Tuesday, June 07, 2011

A "Necessary Connection"?

At bottom, Objectivist epistemology is based on a fundamental fallacy which leads at first to verbalism and then to an authoritarian turn. This fallacy, often difficult to unpack, is clearly expressed here (my italics):
"And, lastly, I suggest that you try to project what would have happened if, instead of Annie Sullivan, a sadist had taken charge of Helen Keller’s education. A sadist would spell “water” into Helen’s palm, while making her touch water, stones, flowers and dogs interchangeably; he would teach her that water is called “water” today, but “milk” tomorrow; he would endeavor to convey to her that there is no necessary connection between names and things, that the signals in her palm are a game of arbitrary conventions and that she’d better obey him without trying to understand."
- Ayn Rand, “Kant Versus Sullivan,” Philosophy: Who Needs It p90.
A truly remarkable thing to write, and to think.


Lloyd Flack said...

Yes, Rand confused symbol and concept. But further she seems to make the claim that consciousness deals with sensory data. It does not. It deals with representations of the world that the brain creates on the basis of sensory data.

An Example. You see a blue bowl. The bowl reflects light of certain wavelengths, what we describe as blue light. But the sensation of "blue" exists in your mind. It is not a property of the bowl. It is not a property of the light. It is not a property of the sensory receptors and the nerves that carry their signals. It is a property of the internal representation of the bowl that the optical sections of your brain create. If someone is colour-blind or has a shifted visual spectrum then their perceptions of the bowl will be different. Neither you nor they will be right or wrong. You will just perceive things differently.

Rand seemed to overlook this additional stage between consciousness and the senses.

Lloyd Flack said...

And in this lies the reason for the counter intuitive nature of modern physics. Our brains internal representations of the world have evolved to allow us to deal with the velocities and scales that our ancestors needed to handle. They did not evolve to deal with the speeds at which relativistic effects become important. And they did not need to handle the very small scale at which quantum effects become important.

As a result for these scales and velocities our brains' representations of reality do not match reality. We can't visualize what is happening very well. This is difficult enough for everyone. But I think for some Objectivists it is especially difficult.

Neil Parille said...


Perhaps you are giving too much weight to the "necessary connection" portion of this. I don't think Rand is saying there is a necessary connection between H2O and "water" such that H2O couldn't be called "agua."

I think she could have (and should have) left the italicized phrase out and the meaning would stay the same.

-Neil Parille

Rey said...

I guess it hinges on what one means by "necessary." Consistent connections between between names and the things they represent are necessary for communication, but there's no intrinsic reason we couldn't call hot "cold" or stone "wood" so long as we all agree to it.

Ken said...

The words/sounds we use to represent different concepts are arbitrary, or we wouldn't have different languages. Within any language there are standard definitions of concepts for words, or standard words to represent concepts, depending on which way around you want to view it.

However, I wouldn't go so far as to say that there is a single unambiguous meaning for any given word, even within a language. My notion of what a word means can and will differ from those used by other people. Consider for example "father", which has quite a different set of connotations for me than it would to the child of an abuser.

These differences don't preclude communication, but they mean it's not unambiguous. I've worked in software specification writing, and even within such a restricted domain with carefully-defined jargon you still end up with ambiguities. To imagine that a piece of literature could have only a single meaning, or that all readers will or should react in exactly the same way to it, is ridiculous.

Rey said...

First of all, this excerpt Daniel quoted sounds like Rand's typical banality wrapped in an aura of profundity. Basically, she's saying, "Words mean things and we should use in consistent manner for the sake of clarity."

However, Ken brought up connotation and denotation
and I vaguely remember Rand (or one of her followers) railing against the very idea of connotation, perhaps in ITEO, but I'm not sure. Anyone know for sure how she addressed the connotation-denotation issue?

I went to the Ayn Rand Lexicon online and didn't see anything, but I did come across this in her definition for "Definition":

It is often said that definitions state the meaning of words. This is true, but it is not exact. A word is merely a visual-auditory symbol used to represent a concept; a word has no meaning other than that of the concept it symbolizes, and the meaning of a concept consists of its units. It is not words, but concepts that man defines—by specifying their referents.(http://aynrandlexicon.com/lexicon/definitions.html)

Anyone care to parse this for me? She seems to concede that a words are symbols...but they are symbols for "concepts." She almost sounds like a Neo-Platonist.

