Issues with the DIM Hypothesis -- #1 Integration or Identification? Options

[Betsy Speicher]

Dr. Peikoff discussed the essential and foundational ideas underlying his DIM Hypothesis in the first two lectures of his 2004 course. (The lectures are currently available free on ARI's web site.) The main difficulty I have had while listening to these lectures is that I have issues with some of the basic premises of the DIM Hypothesis.

ISSUE #1 - INTEGRATION OR IDENTIFICATION?

Dr. Peikoff's foundational premise is that INTEGRATION is the essence of cognition and the basic function of consciousness. The title of Lecture 2 is "Integration. The One in the Many. Integration as the essence of human cognition from start to finish." In Lecture 2 (Part 1 27:53) he stated, "Integration is the essence of human cognition. It is the basic activity of a conceptual consciousness."

While it is true that integration is a crucial and fundamental part of the cognitive process, it is not the essence of human cognition, nor the basic activity of a conceptual consciousness, nor all of cognition "from start to finish" -- at least, not according to my understanding of cognition and of Ayn Rand.

Ayn Rand wrote that "Consciousness is identification" and "Consciousness is the faculty of awareness -- the faculty of perceiving that which exists," I take that as meaning that IDENTIFICATION of the facts of reality -- rather than integration alone -- is the essence of cognition and basic function of consciousness.

Dr. Peikoff stated that "all the key issues that epistemology raises, every one really involves telling you how to integrate." (Lecture 2, Part 1 27:53) In contrast, Ayn Rand wrote that "Epistemology is a science devoted to the discovery of the proper methods of acquiring and validating knowledge" with "the fundamental concept of method" being logic -- the art of non-contradictory IDENTIFICATION.

When it comes to the nature of knowledge, I see a similar integration vs. identification split. Peikoff states, "The essence of our form of knowledge is the establishing of connections, the establishing of relationships. Or you can put it another way: Human knowledge is the continual quest for the one in the many and that quest is what makes our knowledge a unit."

While acknowledging the importance of integration, Ayn Rand also gives equal weight to the role of differentiation in this passage from ItOE.

Starting from the base of conceptual development -- from the concepts that identify perceptual concretes -- the process of cognition moves in two interacting directions: toward more extensive and more intensive knowledge, toward wider integrations and more precise differentiations. Following the process and in accordance with cognitive evidence, earlier-formed concepts are integrated into wider ones or subdivided into narrower ones.

Often the evidence requires that we differentiate and not integrate our knowledge into a unit, but if all of cognition is integration, it is hard to see where differentiation fits in. If the essential is identification, there is no problem because both both integration and differentiation are really types of identification. Integration is the identification of similarities or causal connections in reality. Differentiation is the identification of differences in reality.

If integration is not the essential function of consciousness, purpose of epistemology, or basis of knowledge, this has a major impact on the validity and the applicability of the DIM Hypothesis.

[PhilO]

The first thought that comes to mind is to ask: is integration fundamentally an assistant to the process of identification? I think of identification in computer terms, by analogy: basically a process of pattern matching, of mapping a complex set of concretes into a single output that represents the identification. (We can for instance take a widely variable percept of the letter 'A' and still identify it as the letter 'A', regardless of thousands of different font variations.)

Integration - as a tentative thought - is a process of bringing seemingly disparate information together to make an identification. e.g., one may observe a great number of concrete occurances in a society, then integrate them together to make the identification that the society has a socialist government, a broad abstraction in political philosophy.

[Paul's Here]

The first thought that comes to mind is to ask: is integration fundamentally an assistant to the process of identification? I think of identification in computer terms, by analogy: basically a process of pattern matching, of mapping a complex set of concretes into a single output that represents the identification. (We can for instance take a widely variable percept of the letter 'A' and still identify it as the letter 'A', regardless of thousands of different font variations.)

There are many fonts in which the idenfication of letters is extremely difficult, especially the variety of script fonts as well as old english fonts. One literally has to be told that "this" is the letter "a". The identification of capital "A" and small "a" as representing the same letter is not something that is done perceptually.

-------------

[Paul's Here]

Dr. Peikoff discussed the essential and foundational ideas underlying his DIM Hypothesis in the first two lectures of his 2004 course. (The lectures are currently available free on ARI's web site.) The main difficulty I have had while listening to these lectures is that I have issues with some of the basic premises of the DIM Hypothesis.

ISSUE #1 - INTEGRATION OR IDENTIFICATION?

