The Analytic-Synthetic Dichotomy

[Daniel Barnes]

Mike:

>Imagine that I am seeking knowledge to build a thinking system and not just a word game as I ask again: What is your view of the nature of concepts (or ideas)?

Mike, what you don't realise is that this kind of question is itself an invitation to a word game. That is the fundamental problem with "what is the true nature of...?" questions, and is a fundamental problem that is widespread in philosophy in general, and Objectivism in particular. This style of thinking is called "essentialism", and is the is the whole target of Popper's critical essay, "Two Kinds of Definition." You need to absorb this point - it is an important one. Read the essay again if necessary. I simply do not think this is a productive type of question, regardless of the common assumption that it is intellectually important. I disagree, for the reasons you will hopefully have read twice by now.

Now, you will find that my criticisms of Rand's epistemology focusses on concrete issues and avoids any discussion of the "true nature" concepts. Here's some typical points you will recognise.

1) Her theory includes a violation of the Law of Identity at its first step

2) Where it does not violate the LOI it merely restates it, adding nothing original except some verbiage

3) Her theory of universals begs the question

4) Her theory fails to recognise that words are fundamentally not the same as numbers or algebraic symbols, thus cannot achieve anything like the same precision

5) Her claim that it takes two or more existents to form a concept I believe may be empirically testable; and I believe will almost certainly fail such a test.

6) Her theory relies on fundamentally self-contradictory statements for support ie: her oxymorons like "contextual absolute."

And so on and so forth. While all these points have fuller arguments behind them, you will note they are either logical or (potentially) empirical criticisms, and none of them rely on vague my-true-nature-of-concepts vs your-true-nature-of-concepts type waffle. You can get to a true-or-false with them. Which you cannot do with concepts, or their non-mind-reading equivalent, definitions.

Edited June 4, 2007 by Daniel Barnes

[Gregory Browne]

It may clarify debate on the ASD if I made some points about meaning and convention.

These are my views, but I believe that the great majority of them are compatible with Peikoff's views, and I believe that most of them are actually implied by his views in the ASD, whether explicitly stated there or not.

My belief about meanings is (roughly) that the meaning of a term (at least a noun term) is the same as its reference---that is, the collection of its referent(s)--the beings the term refers to. To the meaning of 'giraffes' is giraffes. I think that ordinary people take

'What does 'giraffe' mean?'

to mean the same thing as

'What is a giraffe?'

The reference is also called the "extension" of a concept, in contrast to the "intension", which is the attributes (or common attributes) of its referents---in the case of giraffes, having a long neck, having long-legs, being yellow with brown spots, having an even-number of hooves, being a mammal, etc.

However, many philosophers, apparently going back at least to the Port-Royal Logic of the 1600s (co-written by Arnauld), have believed the meaing is in fact the same as the intension. This view was adopted in the late 1800s by Frege (who called meanings 'senses' and contrasted them with references) and was carried on by the Logical Positivists and, with modifications, by the Ordinary Language philosophers.

Later (1970s onward) the need to acknowledge the role of reference was recognized by Putnam (Twin Earth, etc.), and so many now think that both the extension and the intension are part of the meaning.

Peikoff definitely takes the extensional view, and I agree with him.

However, I admit the intension may have a role in identifying the extension, and Peikoff has not yet denied this. The role differs with different types of reference:

in the case of individuals, we can directly refer to them extensionally: by ostension (pointing), though partial descriptions (such as 'the President of the U.S.') may help, and they are intensional;

in the case of what I call Narrow Classes (which are random collections of things, such as Bill Clinton, the Eiffel Tower and Mt. Everest, and not important), their referents can be identified by pointing or by listing;

at the other extreme from individuals, what I call Shallow Kinds are normally identified intensionally, by giving a definition or description--for example, the referents of 'bachelors' are identified by a definition: unmarried, male humans of marriageable age;

in the middle, Deep Kinds are identified in a more complex way that combines the extensional and intensional methods: first we apply the kind term to some individual beings that we thereby make references (for example, we come upon some samples of a mineral and call it" gold"), which thus become exemplars or 'paradigm' cases of the kind, and then we study them to ascertain their complete intension (in this case, we eventually discover its atomic structure, and then use this intension to determine what additional beings belong to the extension--i.e., what is a member of the kind (thus ruling out Fool's Gold as gold) (though provisionally we use the intensional attributes we have to provisionally determine what belongs to the extension: until modern chemistry is invented, or until we get to the assay office, we relay on outward appearances). So here the order is: partial extension--intention--remainder of the extension.

Now many people think that meanings, or definitions, are always either convetional or stipulative--in short, arbitrary. This is standard among Logical Positivists and Ordinary Language Philosophers, and, while Quine has questioned much of this doctrine, in other ways he and those influenced by him are apt to spread the conventionalistic doctrine even further. So it is important to be clear on where arbitrariness into all of this.

I deal with 3 states regarding our choices: in some cases are choices are either correct or incorrect (or a mixture of these); in other cases there is no incorrect choice, so our choices can be completely arbitrary; in still other cases, there may be several correct choices but some may be of greater practically value than others, in which cases I say that the choice is "pragmatic". (This seems to be the sort of thing Michael talked about.) So the choice may be arbitrary, pragmatic, or a matter of correctness and incorrectness.

In the case of individuals and Narrow Classes, it is arbitary or pragmatic which we select to refer to, but then this determines the intension.

In the case of Shallow Kinds, it is arbitrary or pragmatic what attributes we select to put into our intenstion, but then this determines the extension.