If I say, "I ate an apple," I'm talking about a particular apple that I consumed, not the concept "apple." (One might argue that I specified such when I said "an," but there were and are languages that don't have articles (Latin and Japanese (Japanese doesn't even have plural forms of nouns), for instance) and rely on context (and an attentive reader/listener) or specific numbers to communicate the meaning.

Even if I make a general statement "I eat apples," everyone knows I'm talking about the fruit of the apple tree...not the idea of the fruit of the idea of an apple tree.

Am I reading too much or too little into this?

Jeffrey said...
This comment has been removed by the author.
Jeffrey said...

Couple things

1)Not that it matters, but it's page 127 in my paperback version

2)Ken is right in that a 1:1 relationship isn't the right way to understand language, but I don't think it's fair to call it a "fundamental fallacy which leads at first to verbalism and then to an authoritarian turn". Put in the broader context (hah, just proves my point that Daniel Barnes is an unscrupulous context-dropping un-Objectivist!), Rand is trying to prove how higher concepts must be reduced somehow to the sensory level, which I think she does very well in that essay.

3) Mr. Flack: I don't know if Rand ever got around to defending sense perception, but Peikoff actually does a pretty good job of it in chapter 2 of OPAR; he states that we see reality by means of out senses, and it would be invalid to call this an indirect means of perceiving it, because direct perception would be perception without any means of perception (p. 51). He also claims that "differences in sensory form do not matter": the color-blind person and the person with normal sight see the same reality, they just do it in different forms (p.42).

Sorry guys, but Peikoff and Rand win this battle! Mozart may not be a red, but that doesn't mean EVRYTHING Rand said was wrong.

Daniel Barnes said...

@Neil, the currents in Rand's thought often run in contrary directions, but as you step back I think you can see what are the eddies and what is the main flow of the stream. I think there's a real lacuna here on the part of Rand and her followers; an un- or semi-conscious underlying assumption that words can have "true" or "false" meanings, which leads to all kinds of problems for her epistemology and her philosophy in general. Sure, in the ITOE she briefly asserts that words are mere "labels" - she has a what I'd describe as a spasm of nominalism - but this runs counter to just about everything else she says and, more importantly, does. But I'll write a bit more about this tonight.

@Jeffrey, have you read the ARCHNblog required text on essentialist definitions vs nominalist definitions?...;-) It's here. It's a good start to understanding the problem, as even though Rand didn't subscribe to Aristotle's "essences" as such (ie metaphysical essences), she subscribed to his methodological approach; and thus inherits some of his important errors. As to the "authoritarian turn", which I see as a consequence of the aforegoing, I'll talk a little bit more about that later.

Michael Prescott said...

"He also claims that 'differences in sensory form do not matter': the color-blind person and the person with normal sight see the same reality, they just do it in different forms (p.42)."

Doesn't that beg the question? The issue is whether the senses provide an accurate picture of reality, or whether "true" reality differs from the appearance provided by sense perceptions. Peikoff takes an instance where two people's senses provide different impressions, and merely asserts that the differences don't matter because the underlying reality is the same in both cases. But that's the whole point at issue!

I'm not saying the senses aren't generally reliable, at least for practical purposes, but this particular argument seems to be fallacious.

gregnyquist said...

but I don't think it's fair to call it a "fundamental fallacy which leads at first to verbalism and then to an authoritarian turn"

Even if it's not entirely fair, it's not entirely wrong. Perhaps claiming it's a fallacy leading to verbalism may be stretching things a bit, but there's definitely a problem here, which is manifested in Rand's belief that there are "true" definitions and that at least some words (e.g., selfishness, altruism) mean only what Rand meant by them, regardless of how others use them. The only reason why it would be an exaggeration to say that this sort of attitude toward words and definitions doesn't lead to verbalism is because it is verbalism. Once you begin arguing about the meanings words (rather than the meanings intended by the people using them) you have lapsed into full-throated verbalism; and nothing more strongly manifests the symptoms of verbalism than a belief that words have true or "proper" definitions.

Lloyd Flack said...

My point was that when you perceive a blue bowl the "blue" that your consciousness perceives is no more a property of the bowl than is the label "bowl". The "blue" is a construction of your mind. Your consciousness examines the brains internal representation of external reality, not that external reality itself. Most of the time what you find by examining that representation is something that is there in the external reality. Occasionally it is not. Just study some optical illusions.

Neither the internal representation of a normal or a colour blind person is correct. However that of a person with normal vision is more useful and more informative.

Daniel Barnes said...

Ok, expanding on my earlier comment, let's try to see how this fallacy of "true" meanings of words gets embedded in Rand's thinking.

