What is it to know?

lfen,I think one crucial matter in asking "what is knowledge?" is to define "what is mind?" or "what is being?" I think the dualist's understanding of knowledge will be fundamentally different from the reductionist's understanding of knowledge.For this thread, do you prefer to talk about one over the other? Or just to hear the thoughts from both dualists and reductionists?I'm formulating a post in my head in the meantime...Being -> What is it to know -> What is knowledge... got it...Ben

In talking about 'knowledge' in a very very limited scope--that of neural implementation of memory--this is certainly the path that has been taken. Neural network / PDP models have (IMHO) certainly shown their power over other types of 'dualistic' models of knowledge. One of the powers of the PDP models is the 'greyness' of THIS 'version' of knowledge.Ben

lfen,I'm not really familiar with fuzzy logic. I took a look at a webpage (the first one Google returns on 'fuzzy logic')...I'm not sure I'd say I favor fuzzy logic over two-value logic. I'd have to know a lot more about fuzzy logic first. From my 2 minute understanding of fuzzy logic, it still seems that they are operating on high-level information.I would favor PDP models that operate on MUCH lower levels (chemical and physical) of information than the ones that fuzzy logic and two-valued logic work.So to summarize... I think two-valued logic has very poor explanatory power. Fuzzy logic... I'm not sure about. I'll just detail a view on one reductionist's view (mine!) on 'knowledge' in a bit (working on typing it up), and hopefully that will answer your quesitons better.Ben

lfen,At this point, I've written a bit about 'what is knowledge' on my local machine. At this point, I'm getting stuck on what facet of 'knowledge' you're interested in investigating. "Knowledge" is just a word--humans use it as a symbol to refer to some set of 'things'. Can you describe for me what part of 'knowledge' you're interested in investigating?My answer, to this point, is simply that "knowledge doesn't exist," that it's a term coined based on a dualist / mind-body separated reality. I deny that reality; in my reality there is nothing that the term 'knowledge' can refer to. Yet, at the same time, there are things that are related to facets of this classical knowledge (such as long-term storage, affect / causation, etc).Can you try to expound a little more on what YOUR direction of 'knowledge' is intented to be? A word is, in many ways, meaningless. Only a common 'cultural literacy' can 'give' a word a meaning, and I don't think we have that here. So.... throw me a bone Regardless, I'm continuing to expound on the 'classical notion' of knowledge on my local machine, as well as what parts of my current model seem somewhat related to this classical notion.Thanks!
Ben

I think the entire concept of 'knower' and 'known' comes from how we perceive and understand, or model, the world. I think the model that postulates 'knowledge' is fundamentally dualistic in nature. I think if you eliminate the dualistic part of that model, 'knowledge' simply ceases to be a useful description of what's going on.Instead, I think a bottom-up approach using emergence is the one that models the world more consistently and usefully. In this model, there are parts that have aspects of what is called 'knowledge', but nothing that's really THAT close.I think you'll have a hard time finding an acceptable set of primitives to describe knowledge because, fundamentally, it derives from a model that has other explanatory problems. I was trying to ask what are the aspects of knowledge that you are interested in, so that I could try and extract those things while avoiding the 'knowledge'. I didn't mean to try and make you take a position--I always appreciate people who are interested to know other people's thoughts. I was just trying to get you to describe your question more, so that I could provide a better answer.In my reductionist model, you can give a basic model of things as organisms and environment (although this model does have problems, when it comes to determining 'being' and 'identity'). You can even pretty reliably model the organism as having a set of organs, one of which is this brain with a bunch of neurons, or computational units. That kind of works.In this model, you have all sorts of 'knowledge'. Here are some properties of 'knowledge':
- persists over time
- 'useful' for 'solving problems' with regards to 'the organism' or 'the environment' (ambiguous!)In this model, what levels or 'things' have these properties?- Cells, Genes, Organs--they all can be modeled to 'solve problems,' they all certainly come from 'knowledge' that is persisting over time (given that they all develop in such a consistent way, and so the same for offspring). This thought of (i.e. modeled) as 'hard-coded' knowledge... although if you include the idea of mutation, you can also model this as 'software' (just change the timescale; that's Darwinian evolution baby!)- Neurons--neurons interact in a Hebbian way. They store 'knowledge'--neurons change connectivity based on firing patterns of other neurons, based on cell death (programmed or through lack of nutrition or trauma). At this level it would be hard to model 'problem-solving', but maybe you could if you really want to. I've no reason to try, so I'll move on.- Networks of neurons--My current understanding of the brain is that there's many ways neurons can interact. Just because you have a bunch of neurons doesn't give you intelligence. It is the connectivity and the pattern of activation that gives rise to what we call intelligence. Networks of neurons still play off of the storage and 'learning' in individual units. Furthermore, we can think of a network to have a 'meaningful input' and 'meaningful output'. This depends on where you decide to 'cut' the network for analysis--what you define as the input layer, and what you define as the output layer.At this level, we can talk about language, memory, etc. However, remember, everything going on here comes from our own analysis and investigation. Even the separation between being and environment. None of this is 'real'--it's our model, our understanding. I think THIS level (the level of memory, language, etc.) is VERY problematic, and not a good abstraction. The next level (the abstraction of an individual) is even MORE problematic. Regardless of the appearance of the world through our own consciousness, our own folk psychology, this very model that is so apparent to us is NOT useful for scientific investigation into ourselves. To me, it's too inconsistent a model to be useful.So I will stop at the level of networks of neurons. Some people try to go from our emergent 'individual' systems, and work backwards to derive the 'pieces of the system.' It just doesn't work. That's cognitive science.I think I'll end the post here, and gather my thoughts again. This will all be easier to discuss after I've really spent time to write about it more. That's what I'm doing these days (well at least, after I finish my darn statement of purpose!)Thanks!
Ben

