor behavioral capacities of the human phenotype.

DNA replaced RNA and writing replaced speech for sequence storage, enabling the systems to improve their physical stability and replicative fidelity. This made possible random access and open-ended, scalable self-reference, the hierarchical white-collar regulation of sequences by other sequences. And it made the sequences abstract and rate-independent by purging the entangled dynamics.

By developing protein biocatalysts to supplement ribozymes and technologies of measurement and manipulation to supplement the human phenotype, the systems were able to increase the power, accuracy, and versatility of their interactors. Both established a division of labor between sequence storage/ replication and biocatalysis. They could recognize a larger set of environmental affordances and respond with the precise application of unprecedented power. Having crossed the threshold of complication, they outcompeted and established hegemony over the remnants of the precursor systems. RNA and speech retain fundamental roles in their respective systems, but they are overshadowed by the elaborate, cumulative effects of DNA and writing.

To be sure, locating von Neumann’s threshold solves neither the underlying origin problem of life or of human culture, but it does break each problem into two parts: pre-threshold and post-threshold. We do not know how the first sequences of RNA and the first ribozymes emerged from the rate-dependent physical world, nor do we know how human speech and toolmaking emerged from the social milieu of our primate ancestors. These remain profound problems.

If stable, abstract sequences and more precise and powerful interactors had not come along, who knows what might have happened? Suppose the RNA world had not discovered DNA sequences and protein. Could that world have persisted for four billion years in its primitive state? Assuming it achieved sufficient scale to survive geologic upheavals, why not? If an RNA world were found on Mars, astrobiologists would triumphantly announce that life exists elsewhere in the universe. Yet it would be life which stops short of crossing von Neumann’s threshold: complex, potentially, but not explosively so.

And suppose humans had discovered neither written sequences nor the technologies of measurement and machines. The preliterate cultures of Africa and the Americas had urban centers, agriculture, ceramics, weaving, trade, art, and political quasi-states encompassing large territories. Could those cultures have persisted indefinitely? Absent major environmental shocks, like European conquest, there is no reason to think they couldn’t. Yet they, too, would have stopped short of von Neumann’s threshold.131 “The evolutionary process of social differentiation, cultural complexification, and political stratification appears to reach a dead end unless writing is developed or adopted,” says linguist Paul Kay.132

Gerald Joyce is deploying a metaphor when he writes of the RNA world: “It is as if a primitive civilization had existed prior to the start of recorded history, leaving its mark in the foundation of a modern civilization that followed.”133 I think he nails it, and more than metaphorically. Recorded history was enabled by the invention of writing, which was necessary for preliterate human culture to cross the threshold of complication.

