Howdy,

I believe there should be a discussion topic opened prior to each meetup. I do want to encourage discussion in as many ways as possible. We're grappling with stuff that bears multiple encounters of multiple perspectives. Please join me here, before and after the talk.

Randall Reetz writes:

Mark, Would you please state in clear terms, prior to your presentation, the reasons you are attracted to C. S. Peirce's ideas and work (in contrast to the work of other thinkers with which you are not so attracted)? Much work since Peirce seems to better express the scientific ideas he explored. In particular, the fields of information theory, cryptography, signal processing, thermodynamics, and linguistics have formalized most of the notions Peirce only sketched.

Whoa there, buddy! That's a tall order. Let my try to take that bit by bit.

>state in clear terms [...] reasons [...] attracted to C. S. Peirce's ideas and work
please see next post

>the scientific ideas [Perice] explored
I'm not covering 99.99% of Peirce's vast work. I'll mention some top level categories in passing, in the hopes that others are inspired to further investigate their own passions in relation to his source material.

>work since Peirce [...] better express[ing] [...] notions Peirce only sketched
I agree, and also believe Peirce should be read as a source in addition. I don't have the resources to argue the latter point. Someone else was to have given a more general presentation on Peirce in advance of my talk.

My talk is specifically about Dr. Terrence W. Deacon's [better to my own purposes(?)] expression of Peirce's triadic analysis of sign relations: iconic, indexical, symbolic. I don't have the resources to argue how expressions are better/worse. The Preparation links in "talk about this meetup" do illustrate this: II. Icon, index, symbol vs. Memes as Signs. Please someone else do a survey of comparative explanation.

>attracted to [...] in contrast to [...] not so attracted [to]
It's a both/and kinda thing: Deacon and Peirce, Sowa and Peirce, Shannon and Peirce, etc. I don't have the resources to go into every thinker I'm not attracted to, but here's one example. I think, in the nature of theoretical epistemological sophistication, compared to Peirce, Foucault was sooo naive.

OK, I am after a direct that leads from your personal motivations to those that you attribute to Peirce. It would provide context to your talk. Context I find necessary. Among other things, the exposure of such context goes a long way to help me understand both Perice and his disciples. As example, if I come across a person building (or disassembling) a wall, it is instructive in the understanding of the wall to ascertain an answer to the question "Why?". Often, aspects of the "wall" aren't nearly so important when weighed against the reason it is being built. Why was Peirce building his "wall"? What is it about Peirce's work that so resonates with your own directions and interests?

As for the content of Peirce's work, will you be exploring the relationship (if any) between Peirce's triples and the more contemporary semantic web data structure of the same name?

Randall Reetz

>state in clear terms [...] reasons [...] attracted to C. S. Peirce's ideas and work
please see next post

Randall Reetz persistently asks essential questions regarding, what are the values of motivation of inquiry:

This is why I'm giving this talk:

Peirce is so big and influential I've cut things down to the core of what I don't know: passionately.

Gregory Bateson [in http://www.oikos.org/...] explains a different process going on in biological communication that's not only propositional logic. In the last year I've seen a little, tiny, bit of semiosis. Both use the term "symbolic," but with significant difference in the operations of meaning.

John Sowa [in http://www.jfsowa.com...] is talking about logic, which is essential. I'm trying to track something else, which is also essential, but which I don't have the words for.

I have a simple tool I've shown the group before [http://mind.mx...] that allows a "X goes-with Y" event. And that makes the simple automatic crosslink behind the scenes: "I'll also remember that Y goes-with X, even though that was not explicit. So if a stimulus Y appears, a possible event will be it goes-with X."

In logic, the "goes-with" part is a formal relationship: X than Y, X than not Y, X is a part of Y, X is everything but Y, X = Z Y^2, etc.

In the other situation I don't really know how to talk about yet, but looks like the analysis of sign relations in semiosis, "goes-with" is a variable, in both directions, depending on context: which is also a variable, or in terms of coupling to a particular subset of all X <--> Y relations, a metavariable.

My bullshit-meter question is: "how can I code this?"

In short, I'm giving this talk because Deacon's expression of Peirce's analysis of sign relations has a family resemblance to Bateson's "syllogism in grass," i.e., relational process in the biological world (and poetic metaphor, and everyday life). And I think I could code this "Had we but world enough, and time." But I cannot yet tell others how to code this, because I don't have money enough, and time. And this family resemblance is my passionate confusion.

Randall's reply slipped in. [i.e., his reply was posted while I was writing my own post. And I didn't see his reply until after I posted.]

Your text-string-as-object-ontology storage and retrieval tool seem to exploit twoples (firsts and seconds relations) where Peirce and later semantic object storage work points strongly towards triples (thirds). Am I wrong?

