2
Crick (1966, p. 10) and Jacob (1974, p. 320).
3
Monod (1971, pp. 30-31).
4
Jakobson (1973, p. 61). He also writes: “Consequently, we can say that, of all the information-transmitting systems, the genetic code and the verbal code are the only ones that are founded on the use of discrete elements, which are, in themselves, devoid of meaning, but which are used to constitute the minimal units of significance, namely the entities endowed with a meaning that is their own in the code in question” (p. 52). See Shanon (1978) on the
differences
between the genetic code and human languages.
5
Calladine and Drew (1992) write: “The mass of DNA is surrounded in most cells by a strong membrane with tiny, selective holes, that allow some things to go in and out, but keep others either inside or outside. Important chemical molecules go in and out of these holes, like memos from the main office of a factory to its workshops; and indeed the individual cell is in many ways like an entire factory, on a very tiny scale. The space in the cell which is not occupied by DNA and the various sorts of machinery is filled with water” (p. 3). De Rosnay (1966) writes: “The cell is, indeed, a veritable molecular factory, but this âmiracle' factory is capable not only of looking after its own maintenanceâas we have just seenâbut also of building its own machines as well as the drivers of those machines” (p. 62). Pollack (1994) compares a cell to a city, rather than to a factory: “A cell is a busy place, a city of large and small molecules all constructed according to information encoded in DNA. The metaphor of a city may seem even more farfetched than that of a skyscraper for an invisibly small cell until you consider that a cell has room for more than a hundred million million atoms; that is plenty of space for millions of different molecules, since even the largest molecules in a cell are made of only a few hundred million atoms” (p. 18). In his book
The machinery of life,
Goodsell (1993) writes: “Like the machines of our modern world, these molecules are built to perform specific functions efficiently, accurately, and consistently. Modern cells build hundreds of thousands of different molecular machines, each performing one of hundreds of thousands of individual tasks in the process of living. These molecular machines are built according to four basic molecular plans. Whereas our macrosocopic machines are built of metal, wood, plastic and ceramic, the microscopic machines in cells are built of protein, nucleic acid, lipid, and polysaccharide. Each plan has a unique chemical personality ideally suited to a different role in the cell” (p. 13). De Rosnay (1966, p. 165) compares enzymes to “biological micro-computers” and to “molecular robots,” whereas Goodsell (1993, p. 29) calls them “automata.” Wills (1991) writes: “The genome is like a book that contains, among many other things, detailed instructions on how to build a machine that can make copies of itâand also instructions on how to build the tools needed to make the machine” (p. 41). For discussions of DNA as a “language” or a “text,” see, for example, Frank-Kamenetskii (1993, pp. 63-74), Jones (1993), or Pollack (1994). Atlan and Koppel (1990) reject the classical metaphor of DNA as a “program” and suggest instead that it is better understood as “data to a program embedded in the global geometrical and biochemical structure of the cell” (p. 338). Finally, Delsemme (1994, p. 205) writes that “we can consider with complete peace of mind that life is a normal physicochemical phenomenon.”
6
Piaget (1975) writes: “Thus the most developed science remains a continual becoming, and in every field nonbalance plays a functional role of prime importance since it necessitates re-equilibration” (p. 178).
7
Scott quoted in Freedman (1994), whose article inspired this paragraph. Goodsell (1993) writes that “proteins are self-assembling machines,” which, among other functions, “form motors, turning huge molecular oars that propel bacterial cells” or “specific pumps [that] are built to pump amino acids in, to pump urea out, or to trade sodium for potassium” (pp. 18, 42).
8
Calladine and Drew (1992, p. 37). See Wills (1989, p. 166) on the speed of carbonic anhydrase. See Radman and Wagner (1988, p. 25) on the minute rate of error of repair enzymes.
