Abduction, not kidnapping but guessing

Kenneth Ketner b9oky at TTACS.TTU.EDU
Mon Sep 16 16:29:54 EDT 2002

Company, 1910, page 1:
Abduction, n. 3. In the logical system of C.S. Peirce, reasoning from
consequent to antecedent; the acceptance on probation (or more
absolutely) of a hypothesis to explain observed facts; the deriving of a
suggestion from observation.

	Peirce argued that all reasoning can be successfully classified under
three main types: abduction, deduction, induction. Basically abduction
is guessing. He argued that humans have a natural instinct for guessing
(abduction) and that all the findings of science have to begin as
abductions of individual persons. The result of an abduction is a
hypothesis, which is nothing in itself and requires vigorous testing.
One uses deduction and previous knowledge to infer possible observable
consequences of the hypothesis plus previous knowledge. Then one uses
these possible observable consequences and previous knowledge to design
an experiment to test the hypothesis, thus allowing nature to give the
answer to the hypothesis under test. Formally a hypothesis is a
declarative sentence, but held only provisionally, so logically a
hypothesis is fundamentally an open question. If one decides an open
question by consistency with previous beliefs, that is what Peirce
called the Method of Tenacity; if one decides a question by appeal to
consistency with a respected person or institution (solely on the basis
of respect, or fear if one is oppressed) that is what Peirce called the
Method of Authority; if one decides a question by way of consistency
with a fashion, that is what Peirce called the A Priori Method; if one
designs experiments to allow nature to answer the question in a communal
and public way, that is what Peirce called the Method of Science. There
is an Abduction/Deduction/Induction cycle (ADI), in that typically one
continues the process recycling upon itself until a confirmed hypothesis
emerges. Perception is this process at a fast rate and unconsciously
done. There is a big connection between abduction and Semeiotic,
Peirce's theory of signs (see PERCY'S article in the last issue of
DOUBLETAKE MAGAZINE, winter 2002 issue).
	So in everyday language Abduction is Guessing
		Deduction is Necessary Reasoning
		Induction is Experimental Design.

In a rare bit of consistency with previously established terminology (in
line with Peirce's Ethics of Terminology), the computer science world
has adopted abduction as the term for the hypothetical aspect of
Artificial Intelligence research, one of the hardest nuts to crack. That
may be because Peircean Abduction assumes there is already an
intelligence present (otherwise known as a human, or The Scientist, or
The Guesser), so it may take a while to get a fully intelligent
artificial device. Anyhoo, there is a lot on abduction in the computer
science world, and they usually give some credit to Peirce even though
sometimes it is misguided or downright wrong: as in their often heard
slogan that "abduction is inference to the best explanation (of a
puzzling event)." The problem there is that after abduction only, we
don't know if this hypothesis is the best - that would only come after
completing the ADI cycle.

There is a good deal on this in the book Percy "made" ;) me write, HIS

> "Karey L. Perkins" wrote:
> Thanks for the response -- I have a couple of thoughts/questions, one
> related to this, and one not:
> (a) This is for Ken, but anyone who is familiar with Peirce can
> respond:  What is "abduction"?  Percy talks about it some, contrasts
> it with induction and deduction, but I'm still not clear on it.  The
> dictionary was little help.  (I haven't gotten to reading Peirce yet,
> which I imagine should clarify it.)

Kenneth L Ketner
Paul Whitfield Horn Professor
Institute for Studies in Pragmaticism
Texas Tech University
Lubbock, TX 79409-0002
806 742 3128
		Office email: b9oky at ttacs.ttu.edu
		Home email: ketner at arisbeassociates.com
				Office website: http://www.pragmaticism.net
				Personal website: http://www.wyttynys.net

More information about the Percy-L mailing list