Notes

  • 1. Parts of this chapter are based on Waters 2012a.
  • 2. Von Neumann 1966
  • 3. Von Neumann 1966
  • 4. Von Neumann 1966
  • 5. Maynard Smith & Szathmary 1995; see also Szathmary 2015.
  • 6. Woese 2002
  • 7. Ban et al. 2000; Nissen et al. 2000
  • 8. Doudna & Cech 2002
  • 9. Atkins et al. 2011; Joyce 2002; Pearce et al. 2017
  • 10. Joyce & Orgel 2006
  • 11. Fica et al. 2013
  • 12. Joyce 1989.
  • 13. Rich 1962; Crick 1968; Orgel 1968
  • 14. Gilbert 1986
  • 15. Robertson & Joyce 2011
  • 16. Takeuchi et al. 2011. The primeval replication of RNA sequences by ribozymes did not entail reproducing a complete genome all at once; it was a piecemeal process: “the earliest RNA polymerases need not have been responsible for replicating entire RNA genomes, but merely for generating RNAs that enhanced the fitness of pre- RNA-based life” (Joyce 2002). The difference between piecemeal and all-at-once replication will be discussed further in Chapter 8.
  • 17. Joyce 2002
  • 18. Cech & Steitz 2014
  • 19. Cech & Steitz 2014. The menagerie of recently discovered RNAs is large. For reviews see Breaker and Joyce (2014), Huang and Zhang (2014), and Morris and Mattick (2014). The most prominent new classes of RNAs include microRNAs (miRNAs), short, non-coding sequences which play an important role in gene expression in eukaryotes, and riboswitches, segments of mRNA that can bind allo- sterically to metabolites to self-regulate their own expression. They also include long non-coding RNAs (IncRNAs), longer sequences which play a role in differentiation and development, and in mammals are chiefly expressed in the brain.
  • 20. Morris & Mattick 2014
  • 21. In another variation on Dawkins’s extended phenotype, RNA sequences have even been weaponized to neutralize genes in other organisms (LaMonte et al. 2012; Cai etal. 2018).
  • 22. Morris & Mattick 2014
  • 23. Steven Benner and colleagues view “modern macromolecular catalysis as a ‘palimpsest’ of an earlier metabolic state, with features that arose recently (‘derived traits’) superimposed upon features that are remnants of ancient life (‘primitive traits’)” (1989).
  • 24. Joyce 2002
  • 25. The genetic material of many viruses contains only RNA, but viruses are not free- living. They survive and replicate only by borrowing the translation and replication equipment of cells. Viruses will be taken up in Chapter 8.
  • 26. С. H. Waddmgton anticipated this difficulty in principle in 1969: “In practice—and perhaps because of a profound law of action-reaction—it is difficult (impossible?) to find a [material structure] which is stable enough to be an efficient store [of information] and at the same time reactive enough to be an effective operator [catalyst]” (Waddington 1969a).
  • 27. Benner et al. 2011, Eigner et al. 1961
  • 28. Joyce 2012
  • 29. Leu et al. 2011
  • 30. Sterelny et al. 1996
  • 31. The point at which the error rate in replication is high enough for mutated copies to overwhelm functional copies is called the error threshold, and it imposes “an upper limit to the frequency of copying errors that can be tolerated by a replicating macromolecule” (Robertson & Joyce 2011). Today, for example, viruses made solely of RNA operate close to the error threshold of “roughly one mutation per genome replication” (Leu et al. 2011).
  • 32. Joyce & Orgel 1999
  • 33. Benner et al. 1999
  • 34. Pressman et al. 2015
  • 35. Robertson & Joyce 2014
  • 36. Maynard Smith and Szathmary did recognize the shift from “RNA as gene and enzyme” to “DNA + protein (genetic code)” as one of their eight major transitions (1995).
  • 37. Like the first sequences, the first modern cells emerged from a world in which they did not exist. Woese calls this transition the Darwinian Threshold. “There would come a stage in the evolution of cellular organization where the organismal genealogical trace (recorded in common histories of the genes of an organism) goes from being completely ephemeral to being increasingly permanent,” he writes. “On the far side of that Threshold ‘species’ as we know them cannot exist. Once it is crossed, however, speciation becomes possible” (2002). What is the relationship between Woese s Darwinian Threshold and von Neumann’s Threshold of Complication? Perhaps they are the same threshold approached from different angles. Woese emphasizes heredity and the origins of the three kingdoms of life, Archaea, Bacteria, and Eukarya. Before the modern cell, there was no cell division, no organismal reproduction as we know it. The earliest cell, says Woese, “was more or less a bag of semi-autonomous genetic elements” (1998). Individual sequences replicated, but evolution was communal, driven, in the words of virologist Patrick Forterre, “by exchange of experiences between individuals, via lateral gene transfers” (2012).
  • 38. Hull 1988
  • 39. Leu etal. 2011
  • 40. RNA can also exist in a double-stranded form, but the double-stranded version does not fold into catalytic ribozymes; only the single-stranded version folds. It is unlikely the double-stranded version played a meaningful role in the RNA world.
  • 41. Joyce 2002. One of the critical events in the transition from the ancient RNA world to the contemporary DNA-RNA-protein world was what Joyce calls “The Great Reverse Transcription” (2012). Sequence information stored in RNA was transcribed into DNA, and there it has remained ever since (Leu et al. 2011).
  • 42. Woese 2002
  • 43. Benner et al. 1999
  • 44. Joyce 2002
