Jon S, Ben, Jeff, List,

I'm going to take one last shot at explaining why I think the 3 inference
patterns should follow the paths (or vectors) I've argued for in recent
posts (and for the past decade and a half). But first some general
vectorial preliminaries which I hope will be helpful in following my
argumentation.

In the chart below the six possible vectors are reproduced from a 2012
paper of mine, arranged simply by the category from which they commence.
So, working from top to bottom, 2 begin at 1ns, 2 at 3ns, and 2 at 2ns. In
addition, I  will refer to those vectors which reverse a categorial path as
in "inverse vectorial order." So, for example, at the very top of the
chart, 'process' (1ns/3ns/2ns) is in 'inverse order' to 'aspiration'
(2ns/3ns/1ns) at the bottom. (Additionally, a vector may 'mirror;
another--such as 'representation' and 'analysis' do at the middle right of
the chart--while mirroring has little bearing on the present discussion,
'inverse order' most certainly does).

[image: Inline image 1]

The first two vectors at the top of the chart, process and (dialectical)
order, commence at 1ns.

1. 1ns/2ns/3ns, *vector or order* (more precisely, dialectic or, as Peirce
terms it, "dilemmatic" order, see CP 1.491). In one sense, which Peirce
outlines in "The Logic of Mathematics," Hegel's dialectical order
(1ns/2ns/3ns) can be seen as a kind of *ur*-vector. Peirce even suggests in
some places that his own categories were inspired by Hegel's order, but
appear "in new costume." The Hegelian order (thesis/antithesis/synthesis)
is, in Peirce, something/other/medium. As I see it, this is not an
inferential path, but logically and categorially much more primitive.

2. 1ns/3ns/2ns, *vector of process*: While in "The Logic of Mathematics"
Peirce refers to the vector just discussed as Hegel's *evolutionary order *(CP
1.490), Peirce's own order for biological evolution has 3ns (not 2ns) as
the second step or stage. I have also discussed how this is as well the
order of at "complete inquiry" (N.A.) with hypothesis generation (1ns)
followed by the deduction of the implications of the hypothesis for the
purpose of testing it (3ns) leading to the divising of an inductive test
and the actual experimental testing of the hypothesis (2ns).

As important as both of these vectors clearly are, as I see it, none of the
3 inference patterns follow either of the above two paths.

The next two vectors at the middle right of the chart,* representation* and
*analysis*, commence at 3ns.

3. 3ns/1ns/2ns, *vector of representation*: As I pointed out in my first
paper on Peirce's trichotomic (category theory)
http://www.iupui.edu/~arisbe/menu/library/aboutcsp/richmond/trikonic.htm ,
along with Peirce's discussion of Hegel's order and it's reverse (the
vector of analysis, or, involution, to be discussed below), the secondary
source material which got me most keenly interested in categorial vectors
was Robert Parmentier's "Signs' Place in Medias Res: Peirce's Concept of
Semiotic Mediation" (1985). In it Parmentier argues that, in his opinion,
Peirce has placed too much emphasis on the *vector of determination
*(2ns/1ns/3ns)
as opposed to its inverse, the *vector of representation
*(3ns/1ns/2ns)*.* While
I did not then, and do not now agree, with Parmentier on this point of
vectorial emphais, I did lift from that paper theterminology he used for
these two paths (i.e., vectors). Parmentier saw the *vector of
representation* as that path taken by the scientist in hypothesis formation
(or the artist in creating a new work), and on this point I have always
agreed with him.

So, as I see it, this is the path of one of the patterns of inference,
namely, abduction. In this variation of the bean example, I will *presume*
that all the beans from this bag are white *because* I see a handfull of
white beans lying near that bag, and so I suppose that a sampling of the
bag will show them *possibly* to be white. (But, in fact, the present bag
of beans may turn out to be black, and the bag of white beans from which
they came had, in face, been earlier removed, and, although it has no
actual bearing on the inference, perhaps, later found to be the missing bag
from which the handful came).

4. 3ns/2ns/1ns, *vector of analysis* or, better, *vector of involution*:
Although Peirce uses both analysis and involution to describe this path,
which he offers as the inverse of Hegel's dialectical 'evolution', I have
more or less come to prefer the expression *vector of involution *as it
suggests how each of the categories *necessarily *involves the others when
analyzed: 3ns involves 2ns and 1ns and, in turn, 2ns involves 1ns, but 1ns
'involves' only itself. This is offered in "The Logic of Mathematics" as a
mode of derivation of the categories themselves, peculiar in that 'first'
one has 3ns (Peirce remakrs that it may be difficult for certain minds,
indeed, some of the very best, to grasp this involutional order).

I argue that it is also the path of deductive inference since, as I see it,
the rule (a law) involves a case (the sampling, as of the beans) which in
turn involves the result (a character, in our example, *whiteness*). That
character will necessarily follow from the rule. [Rule: All the beans in
this bag are white; Case: this *sample *of beans is from this bag; this
sample will *necessarily* be white.]

