confidence

[Eric Holman]

This is in response to the question from Jorge Soberon Mainero about
empirical data on errors in taxonomic determinations.  I came across
some relevant data in the following paper:

Taylor, Paul Michael.  1990.  The folk biology of the Tobelo people:
A study in folk classification.  Smithsonian Contributions to
Anthropology 34.

Taylor tried to find the scientific names of the plants and animals
included in the Tobelo folk classification.  He sent plant vouchers to
herbaria for identification; and for 65 specimens, he sent a voucher
to each of the two herbaria (Bogor and Leiden) that are preeminent for
material from the region (in Indonesia).  Of the 65 pairs of vouchers,
both were identified to species in 41 cases, one to species and one to
genus in 12 cases, and both to genus in 12 cases.  The consistency of
the identifications is summarized below; entries in the table are
numbers of pairs.

                             Both to sp.   One to sp.,   Both to gen.
                                           one to gen.

Same species                      24            -             -

Same genus, diff. species         12            2             4

Same family, diff. genera          2            8             4

Different families                 3            2             4

This implies that fewer than half (24/65) of the vouchers from the same
specimens were independently classified in the same species.  Also, the
vouchers not classified in the same species were more likely to be
classified in the same genus if both were identified to (different)
species (12/17) than if one or both were identified only to genus
(6/24).  These plants were not cultivated, but they were probably fairly
conspicuous because they were in the folk classification.

I hope these data are of some interest.  If anyone knows of similar
data, I would love to hear about them.

Eric Holman
Psychology Dept., UCLA, iap8ewh at mvs.oac.ucla.edu