Prerequisites for the Concept of Knowledge

[Vik]

Again, I don't think the deductive mode is going to pay off.

Consider the behavior of a kitten who encounters a mirror for the first time. He bats and paws at his image, treating it as he would another kitten. But he just touches a flat glass. After a while, he quits reacting to his own image at all, and (usually) never does again. It seems as if he has learned that certain appearances of cats are not knowledge of the presence of a cat.

I'm not proposing that the kitten does arrive at such knowledge. My point is that experiences that simple provide data to support a first notion of knowing. I would also claim that actions of animals and infants in trying to see or otherwise perceive something amounts to a primitive notion of knowledge.

BTW - I'm talking about *reducing* a concept to less abstract concepts and perceptual concrete referents. I'm not suggesting you deduce "knowledge" from first principles. That's rationalism, not Objectivism.

[Mindy]

I've seen adult cats continually leap at the television screen, figure it out, and then leap at it again when a different program is on. They do NOT generalize. At best, the mighty television-hunters have knowledge of *particular facts*.

If there are cats that are able to omit the particular program AND the particular television screen, that would prove they can abstract from particulars but that does NOT prove they can abstract from abstractions. Saying that cats learn is a far cry from saying they have a "notion" of knowledge.

Most importantly, humans grasp the difference between reality and projection around the same time: 2 years.

Cats, if they ever grasp the difference, figure it out in adulthood across a wide variety of ages.

I didn't say TV programs, I said mirrors. It matters.

An animal will make all kinds of different adjustments to be able to see or hear something they have detected. They make the adjustments the situation requires, and the specific behavior involved varies greatly. The goal of this behavior is seeing or hearing. They act to be able to see, etc. Seeing is the goal of that, instrumental behavior. Isn't it true, then that they have some primitive grasp of what it is to see, versus not being able to see, something they have detected? I think so.

And, as we share that animal level of organization, our well-defined concept of knowledge is likely to have its roots in that very primitive grasp. That's why I mentioned the behavior of infants who would not be considered conceptual.

Are you comfortable dismissing such behavior entirely?

-- Mindy

[knast]

How to reduce the concept of "knowledge"? To know what facts give rise to this concept I must know what knowledge is. Ayn Rand defines knowledge as: "a mental grasp of fact(s) of reality, reached by perceptual observation or by a process of reason based on perceptual observation".

So what do we have to know to form the concept of knowledge? We have to know the difference between the two types of awareness of reality, namely perceptual and conceptual. We have to know of the process of reasoning. We have to know the concept of consciousness and reality.

Are we at the perceptual level? Yes, more or less: all these concepts are either axiomatic, and thus perceptual, or very close.

The distinction between perceptual and conceptual is not axiomatic, but once you have formed some concepts then you can introspectively observe the differences between your abstract ideas and your concrete percepts. (Plato discovered many such differences through introspective observations, e.g., concepts are general but percepts are particular, concepts are eternal but perceptual concretes can go out of existence, etc.) We can also form the concept of reasoning or thinking introspectively by observing what we are doing when we form concepts, generalize, make judgments, integrate, etc and contrast that to other mental actions, such as when we experience emotions or when we are trying to remember something.

Reality or existence is an axiomatic concept. So is consciousness. It is therefore not necessary to reduce them. To _validate_ them only requires direct observation; that you open your eyes and look at reality. So this is were we begin.

[Vik]

How to reduce the concept of "knowledge"? To know what facts give rise to this concept I must know what knowledge is. Ayn Rand defines knowledge as: "a mental grasp of fact(s) of reality, reached by perceptual observation or by a process of reason based on perceptual observation".

So what do we have to know to form the concept of knowledge? We have to know the difference between the two types of awareness of reality, namely perceptual and conceptual. We have to know of the process of reasoning. We have to know the concept of consciousness and reality.

Are we at the perceptual level? Yes, more or less: all these concepts are either axiomatic, and thus perceptual, or very close.