In the case Deep Kinds, it is arbitrary or pragmatic what individuals we select as our paradigms, but then this determines the intension, and that determines the rest of the extension.

In all cases, it is not arbitrary or pragmatic what the definition is, unless we simply start of giving the intension by stipulating a definition.

A final note on the "open-endedness" of concepts that Rand and Peikoff talk about: they use the term in two ways, to refer to two features of concepts, which I will refer to as "open-sidedness" and "open-bottomedness".

Open-sided is the feature of a concept which allows the concept to refer to an unlimited number of referents of the same kind: for example, the concept of dogs is open-ended because it refers not just to all the dogs we know or have known, but to all dogs, past, present and future, and indeed to all possible dogs: if some mother dog who has died had in fact had one more litter then there would be a few more dogs in the world than there actually are now, and these, too, would be referents of 'dogs'. All kinds are open-ended (in contrast to Narrow Classes, which are not, but they are unimportant and interesting).

Open-bottomedness is the feature of a concept by which it can contain more attributes than we know at a given time. It is belief in this kind of open-endedness for which Peikoff has been criticized by several people here. I say that Deep Kinds, Narrow Classes and individuals are open-bottomed, whereas as Shallow Kinds are not

Edited June 4, 2007 by Greg Browne

[Gregory Browne]

Mike:

>Imagine that I am seeking knowledge to build a thinking system and not just a word game as I ask again: What is your view of the nature of concepts (or ideas)?

Mike, what you don't realise is that this kind of question is itself an invitation to a word game. That is the fundamental problem with "what is the true nature of...?" questions, and is a fundamental problem that is widespread in philosophy in general, and Objectivism in particular. This style of thinking is called "essentialism", and is the is the whole target of Popper's critical essay, "Two Kinds of Definition." You need to absorb this point - it is an important one. Read the essay again if necessary.

I glanced at your website and saw some interesting topics, but I think that the major relevant points of it should be reproduced here, so we can quote and reply to passages, and not have to jump back and forth between the websites. Thank you very much.

For now, I would like to know how Popper, or you, define the term 'Essentialism', and what Popper's objection to Essentialism, or your objection, is (I can guess, but I prefer to hear it stated explicitly), and what are the 2 kinds of definition he has in mind.

Literally, 'Essentialism' means belief in essences. 'Essence' has 2 main meanings in philosophy: (1) the set of the necessary attributes of something: the word is used this way by Aristotle (according to one interpretation) and most modern philosophers who discuss essence, such as Putnam and Kripke; (2) a subset of this first set, consisting of the smallest set of necessary attributes from which the others can be deduced; in other words, the fundamental attributes: the word is used this way by Aristotle (according to a second interpretation) and by Rand.

Well even Logical Positivists can believe in essences in this sense: e.g., the essenc of bachelors is humanness, unmarried, maleness and being of marriageable age. But they are not usually called "Essentialists".

Most commonly, it is used to mean those who believe in what I call "Deep Essences", which are the essences of Deep Kinds.

1) Her theory includes a violation of the Law of Identity at its first step

What is the violation?

3) Her theory of universals begs the question

What question does it beg, and in what way?

4) Her theory fails to recognise that words are fundamentally not the same as numbers or algebraic symbols, thus cannot achieve anything like the same precision

Why do you think that they cannot be precise?

6) Her theory relies on fundamentally self-contradictory statements for support ie: her oxymorons like "contextual absolute."

Why do you think it is contradictory?

(Also, an oxymoron is not a contradiction, but rather something that merely sounds contradictory: it means "pointedly foolish": e.g., "They stayed away in droves", "They were conspicuous by their absence".

[Gregory Browne]

Daniel's mention of Popper's 2 kinds of definition reminded me that I was going to say something about the 2 kinds of definition that I think that we need to know about for this forum.

The distinction between them was made by Aristotle. The 2 kinds of definition are Nominal Definitions and Real Definitions.

Nominal Definitions are definitions which we learn when we first learn the meaning of a term (or which can be formulated later from the conventions governing meaning that we learn at this time). For example, the Nominal Definition of water might be: a material that quenches thirst, is liquid at normal temperatures, it tasteless, is colorless and is odorless.

Real Definitions are definitions which list the fundamental attributes of something, from which all its other attributes can be deduced; such definitions thus express the essence, in the second sense of 'essence' I discussed in my last post. For example, the Real Definition of water might be: the chemical element whose structure is expressed by the formula H2O.

In Deep Kinds, Real Definitions usually require a lot of empirical investigation to discover, and are usually different from the Nominal Definition of that kind.

In Shallow Kinds, Real Definitions and Nominal Definitions are the same: e.g., there is nothing more to be discovered about triangles than what is contained in their Nominal Definition (explicitly or implicitly: much was discovered about triangles in trigonometry, but this was deduced from the definition, and so it all had to be implied by the definition, and so implicitily contained in it).

[Michael Stuart Kelly]

Mike, what you don't realise is that this kind of question is itself an invitation to a word game. That is the fundamental problem with "what is the true nature of...?" questions,...

Daniel,

I am going to hold you to my words as I wrote them. I was very precise. I did not ask for the "true nature" of anything. That is your invention. I asked:

"What is your view of the nature of concepts (or ideas)?"

"Your view of the nature," not "true nature." What that means is I want to know what on earth you think a concept or idea is in terms of knowledge. I will give an example that might help, but before I continue, let me address a point you made about the Objectivist theory of concepts (the only one with which I agree).

5) Her claim that it takes two or more existents to form a concept I believe may be empirically testable; and I believe will almost certainly fail such a test.