Does it surprise you that "biology" and "poetics" would be symmetric to "logic"? How could they be ever be different?

Randall Reetz

Slippage.

I edited the above during the slippage to:

In the other situation I don't really know how to talk about yet, but looks like the analysis of sign relations in semiosis, "goes-with" is a variable, in both directions, depending on context: which is also a variable, or in terms of coupling to a particular subset of all X <--> Y relations, a metavariable.

So here's the crux of the difference:

In the world outside of digital computer process, i.e., "symbolic" logic, in the world outside of that:

1) Any X can have an finite but always open number of relations to Y, not just one

2) Peirce's icon/index/symbol is a category system for types of relations

My own (possibly unique) contribution is highlight the fact there's an often-missing perspective:

3) If an X has any relation(s) to a Y, than it is always the case that Y also has some relation(s) to X

---

So let's break that down further:

1) a. If any X - Y relation exists, it's almost always possible to find another type of relation between X and Y

1) a. 1: If we can't produce another possible relation, that usually tells us that we don't really understand either X and/or Y

1) b. The Semantic Web triplet is about unambiguous reference (address) to two different nouns (things) and and unambiguous reference to a single relation (adjective, but perhaps verb) between two things.

1) b. 1: considering only a single relation between two things is based on a conventional simplification in the digital world that is too often imported back into considerations of real world systems.

1) b. 2: considering only a single relation between two things can be terrifically pathological in real world systems.

1) b. 3: Peirce's icon/index/symbol analysis of sign relations allows for multiple, multi-layer, open-ended sign relations. That is, a sign can have, at the same time, multiple iconic and multiple indexical and multiple symbolic relations. This is similar, but different from what I'm saying in 1). Peirce is saying X can be interpreted as Y, Z, A, B, C, etc, depending on the relation of X as a sign to its object, by an interpreter.

2) a. Peirce's icon/index/symbol category system of sign relations is one part within a much larger semiotic system, about both analysis and communication, with very unfamiliar yet precise terminology.

2) b. I certainly don't know if the icon/index/symbol analysis exhausts the possible types of sign relations, or not. And I'd love to see a deeper analysis on this.

3) a. I'd love to see someone else's work on inherent cross-relation. It's probably out there somewhere.

3) b. The inverse is true: if X has no relation to Y, than Y has no relation to X.

3) c. In the realm of the mind, it is almost never the case that any X has no relation to any other Y. There are so many possible relations, and the mechanism of mind actively seeks/makes them. What does this say about nature?

[edited for clarity]

Does it surprise you that "biology" and "poetics" would be symmetric to "logic"? How could they be ever be different?

>surprise you
Thank you for your questions Randall, but I sense you're putting your words in my mouth here. I don't like how that is phrased. I'd prefer you proffered a positive statement of position rather than a personal dispositional question.

The term "symmetric" is, to me, vague in this context. Please explain.

I do believe these two ways of analysis and communication:

• processes of logic: Syllogism
  
• processes of sign relations: Semiosis
  

are complementary, but fundamentally different systems. I believe that, of the two, especially in computer science, semiosis is tremendously under-appreciated and has the more potential for driving human-computer co-evolution. And I can't claim that in both they exhaust the possible most-basic means of analysis and communication, but I'm open to it.

I do believe these two ways of analysis and communication:

• processes of logic: Syllogism
  
• processes of sign relations: Semiosis
  

are complementary, but fundamentally different systems. I believe that, of the two, especially in computer science, semiosis is tremendously under-appreciated and has the more potential for driving human-computer co-evolution. And I can't claim that in both they exhaust the possible most-basic means of analysis and communication, but I'm open to it.

OK now we have a basis for analysis. I am of the opinion that the difference you here point to is only a difference of stratum death and that it does not indicate any real qualitative delta in causality. It is obvious to me that signs are a higher order aggregate and that they are built of logic. Meaning is to my way of thinking simply a meta-grouping or grammatical stacking built ultimately of the same stuff you label "logic". An atom can never be a washing machine, but all washing machines are necessarily built exclusively and only of atoms. Logic primitives can't be meaning, but meaning must be built of logic primitives. Presenting logic and relations (between objects) as siblings is obfuscating at best.

If your vantage is the study of things, is academic, then yes, the study of signs can be thought of as a sibling of the study of meaning. To avoid coming to wrong conclusions, we must remain clear of the "relationship' between any two things. Sorry to belabor this point.

The topic "human/computer co-evolution" is fraught with so much hubris that I hesitate to engage. A more rhetorically clean line of inquiry would be the various ways in which complexity handling could be advanced. Isn't that what evolution "cares" about?