Science
nominated DNA repair enzymes “molecules of the year 1994.” Recently, it was found that these enzymes are highly adaptable and that “repair” enzymes also participate in DNA replication, the control of the cell cycle, and the expression of genes. Similarly, enzymes that splice the double helix can do so in both chromosome recombination and repair operations. Enzymes that unwind DNA can act during transcription of the genetic text as well as repair (see Culotta and Koshland 1994). Wills (1991) writes on the speed of DNA duplication by enzymes called replisomes: “Replisomes work in pairs. As we watch, about 100 pairs of replisomes seize specific places on each of the chromosomes, and each pair begins to work in opposite directions. Since all the chromosomes are being duplicated at once, there are about ten thousand replisomes operating throughout the nucleus. They work at incredible speed, spewing out new DNA strands at the rate of 150 nucleotides per second.... At full bore, the DNA can be replicated at one and a half million nucleotides per second. Even at this rate, it would still take about half an hour to duplicate all six billion nucleotides” (pp. 113-114).
9
Margulis and Sagan (1986, p. 145). Since the time of writing the French original of this book, two articles by Heald et al. (1996) and Zhang and Nicklas (1996) seem to indicate that the dance of chromosomes is orchestrated by spindle microtubules, which function even in the absence of chromosomes. This does not remove the question of intention, however. As Hyams (1996, p. 397) comments: “A great many questions about mitosis remain to be answered. To what extent do chromosomes contribute to spindle formation and to their own movement at anaphase? Do they have a role in positioning the cleavage furrow? What holds sister chromatids together, how are they âunglued,' and what is the signal for this detachment? How do the checkpoints that sense a single detached chromosome or an imperfect one work?”
10
Wade (1995a) writes: “Only DNA endures. This thoroughly depressing view values only survival, which the DNA is not in a position to appreciate anyway, being just a chemical” (p. 20).
11
Trémolières (1994, p. 138) considers that “our human comprehension and intelligence reach their own limits. It seems that our brain is one of the most complex objects that we can find in the universe.” McGinn (1994, p. 67) writes: “We want to know, among other things, how our consciousness levers itself out of the body. We want, that is, to solve the mind-body problem, the deep metaphysical question about how mind and matter meet. But what if there is something about us that makes it impossible for us to solve this ancient conundrum? What if our cognitive structure lacks the resources to provide the requisite theory?”
12
Hunt (1996) writes: “Crow tool manufacture had three features new to tool use in free-living nonhumans: a high degree of standardization, distinctly discrete tool types with definite imposition of form in tool shaping, and the use of hooks. These features only first appeared in the stone and bone tool-using cultures of early humans after the Lower Paleolithic, which indicates that crows have achieved a considerable technical capability in their tool manufacture and use” (p. 249). See Huffman (1995) on chimpanzees using medicinal plants. Perry (1983) writes about ants that herd aphids: “In one species, the ants take fine earth up to the leaves and stems of plants and, using their own saliva, cement together tiny shelters, shaped like mud huts, for their aphid partners. These shelters help to protect the aphids from severe weather and to some extent from predators.... Some ants will round up local populations of aphids at the end of the day, in much the same way that a sheepdog herds sheep. The ants then take their aphids down into the nest for protection from predators. In the morning the aphids are escorted to the required plant for another day's feeding and milking” (pp. 28-29). See also Hölldobler and Wilson (1990, pp. 522-529). Concerning mushroom-cultivating ants, see Chapela at al. (1994) and Hinkle et al. (1994). Wilson (1984, p. 17) compares an ant's brain to a grain of sugar.
13
Monod (1971, p. 18). Wesson (1991) writes: “By what devices the genes direct the formation of patterns of neurons that constitute innate behavioral patterns is entirely enigmatic. Yet not only do animals respond appropriately to manifold needs; they often do so in ways that would seem to require something like forethought” (p. 68). He adds: “An instinct of any complexity, linking a sequence of perceptions and actions, must involve a very large number of connections within the brain or principal ganglia of the animal. If it is comparable to a computer program, it must have the equivalent of thousands of lines. In such a program, not merely would chance of improvement by accidental change be tiny at best. It is problematic how the program can be maintained without degradation over a long period despite the occurrence from time to time of errors by replication” (p. 81). On the absence of a goal, or teleology, in nature, Stocco (1994) writes that “biological evolution does not proceed in a precise direction and aims at no particular goal” (p. 185), and Mayr (1983) writes: “The one thing about which modern authors are unanimous is that adaptation is not teleological, but refers to something produced in the past by natural selection” (p. 324). According to Wesson (1991): “For a biologist to call another a teleologist is an insult” (p. 10).