I agree with this. I believe that a concept of a single existent can be made. However, this corresponds to the number 1. Inherent in the number 1 is the possibility of being a unit or having unit(s) and it needs at least one or more numbers for this to become concrete. But that still does not mean that it is not a number if that is all there is. So although a concept of a single entity only has one unit, it has the capability of compressing more units into itself if more existents with the proper characteristics appear. I see no rational reason to exclude the denomination of concept from this mental abstraction.

Now back to the Barnes notion of concepts. I fully understand the Aristotelean essentialism part of Popper's essay, so much so that I identified it almost immediately with Rand's categories of schools of thought on universals even though the wording was different (she called it "moderate realism" for some reason).

But I have an example that I think might help. Under Objectivism (and even Popperism), reality exists independently of the awareness of it. According to reality, there are two facts that are inescapable without calling them "essences." The first is that individual entities exist. You, the person Daniel, exist as a separate thing from me, Michael. The second fact is that reality exists in such a manner that different entities (and attributes, actions, and relationships) have similarities and differences that can be perceived, identified and mentally categorized. This is not a mental game. These similarities and differences exist irrespective of any mind to perceive them. If they did not exist, it would be impossible to perceive them.

This is not "essence" in the manner Aristotle was saying, where metaphysical spirits or energies or forces (essences) made concrete things what they are.

And even so, a case could be made for species of living things being some kind of force or essence, since birth is the process of life. Species do exist and vast numbers of individuals create vast numbers of similar individuals in most all major respects, while leaving small differences. That is more than random coincidence at the individual level.

In Objectivism, mental symbols for similarities and differences are what is called knowledge. Knowledge rests on recognizing the fact that individual entities exist with the feature of having similarities and differences with other existents. My problem is that I don't understand yet what Daniel Barnes calls "knowledge."

Now my example: Daniel himself. Yep. You.

We all agree that Daniel exists and that there is only one of him. But in order to understand what Daniel is in his entirety would require a mind that would have awareness of everything past, present and future, plus all the gory details down to the tiniest subparticles. Still, I say "Daniel" and I mean you. I know you are a man. However, I have not even seen you. My idea of Daniel right now is very limited because we have not met yet. I have seen words written on a computer monitor and that's about it.

What a medical doctor who knows you would call Daniel would be a very different view than I have. He would see bones and innards and many small details that to him are "man," plus your normal aspect of what you look like and your personality. Most of that to me is either so remote from my normal manner of thinking as to be practically nonexistent or it is not at all within my knowledge or range of awareness. Yet we both refer to the same entity: you.

What a cannibal would call Daniel would be a very different view than either I or the medical doctor would have. He would automatically look at the choice succulent parts before all else.

The same observation about a different view goes for a criminal, or a missionary, or a ballet dancer, or a plastic surgeon, etc. All of these would have vastly different mental images of you, yet they all refer to the same entity: Daniel.

So it is safe to say that the entity, you, exists and the awareness of it (knowledge) depends on many factors. Yet there is a common form of indicating this unique whole entity where everyone knows which entity is being referred to. And this is despite there being many different perspectives, bodies of observations and values involved.

That is knowledge. I can tell you what the components of similarity and difference are and why so many different views all mean the same Daniel, and still everybody understands each other.

That is my view. I can't believe that the Popper view of knowledge is based on lack of the existence of knowledge. That wouldn't make sense, even by his terms. And if knowledge exists, it must have some feature that can be described. If knowledge can be falsified, it has to exist, thus it has to be able to be described.

So from this angle, I repeat my question: "What is your view of the nature of concepts (or ideas)?"

What is knowledge to you?

Michael

[Wolf DeVoon]

words are fundamentally not the same as numbers or algebraic symbols, thus cannot achieve anything like the same precision

This is set theory arrogance and foot-stamping, a variety of Know Nothing nihilism. Its purpose is the destruction of natural language predicate logic, forbidding men to claim knowledge of any kind other than arbitrary lists and caveman finger pointing at ostensible objects.

many people think that meanings, or definitions, are always either convetional or stipulative--in short, arbitrary

Who profits from that assertion? Anyone who feels threatened by reasoned inquiry. It may have seemed that our sidebar on defining an 'emergency' was mere burlesque. I suggest you think twice. Without abstract concepts and predicate logic, man is helpless to posit, isolate and distinguish, refute, clarify or remember any proposition beyond the spottedness of giraffes. All the numbers, symbols, sets and conventions in the world will not help you define 'due process of law' or 'justice.' When a man attempts to define abstract terms such as those -- or an 'emergency' that exempts government from both -- it is possible to challenge those definitions and propose alternate theories.

The definition of justice and implications of due process are not idle banter in a child's sandbox. The welfare and moral rectitude of nations is at stake.

Go ahead. Tell me that 'moral rectitude' is an arbitrary, fuzzy concept incapable of precise communication to our innocent progeny.

W.

Edited June 4, 2007 by Wolf DeVoon

[Daniel Barnes]

Greg:

>For now, I would like to know how Popper, or you, define the term 'Essentialism', and what Popper's objection to Essentialism, or your objection, is (I can guess, but I prefer to hear it stated explicitly), and what are the 2 kinds of definition he has in mind.

Hi Greg

I reproduce Popper's essay here.

>What is the violation?

"The first stage is a child's awareness of objects, of things-which represents the (implicit) concept "entity." The second and closely allied stage is the awareness of specific, particular things which he can recognize and distinguish from the rest of his perceptual field-which represents the (implicit) concept 'identity.'" - Ayn Rand, ITOE

As the Objectivist site Importance of Philosophy puts it, "An entity without an identity cannot exist because it would be nothing." Hence there can be no first stage.