I understand. This is such a cognitive perspective.From this perspective, fuzzy logic seems better than a two-valued approach. Here's a list of questions that I think are important in talking about this perspective of knowledge:What is it that you know?
How quickly can you state it in words?
In how many situations can you apply it?
What cues will make you remember it?
What is your variability when stating it?
How often do you state it or apply it incorrectly?Like in math, some cases are so biased in one direction, they really hide the number of questions. 'what is your name?' is such a question, and by reducing the number of questions, it is simple and obvious. You can even say, 'yes, i know my name!' and get away with it (for the most part... just don't dig too far!).In this way of thinking, I think there's all sorts of knowledge, not just one.
- knowledge of HOW to do things
- knowledge of the relation of internalized 'concepts' (such as logic, math, and 'abstract thought'
- knowledge of the past (memory) (which is notably 'reconstrucive', not like computer memory).When you move into the arena of PDPs and distributed systems, there's just no 'knowledge' like that. I hope this post offers something for you . I'll be interested to hear your thougths.BenP.S. Thanks for bearing with me... I'm doing my best to understand where you're coming from and to address your questions with my own thoughts.

lfen,One primitive understanding that I always work is that we're NEVER talking about 'truth' in the absolute sense. There is know such 'knowledge' available. You can read my (fairly short) post here that might explain my view better. Basically, my primitive is that all understanding is MODELING. There is nothing more, nothing less. All concepts are simply ways of understanding, models. Identity, consciousness, knowledge... they are concepts to try and help us understand, to survive, to 'make due.' Nothing more, nothing less.Models (and therefore understanding) are about two things: logical consistency and explanatory power. Those are the two criterion for any set of concepts (i.e. a model) that I judge.I understand where you're coming from. I think the top-down approach (taking consciousness as a primitive, and trying to derive answers from there) doesn't work; it's not a good model.I'm currently trying a bottom-up approach, and I'm getting somewhere.PDP / neural network models are pretty simple and very powerful. It might take a little time to feel comfortable with them, but I think they're a really powerful tool for understanding. Beyond that, I think you "know" just as much about this stuff as I do. I think I'm just confusing you with my thoughts.I think your writing about identity and 'what it is to be a river' is on the right path. Half of what I was trying to write (but didn't post) was a description of WHY the top-down, dualist approach to knowledge is problematic. You're touching on it here. Maybe if I try to flesh that out, and we take that direction, that would be useful.Anyway, I think I understand where you're coming from. If you're interested in pursuing any of these lines of discussion, I think that would be interesting. If you want to take a different direction, that's cool too. Not too many people are interested in following these directions of investigation (in my experience).Thanks!
Ben

lfen,Everything made sense to me until you said, Who is we? Do you mean the individual knows it's own functioning? Or did you mean that WE, scientists, know only how an animal functions, and not the "reality." If you mean this, what sense of reality do you mean?My first thought is that calling behavior "knowledge" is not quite right; it is like the paradigm, or data, for knowledge. Most people think of knowledge as something static--the thing that produces the behavior, the more general storage and capacity behind the behavior. At least, that's true for organisms.For something like gravity, some people might call THE GRAVITY itself "knowledge." This I think goes back a bit to what Searle was saying, about the two types of causality.Ben(edited to change the subtitle to reflect my real EvC handle... right "Ifen"? )This message has been edited by bencip19, 12-12-2004 09:39 PM

haha just a joke; just taking the opportunity to (subtly) wonder why people call you Ifen, and why you don't seem to mind.Take it easy. It's almost noon here, I'm at a big advantage Until next time!
Ben

forgot to mention... LOL!