14
According to several recent studies, non-coding DNA might actually play a structural role and display the characteristics of a language, the meaning of which remains to be determined. See Flam (1994), Pennisi (1994), Nowak (1994), and Moore (1996).
15
The twenty amino acids used by nature to build proteins vary in shape and function. Some play structural roles, such as making a hairpin turn that folds the protein back on itself. Others make sheet-like surfaces as docking sites for other molecules. Others form links between protein chains. Three amino acids contain benzene, a greasy compound that is the molecular equivalent of Velcro and that can hold certain substances and then release them without modifying its own structure. One finds these benzene-containing amino acids at exactly the right place in the “lock” of nicotinic receptors, where they bond molecules of acetylcholine or nicotine (see Smith 1994). Couturier et al. (1990) provide the exact sequence of the 479 amino acids that constitute one of the five protein chains of the nicotinic receptor. My estimate of 2,500 amino acids for the entire receptor is an extrapolation based on their work. See Lewis et al. (1987) regarding the presence of nicotinic receptors among nematodes.
17
Trémolières (1994, p. 51). He adds: “We know that more than 90% of the changes affecting a letter in a word of the genetic message lead to disastrous results; proteins are no longer synthesized correctly, the message loses its entire meaning and this leads purely and simply to the cell's death. Given that mutations are so frequently highly unfavourable, and even deadly, how can beneficial evolution be attained?” (p. 43). Likewise, Frank-Kamenetskii (1993) writes: “It is clear, therefore, that you need a drastic refitting of the whole of your machine to make the car into a plane. The same is true for a protein. In trying to turn one enzyme into another, point mutations alone would not do the trick. What you need is a substantial change in the amino acid sequence. In this situation, rather than being helpful, selection is a major hindrance. One could think, for instance, that by consistently changing amino acids one by one, it will eventually prove possible to change the entire sequence substantially and thus the enzyme's spatial structure. These minor changes, however, are bound to result eventually in a situation in which the enzyme has ceased to perform its previous function but it has not yet begun its ânew duties.' It is at this point that it will be destroyedâtogether with the organism carrying it” (p. 76).
19
See Wesson (1991, p. 52). He adds: “By Mayr's calculation, in a rapidly evolving line an organ may enlarge about 1 to 10 percent per million years, but organs of the whale-in-becoming must have grown ten times more rapidly over 10 million years. Perhaps 300 generations are required for a gene substitution. Moreover, mutations need to occur many times, even with considerable advantage, in order to have a good chance of becoming fixed. Considering the length of whale generations, the rarity with which the needed mutations are likely to appear, and the multitude of mutations needed to convert a land mammal into a whale, it is easy to conclude that gradualist natural selection of random variations cannot account for this animal” (p. 52). Wesson's book is a catalogue of biological improbabilitiesâfrom bats' hypersophisticated echolocation system to the electric organs of fishâand of the gaping holes in the fossil record.
20
Mayr (1988, pp. 529-530). Goodwin (1994) writes: “New types of organism appear upon the evolutionary scene, persist for various periods of it, and then become extinct. So Darwin's assumption that the tree of life is a consequence of the gradual accumulation of small hereditary differences appears to be without significant support. Some other process is responsible for the emergent properties of life, those distinctive features that separate one group of organisms from another, such as fishes and amphibians, worms and insects, horsetails and grasses. Clearly something is missing from biology” (p. x).
22
Mycoplasma genitalium
is the smallest genome currently known, at 580,000 base pairs. Mushegian and Koonin (1996) compared it to the genome of bacterium
Hemophilus influenzae,
which contains 1,800,000 base pairs, and concluded that the minimal amount of genetic information necessary for life is 315,000 base paris. This is still an enormous amount of information.