>What question does it beg, and in what way?

The "Conceptual Common Denominator" seems to be little more than a vague buzz-phrase meaning a similarity. (refer here for Will Thomas of the Objectivist Centre's examples of CCDs - scroll to the end) As the problem of universals is, simply, "why are different things similar?" then arguing that this is because they have similarities (or CCDs) begs the question.

>Why do you think that they cannot be precise?

I discuss the situation here.

>Why do you think it is contradictory?

Because an "absolute" is something that is invariable, no matter where it is in space or time (like an absolute law of physics) ie no matter what context.

Edited June 4, 2007 by Daniel Barnes

[Daniel Barnes]

Mike:

>That is my view. I can't believe that the Popper view of knowledge is based on lack of the existence of knowledge.

That's because it's not.

In Critical Rationalism, knowledge certainly exists. In fact you can have two flavours:

1) subjective knowledge (your consciousness: emotions, feelings, imagination, intuition, memory etc that no-one else can know)

2) objective knowledge (any knowledge that exists outside of yourself: literature, art, music, language, arguments, theories etc)

Both kinds of knowledge can err. So concepts or ideas or whatever you want to call them exist. But their analysis leads to unproductive verbalism. What you need to do is focus on testing what are broadly called ideas, which is best done by testing the truth of statements, or theories, and the viability of plans or proposals etc.

(The argument that the above are composed of words, thus words should be analysed first, is countered by the argument that words in turn are composed of letters - do you therefore need to analyse them?)

>What is knowledge to you?

So you can see it is a fairly broad term as far as I am concerned.

[Michael Stuart Kelly]

In Critical Rationalism, knowledge certainly exists. In fact you can have two flavours:

1) subjective knowledge (your consciousness: emotions, feelings, imagination, intuition, memory etc that no-one else can know)

2) objective knowledge (any knowledge that exists outside of yourself: literature, art, music, language, arguments, theories etc)

Both kinds of knowledge can err. So concepts or ideas or whatever you want to call them exist. But their analysis leads to unproductive verbalism. What you need to do is focus on testing what are broadly called ideas, which is best done by testing the truth of statements, or theories, and the viability of plans or proposals etc.

Daniel,

This confuses me. In your two flavors, you describe what in Objectivism would be called the referents of knowledge. For instance, you say that music can err? Language can err?

People err, not the referents.

An emotion doesn't err. It just is. How people interpret it or apply it to another referent is where the error can happen, but that is not the emotion itself erring. The emotion is a referent itself here.

Then you go on to make an extraordinary claim that examining concepts "leads to unproductive verbalism," but then apparently it is this "unproductive verbalism" that you find necessary to test. What else is the actual knowledge in your mind if not "unproductive verbalism"? (I will ignore for now that this begs the concept/definition/word differences.)

I am starting to get the impression that Popper's theory takes knowledge for granted as a metaphysical given like a not-to-be-questioned axiom and merely prefers to skip over it. One thing I have noticed so far with Popper is that he provides a good theory of what isn't. He is good at telling you what is not knowledge. I am having a much harder time trying to discover what he proposes that is.

So far I have discerned reality as a constant. Then it gets really fuzzy from there.

Michael

[Wolf DeVoon]

The "Conceptual Common Denominator" seems to be little more than a vague buzz-phrase meaning a similarity.

This is a misunderstanding of Rand's CCD. Table is a flat surface to hold other objects, a kind of furniture. Both furniture and table are defined by shape, an Aristotelean category. Has nothing to do with 'a similarity' of table and furniture.

W.

[Michael Stuart Kelly]

Daniel,

A CCD (conceptual common denominator) is a feature of an entity (attribute, action or relationship) that can be measured. This means that a unit of measurement can be extracted from it. Notice that this applies to everything on Will Thomas's list in the Appendix at the end of his article, Ayn Rand’s Theory of Concepts: A Brief Overview, that you linked to earlier.

That's a little bit more substantive than a "vague buzz-phrase meaning a similarity."

Michael

[Daniel Barnes]

Mike:

>That's a little bit more substantive than a "vague buzz-phrase meaning a similarity."

I disagree. I think it's the same thing just dressed up in intellectual-sounding verbiage. Mike, what do you think 'common denominator' means? Well, it's an attribute that a group of different things have in common - that is, a similarity. You can measure similarities, but so what? A theory of universals is supposed to explain them.

>People err, not the referents.

Here is the fundamental confusion, right here. I've said this before, but I'll say it again. We need to draw a distinction between

1) The objective facts of reality outside of our consciousness, or what I presume you are calling "referents"

2) Our knowledge of them. In Critical Rationalism, this becomes:

2a. Our internal, personal knowledge inside our consciousness (our thoughts, feelings etc, much of which is inexpressible)

b. Our conscious knowledge that is transferred into some physical medium, such as language (eg: writing on paper, or speech etc) so that others can understand it.

1) obviously cannot err, but this because it's not knowledge in the first place! 2)a. and b. can.

>music can err? Language can err?...An emotion doesn't err. It just is.

As if there had never been a bum note or a spelling mistake! And, as the Buzzcocks would say: Ever fallen in love with someone you shouldn't have fallen in love with?

I come back to this point:

>People err, not the referents.

This is why it is important not to muddle the two.

Edited June 4, 2007 by Daniel Barnes

[Michael Stuart Kelly]

You can measure similarities, but so what? A theory of universals is supposed to explain them.