  • 45. Robertson & Joyce 2011. Some origin-of-life researchers argue the RNA world probably had a simpler and even-earlier precursor with sequences composed of something other than nucleic acids, perhaps something like a mineral chemistry. These researchers are seeking “a glimpse of the widely accepted, though hypothetical, initial replicator, whose biomolecular activity initiated Darwinian evolution on Earth” (Yarus 2011; see also Robertson & Joyce 2011).
  • 46. Cech 2011. Interestingly, Pattee anticipates the ribozyme from first principles in 1966: “The memory and transcription functions must occur at different times; that is, the sequence of copolymers cannot act as a degenerate memory and a nondegenerate selective structure at the same time. Therefore there must be two states of this copolymer; perhaps a straight chain which serves as a memory, and a folded conformation which can selectively interact with the environment.”
  • 47. Boza et al. 2014
  • 48. Replicating a ribozynte requires several steps. First, the ribozyme has to unfold. However, the unfolded sequence cannot be copied directly to create a second ribozyme. The ribozyme is made of single-stranded RNA. As a result, the ribozyme sequence is first copied into its complementary sequence, which in turn is copied back to the original ribozyme sequence. The existence of this intermediate template adds a layer of complexity that will not be dealt with here (Takeuchi et al. 2011; Boza et al. 2014).
  • 49. Jose et al. 2001; Breaker 2002; Breaker 2018
  • 50. Another argument for dividing the labor by splitting replication and interaction into separate molecules is made by Benner: “Genetics versus catalysis place very different demands on the behavior of a biopolymer. According to theory, catalytic biopolymers should fold; genetic biopolymers should not fold. Catalytic biopolymers should contain many building blocks; genetic biopolymers should contain few. Perhaps most importantly, catalytic biopolymers must be able to catalyze reactions, while genetic biopolymers should not be able to catalyze reactions and, in particular, reactions that destroy the genetic biopolymer” (2014, emphasis his). The assignment of genetic functions to DNA and catalytic functions to protein gave the system “the potential to catalyze hard-but-desired reactions without the potential to catalyze easy-but- undesired reactions” (Benner 2014).
  • 51. Harris 1986
  • 52. Harris 1986
  • 53. Goody & Watt 1963
  • 54. As historian Patricia Crone writes, “We all take the world in which we were born for granted and think of the human condition as ours. This is a mistake. The vast mass of human experience has been made under quite different conditions” (1989).
  • 55. Levinson A Holler 2004
  • 56. Technologies of vocal communication overcome these limitations. Telephone conversations are not space bound and voicemail is not time bound.
  • 57. Olson 1994. Simply by using language, preliterates were evincing some implicit “knowledge” of phonology and grammar, but this is not the same as the explicit, conscious awareness of language as understood by literates. As Stevan Hamad points out: “Wittgenstein emphasized the difference between explicit and implicit rules: It is not the same thing to ‘follow’ a rule (explicitly) and merely to behave ‘in accordance with’ a rule (implicitly)” (1990).
  • 58. Enquist et al. 2008
  • 59. Dyson 1979
  • 60. How large-scale social organization emerged remains a significant topic of research in anthropology and archaeology. See, for example, Kosse 2000, Powers et al. 2016, Turchin et al. 2018, and Shin et al. 2020.
  • 61. Another Hockett design feature is displacement. Speech allows discussion of “things remote in time or space, or both, from the site of the communication” (1966). Writing adds another hierarchical level, what we might call second-order displacement; not only are the things or events under discussion remote in time and space, but so are the writer and reader.
  • 62. Goody 2000. For some this is a feature, not a bug. “The wonderful thing about language is that it promotes its own self-oblivion,” writes Maurice Merleau-Ponty (2002). Corporate malfeasance is often accompanied by an injunction to leave no paper trail.
  • 63. Madden et al. 2006
  • 64. Schmandt-Besserat 1987
  • 65. Goody & Watt 1963
  • 66. See also Olson 2016, 2020; Davidson 2019.
  • 67. Scalability was also limited by the conversational demands of speech production and reception. There is a practical upper limit on the number of people who can participate simultaneously in one conversation. Evolutionary psychologist Robin Dunbar and colleagues conclude “a maximum clique size of around four is an inherent property of human speech mechanisms” (1995). Larger speech-based events, such as lectures, plays, or town meetings, require that turn-taking and other pragmatic norms be agreed to in advance.
  • 68. Goody 2000
  • 69. Eerkens & Lipo 2007
  • 70. Dyson 2013
  • 71. Entanglement is widespread and eclectic. Systems of sequences in the natural world do not depend on universal hardware, like digital computers. They depend on a menagerie of entangled, idiosyncratic mechanisms.
  • 72. Levinson & Holler 2004
  • 73. Goldin-Meadow 1999
  • 74. Kendon 2004