I very much appreciate your willingness to dig deep and take the time to expose your ideas and to address my questions. However, you do not seem to have talked to my central question of motivation. What is it about Peirce that is more intriguing than the general concepts of communication you have excerpted from his work? Certainly, these concepts could be discussed unhindered by the attachment to a person.

Why should this matter? Because we humans so easily over or under value a concept or measurement based on agreement with some other aspect attached to that content. One of these aspects is origin. Did Lincoln say it? Einstien? Peirce? How could origin effect validity? Does authority influence truth?

In particular, I have read many libertarian and supply-side political/economic manifestos that quote Peirce. The entire free-will foundational argument that has been built to support and add some semblance of theoretical bases for libertarian philosophy is itself based on (a reading of) Peirce's work. This rhetorical minefield worries me. Please at least address it, show that you see it, explain how or why you would choose to explain so complex a subject through an example space so fraught with emotional/cultural baggage.

Back to the subject of the formalism of the storage and processing of the relationships from which meaning precipitates:

My understanding of the use of triplets (and other formal methods) is that as a relationship-notating formalism it in no way restricts the number or direction of relational constructs. When a triplet is used to store the member-of relationship of "bat" to the larger category "baseball equipment", it puts no restraint on the establishment of an infinite number of triplets that simultaneously store other relationships between that same bat and that same equipment set. The formalism was established as a base unit. The reasoning behind a triplet as a viable formalism for the storage of semantic relationships is not that an object's relationships to another object can be described with three links, but that no relationship falls outside of the descriptive capacity of a formalism built of the triplet as a syntactic mechanism. Those who advance the use of triplets argue that the triplet is the minimal or atomic relation holder. That all descriptions are ultimately built of triplets and that no relationship needs more than the triplet formalism to capture its effect on and from everything else.

So, from a semiotics perspective, it might well be instructive to set up a higher-order categorization scheme that considers relationship taxon pertinent to human centered relations (social value, power, access to resources, sexual or reproductive status, child rearing, authority, etc.), the logicians and computationalists will argue that each of these forms of relation (no matter how complex) can be described by a network or graph made up of linked triplets. In fact, logicians will argue that the the real value of an atomic relationship formalism like a triplet is that it can act as the atoms from which one can build all three sides of Peirce's holy trinity (object, sign, interpreter). Teach a computer how to assemble and parse a triplet and you can teach a computer how to build any system.

Randall Reetz

There is a difference, a huge difference, between that which is optimal and that which is practical. For instance, if one had access to the contents of a warehouse full of clothing, than one might never choose to use (or invent) a washing machine.

In the same way, we in computing work within the historical confines dictated by the legacy stack that we call computing. Within this rarefied world, we frequently confuse effectiveness with truth. If our CPU/Bios/OS/Application has been optimized for the processing of tables and arrays, we not only avoid n-dimentional graph processing, we tend to confuse truth within the larger universe of logic with truth within the limited but omnipresent domain that is legacy computing.

A good logician says, "Fuck the computer!" and shield his or her eyes from its influence on their thinking. Those of us not so lucky, those of us that must make working systems, we have to hold our breath, jump in, and hope we can navigate reality while maintaining an unbiased understanding of the physics of logic.

Systems abstracted as relationship primitives in an n-dimentional graph are systems who's description might be idealized towards logical elegance, but they most definitely are not systems idealized towards execution on legacy processing architectures. This is why you will get two very distinct and apposing answers when you ask an engineer and a theorist whether or not a particular simulation is elegant.

Today's systems don't like n-dimentional graphs. They don't like them from the perspective of hardware any more than they do from a perspective that considers software optimization. But the world of things (and memes) doesn't really care about legacy computing. The world of relationships will win the battle over legacy. Over time, both computers and processing architectures will come more and more to reflect the actual world of the logical relationships that make up systems.

Of greater import is the determination and filtering towards those relationships that matter more than others. This metric is locally situational and globally weighted towards complexity handling capacity building. Atoms matter, but they don't matter to the extent that atom assembly schemes matter. We need to identify the atoms of logic, and then move the hell on. We need to learn not to confuse structures from the parts those structures are built. We need to acquire the cognitive and technical tools with which to manipulate the stacking of systems (grammars) and we need to automate the optimization of grammatical stacking.

If you are disappointed with the direction computing has taken, god knows I am, then let go of legacy and build computing the way it should have been built. Don't fault the founders, or the suits or the happy nerds, they were (are) trying to solve local processing/market challenges and remain un-incumbered by the need to understand processing from the perspective of computation's ultimate future.

Randall Reetz