So on p40 of the ITOE, Rand says "A word is merely a visual-auditory symbol used to represent a concept...It is not words, but concepts that man defines...".

Then later, in the ensuing discussion p175, she says "But a concept is only a mental unit, a symbol, for a number of concretes of a certain kind."

So a word is in effect a symbol of a symbol. Faced with this increasing complexity, one of Rand's interlocutors Prof. D shoots for a simplification:

"Granted, then, that concepts denote objects in reality and that the concept is a mental unit, I wonder whether it isn't in some kind of indirect sense that words denote objects - indirectly, via the concept - with the direct meaning of the word being the concept, the integration."

Rand replies: "I don't think we can make that distinction. A word which is not a proper name does refer directly to an indefinite number of concrete objects. A concept, in the form of a word, refers to them directly, not indirectly."

Notice how, lo, the word and the concept have now taken the same form.

Now just to drive it home, Prof D asks:
"The word refers to the objects directly so the objects directly constitute the meaning of the word?"

Rand:"That's right."

So now the word is no longer a mere "label" or symbol for the concept, but directly integrated with it. And this is of course consistent with Rand's earlier claim (p19) that "The process of forming a concept is not complete until its constituent units have been integrated into a single mental unit by means of a word." Note how somehow the visual-auditory word is now vitally involved in completing the mental concept's formation. So in fact the distinction between a concept and a word has been thoroughly blurred, with Rand's express approval. Now that we have the two firmly fused in Rand's mind at least, in my next comment I'll discuss how upon this rock she builds her church.

Lloyd Flack said...

A area where I think Rand's epistemology is weak is in handling gestalts. She focuses on the trees and often does not see the forest. Or if she see the forest she does not realize that it is a gestalt of trees. She often sees just a collection where there is a gestalt.

Where this is of most importance is in how she deals with the concept of society and with social concepts. But I think she does not recognize things as systemic properties in other areas. For example life is a property of a cell as a whole system. It is not a property of any part of the cell. While this is not an example she dealt with it is something that I do not think fits in well with her treatment of concepts.

If she can recognize life as property of a whole system which is found in none of its parts why can she not see a society as having properties not found in individuals. If she wants to attack ideas treating a society as some sort of larger scale person then I would say that she is probably right. If she wants to say a society is nothing but a collection of individuals then I would say she was probably wrong.

Xtra Laj said...


Lucid summary of the main problem with refusing to accept that groups can have collective properties not reducible to the property of each individual. It's also a hot topic in evolutionary biology, so I wouldn't blame Objectivism exclusively for the oversight.

One of the moments when I started having doubts about libertarianism was when I reviewed (with less allegiance to the "capitalism is the cure for all social ills" solution) the analysis of "the tragedy of the commons" and "externalities" in libertarian economics. It seemed to me that the values underlying the libertarian treatment of such issues were not the values shared by most people in the world today.

Ken said...

The comments raise a couple of issues for me. Some of them can be illustrated by borrowings; consider for example deja vu, or Schadenfreude, or even philtrum.

The referents for all of these are probably familiar to most people, but no English speakers ever bother inventing a term for them. Under Rand's understanding of the relationship between word and concept, does the fact that English had to borrow these mean that no English speaker had ever developed these concepts, despite familiarity with their referents? That is, does the lack of a word mean the lack of the concept?

Continuing that chain of thought, English may not have had a word for these, but it is certainly capable of describing them more verbosely - that is after all what a dictionary definition is. So, what does Rand mean by "word"? English didn't have schadenfreude, but people could still say "that lovely nasty feeling I get when other people are unhappy." Does that suffice to show that the concept has been "formed" in the speaker, even lacking a word? If so, is there some limit on the allowed length of the definition?

In that light consider "quantum mechanics." That is probably a "word" in Rand's sense, and there is a concept behind it - but that concept can only really be understood after a "definition" that encompasses several years of study. Does the concept formation lie in the lectures, or in the mere coining of the word?

Finally there's the quote of Rand provided by Daniel Barnes. That seems to indicate that Rand thought concepts referred to objects, and even "concrete" objects. Now to my understanding, deja vu is not a object, concrete or otherwise. It is in fact a thing that I can only perceive - though not with any sensory organ that English seems to have named - within my own mind. The same is true of other mental states and emotions, and for that matter a lot of quantum mechanics. Are these "concrete objects" in Rand's sense of the term?