Dr. Peikoff's foundational premise is that INTEGRATION is the essence of cognition and the basic function of consciousness. The title of Lecture 2 is "Integration. The One in the Many. Integration as the essence of human cognition from start to finish." In Lecture 2 (Part 1 27:53) he stated, "Integration is the essence of human cognition. It is the basic activity of a conceptual consciousness."

---------------------------
If integration is not the essential function of consciousness, purpose of epistemology, or basis of knowledge, this has a major impact on the validity and the applicability of the DIM Hypothesis.

I listened to several of the lectures and had similar questions about the basis of the DIM Hypothesis. After listening to the development of his ideas and the many qualifications and exemptions that Dr. Peikoff addressed, I found it confusing to try and grasp how to apply his hypothesis. I am going to wait until he comes out with his book so that I can read his final thinking on this subject, rather than grasp many of the less than fully formed formulations, what he really means, and how would I apply the ideas. My biggest difficulty was in integrating DIM with my understanding of Objectivist epistemology, which is an issue that he barely touches on.

One of the biggest issues for me was how the concept of truth applied to the DIM. According to Dr. Peikoff, DIM is not concerned with the truth of an idea, only with how well the idea has been integrated. A person can be integrated and false as well as misintegrated and true in some respects. So what is the point of arguing whether someone is misintegrated or integrated? Is not truth and understanding reality the fundamental focus of cognition? The main focus of DIM seems to be on people not the ideas. It seems to be a psychological theory to me.

The other main issue is that "the DIM theory provides a means of identifying, inductively, the cultural essence of a society, and thus a means of understanding the progression of Western societies from Greece to the present." (http://www.leonardpeikoff.com/) If that is the context, then how does it apply to those people not involved in the essences of society? Should I be using DIM to judge people at work, or my friends, or acquaintances? Does DIM only apply to those philosophers and intellectuals who guide society? Should I focus on the truth of someone's ideas or how well integrated they are? I'm not sure.

[Arnold]

While acknowledging the importance of integration, Ayn Rand also gives equal weight to the role of differentiation in this passage from ItOE.
Often the evidence requires that we differentiate and not integrate our knowledge into a unit, but if all of cognition is integration, it is hard to see where differentiation fits in. If the essential is identification, there is no problem because both both integration and differentiation are really types of identification. Integration is the identification of similarities or causal connections in reality. Differentiation is the identification of differences in reality.

If integration is not the essential function of consciousness, purpose of epistemology, or basis of knowledge, this has a major impact on the validity and the applicability of the DIM Hypothesis.

Interesting. I suspect that he sees these as the two methods of the way one integrates. That in the end, one must integrate the differentiations as well.

[PhilO]

There are many fonts in which the idenfication of letters is extremely difficult, especially the variety of script fonts as well as old english fonts. One literally has to be told that "this" is the letter "a". The identification of capital "A" and small "a" as representing the same letter is not something that is done perceptually.

Well, the very process of associating any perceptual data with the letter "A" *is* a higher level abstractive process. My point is that, mechanistically, it involves taking a wide variety of perceptual data and learning to associate it with a single "output": identification of the letter "A". This is of course a learned process involving memory and other cognitive processes. Identification *is* an abstractive process, it maps the many into the one. Human consciousness is still much better than computers at this task, which is why you see the heavily mangled letters and numbers that you have to identify and enter input a field for human validation on some web forms; handling that level of mangling is still beyond OCR programs.

Identifying an instance of a concept means knowing (implicitly or explicitly) the definition of the concept and being able to identify that instance by omitting irrelevant details while focusing on the essential qualitative characteristics: the very essence of human conceptual ability. However, I think that concept formation and conceptual identification is a subset of our abstractive ability, though the same general mechanism (measurement omission) is involved.

[Ed from OC]

I think of identification in computer terms, by analogy: basically a process of pattern matching, of mapping a complex set of concretes into a single output that represents the identification.

Why think of something so basic as identification in such advanced, complex terms? It would be like thinking of addition in terms of set theory or other advanced math terms.

Identification is a process of relating consciousness to reality. It is awareness as such of reality.

[Paul's Here]

Why think of something so basic as identification in such advanced, complex terms? It would be like thinking of addition in terms of set theory or other advanced math terms.

Identification is a process of relating consciousness to reality. It is awareness as such of reality.

Yes, as Miss Rand put it, "consciousness is identification."

