[asa] Proof That Common Descent is NOT Begging the Question

One of the complaints made by strong ID is that common descent is assumed
and then we find it and that continuous creation is just as plausible. A
novel computer algorithm is presented in this week's PNAS shows that an
unbiased search for structure shows otherwise.

http://www.pnas.org/content/105/31/10687.full

*The discovery of structural form*

Algorithms for finding structure in data have become increasingly important
both as tools for scientific data analysis and as models of human learning,
yet they suffer from a critical limitation. Scientists discover
qualitatively new forms of structure in observed data: For instance,
Linnaeus recognized the hierarchical organization of biological species, and
Mendeleev recognized the periodic structure of the chemical elements.
Analogous insights play a pivotal role in cognitive development: Children
discover that object category labels can be organized into hierarchies,
friendship networks are organized into cliques, and comparative relations
(e.g., "bigger than" or "better than") respect a transitive order. Standard
algorithms, however, can only learn structures of a single form that must be
specified in advance: For instance, algorithms for hierarchical clustering
create tree structures, whereas algorithms for dimensionality-reduction
create low-dimensional spaces. Here, we present a computational model that
learns structures of many different forms and that discovers which form is
best for a given dataset. The model makes probabilistic inferences over a
space of graph grammars representing trees, linear orders, multidimensional
spaces, rings, dominance hierarchies, cliques, and other forms and
successfully discovers the underlying structure of a variety of physical,
biological, and social domains. Our approach brings structure learning
methods closer to human abilities and may lead to a deeper computational
understanding of cognitive development.

With respect to biological relationships the paper notes:

For centuries, the natural representation for biological species was held to
be the "great chain of being," a linear structure in which every living
thing found a place according to its degree of perfection
(16<http://www.pnas.org/content/105/31/10687.full#ref-16>).
In 1735, Linnaeus famously proposed that relationships between plant and
animal species are best captured by a tree structure, setting the agenda for
all biological classification since.

In Figure 3 of the paper (
http://www.pnas.org/content/105/31/10687/F3.large.jpg) the computer program
looked at the following structures:

Structures learned from biological features (*A*), Supreme Court votes (*B*),
> judgments of the similarity between pure color wavelengths (*C*),
> Euclidean distances between faces represented as pixel vectors (*D*), and
> distances between world cities (*E*). For *A–C*, the edge lengths
> represent maximum *a posteriori* edge lengths under our generative model.
>

The Supreme Court decision produced the left to right linear structure one
would expect by doing a political analysis of the court. Did the computer
program produce the tree structure predicted by Darwin when looking at
biological features? Yes, it did! The caption to figure 5 (
http://www.pnas.org/content/105/31/10687/F5.large.jpg) explains this in more
detail:

Developmental changes as more data are observed for a fixed set of objects.
> After observing only five features of each animal species, the model chooses
> a partition, or a set of clusters. As the number of observed features grows
> from 5 to 20, the model makes a qualitative shift between a partition and a
> tree. As the number of features grows even further, the tree becomes
> increasingly complex, with subtrees that correspond more closely to adult
> taxonomic intuitions: For instance, the canines (dog, wolf) split off from
> the other carnivorous land mammals; the songbirds (robin, finch), flying
> birds (robin, finch, eagle), and walking birds (chicken, ostrich) form
> distinct subcategories; and the flying insects (butterfly, bee) and walking
> insects (ant, cockroach) form distinct subcategories. More information about
> these simulations can be found in *SI Appendix*<http://www.pnas.org/cgi/data/0802631105/DCSupplemental/Appendix_PDF>.
>
>
This is a computer program so it shows that the structure assumed by Darwin
in fact is the structure that comes naturally out of the data itself. Thus,
evolutionary biology is not begging the question after all but is just
following good old fashioned human pattern matching.

Rich Blinne
Member ASA

To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Fri Aug 8 19:50:53 2008