interactive keys

Mike Dallwitz M.J.Dallwitz at NETSPEED.COM.AU
Fri Mar 10 14:04:00 CST 2006

Christian Thompson wrote:

> I would like to see the public embrace interactive keys. But having produced
> one (fruit fly key in Intkey and Lucid) ... I am sorry to say after 8 years
> the Public has not come to my door.

Anita Cholewa wrote:

> With respect to the public and online interactive keys ...  if it's not
> simple to use they'll leave the site quickly.

Lyn Craven wrote:

> For producers of interactive keys, the Windows version of DELTA is a lot
> easier to use than the original. ... If I can use an interactive package to
> produce a key, anyone can.

It is harder (both technically and taxonomically) to _produce_ an
interactive key than to use one. But, as Lyn points out, producing
interactive keys with modern software is not particularly difficult,

The technical basics of using Intkey for identification are extremely simple
and intuitive.

(1) The available characters are automatically displayed in the 'best' order
for identification. Click on any character.

(2) The character states, with illustrations if available, are displayed.
Click on a state and press OK (or Cancel if you decide you don't want to use
the character after all).

(3) Repeat until only one taxon remains.

Most other interactive-key programs are not quite this easy to use, but are
not much more difficult. Even young children can do it if the characters are
simple enough (see, for example, 'Butterflies and Moths: Demonstration of an
Interactive Key'

Unfortunately, if you want to make the _most effective_ use of interactive
identification, you need a program with certain features (see 'Principles of
Interactive Keys', and,
horror of horrors, you need to learn a few techniques. The most important are:-

(1) Use the characters near the top of the 'best' list, unless you have a
reason for doing otherwise.

(2) If you are uncertain which state applies to your specimen, select all
the states that you think might apply, or use another character. Don't guess
except as a last resort. (This is particularly difficult for people who are
used to conventional keys, where guessing is an important part of the

(3) Check your identification against the available material (e.g. full or
diagnostic descriptions, illustrations). If it seems to be incorrect,
increase the error tolerance and continue the identification as before.

(4) If you used error tolerance to reach an identification, press the button
provided (in Intkey) to find out what your mistake was (or, possibly, what
the mistake in the database was).

This is not rocket science - you would think that anyone could learn it in a
few minutes. But apparently they can't or won't.

Extracting information from a database is just as easy, but people are
reluctant to do so, preferring to try to extract it laboriously from
natural-language descriptions generated from the database. This is such a
problem that we now place a warning at the bottom of each description, along
the following lines.

"This description is offered for casual browsing only. We strongly advise
against extracting comparative information from it. This is much more easily
achieved using the interactive key, which allows access to the character
list, illustrations, full and partial descriptions, diagnostic descriptions,
differences and similarities between taxa, lists of taxa exhibiting
specified attributes, summaries of attributes within groups of taxa,
geographical distribution, genera included in each family, classifications
(Dahlgren; Dahlgren, Clifford, and Yeo; Cronquist; APG), and notes on the
APG classification."

It remains to be seen whether this will have any effect.

Mike Dallwitz
Contact information:
DELTA home page:

More information about the Taxacom mailing list