Daniel,

Sez who?

A theory of universals is supposed to explain the relationship of facts to knowledge.

But going on your premise, what do you think measurements (both identifying measurable features and the measurements themselves) do if not explain similarities and differences?

Also, are you not understanding on purpose? We both seem to agree that people err (the knowledge), not things (the facts). So why state it as if I were saying the contrary?

When I ask you a question, you have not really answered it in these last few posts. I have a good idea of what you don't mean by knowledge. I have very little idea of what you do mean by knowledge.

Let's do it another way. Is knowledge only something in the head or is there a correlation between what is in the head and facts? If a correlation exists, what is it in CR?

Michael

[Wolf DeVoon]

Will Thomas's list in the Appendix conflates distinguishing characteristics and CCDs, which I doubt he understood very clearly. You have to imagine concept formation in three dimensional holography. Defining characteristics are reflected in the concept, but it can't be recorded unless there is a 'reference beam' that creates an interference pattern.

W.

[Dragonfly]

Now we are getting closer to the hear of our disagreement. The reason why we don't check for married bachelors is that the bachelor kind is what I call a Shallow Kind, which means that it has only a few kind attributes (i.e., attributes its members have in common) and so has a simple concept and can be simply and easily defined. So when we first learning the meaning of term 'bachelor' we can learn all of these attributes and do, and know that these are all of the common attributes there are, so there is no need to look for any more.

That is all there is to it: individuals and classes, some of which are kinds, having varying degrees of depth (i.e., varying numbers of common attributes) (with different fields of study studying kinds with different degrees of depth: biological kinds are deeper than chemical kinds which are deeper than mechanical kinds, etc), some of which kinds having such a small number of common attributes that they can be learned when we learn the meanings of the term referring to the kinds.

These are the realities of the situation, but they have led many philosophers to believe that truths expressing attributes of Shallow Kinds are fundamentally different from other truths, to imagine that they are known by non-empirical means, and even to imagine that they are created by us (are products of our minds or our language) and so say nothing about reality. But these latter beliefs are simply illusions, and in general unnecessary suppositions, as the facts about the truths are explained adequately by the differences between individuals and classes, between Narrow Classes and Wide Classes (Kinds), by the differences between kinds having varying degrees of depth (and differences between fields of study studying kinds with different degrees of depth), without positing these dichotomies of truths. There are truths about different subject matters, but their differences are not differences in their truth.

Analytical vs. synthetic statements have nothing to do with your classification of shallow kinds vs. deep kinds. The point is that in some cases the definition of a concept is unequivocal, which makes it easy to construct a recognizable analytic statement. An example is the 'bachelor'; while the exact wording may vary, all definitions boil down to "an unmarried man". So we don't have to give the definition explicitly to conclude that "a bachelor is not married" is an analytical statement, as everybody assumes the same definition. In other cases this is not so clear however, I refer to the ice example in my original post, I won't repeat it here. We could also create analytical statements about what you call the deep kinds. Whether such a statement is analytic depends on the definitions of the elements in that statement however, so depending on the use of different definitions, the same proposition may be analytic or not. Of course this is a rather useless exercise, as it doesn't give us any new information, in contrast to the analytical statements of mathematics, which can be extremely useful (I won't repeat here my earlier arguments about that).

They are not assumptions, but elements of stipulative definition (equivalent to a conventional definition, except made by only one person). But I think that's what you meant, because you want to say that my answer is a logical truth that depends on arbitrary linguistic choice (convention or stipulation).

However, I reply by pointing out that ALL truth depends, in part, on linguistic choice. To use Quine's example, even the truth 'Brutus killed Caesar' (which is factual, empirical, probably contingent and supposedly synthetic, and moreover it about individual rather than a class) depends for its truth on the fact that Brutus did kill Caesar but also on the linguistic convention which assigns them their names in our language: had those conventions been different the sentence 'Brutus killed Caesar' could easily be false. And Peikoff and I, and Quine, say that it a mistake to think that any truth is true solely by convention.

It would be only an analytical truth if you define Brutus as the man who killed Caesar, in other words if "Brutus" is the synonym for "murderer of Caesar". But that wouldn't be a sensible definition. Suppose we happen to find new evidence that shows clearly that it wasn't Brutus but Marcus Antonius who killed Caesar, then you would logically have to conclude that Marcus Antonius is Brutus, as "Brutus" is in your definition the man who killed Caesar. That is a good indication that your statement is a synthetic statement, as it can be falsified by empirical evidence, which will never happen with the statement "a bachelor is not married" or "2 + 2 = 4". "Brutus didn't kill Caesar" may be a false statement as far as we know, but it is not self-contradictory, while "a bachelor is married" is self-contradictory.

There is no Platonic realm or purely abstract entities or Platonic Forms or Ideas. If you think there are you must argue for them. I, and most philosophers since the 1200s, and most modern scientists who think about philosophy, will need to be persuaded of their existence.

By the way, I am glad to see that you think that the dichotomies you believe in need to rest on Platonic doctrines. Most people in modern times who believe as you do, such as the Logical Positivists, see themselves as poles apart from Platonism. But your stand supports Peikoff's belief that those poles are really two sides of the same coin, and his tracing of the dichotomies back to Platonism.

You shouldn't jump to conclusions. When I use the metaphor "Platonic realm" you shouldn't try to pigeonhole me immediately as a Platonist, that's a bad Objectivist habit. I just mean that the content of mathematical statements is independent of the physical world, that you can consider it as an abstract world with its own truths that cannot be falsified by empirical evidence. The value of pi will always be the same, independent of the universe we live in.