The distinction between perceptual and conceptual is not axiomatic, but once you have formed some concepts then you can introspectively observe the differences between your abstract ideas and your concrete percepts. (Plato discovered many such differences through introspective observations, e.g., concepts are general but percepts are particular, concepts are eternal but perceptual concretes can go out of existence, etc.) We can also form the concept of reasoning or thinking introspectively by observing what we are doing when we form concepts, generalize, make judgments, integrate, etc and contrast that to other mental actions, such as when we experience emotions or when we are trying to remember something.

Reality or existence is an axiomatic concept. So is consciousness. It is therefore not necessary to reduce them. To _validate_ them only requires direct observation; that you open your eyes and look at reality. So this is were we begin.

That forms the basic skeleton of the reduction of the concept of "knowledge" to the perceptual level, yes.

It should be fleshed out with instances of each concept. Definitions should be adjusted when necessary.

My fear is that if a careless reader focuses on definitions and ignores the necessity of examining instances, they'll end up as another victim of the analytic-synthetic dichotomy.

[Vik]

I didn't say TV programs, I said mirrors. It matters.

An animal will make all kinds of different adjustments to be able to see or hear something they have detected. They make the adjustments the situation requires, and the specific behavior involved varies greatly. The goal of this behavior is seeing or hearing. They act to be able to see, etc. Seeing is the goal of that, instrumental behavior. Isn't it true, then that they have some primitive grasp of what it is to see, versus not being able to see, something they have detected? I think so.

And, as we share that animal level of organization, our well-defined concept of knowledge is likely to have its roots in that very primitive grasp. That's why I mentioned the behavior of infants who would not be considered conceptual.

Are you comfortable dismissing such behavior entirely?

-- Mindy

I know what you said.

My point is that if cats can't generalize about the television, they have NOT learned the difference between an appearance and a physical presence.

The mirror example tells us that they can distinguish an appearance *in a mirror* from physical reality. But it doesn't prove that they are capable of isolating the characteristics of appearance as such. And acquiring knowledge means learning about the characteristics, properties, potentials, etc. of things. So while cats certainly have some sort of "primitive grasp", what they have "learned" doesn't qualify as knowledge. It's something more basic.

The way I see it, primate evolution took a radically different turn with cognition than feline evolution. Gorillas and chimpanzees will happily watch television. Only a cat will bat at it.

The cat cases reinforce the idea that we need to clearly distinguish *our* kind of cognition from the other, dare I say more primitive, types of animal cognition.

The fact that cats can grasp one subcategory of appearance *but not another* is very interesting. It suggests that they can abstract from perceptual concretes under certain circumstances, but they cannot abstract from abstractions. Perhaps they cannot grasp "appearance" as such. Perhaps they can grasp differences *only* when virtually everything else is the same but not when multiple things are different.

[knast]

That forms the basic skeleton of the reduction of the concept of "knowledge" to the perceptual level, yes.

It should be fleshed out with instances of each concept. Definitions should be adjusted when necessary.

My fear is that if a careless reader focuses on definitions and ignores the necessity of examining instances, they'll end up as another victim of the analytic-synthetic dichotomy.

I only aimed at giving the "skeleton".

[Vik]

I only aimed at giving the "skeleton".

It's an excellent one. Thank you very much.

[Mindy]

I know what you said.

My point is that if cats can't generalize about the television, they have NOT learned the difference between an appearance and a physical presence.

The mirror example tells us that they can distinguish an appearance *in a mirror* from physical reality. But it doesn't prove that they are capable of isolating the characteristics of appearance as such. And acquiring knowledge means learning about the characteristics, properties, potentials, etc. of things. So while cats certainly have some sort of "primitive grasp", what they have "learned" doesn't qualify as knowledge. It's something more basic.

The way I see it, primate evolution took a radically different turn with cognition than feline evolution. Gorillas and chimpanzees will happily watch television. Only a cat will bat at it.

The cat cases reinforce the idea that we need to clearly distinguish *our* kind of cognition from the other, dare I say more primitive, types of animal cognition.

The fact that cats can grasp one subcategory of appearance *but not another* is very interesting. It suggests that they can abstract from perceptual concretes under certain circumstances, but they cannot abstract from abstractions. Perhaps they cannot grasp "appearance" as such. Perhaps they can grasp differences *only* when virtually everything else is the same but not when multiple things are different.