  • 75. See Hewes 1973, Corballis 2003, Armstrong & Wilcox 2007, Tontasello 2008, and McNeill 2012. There is also evidence the relationship between sound and meaning in speech is not entirely arbitrary, which adds another level of entanglement. “Despite the immense flexibility of the world s languages,” say linguist Damian Blasi and colleagues, “some sound—meaning associations are preferred by culturally, historically, and geographically diverse human groups” (2016). They continue: “A substantial proportion of words in the basic vocabulary are biased to carry or to avoid specific sound segments, both across continents and linguistic lineages. Given that our analyses suggest that phylogenetic persistence or areal dispersal are unlikely to explain the widespread presence of these signals, we are left with the alternative that the signals are due to factors common to our species, such as sound symbolism, ico- nicity, communicative pressures, or synesthesia” (2016).
  • 76. Olson 1994; Note that the goal of speech recognition software is to ignore the dynamics and extract only the rate-independent sequential information from the speech stream.
  • 77. Vygotsky 1962
  • 78. Olson 1977
  • 79. Brenner 2009
  • 80. Olson 1977
  • 81. Gee 2006. In his book Class, Codes and Control (1971), sociologist Basil Bernstein uses the terms elaborated code and restricted code to distinguish speech which is more writing-like from that less so. Speakers using informal restricted code assume a shared context and implicit understandings between speaker and listener, whereas speakers using formal elaborated code make no assumptions and as a result need to make such context and understandings explicit. Speakers using restricted code make many assumptions about the listener and speakers using elaborated code make few. See also Kay 1977.
  • 82. innis 1950
  • 83. Daniels 1996. See also Olson 1977.
  • 84. Hyman & Renn 2012. It is the rare preliterate, tribal language that even has a word for “word” (Dixon & Aikhenvald 2002).
  • 85. Olson 1994
  • 86. Tomasello 2008
  • 87. Olson 1994. In his 2005 book The Written Language Bias in Linguistics, linguist Per Linell argues that its emphasis on text-based analysis has caused modern linguistics to view language as “inventories of abstract forms, rather than as aspects of meaningful action, interaction, and practices in the world.” This point is echoed by Talmy Givon: “there is something decidedly bizarre about a theory of language (or grammar) that draws the bulk of its data from well-edited written language” (2002).
  • 88. Olson 2006
  • 89. Harris 1986
  • 90. Goody 1977
  • 91. Olson 1994
  • 92. Olson 1994
  • 93. Bazerman 2006
  • 94. Goody 2000. One aspect of “the power of the written word” that impressed Goody was “the power it gives to cultures that possess writing over purely oral ones, a power that enables the former to dominate the latter” (2000). This echoes Benner and colleagues when they write that “a life form that did not exploit proteins as catalysts could not have competed with life that did” (1999).
  • 95. Mullins et al. 2013
  • 96. Mullins et al. 2013
  • 97. Von Uexkiill 1957
  • 98. As philosopher Karl Popper writes, “man, instead of growing better eyes and ears, grows spectacles, microscopes, telescopes, telephones, and hearing aids. Instead of growing swifter and swifter legs, he grows swifter and swifter motor cars” (1972).
  • 99. “We can cope with thinking about the world when it is of comparable size to ourselves and our raw unaided senses,” says mathematician Richard Hamming, but “when we go to the very small or the very large then our thinking has great trouble. We seem not to be able to think appropriately about the extremes beyond normal size” (1980).
  • 100. James Gibson calls this the “more or less direct perception of the very distant and the very small by means of simple instruments” (Gibson 1982).
  • 101. For Gibsons take on various kinds of what he calls “indirect apprehension,” see Gibson (1982) in Reasons for Realism, a collection of his essays and miscellaneous writings.
  • 102. Michailidou 2010. Similarly, with another important measure used by merchants, “the beveled-rim bowls common from Syria to Iran during the Uruk period, 3500- 3000 BC, may provide the first evidence for the standardization of units of capacity” (Schmandt-Besserat 2010).
  • 103. Michailidou 2010
  • 104. Michailidou 2010
  • 105. Morley 2010
  • 106. Scott 1998
  • 107. Gibson 1979
  • 108. A modern molecular biology laboratory is a case in point. Its inputs are sequences (the existing scientific literature) and biological samples. The samples are processed and measured, with the measurements output as sequences—in many cases sequences of text that represent sequences of nucleotides in the DNA of the sample. These sequences are then processed further and the lab outputs even more sequences in the form of scientific papers (see Latour & Woolgar 1986).
  • 109. Herva et al. 2014
  • 110. Measurement and numeracy also brought new precision to the description of affor- dances by language. “Language does not typically represent the dimensions of objects in analog fashion, but rather digitizes them,” say Barbara Landau and Royjackendoff. “Thus, dimensional adjectives such as big/small, thick/thin, and tall/short refer to continuous dimensions of size, but the linguistic terms bifurcate these dimensions into pairs of relative contrasting terms.” They continue: “Precise metric information is simply not encoded in the language’s stock of spatial terms. . . .It is possible to be precise in expressing distances and orientations, but to do so, one must invoke a culturally stipulated system of measurement that operates by counting units such as meters or degrees (go 30 meters, turn 30 degrees)”(1993, emphasis theirs).
  • 111. Schmandt-Besserat 1996. Just as writing emerged for purposes of accounting, evolutionary biologist Stephen Jay Gould has argued DNA s chief function in the cell is