5. 2ns/3ns/1ns, *vector of aspiration*: As I see it, this vector is not
involved in any of the inference patterns, but represents, for example, a
community of, say, scientific interest, pooling its intellectual resources
towards the accomplishment of some shared goal.

6. 2ns/1ns/3ns, *vector of determination*: Perhaps the most familiar of the
vectors (although Hegel's dialectical order may be a close second). This,
of course, is the path of semiosis itself: the object (2ns) determines the
representamen, or sign (1ns) which, in turn, determines the interpretant
(3ns) to stand in the same relation to the object which it (the
representamen) stands.

This appears to me to be the path which inductive inference follows. A
large sample (2ns) of beans from a bag reveals a character (all white)
which suggests that *probably* the entire bag is white (the rule).

Comment:
​As I see it,
 3., 4. and 6. above are
​the only ​
vectors involved in inference.
​Most importantly, t
o me, 2ns is
​, in particular,​
associated with *case* principally because of the idea of *sampling*. Now
​a
 sample,
​ ​
in at least physically based experiments (but one can extrapolate to other
types of experiments
​ as well​
)
​, is an​
*existential* something
​:
 Sampling is, as I read Peirce, associated with 2ns
​,​
*not* with 1ns.

​So this is still how I see things for the three inference patterns:


​1ns character (for, as I see it, 'result' only applies to deduction)
|> 3ns rule
2ns case​ (i.e., existential case, a sampling, for prime example)

​So, i​
n the bean example, in the case of *deduction *the actual sample
​ (2ns)​
will *necessarily* be
​ of​
the color (character, 1ns) of the bag of beans which represent
​s​
the rule (3ns). In *induction*, one begins with the existential sample
(again, 2ns as I analyze it),
​then ​
finds the sample to have a certain character,
​viz., ​
whiteness (1ns), and posits that the rule (3ns) will *probably *be that all
the beans in the bag will
prove to ​
be white. In
​ *abduction*,​
my abductive variation o
​f
 the bean example
​,​
one needs in a
​n important​
way to see all three
​phases
 all-at-once-together (as Matthias Alexander might have put it
​; or as Ben Udell recently wrote, "you have to look at the inference as a
whole"​
), so that I *presume* a rule (3ns) is in effect,
​that is, ​
that all the beans in this bag are white, *because* I see a handful of
white (1ns) beans
​nearby which
I imagine to *possibly* be from that bag *were* a sample (2ns) to be taken.
​[​
As a further step in my inquiry, I
​might ​
take that sample and find that all the beans are
​, in fact, not white but​
black
​.
 I now look for another explanation and
​discover
 that some of the bags of beans were earlier removed including the one with
all white beans
​; in this case my hypothesis turned out to be incorrect​
.
]


If one reverses 1ns and 2ns
​​
as Jon has suggested, making 1ns the case and 2ns the resul
​t​
​, ​
​
then one arrives at very different, and as I see it,
​virtually impossible
 vectorial paths.

This is how Jon may be seeing it:

​1ns case
|> 3ns rule
2ns result (perhaps, 'fact')

Following Jon's suggestion, deduction would be 3ns/1ns/2ns, following
the *vector
of representation* (rather, than what I'm arguing, the *vector of
involution, or analysis*). But while a clear case can be made. I think, for
the rule *involving* a sample (2ns) *revealing* a character (1ns), I don't
see how representation , at least as seen by Parmentier and me, makes any
sense for deduction. Indeed I have argued rather pretty much the
'opposite': that abduction--and* not *deduction--follows the *vector of
representation*. For in abduction one imagines a rule (forms a hypothesis)
which may or may not represent the facts of the matter *when* a sampling is
*actually* taken. So, from my standpoint Jon incorrectly reverses the
categorial order for deduction and abduction.

Similarly, if one reverses case and result for induction, one is following
the Hegelian *vector of order*, and, rather than starting at a sample and
arriving at a *probable* rule, one starts at the result (1ns, a character
in my analysis). I must say that this just doesn't make any sense to me. As
I see it, you have to *start* with an actual, existent sample in induction.
And, to repeat, Hegelian dialectic seems far too primitive (basic) a vector
than to be any given inference pattern.

I agree with Jon that we may not reach agreement on this matter. But this
is how I continue to see the 3 inferences patterns vectorially. I have
reviewed his (and other comments) and find that I'm content to stick with
my original vectorial associations for the 3 inference patterns. On the one
hand, this might not amount pragmatically to a hill of beans. But on the
other, consistency ought to count for something in the vectorial part of
Peirce's trichotomic.

Best,

Gary R




[image: Gary Richmond]

*Gary Richmond*
*Philosophy and Critical Thinking*
*Communication Studies*
*LaGuardia College of the City University of New York*
*C 745*
*718 482-5690 <718%20482-5690>*
-----------------------------
PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON PEIRCE-L 
to this message. PEIRCE-L posts should go to [email protected] . To 
UNSUBSCRIBE, send a message not to PEIRCE-L but to [email protected] with the 
line "UNSubscribe PEIRCE-L" in the BODY of the message. More at 
http://www.cspeirce.com/peirce-l/peirce-l.htm .




Reply via email to