(By the way, there is interesting research on deja vu and its neurological causes which could be the basis of a whole discussion of Rand's model of thought and mind. If the research is correct, it is yet another of those cases where your brain has already done something without conscious thought - and has, in fact, made a mistake, in some ways like an optical illusion - and your mind is trying to provide an after-the-fact rationalization of the result.)

Ken said...

On re-reading my last post, I see that there is some ambiguity in "The same is true of other mental states and emotions, and for that matter a lot of quantum mechanics." Because of the previous sentence, it can be read as saying that quantum mechanics (like emotions) is something that can only be perceived within your mind. I did not mean that; I only meant to include QM as one of those concepts which does not refer to concrete objects, at least by my understanding of concrete.

gregnyquist said...

My point was that when you perceive a blue bowl the "blue" that your consciousness perceives is no more a property of the bowl than is the label "bowl". The "blue" is a construction of your mind. Your consciousness examines the brains internal representation of external reality, not that external reality itself.

This is exactly right; there is no "direct" perception, and the Objectivist solution to this issue seems based on little more than equivocating on vague phrases, such as "form of perception."

While Rand recognized that words are symbols and that concepts are not identical to what they represent, she failed to fully realize that percepts are also merely symbolic, signs of outward things rather than copies of their "form." Rand was too obssessed with "identity," which she often seems to hold as a kind of the ideal in cognition. In her "validation" of concepts, she appears to be trying to show in what way concepts are identical to what they represent, because, presumably, it's by demonstrating this thread of identity that she seeks to "validate" the connection between concepts and reality. But this approach merely reinforces the strong tendencies toward verbalism and rationalism that already existed in Rand from the start: for it distracts her (and her followers) from the important questions involving practical and experimental consequences of theories, rather than insignficant issues dealing with the "validation" of concepts and their definitions.

Of course, if you hold theories that don't accord with fact, nothing can be more to the purpose than adopting cognitive approaches that steer one away from pratical and experimental consequences of such theories.

Ken said...

Speaking of the interpretation of sensory data, here are a few more data points: http://illusioncontest.neuralcorrelate.com/cat/top-10-finalists/2011/

I was alerted to this through the Bad Astronomy blog. He also referenced one of my favorite color illusions, which seems relevant to the question of whether a bowl is blue or not: http://blogs.discovermagazine.com/badastronomy/2009/06/24/the-blue-and-the-green/

Mr. A said...

Peikoff also handles the issue of illusions pretty well, see page 40; the senses only tell us that something in, what it is we have to figure out on our own.

But what Peikoff doesn't cover is the bigger issue of hallucinations: seeing and hearing something that isn't there at all. Seems to be an extremely important oversight if you ask me; how can a schizophrenic be sure that his senses are real? How do we know we aren't hallucinating? I don't think these are unsolveable issues, but you'd thing they'd deserve mention.

Reading Rand and Peikoff is like prospecting: there's gold nuggets in there somewhere, but you have to sort through a lot of silt to get there.

gregnyquist said...

Reading Rand and Peikoff is like prospecting: there's gold nuggets in there somewhere, but you have to sort through a lot of silt to get there.

I've been sifting through Rand and Peikoff for quite some time and have little luck with prospecting for gold nuggets. Occasionally, one comes across something glittering amongst the reams of silt that looks like it could be gold. But closer inspection reveals that it is mere fool's gold. Indeed, except perhaps Rand's "unit economy" (hardly an original insight on her part), I can't think of anything in either Rand or Peikoff that rises to the gold standard. Merely a lot of mediocre to bad reasons for believing in a common sense view of reality, an egoistic ethics, and laissez-faire capitalism.

Alex said...


Anyone care to parse this for me? She seems to concede that a words are symbols...but they are symbols for "concepts." She almost sounds like a Neo-Platonist

Ayn Rand was a Neo-Platonist rationalist masquerading herself as an Aristotalian empiricist. I don't even think she knew that, because she never bothered to read philosophy deeply enough.

Dragonfly said...

Greg: In her "validation" of concepts, she appears to be trying to show in what way concepts are identical to what they represent, because, presumably, it's by demonstrating this thread of identity that she seeks to "validate" the connection between concepts and reality.

Her thinking about concepts is terribly muddled. On the one hand she writes: "Concepts represent condensations of knowledge, which make further study and the division of cognitive labor possible", i.e. concepts are human constructs representing their knowledge about something. But then she writes "It is crucially important to grasp the fact that a concept is an 'open-end' classification which includes the yet-to-be-discovered characteristics of a given group of existents", or as Peikoff (with Rand's blessing) writes: "The fact that certain characteristics are, at a given time, unknown to man, does not indicate that these characteristics are excluded from the entity—or from the concept", trying thus to reason away the essential difference between analytical and synthetic statements.