Remember the old confusion of Phosphorus and Vesperus? The morning star and the evening star? They "appeared" to be different stars, though they weren't. Penetrating appearances is not a given, it involves a discovery. It is not some level of cognition. Science is full of discoveries of the sort.

So the fact that cats do not realize TV is TV is irrelevant. (As far as I know, cats do know TV is illusion, but, like so many of us, they can interact entertainingly with that illusion.)

Now, your claim that to realize something is an appearance requires that "...they are capable of isolating the characteristics of appearance as such," confuses the phenenomen with its conceptualization, and inverts the process of conceptualization. First comes experience, then concepts pertaining to it.

Animals, some, at least, and infants operate with a notion of "see," and "hear," etc. if you want to understand how we come to understand knowledge per se, I think you need to embrace that evidence.

"[A]cquiring knowledge means learning about the characteristics, properties, potentials, etc. of things." Isn't seeing acquiring knowledge? What are concepts, propositions, and reasoning, independent of their sensory roots?

-- Mindy

[Vik]

Remember the old confusion of Phosphorus and Vesperus? The morning star and the evening star? They "appeared" to be different stars, though they weren't. Penetrating appearances is not a given, it involves a discovery. It is not some level of cognition. Science is full of discoveries of the sort.

Equivocation on "appearance". We were discussing the difference between a projection and reality, NOT whether two observations happen to correspond to the same object.

Those are VERY different acts of cognition.

But moving on.

Discovery of general truths requires a level of cognition capable of identifying general truths.

Discovery of particular truths requires only a level of cognition capable of identifying particular truths.

I do not deny that cats can grasp particular truths about mirrors.

But telling me that humans have difficulty with even higher levels of abstraction doesn't make my point about lower level abstraction irrelevant.

So the fact that cats do not realize TV is TV is irrelevant.

Cats have far more difficulty with televisions than mirrors.

Humans have little difficulty with either.

There's a major, highly relevant difference there.

(As far as I know, cats do know TV is illusion, but, like so many of us, they can interact entertainingly with that illusion.)

If they do, they forget far more often than console gamers do.

And you need to rule out alternate explanations for NOT attacking a television, such as the possibility that cats think there's an inaccessible animal behind an impenetrable shield.

The fact of the matter is that I don't bat at a television screen. That's a *qualitative* behavioral difference, not a quantitative one involving levels of knowledge. There's a *qualitative* difference between cat cognition and human cognition, one that cannot be explained by any amount of gradation of "discovery".

I don't doubt that cats can discover particular facts about particular objects such as mirrors and televisions.

I even grant that they can (eventually) distinguish between mirrors and reality.

But it would take an enormous amount of evidence to convince me that they can go beyond that and grasp the idea of "illusion" as such and apply it to EVERY type that we do.

Now, your claim that to realize something is an appearance requires that "...they are capable of isolating the characteristics of appearance as such," confuses the phenenomen with its conceptualization, and inverts the process of conceptualization. First comes experience, then concepts pertaining to it.

What I claimed was that

• grasping appearance as such requires a HIGHER-LEVEL abstraction
  
• evidence of grasping LOWER-LEVEL types of appearances, such as mirrors, does not in itself qualify as grasping appearance as such, particularly when we have evidence that ca
  

Animals, some, at least, and infants operate with a notion of "see," and "hear," etc. if you want to understand how we come to understand knowledge per se, I think you need to embrace that evidence.

Which animals and what evidence?

As for infants, it's debatable.

"[A]cquiring knowledge means learning about the characteristics, properties, potentials, etc. of things." Isn't seeing acquiring knowledge? What are concepts, propositions, and reasoning, independent of their sensory roots?

I do not deny that perception is required before you can have concepts, propositions, and knowledge of anything.

But I deny that perception alone will give you knowledge.

If all you had was perception, if you couldn't integrate any of it, if you had no ability to do logic, if you couldn't validate any of it, you'd have a loosely associated pile of perceptual concretes.

Would you call THAT "knowledge"?

