RE: SUO: 2000-7-26 Peircean categories
On Fri, 31 Aug 2001, pat hayes wrote:
> Yes, I have to agree with Adam on this one. Its not so much that they
> don't belong, but that they don't seem to *exist*. I (still) have
> never seen an account of these ideas of firstness, secondness,
> thirdness and fourthness - whoops, sorry, strike that last - which
> come anywhere close to warranting their inclusion in any practical
> ontology, and I have absolutely no idea what anyone would want to do
> with them. Maybe if a Peircian partisan could actually SAY (without,
> please, quoting the Master, whose pronouncements on this topic are
> completely impossible to even parse coherently, let alone understand
> or formalize) what these are intended to mean, what relations they
> have to other things, etc., maybe the topic could move forward. But I
> have my doubts.
> Bear in mind, whoever tackles this task, that the final container for
> this intended meaning must be some set of axioms in a formal logical
> language (or equivalent). Documentation is fine, but if there is some
> topic that cannot, by its very nature, even begin to be captured in
> this way, then it is probably best left in silence. For example, I
> wouldn't suggest that we include an ontology for subjective
> impressions of colors.
I just joined this list (having just started work at Cycorp) and am
lurking and learning a great deal. Despite my newness, I feel moved to
comment on this topic as my PhD dissertation (in philosophy) was largely
about firstness, secondness and thirdness.
Peirce says somewhere (fear not, this is an indirect quotation) that after
some exposure to his categories, one may find that they naturally acquire
a structuring role in all one's thought, and I have found this to be true in
my own case.
Before they are applied to logic or metaphysics, however (in which they do
have a great deal to offer), the categories are phenomenological, and
this is the best way to get the 'feel' for them which is required to go on
and apply them elsewhere. Here is a brief phenomenological approach to the
3 (and I apologise if I am telling anyone more than they want to hear on
For Firstness, think of the smell of fresh coffee, its uniqueness, the
way in which it cannot be captured in words, and the way in which it
expands one's 'sensorium' to smell it if one has never smelt it before.
Now, don't *identify* Firstness with coffee-smelling or even qualia in
general, but see if you can "prescind" (abstract) out of the experience
a notion of something distinctive which is as it is independently of any
relation to anything else.
For Secondness, think of the way you can be walking through a room feeling
quite happy and suddenly you bang your shin on a coffee-table. You
have come into sharp, immediate contact with the world (this is often a
bit shocking). Now don't identify secondness with a bang on the shin, or
even efficient causation in general, but see if you can prescind a notion
of two things directly interacting with one another in some fashion as yet
For Thirdness, think of any interpretative or 'seeing-as' experience.
Wittgenstein's duck-rabbit is good for this. When one sees this
interesting drawing first as a duck and then as a rabbit, one is
connecting the drawing first of all with ducks (whether real or
imagined, it doesn't matter here) and secondly with rabbits. These are
different Thirdnesses. Note how each interpretation falls naturally into
i) the drawing
ii) ducks or rabbits
iii) the viewer who takes i) as a sign for ii)
Though thirdness may be 'prescinded from' interpretation, it is not
restricted to it. It is also found in habits, laws of nature,
As an example of the structuring power I see in the categories - they are
superb for capturing any evolutionary process. 1. Something appears, as if
from nowhere, something new and distinctive (genetic mutation, new
hypothesis, whatever) . 2. This thing then meets some kind of
intransigence external to it (survival of the fittest, scientific
experiment, whatever). 3. If it survives, new habits or adaptations (or
interpretations of the world) will emerge, more or less stable....and the
cycle begins again.
Pat argues that it is difficult to think of how to formalise
> PS. As an example of the kind of thing I have in mind, start with
> Firstness. I take it that this is a property of something (since it
> ends in "ness"), and I gather in fact that it is a property of
> properties, characterizing those which attach to their object in some
> very 'basic' way, as it were directly, without reference to any other
> thing; what some people call 'essential' or 'intrinsic' properties.
> If this is more or less right, then we might expect to see axioms
> like this (?):
> (and (Firstness P) (P ?x)) implies ...
> Now, what such axioms could be written? Are there any general facts
> about P and ?x that might be inferrable from the firstness of P that
> would not be inferrable from the simple truth of P applying to ?x ?
The thing about these categories is that they are not properties, but
so-called 'modes of being'. What is the difference between the two? One
difference is that where properties 'crowd each other out' (fall into
sibling disjoint collections?), e.g. the same thing cannot be red and
green in the same place at the same time, modes of being don't. Any real
thing will partake in all three of Peirce's categories. Take, just to pick
a random example, the First World War. It has firstness insofar as there
was a particular 'feel' to that whole time (still accessible to some
degree via documentary footage, novels etc). It has secondness insofar as
it tore up the lives of a generation, Europe was never the same again, etc.
It has thirdness insofar as it was a war, like every war that has ever
been and ever will be (and also because one can interpret what went on
there in terms of general military strategic lore etc.)
Thus I see the categories more as a high-level tool for an ontologist to
evaluate and guide their own ontologising than as an element of the
ontology itself. Most importantly, (in philosophy) I have found that their
simple but very general structuring power leads me to ask powerful
Here are some 'category-inspired questions' with respect to ontologising
that occur to me just off the top of my head (and pardon me if any of
these seem terribly naive from a formal ontology perspective, with respect
to which I am very much a beginner.)
- Does an ontology have room to express the discriminations between
feelings on which so much human thinking depends?
- Has it any room for chance/randomness?
- How does an ontology make contact with things 'out there'?
- How does one formalise growth?
- Could one create an ontology that 'grew itself' (or is this the holy
grail of AI)?
Again, I hope I haven't rambled on too long! Looking forward to learning
more from this list.
Cathy Legg, Phd Cycorp, Inc.
Ontologist 3721 Executive Center Dr., ste 100
www.cyc.com Austin, TX 78731-1615