So one moment she states that concepts represent human knowledge, but the next moment we hear that also that what is not yet known, thus non-knowledge also belongs to that same concept. Suppose now tomorrow some asteroid hits the earth and destroys all life on it. Then there won't be any new knowledge about those concepts, or should we apply counterfactual reasoning, implying that the discoveries that we could have made if that pesky asteroid hadn't interfered also belong to those concepts? However, that would in fact mean that Rand equates concepts not with human knowledge, but with reality itself, with "das Ding an sich".

She is so confused that she doesn't see that she's equivocating here between the concept of "human concepts" and "reality". Of course it would be much more useful to view concepts as actual human knowledge. In that view concepts can grow, be deepened and extended with increasing knowledge in the course of time, but also "die" if they turn out not to be useful representants of reality (like for example phlogiston, élan vital or polywater).

Anonymous said...

So if Hickman had sown epistemological confusion in the psyche of his little girl victim, instead of kidnapping, murdering, and mutilating her, why then Rand would have seen occasion to view him with unqualified opprobrium.

Daniel Barnes said...

@Alex, we agree.

Following from my earlier comments, we now see how Rand treats concepts and words as basically one.

This is hardly surprising. Conceptualism, conceptual analysis etc usually leads to verbalism for the simple reason that as humans don't have the ability to mind-meld, we have to mediate any such analysis or comparison of such purely mental abstractions through a medium such as language.

Now we look at Rand's basic epistemological assumptions, and we find that they are very traditional. That is, she assumes the old fallacy of the secure starting point. That is, if one is to begin the journey to discover truth, one must start from an entirely secure, indubitably true starting point (Which at first sounds plausible but is clearly a logical error, as it would be an infinite regress. Which is why, quelle horreur, we actually start in error and try to get to the truth). Further, due to her claim that the conceptual process is supposedly mathematical - "the algebra of cognition" as she calls it - she can't allow any false (or "invalid") concepts into her conceptual "equation", otherwise the resulting conceptual "sum" will be erroneous. So establishing the truth or falsity of concepts becomes critical to her system. And as she has already established that concepts are in effect words, the issue then becomes establishing the truth or falsity of words, which in turn inexorably must become establishing the truth or falsity of definitions.

But doh! We now know courtesy of Popper (in the latter part of our ARCHNblog required reading) that in any dispute over meaning, there is no way of logically establishing which is "true" or which is "false" other than convention, which Rand has rejected firmly. So to preserve her system, this is where she is forced to take an authoritarian turn, which I'lll expand upon in my next comment.

Daniel Barnes said...

So now to recap, we have seen that Rand considered words and concepts basically welded together as one. We also see, from the main post, that Rand implicitly seemed to assume a kind of "necessary connection" between words and things - only a "sadist" could think otherwise!

Such appears to be the way she tries to "ground" her conceptual chains in reality - by assuming (in a kind of semi-conscious doublethink) such a connection exists of necessity, rather than by convention. This would also explain her repeated, heated rejections of the idea that words are conventions - this would sever her conceptual chain at its first link.

But the problem is that words are conventions. They do not spring forth of necessity from objects or actions, as if the universe had tagged its Facebook photo album. Further, contra Rand's claims in the ITOE, if someone disagreed with her on the meaning of a word, logic could not resolve it as Popper pointed out.

How then to anchor the chain? At this point, the ultimate authority is introduced: philosophy. On p74 of the ITOE we discover that it is philosophy who, among other things, "is to keep order in the organization of man's conceptual vocabulary, suggest the changes or expansions of definitions..." She gives us some actual examples of this in action later on p290, saying how, in studying the mind/brain problem, "Philosophy would have to define the terms of that question. In asking what's the relationship between "mind" and "brain", scientists have to know what they mean by the two concepts. It's philosophy that would have to tell them the [general] definitions of those concepts. But then to find the specific relationship, that's a scientific question."

So philosophy's job, apart from reminding us now and again that A=A, is to tell everyone else what words - even ordinary words like "mind" and "brain" - mean.

Back on p74, Rand commits a strange slip. "Who, then, is to keep order [...etc]? - the answer is philosophy. But "philosophy" is a what, not a who. Clearly she has in mind not just a suitably authoritative philosophy to maintain the proper order and meaning of language, but such a philosopher.