[Grames]

If all you had was perception, if you couldn't integrate any of it, if you had no ability to do logic, if you couldn't validate any of it, you'd have a loosely associated pile of perceptual concretes.

Would you call THAT "knowledge"?

If all you had was perception, if you couldn't integrate any of it, if you had no ability to do logic, then you would have no need to validate anything because you would be a cat.

Perceptual concretes are knowledge, they are the basis of all knowledge. Chapter 6 of Kelley's "The Evidence of the Senses" covers the two contending theories of the justification of knowledge, foundationalism and coherentism.

Knowledge has structure. Most knowledge is acquired and validated by inference from prior knowledge. Where does this prior knowledge come from? This chapter is about the foundationalist and coherence theories of the justification of knowledge and the realist theory of perception affects that debate.

Definitions

I.Foundationalism

Basic knowledge is necessary for any other kind of knowledge, for the truth of any inferential knowledge depends upon the truth of its premises.

Basic knowledge is possible. There are judgements justified without prior conceptual knowledge of other facts.

Perception provides nonpropositional, non-inferential basic knowledge. Denying percepts are knowledge cuts off any possibility of justifying knowledge by tying it to reality.

Knowledge as an abstraction refers to units, those units include in their most elementary form the percepts delivered by the senses. Cats do not have the facility of abstraction to possess the concept of knowledge, but they have the referents of that concept.

[Mindy]

Equivocation on "appearance". We were discussing the difference between a projection and reality, NOT whether two observations happen to correspond to the same object.

Response: No, we were not discussing the difference between a projection and reality. I brought the cats into this, and it was regarding the difference between their behavior when they first encounter their own image in a mirror, and their behavior when they become accustomed to it. So, you see, we WERE discussing whether two observations happen to correspond to the same object. Since you responded in terms of TV projections, you yourself brought an additional version of "appearance" into the discussion. That's not an equivocation at all, but certainly not one you can pin on me.

Discovery of general truths requires a level of cognition capable of identifying general truths.

Response: Animals learn how to do things such as open a latch. They then get faster and faster at opening increasingly complex latches. No generalization? Perceptual abstractions are recognized as such because they lead to generalized behavior.

Discovery of particular truths requires only a level of cognition capable of identifying particular truths.

Response: You are way, way beyond me here. I don't suppose animals or infants formulate propositions, which are the only intellectual product subject to the judgment of true/false. Funny that you take that to be a simpler process.

I do not deny that cats can grasp particular truths about mirrors.

Response: I do deny it.

But telling me that humans have difficulty with even higher levels of abstraction doesn't make my point about lower level abstraction irrelevant.

Response: This refers to what? I haven't brought up any difficulty humans have...

Cats have far more difficulty with televisions than mirrors.

Response: If you say so. It isn't a point I addressed, and I don't see its relevance.

Humans have little difficulty with either.

There's a major, highly relevant difference there.

Response: Yes, there's a gigantic difference between cats and humans. That doesn't mean they have nothing in common. People grasp the phenomena of appearances more readily, and in more subtle ways than animals/infants do. That does not mean that animals and infants do not have any such grasp. Which is the point.

And you need to rule out alternate explanations for NOT attacking a television, such as the possibility that cats think there's an inaccessible animal behind an impenetrable shield.

The fact of the matter is that I don't bat at a television screen. That's a *qualitative* behavioral difference, not a quantitative one involving levels of knowledge. There's a *qualitative* difference between cat cognition and human cognition, one that cannot be explained by any amount of gradation of "discovery".

Response: I'm glad to hear it. To claim, however, that your not batting at your TV is a difference from cats' behavior explained by a qualitative difference, not one of degree of sophistication, is to beg the question outright.

I don't doubt that cats can discover particular facts about particular objects such as mirrors and televisions.

I even grant that they can (eventually) distinguish between mirrors and reality.

But it would take an enormous amount of evidence to convince me that they can go beyond that and grasp the idea of "illusion" as such and apply it to EVERY type that we do.

Response: That straw man is safe from me. ("...and apply it to EVERY type that we do."--You stuffed him full.) (When did we change to "illusion?")