“bookkeeping” (2002), as have philosopher William Wimsatt (1980) and evolutionary biologist George Williams (1985). See also Basu and Waymire (2006).

  • 112. The abstract split between tokens for quantity and tokens for things being counted was not always complete. Linguist Igor Diakonoff reports on a language with different numeral systems for different classes of things being counted: “The most curious numeral system which I have ever encountered is that of Gilyak, or Nivkhi, a language spoken on the river Amur. Here the forms of the numerals are subdivided into no less than twenty-four classes, thus the numeral ’2’ is тех (for spears, oars), mik (for arrows, bullets, berries, teeth, fists), meqr (for islands, mountains, houses, pillows), merax (for eyes, hands, buckets, footprints), min (for boots), met’ (for boards, planks), mir (for sledges) etc., etc.” (1983, emphasis his).
  • 113. Schmandt-Besserat 2010
  • 114. Goody 1977; “Throughout the world, local developments of writing and arithmetic have interacted with each other in various ways,” says science historian Peter Dam- erow. “In the case of arithmetic, the final outcome is a relatively unified system of arithmetical notation and calculation methods” (2012).
  • 115. “Today, as a result of globalization processes, writing is used all over the world,” says Damerow, “but neither the languages nor the writing systems have been unified by these processes” (2012).
  • 116. “One of King Darius’s greatest achievements,” writes Schmandt-Besserat, “was to give some uniformity to the weights and measures within the Persian Empire” (2010).
  • 117. Renfrew & Morley 2010
  • 118. Schieber & Santello 2004
  • 119. Marzke 2013
  • 120. Shumaker et al. 2011
  • 121. Ambrose 2001. Humans not only fabricate tools. We also exchange tools and parts of tools with each other, something no other animal does. In Robert Aungers words, “animals actively share food and other resources, but not made things” (2010a, emphasis his).
  • 122. Ambrose 2001
  • 123. Washburn 1960
  • 124. Aunger 2010b
  • 125. Aunger 2010b
  • 126. Ambrose 2001. Some researchers have argued human speech evolved hand-in-hand with tool manufacturing (Gibson & Ingold 1993). “The very intimate way in which gesture is integrated with speech,” says Kendon, “could suggest that speech itself is intimately linked to manipulatory activity” (2004).
  • 127. Lind et al. 2013
  • 128. Scott 2017
  • 129. Diamond & Bellwood 2003
  • 130. See Cotterell & Kamminga 1990.
  • 131. The tension between the accomplishments of pre-threshold and post-threshold systems is captured by anthropologist Claude Levi-Strauss. Writing, he says, was the “essential acquisition of culture” and the “source of our civilization.” Nonetheless, he continues, “we must never lose sight of the fact that certain essential forms of progress, perhaps the most essential ever achieved by humanity, were accomplished without writing” (1969).
  • 132. Kay 1977
  • 133. Joyce 2002
 
Source