All geometries, if they have any truth in them and any value of them, say something about possible ways the world can be.

What is "possible"? That is a very vague term that you'll have to define. In daily life we use it to express our ignorance about something. It means that we see no compelling reasons to conclude that a certain event cannot happen, for example while it doesn't seem to violate the known physical laws. But it may be that that event doesn't happen in fact, while it cannot happen. You might say: "under different circumstances it would have happened". But the circumstances were in fact not different; with enough information you'd have known that what seemed possible was in fact not possible.

How do you know that a certain geometry is "possible" in the world? The laws of physics might make such a geometry impossible. You may imagine a world with different laws of physics, in which such a geometry would be "possible". But the word says it all: it is an imaginary world, not the real world. You shouldn't put imaginary worlds on the same footing as the real world. In an imaginary world unicorns, tooth fairies and flying pink elephants may exist, but that doesn't imply that they do exist in the real world.

[Michael Stuart Kelly]

Will Thomas's list in the Appendix conflates distinguishing characteristics and CCDs, which I doubt he understood very clearly. You have to imagine concept formation in three dimensional holography. Defining characteristics are reflected in the concept, but it can't be recorded unless there is a 'reference beam' that creates an interference pattern.

Wolf,

For physical existents, I can certainly go with that. Reference beams are for measuring. (Even alignment is a form of measurement.)

Michael

[Daniel Barnes]

Wolf:

>Will Thomas's list in the Appendix conflates distinguishing characteristics and CCDs, which I doubt he understood very clearly. You have to imagine concept formation in three dimensional holography.

So the Director of Programs at the TOC doesn't understand a central component of Rand's epistemology. Why don't you write and correct him before he spreads the wrong ideas any further?

[Daniel Barnes]

Mike:

>Sez who?

Sez you:

>A theory of universals is supposed to explain the relationship of facts to knowledge.

>But going on your premise, what do you think measurements (both identifying measurable features and the measurements themselves) do if not explain similarities and differences?

They er, measure them.

>Also, are you not understanding on purpose? We both seem to agree that people err (the knowledge), not things (the facts). So why state it as if I were saying the contrary?

Because you then went on and said the contrary! "Do emotions err? Does language err?" etc

>When I ask you a question, you have not really answered it in these last few posts. I have a good idea of what you don't mean by knowledge. I have very little idea of what you do mean by knowledge.

Here's a small selection, once again, of what I mean by human knowledge: theories, arguments, plans, proposals, music, art, myths, science etc etc etc. Once again, its a huge and varied category. How much clearer can I be? Now, at what point do you cease blaming me for this failure to understand, and take a little responsibility yourself?

Edited June 5, 2007 by Daniel Barnes

[Michael Stuart Kelly]

Here's a small selection, once again, of what I mean by human knowledge: theories, arguments, plans, proposals, music, art, myths, science etc etc etc. Once again, its a huge and varied category. How much clearer can I be? Now, at what point do you cease blaming me for this failure to understand, and take a little responsibility yourself?

Daniel,

Every one of these selections are plural. What are the units?

Or is it in Popper's theory that you can have many without one existing? Wouldn't that be false?

Michael

[Daniel Barnes]

Mike:

>Every one of these selections are plural. What are the units?

Well, amazingly, they are: a theory, an argument, a proposal, a plan, a piece of music, a myth, etc.

I confess I am at a loss as to the point of your question.

Edited June 5, 2007 by Daniel Barnes

[Michael Stuart Kelly]

Daniel,

A theory (as knowledge) is made of units of knowledge. It is a composite of many different knowledges.

That goes for the other things you mentioned, too.

Do you need further explanation?

With Popper's theory of knowledge, do composites exist without units?

Michael

[Gregory Browne]

Something is wrong here.  My [quote name]s and [/quote]s come out even, but still I cannot get the white and blue right.

That is all there is to it: individuals and classes, some of which are kinds, having varying degrees of depth (i.e., varying numbers of common attributes) ...some of which kinds having such a small number of common attributes that they can be learned when we learn the meanings of the term referring to the kinds.

These are the realities of the situation, but they have led many philosophers to believe that truths expressing attributes of Shallow Kinds are fundamentally different from other truths, to imagine that they are known by non-empirical means, and even to imagine that they are created by us (are products of our minds or our language) and so say nothing about reality. But these latter beliefs are simply illusions, ...There are truths about different subject matters, but their differences are not differences in their truth.

Analytical vs. synthetic statements have nothing to do with your classification of shallow kinds vs. deep kinds.

My classification explains why people are tempted to believe that some truths are synthetic, some are a priori and some are non-factual.

The point is that in some cases the definition of a concept is unequivocal, which makes it easy to construct a recognizable analytic statement. An example is the 'bachelor'; while the exact wording may vary, all definitions boil down to "an unmarried man". So we don't have to give the definition explicitly to conclude that "a bachelor is not married" is an analytical statement, as everybody assumes the same definition.

In other cases this is not so clear however, I refer to the ice example in my original post, I won't repeat it here.

Where was the unclarity there? Peikoff believes that the definition of 'ice' is solid water, and you don't deny that, do you? Your disagreement was whether ice's floating on water is part of the concept of ice.

