[Taxacom] source of quote about keys: PS

[Stephen Thorpe]

In my experience, some biosecurity people (in particular) tend to think that the ONLY way to identify something reliably is using a key! It provides them with a "protocol" for identification. The irony is that quite often the available keys are pretty useless and vague! What is the best way to identify something? Familiarity and experience. But people whose job (like biosecurity) demands rapid routine identifications need a short cut tool - keys. They often look to systematists for those keys, but what they get is sometimes a "phylogenetic" key. What we need are more diagnosticians - people whose job it is to facilitate identification. Good systematists ain't necessarily good diagnosticians!

________________________________________
From: Neal Evenhuis [neale at bishopmuseum.org]
Sent: Tuesday, 1 September 2009 10:58 a.m.
To: Stephen Thorpe
Cc: Robin Leech; Taxacom at mailman.nhm.ku.edu
Subject: Re: [Taxacom] source of quote about keys: PS

In my experience, the lament of the quote is:

1. the person writing a key "obviously" knows his/her organisms and
really doesn't need a key; and

2. people don't use them because they really don't want to have to
read something to identify it. They would rather a) go to a
collection and picture match; or more commonly b) go online and ask
people (on taxacom, for example [I thought this was a discussion list
but is has lately been an identification service]) what something is
based on a photo.

-Neal

At 10:44 AM +1200 9/1/09, Stephen Thorpe wrote:
>Robin,
>
>think of it another way: yes, keys are used only to IDENTIFY things,
>but there are 2 quite different cases:
>
>(1) identifying (in the sense of placing it within a classification)
>a new taxon for the very first time; and
>
>(2) routine identification (in the sense of putting a name to it, or
>just placing it within a family, etc.) over and over again by
>various people
>
>both cases involve keys, but the keys for (1) will be more difficult
>and use less "convenient" characters, but it only has to be done
>once (or very few times)