OK, let me try... I've re-ordered some of your comments for my own convenience. Just FYI. Actually, distinguising between knowledge / knowing and information is important to me. Information is a term based on 'value' or 'function'; weird term. Knowledge, to me, just means "what is known." It is the physical structure supporting whatever "to know" is. Weird. Weird! Weird!! Weird!!!...But serously, I think I'm starting to understand. You're using these words in a very colloquial sense I think, and that causes trouble in communication. What you call knowledge is more like what I call information.I think cognitive scientists would agree with the direction of your statements (a reduction of knowing to something else), but not with the specifics. You want to define knowing in purely physical terms, to kind of remove the consciousness part of it. Let me lay out what "knowledge" and "knowing" are from a cognitive perspective, and you can let me know how that jives with you.Speaking broadly and making up generalizations as I go, information is stored in the brain in a few different ways.

1. There's the biased processing of the sensory system. "Knowing" a cheese sandwich includes being able to recognize what is a cheese sandwich, what is not a cheese sandwich, and what looks like a cheese sandwich. Even more, being able to imagine a cheese sandwich when one is not around.
​
2. There's the association areas, where information from different modalities are combined; this would suggest that "knowing" a cheese sandwich means being able to visualize a cheese sandwich when you hear the sound of one (haha).
​
3. there's sensory-motor association areas. This would mean, "knowing" what is a cheese sandwich means salivating when you see one, or whatever.
​
4. Finally, there's the purely motor areas. This would mean that "knowing" what a cheese sandwich is entails being able to make one, being able to eat one, etc.

If you're trying to propose that we are "removed" from encountering the "essence" of a cheese sandwich, that our "knowing" a cheese sandwich does not involve our direct encounter with its "essence", then sure, I agree. But I don't think you're going there.Weird!Help me understand Ben

Contracycle,I think these questions are super-important, and I hope I can convince you of my answers. I re-ordered your questions a bit to fit the logical flow I wanted for my response.

• Dr. Ramachandran's studies of consciousness and the generative aspect of our own awareness. His studies clearly and scientifically demonstrate that our 'model' of the world 'inside' our head, what we call mind, is an active construct of the brain, and nothing more.
• Dreaming
• Philosophical studies on the nature of knowledge (epistemology)
• The fallability, and reconstructive nature of our own memories. This has been studied extensively in the memory literature, and very specifically in the eyewitness testimony literature.
• The ability to fractionalize the "I" through different diseases or neural procedures (such as a corpus callosectomy

Another great question. Really, we ALL model the world in our heads as "reality." Is there really any explanatory value in adding this extra layer?I believe the answer is a definitive YES! Here's some reasons why:

• It makes scientific predictions that are different, and that we find. In other words, it gives us a much better understanding of the mind.
• It answers some philosophical questions that can't be answered by the "reality" model.
• It removes the dualist distinction between mind and body, which is a HUGE source of confusion and assumption in scientific and philosophical studies
• It allows you to freely abstract away from "apparent reality," to find explanations for data that go outside of what seems to be "reality." For example, proposing a 5th dimension to explain some phenomena in cosmology, or to explain that there are no particles in the world, only fields--these are more acceptable when we realize that "reality" IS just an abstraction. To explain the apparent reality using other abstractions is then very valid; that is exactly what our mind does. Our mind constructs a model of "reality" using some rules and procedures, no more or less valid than those scientific rules and procedures to describe different aspects of this "reality."
​
Why should I not trust my senses to describe the world?
​
You better. If you don't, you'll get killed. This is not for behavior; this is for understanding. This is science and philosophy, not life. Scientifically, there's no justification for trusting your senses. Practically, however, you've got no choice. Besides to shut it all down, do nothing.
​
Knowledge of all constituent components is unnecessary. Why should 100% complete knowledge be the only type of valid knowledge?
​
It's just a matter of definitions. I think the point is just that knowledge, in the classical sense of the word, is impossible. There are things that are possible--like constructing a pretty consistent model of the world, and knowing (inductive) facts about the model.
​
I think you're asking all the right questions; I hope my answers provide some value to you.
​
Thanks!
Ben