I thought that were going to try to argue that the definitions of what I call "Deep Kinds" are unclear, along with their concepts and meanings. It is this kind of think that led the Ordinary Language philosophers to modify the Logical Positivist doctrine on meanings and definitions by introducing the concepts of "cluster concepts" and "family resemblance concepts" and other "open-textured concepts". This was just a Dichotomist way of dealing with some of the problems of Dichotomism within a Dichotomist framework using only Shallow Kinds. But the correct solution was to acknowledge the existence of Deep Kinds (what most philosophers today call "natural kinds").

We could also create analytical statements about what you call the deep kinds.

All statements about Deep Kinds are analytic, and indeed all statements are.

Whether such a statement is analytic depends on the definitions of the elements in that statement however,

If you define "analytic truth" as "definitional truth" , OK (this is what Logical Positivists really mean by their definitions, though that is not what their definitions say), but then that gets you into the problem in the passage:

so depending on the use of different definitions, the same proposition may be analytic or not.

Yes. So I ask again: what good is the analytic-synthetic distinction then?

Of course this is a rather useless exercise, as it doesn't give us any new information, in contrast to the analytical statements of mathematics, which can be extremely useful (I won't repeat here my earlier arguments about that).

Why is this true only about statements of math and not about statements about Deep Kinds?

They are not assumptions, but elements of stipulative definition (equivalent to a conventional definition, except made by only one person). But I think that's what you meant, because you want to say that my answer is a logical truth that depends on arbitrary linguistic choice (convention or stipulation).

However, I reply by pointing out that ALL truth depends, in part, on linguistic choice. To use Quine's example, even the truth 'Brutus killed Caesar' (which is factual, empirical, probably contingent and supposedly synthetic, and moreover it about individual rather than a class) depends for its truth on the fact that Brutus did kill Caesar but also on the linguistic convention which assigns them their names in our language: had those conventions been different the sentence 'Brutus killed Caesar' could easily be false. And Peikoff and I, and Quine, say that it a mistake to think that any truth is true solely by convention.

It would be only an analytical truth

I was not talking about whether it was analytic, but whether it was true by convention (and even those who equate these two, such as Logical Positivists, consider them to have different definitions). My point was that all truths owe their truth partly to linguistic convention, and would no longer be true if we changed the conventions. Therefore it is appropriate to believe that some and only some truths are "true by convention". Rather, I say that all truths are true both because of the fact they express and the conventional meanings of the words used to express the fact.

Brutus didn't kill Caesar" may be a false statement as far as we know, but it is not self-contradictory,

while "a bachelor is married" is self-contradictory.

No. Both are self-contradictory, if false. If he first statement is false then it is true that Brutus killed Caesar, and killing Caesar is one of the predicates which is true of Brutus and so the predicate expresses an attribute of Brutus, and therefore it is part of the meaning of 'Brutus' (even though it is only contingently part of the meaning) and therefore it contradicts the meaning of 'Brutus' (even though only contingently). As Peikoff would say, the meaning of 'Brutus killed Caesar' is 'Brutus (with all of his attributes, including--if the sentence is true--being a killer of Caesar) has the attribute of having killed Caesar', whereas to say that 'Brutus did not kill Caesar' is to say 'Brutus (with all of his attributes, including--if the sentence is true--being a killer of Caesar) does not have the attribute of having killed Caesar', which is self-contradictory.

There is no Platonic realm or purely abstract entities or Platonic Forms or Ideas. If you think there are you must argue for them. I, and most philosophers since the 1200s, and most modern scientists who think about philosophy, will need to be persuaded of their existence.

By the way, I am glad to see that you think that the dichotomies you believe in need to rest on Platonic doctrines. Most people in modern times who believe as you do, such as the Logical Positivists, see themselves as poles apart from Platonism. But your stand supports Peikoff's belief that those poles are really two sides of the same coin, and his tracing of the dichotomies back to Platonism.

You shouldn't jump to conclusions. When I use the metaphor "Platonic realm" you shouldn't try to pigeonhole me immediately as a Platonist, that's a bad Objectivist habit.

It is not exclusively Objectivist and it is not bad: if you claim to believe in something you call "Platonic" your reader has a right to call you are a "Platonist" on that issue. And there are many people who believe in some part of Plato's doctrine on a give issue, but do not believe in most of his philosophy, who are quite willing to say that they are "Platonist" or "platonist" on that specific issue. If you want to say that just because you believe in an abstract realm with its own truths that is independent of the physical world you are not Platonist or Platonic, fine, but then you go on to call it "a Platonic realm", and with a capital 'P'.

In any case that ignores my point: if you think that there is a Platonic realm or purely abstract entities you must argue for them, since they are not believed in by me, nor by Peikoff or Rand, nor by most philosophers since the 1200s, nor by most scientists (whom you quote as authorities for believing in the analytic-synthetic dichotomy).

I just mean that the content of mathematical statements is independent of the physical world, that you can consider it as an abstract world with its own truths that cannot be falsified by empirical evidence. The value of pi will always be the same, independent of the universe we live in.

It is still saying something about this universe (as well as every other possible universe), and so it is factual.

All geometries, if they have any truth in them and any value of them, say something about possible ways the world can be.

What is "possible"? That is a very vague term that you'll have to define.

Yes, but your own position uses that concept, if you believe in all of the dichotomies that Peikoff discusses, because then you believe in the necessary/contingent dichotomy, and that is standard defined in terms of possibility:

a necessary truth is a truth whose opposite is impossible;

a contingent truth is a truth whose opposite is possible;

a clearer way of putting it is:

a necessary truth is a truth such that it is (and always was and always will be) impossible for its negation to be true;

a contingent truth is a truth such that it is (or was or will be) possible for its negation to be true.