What I claimed was that

[*]grasping appearance as such requires a HIGHER-LEVEL abstraction

Response: Higher than something, yes.

[*]evidence of grasping LOWER-LEVEL types of appearances, such as mirrors, does not in itself qualify as grasping appearance as such, particularly when we have evidence that ca

Response: Your own statement is: "Evidence of grasping...appearances,...does not...qualify as grasping appearance..." No? Works for me.

...But I deny that perception alone will give you knowledge.

If all you had was perception, if you couldn't integrate any of it, if you had no ability to do logic, if you couldn't validate any of it, you'd have a loosely associated pile of perceptual concretes.

Would you call THAT "knowledge"?

The question before us is what underlies our attainment of the concept, "knowledge." Like all concepts, what underlies "knowledge" is something more primitive than that concept. You appear to be objecting to my searching for the most basic grasp of "knowledge" we can identify.

-- Mindy

Edited July 24, 2010 by Mindy

[Vik]

If all you had was perception, if you couldn't integrate any of it, if you had no ability to do logic, then you would have no need to validate anything because you would be a cat.

Perceptual concretes are knowledge, they are the basis of all knowledge. Chapter 6 of Kelley's "The Evidence of the Senses" covers the two contending theories of the justification of knowledge, foundationalism and coherentism.

Perception provides the *material* for knowledge, but you should be very careful what you call "knowledge".

You wouldn't call the verdict of your inner ear knowledge of the direction of gravity.

You have to integrate what you inner ear is telling you with other observations before you can determine your orientation with respect to gravity.

Perception provides nonpropositional, non-inferential basic knowledge. Denying percepts are knowledge cuts off any possibility of justifying knowledge by tying it to reality.

Knowledge as an abstraction refers to units, those units include in their most elementary form the percepts delivered by the senses. Cats do not have the facility of abstraction to possess the concept of knowledge, but they have the referents of that concept.

The fact that knowledge can be reached through perceptual observation does NOT mean that undigested percepts qualify as knowledge.

[Vik]

The question before us is what underlies our attainment of the concept, "knowledge." Like all concepts, what underlies "knowledge" is something more primitive than that concept. You appear to be objecting to my searching for the most basic grasp of "knowledge" we can identify.

-- Mindy

What I meant by "prerequisite" was that humans mentally do certain things before they form a concept of "knowledge".

I'm not asking for an evolutionary perspective on the cognitive machinery that gives us knowledge.

Insofar as that is concerned, I submit the possibility that cats can form abstractions about perceptual concretes. But many facts indicate they can't abstract from first-level abstractions.

The fact that cats have more difficulty with televisions than mirrors is one of them.

And you DID bring up a "difficulty" concerning humans:

"Remember the old confusion of Phosphorus and Vesperus? The morning star and the evening star? "

It takes quite a bit to grasp that two bright things in the sky are actually the same object. The ability to do the kind of work that astronomy and mathematics demands involves far more cognitive machinery than it takes to grasp that a reflection in the mirror is oneself.

As for animals opening increasingly complicated latches, it's possible that animals can abstract from perceptual-concretes. I still have yet to see evidence that they can abstract from abstractions. The token-hoarding gorillas come close.

I don't deny that cats and humans share some similarities. But if your goal is to find something more primitive than what we can do, you should distinguish it from what we can do. That was why I kept emphasizing levels of abstraction. That could be part of the continuum you're looking for.

BTW, I didn't mean "truth". I meant something more advanced than "perceptual observation" but less advanced than "proven". I'm still not sure what the right word is. For now, replace with "fact".

[Grames]

Perception provides the *material* for knowledge, but you should be very careful what you call "knowledge".

You wouldn't call the verdict of your inner ear knowledge of the direction of gravity. You have to integrate what you inner ear is telling you with other observations before you can determine your orientation with respect to gravity.

The fact that knowledge can be reached through perceptual observation does NOT mean that undigested percepts qualify as knowledge.

