Bad Science, Worse Philosophy: the Quackery and Logic-Chopping of David Foster's The Philosophical Scientists (2000)
Addendum H: The Reality of Viral Evolution
Richard Daniel has pointed out that the picture is more complex than this. First of all, "evolution doesn't go in a straight line from simple to complex; it wanders all over the place, creating new species as the opportunity arises. Parasites are often less complex than the independent organisms from which they evolved, because their hosts now provide services that the parasite ancestors did for themselves." He is right. Although evolution can produce complexity, it does not have to.
As I understand it, evolution only has the tendency to seize more and more available environmental niches, i.e. all the places in an ecosystem where energy is available for the taking. If increasing simplification facilitates an organism's finding a new source of untapped energy, or improves its ability to compete for energy already being gathered, as is the case for many parasites, then simplification will result. But once these niches are filled, new ways of exploiting energy resources generally require increasing complexity. Likewise, since there are more niches to exploit through complexity than through simplicity, we tend to see the latter more prominently than the former.
Richard Daniel also points out that "viruses don't need to have genes for any of the molecular machinery needed to reproduce, because the bacteria provide that for them." Thus, the T4 is simpler than the bacteria it eats, but only because it depends upon DNA already present in that bacteria. Mr. Daniel adds that, in terms of information quantity, "the smallest known bacterial genome has 580 kb (Mycoplasma genitalium), while viruses range from 3.5 to 280 kb. Likewise, he also notes that "complexity is not a simple function of genome size, because of the large amounts of gene duplication in plants and animals, but there's very little of that in bacteria."
Given these figures ("a kb equals a thousand bases, i.e. nucleotides, often abbreviated kbp for kilo-base-pairs"), the T4 genome is indeed much smaller than that of the smallest bacterium (61.5 kb vs. 580 kb) but it is not the smallest genome among viruses (61.5 kb vs. 3.5 kb). However, Mr. Daniels adds that "when speaking of viruses, though, base pairs can be misleading because some of them use single-stranded DNA or RNA instead of the usual double helix."