lfen,(laughing to myself) I really feel like a failure. I still can't understand exactly your position... the terms seem too ambiguous to me. Or maybe your idea is just different from any of mine, and I'm having trouble moving away from my own thought. I'm really sorry about that.But maybe you know me by know; when it comes to this stuff, I'll always give it a go. Let's see... What does "consciousness" refer to here? Does it mean the actual phenomenon of being conscious, the actual experience itself? Or are you talking about the brain processing that gives rise to consciousness?(pausing to think) I can't imagine what 'consciousness' is, without content. Actually, I meant to write this next thought experiment to you, but forgot. Maybe it's relevant now.Can you imagine what it might be like to be blind? Maybe not. But let's try anyway. I want you to first imagine being blind. What is it like? Try to describe it.Then I want you to imagine being blind, and deaf. What is that like? Try to describe it. Is it different from being blind?Keep eliminating senses, one by one. Describe it.After you've removed all the senses, what is left? What were the steps that got you there?I'll be interested to hear your response, if you're game.Thanks!
BenP.S. I got some more links for video lectures (MIT). Dude, get Real Player. There's some crazy ass stuff there.

Contracycle,Here's my thoughts... I hope I'm able to add some value this time, and not simply state the same thing over again... It depends on what you mean "real." If you mean that the synthetic experience is in some sense "real"--because it's part of our experience, or because it has some corresponding brain state--then sure, I agree. If you mean it's synthetic but representational (i.e. represents something out in the real world), then I completely disagree. There's a wide philosophical literature on the problems with this kind of representation.Plus, personally, I just don't see the need to postulate that at all. Why make two kinds of real? Stick with the one that's for sure--the synthetic one. There's no point in going further--as we both agree, the differences between the synthetic and the postulated underlying reality are, in most practical applications, the same. That's right. The data supports your position and mine as well; I think that's your point.I simply think it's more parsimonious, with better philosophical explanatory power, and better for conceptual understanding of what "knowledge" ACTUALLY means, to support my position. I'm still unclear on lfen's position. I feel like such a moron... haha Let me try to describe this one again, because I think I chose words poorly.I watched the 2004 Nobel Physics Laurelate talk about his findings. In it he mentioned "Einstein's second law":M = E / c^2Which is related, of course, to "Einstein's first law"E = M * c^2Now wait a minute, right? They're the same. Well, really, they're not, at least not to the human mind (and this has been shown in some studies by Dr. Miriam Basso(c?)k, who I actually interviewed with at the Unversity of Washington).The first one is a statement about mass--mass is derived from something (a very new concept!), while the second is the equation of a bomb. Einstein published the first, even though he is known for the second.So what's my point? My point is, at least in my experience, if you take the position that "reality" is unknowable, then the formulation of the intellectual position itself quickly discards some of what previously seemed to be barriers. You quickly see that our experience is fundamentally no different than any other kind of synthetic model. In other words, our mind runs under the same principles as a scientific model.I find this liberating. For example, the existence of extra dimensions loses it's edge like "no, that's too far from reality." "reality" no longer has to be judged by appearances--because those appearances are completely the property of our own minds. They are only one of many models. Models are judged in only one way--their practical use, and their explanatory power.Maybe this doesn't happen for you. It sure did for me. I respectfully disagree. We are ad-hoc creatures. We have an ad-hoc, unconscious understanding of physics, causation, and many other things. These drive our language, our behavior, everything. This is VERY SEPARATE from our intellectual studies of classical physics, mathematics, epistemology, etc.My position isn't that it's useful to focus on "reality" being a construct of the mind when catching a baseball. There's no need for this kind of THINKING when doing. But if I'm going to describe that situation rigorously, or talk about anything intellectually, then the position of "reality" being a construct of the mind is EXTREMELY useful. Just like you wouldn't talk about relativity when catching the ball, or quantum mechanics. Those are not practically relevant to that situation.I hope this helps you understand the value I see in holding the position in question. Like lfen says, if it doesn't hold value for you... that's OK. But you seem to be interested in why anybody would be interested in the position... so I hope this helped you understand me better.Thanks!
Ben

... and I don't know your ancient sources well, even though I'm interested to know what they have to say. So we're both in the same boat I don't know Damasio's work much at all, but I took a look the other day at this distinction. It confused me. I'm getting a lot of that these days Can we view neural investigations into consciousness, then, as investigations into the nature and properties of this emptiness which can be filled?Also, have you read Kant? I don't know him well enough to say well... but he writes about the necessary properties of minds like ours. This seems like a similar type of investigation into the properties of the 'underlying emptiness' and being able to predict the ways in which it can be filled.Ben