Given that I know the equivalence between acceleration and gravity I would hesitate to call the verdict of my inner ear the direction of gravity in some exotic circumstances, but I would never hesitate to call it the direction of local 'down' and that would be knowledge. The same applies to knowing where my foot is without looking at it, or by means of looking at it.

It is not possible to induce or deduce knowledge from non-knowledge. This is the essence of the Kantian attack on Cartesian representationalism and it is a perfectly valid critique. All concepts and propositions subsequent to undigested percepts only qualify as knowledge if percepts are knowledge.

The formulation that "perception provides the material for knowledge" comes from Galt's speech, but the later definition of knowledge Rand provides in IOE is “'Knowledge' is . . . a mental grasp of a fact(s) of reality, reached either by perceptual observation or by a process of reason based on perceptual observation." Because it is later and appears in a technical non-fiction presentation on epistemology the IOE definition has priority as the final word of what Ayn Rand considered knowledge. The phrases "body of knowledge" and "sum of human knowledge" refer to knowledge that can be communicated or recorded; this must refer to conceptual knowledge only as it requires words and percepts are non-verbal and cannot be communicated. These phrases designate subsets of the broad definition of knowledge and are derivatives of the main idea. If it suits your purpose to restrict your inquiry into only conceptual knowledge you are perfectly free to do so, but since you brought up perception in the first post I won't guess at what you had in mind.

[Vik]

Given that I know the equivalence between acceleration and gravity I would hesitate to call the verdict of my inner ear the direction of gravity in some exotic circumstances, but I would never hesitate to call it the direction of local 'down' and that would be knowledge. The same applies to knowing where my foot is without looking at it, or by means of looking at it.

The equivalence principle represents higher-level knowledge. If you use it to interpret perceptual information, you're demonstrating that some perceptual observations require higher-level knowledge before they can be used to tell you what you want to know.

I don't deny that you have knowledge of "local down". I'm just hesitant to apply certain words to certain things.

Explanation below.

It is not possible to induce or deduce knowledge from non-knowledge.

This is the essence of the Kantian attack on Cartesian representationalism and it is a perfectly valid critique. All concepts and propositions subsequent to undigested percepts only qualify as knowledge if percepts are knowledge.

I would say they qualify as knowledge only when you use them properly. Like you said, the inner ear merely reports on "local down", not the direction of gravity. It would be a mistake to regard the former as knowledge of the latter.

We reach knowledge of certain particular, concrete facts through perceptual observation. That secures all higher-level knowledge to a perceptual base. You don't need to invoke "percept".

More on the distinction below.

The formulation that "perception provides the material for knowledge" comes from Galt's speech, but the later definition of knowledge Rand provides in IOE is “'Knowledge' is . . . a mental grasp of a fact(s) of reality, reached either by perceptual observation or by a process of reason based on perceptual observation." Because it is later and appears in a technical non-fiction presentation on epistemology the IOE definition has priority as the final word of what Ayn Rand considered knowledge. The phrases "body of knowledge" and "sum of human knowledge" refer to knowledge that can be communicated or recorded; this must refer to conceptual knowledge only as it requires words and percepts are non-verbal and cannot be communicated. These phrases designate subsets of the broad definition of knowledge and are derivatives of the main idea. If it suits your purpose to restrict your inquiry into only conceptual knowledge you are perfectly free to do so, but since you brought up perception in the first post I won't guess at what you had in mind.

Rand's definition of "knowledge" involves "perceptual observation", not "percept".

And

"A percept is a group of sensations automatically retained and integrated by the brain of a living organism. It is in the form of percepts that man grasps the evidence of his senses and apprehends reality."

~IOE, 1, pg5

I think it's only fair to her views that we don't treat her words as synonyms unless she explicitly said we could treat certain words as synonyms.

If you want to talk about interpretations, I'd say the use of "perceptual observation" rather than "percept" or "perception" suggests that she believed we did something with percepts before we have "perceptual observation".

I'm not saying my interpretation is correct, but I haven't seen anything to contradict it. Although honestly it's a minor point.

What's important is that we reach knowledge of certain particular, concrete facts through perceptual observation. That secures all higher-level knowledge to a perceptual base, thereby avoiding the Kantian attack.