So necessity, contingency, possibility and impossibility are intertwined, such that they need to be accepted or rejected together, and defined together (together, they are called "modal terms").

But we must define one and then define the others in terms of it. I think impossibility is basic, so I will say:

'S is P' is impossible if and only if there is an incompatibility between S and P,

or, if you object to 'incompatible' is itself implying possibility (because of the '-ible') then we can replace it with 'inconsistency', thus

'S is P' is impossible if and only if there is an inconsistency between S and P.

Then we can define 'possibility' as absence of 'impossibility' and define 'necessary' and 'contingent' in the usual way.

In daily life we use it to express our ignorance about something. It means that we see no compelling reasons to conclude that a certain event cannot happen, for example while it doesn't seem to violate the known physical laws. But it may be that that event doesn't happen in fact, while it cannot happen. You might say: "under different circumstances it would have happened". But the circumstances were in fact not different; with enough information you'd have known that what seemed possible was in fact not possible.

Yes. The possibility you talk about with your last 'possible' is possibility in the strict sense, also called "metaphysical possibility", to distinguish the first possiblity you talk about, which is "epistemic possibilty", which is equivalen to uncertainty.

The laws of physics might make such a geometry impossible.

In that case then of course it is not possible.

And I am very glad to see that you admit that the laws of physics can make something impossible, because this entails that they are necessary, which is something those who hold you position deny.

Some, such as Ayer, say that the denial of a law of physics is only physically possible, but not logically possible, as logic and math are. But you say that they can make a geometry impossible--not merely false, but impossible--and so you are implying that the laws of physics are logically necessary. This is great concession!

You may imagine a world with different laws of physics, in which such a geometry would be "possible".

On the contrary, I am less apt to believe that it is possible for most of the laws of physics, or most other scientific truths, to different than you--or at least most believers in the dichotomies--are.

But the word says it all: it is an imaginary world, not the real world. You shouldn't put imaginary worlds on the same footing as the real world. In an imaginary world unicorns, tooth fairies and flying pink elephants may exist, but that doesn't imply that they do exist in the real world.

I never said they did, and I don't consider them even to be possible.

But, unless you agree with Spinoza that everything is necessary, and therefore nothing is possible except what is actual, you must admit that some non-actual things are possible. And those who believe in all the Dichotomies Peikoff reject certainly believe in non-actual possibles.

[Daniel Barnes]

Mike:

>A theory (as knowledge) is made of units of knowledge. It is a composite of many different knowledges.

That goes for the other things you mentioned, too.

>Do you need further explanation? With Popper's theory of knowledge, do composites exist without units?

Obviously all human knowledge is a kind of "composite"; a kind of massive jackdaw's nest of true and false theories, of fiction and non-fiction, of custom and revolution, of art and science etc. Equally obviously, humans use a kind of unit economy to simplify the complex world into abstractions such as language and art, in the process creating a kind of new "world" with a new and intriguing set of problems (for example, mathematical problems, or musical or literary problems) We only have a small lump of flesh inside our skulls with which to take in the universe - we have to simplify. For the same reasons, we also have to imagine, to conjecture.

Just because we use some form of unit economy, however, does not mean that Rand's particular theory of concept formation is correct, or even close to the truth. I personally think there is not much to it. As you know, I can give a straightforward demonstration that her theory of the formation of "units" is merely a violation of the Law of Identity, followed by a couple of confused, verbalist restatements of the LOI that, on examination, add nothing to it. This indicates that there whatever grain of truth her theory might have - and we can find, if we look hard enough, that almost all theories, no matter how mistaken, have some grain of truth to them - this grain is rather small. Further, this empty verbalism of her description of "unit" formation is a quite predictable result, as it is the standard consequence of the Aristotelian method.

As they say in courtroom dramas: Where are you going with this, counsellor? Do you have some specific criticisms of my arguments you want to make? If so, please do. Currently it feels like a protracted fishing expedition...;-)

[Michael Stuart Kelly]

As you know, I can give a straightforward demonstration that her theory of the formation of "units" is merely a violation of the Law of Identity, followed by a couple of confused, verbalist restatements of the LOI that, on examination, add nothing to it.

Daniel,

I must have missed this. Could you please refer me to where you made such a straightforward demonstration? I don't remember it. (If it occurred during a post to another person, this would be a reason. Time constraints do not allow us to read everything by everybody when posts get long.) The fact of you talking about the Law of Identity would have jumped out at me, especially as this is one of the fundaments of knowledge.

Where am I going with this? Merely trying to understand. I am not competing, and I think that is one place where we differ. I don't care if I win an argument or not. If I end up understanding, I win.

I think I am starting to understand, too. It is just a glimmer right now, but I think knowledge for CR (at least in the manner you are presenting it) is a series of virtual entities. Each virtual entity is made up of parts and features, like all entities are, but those parts and features cannot be examined too closely. They are more felt than identified.

For example, a theory is a self-contained virtual entity. That this or that word within the theory actually refers to parts of reality is talked around, not actually dissected and explained. The parts and features of the virtual entity are taken for granted. The closest explanation that I have read so far to explain the parts is "convention." That explains the form, but not the content, so it is still an incomplete explanation.

Michael

[Michael Stuart Kelly]

Something is wrong here.  My [quote name]s and [/quote]s come out even, but still I cannot get the white and blue right.

Greg,

I fixed it. You had a dangling end code ("/quote" in brackets) after the phrase: "that's a bad Objectivist habit." I had to put your initial statement in a code box, too, in order to keep the command from executing (it eliminated text as written).

Michael