[Paul's Here]

Well, the very process of associating any perceptual data with the letter "A" *is* a higher level abstractive process. My point is that, mechanistically, it involves taking a wide variety of perceptual data and learning to associate it with a single "output": identification of the letter "A". This is of course a learned process involving memory and other cognitive processes. Identification *is* an abstractive process, it maps the many into the one. Human consciousness is still much better than computers at this task, which is why you see the heavily mangled letters and numbers that you have to identify and enter input a field for human validation on some web forms; handling that level of mangling is still beyond OCR programs.

Identifying an instance of a concept means knowing (implicitly or explicitly) the definition of the concept and being able to identify that instance by omitting irrelevant details while focusing on the essential qualitative characteristics: the very essence of human conceptual ability. However, I think that concept formation and conceptual identification is a subset of our abstractive ability, though the same general mechanism (measurement omission) is involved.

I'm a little confused by your second paragraph. Are you saying that the letter "A" is a concept or an instance of a concept? If the latter, what concept do you mean? It seems to me that the concept here is "font" with different A's being instances of "font" and the various A's being concretes. I don't see how "A" can be a concept when it is a perceptual concrete. (Nor would I say that the process is mechanistic.)

[PhilO]

Why think of something so basic as identification in such advanced, complex terms? It would be like thinking of addition in terms of set theory or other advanced math terms.

I suppose because if you focus on the elements that comprise identification, you can see that it isn't simple at all. If integration is indeed fundamentally a mechanism to bring together pieces to make an identification, that just emphasizes the potential complexity of identification at more advanced conceptual levels.

[PhilO]

I'm a little confused by your second paragraph. Are you saying that the letter "A" is a concept or an instance of a concept?

The "output" of our consciously processed perception of a particular instance of "A" is the concept of the letter "A" (the definition being, I suppose: the first symbol in the written English language. You'd probably want to include something about the visual elements comprising "A" (and "a") but it's surprisingly difficult to explicitly do so in a way abstract enough to cover the many font variations. It becomes a mathematical statement, including topology, because the number of times that lines or curve segments touch each other, and closed loops that they form, is an important part of identification of a letter. Of course, we seldom consciously experience that sort of detailed analysis, we long ago automatized the implicit definitions that let us subconsciously and effortlessly identify letters.)

[dondigitalia]

I listened to the first couple of lectures in the course over a month ago, and admittedly didn't give them my full attention, but I didn't understand Peikoff to be discounting differentiation as a fundamental process in cognition. I thought that he was saying differentiation is something that happens automatically, on the perceptual level, and that is the reason why epistemology is concerned mostly with integration.

Upon a superficial introspection, I see that this matches up with what I can observe. When I look out at the world, I see a whole bunch of different things, whether I want to or not. But integrating them requires work and a method.

[Paul's Here]

The "output" of our consciously processed perception of a particular instance of "A" is the concept of the letter "A" (the definition being, I suppose: the first symbol in the written English language. You'd probably want to include something about the visual elements comprising "A" (and "a") but it's surprisingly difficult to explicitly do so in a way abstract enough to cover the many font variations. It becomes a mathematical statement, including topology, because the number of times that lines or curve segments touch each other, and closed loops that they form, is an important part of identification of a letter. Of course, we seldom consciously experience that sort of detailed analysis, we long ago automatized the implicit definitions that let us subconsciously and effortlessly identify letters.)

I'm not sure I can agree with your formulation. Every instance of "A" is a symbol that represents a certain type of sound that people make when talking. As a symbol, what you state as a definition is what I would call a description of the symbol. In a certain sense, I could regard the letter "A" as the word for the concept "aye" that was abstracted from various sounds (other words) that have that particular intonation. That "A" and "A" represent the same letter is because both represent the same sound not similar visual elements. I know that in Chinese writings, similar symbols can have different meanings depending upon the intonation. Notice that there are many sounds that people can make with words (the guttural sounds found in Hebrew or Arabic, or the clicking sounds of some primitive tribes) but have no representation in letters in English because those sounds are not made while talking English.

[Betsy Speicher]

Integration - as a tentative thought - is a process of bringing seemingly disparate information together to make an identification. e.g., one may observe a great number of concrete occurances in a society, then integrate them together to make the identification that the society has a socialist government, a broad abstraction in political philosophy.

I don't quite agree with that characterization of what integration is.

Dr. Peikoff defined "integration" as "an active human process of putting elements together to make a whole." (DIM 2-1 3:24) I have a problem with that as a definition, too, for several reasons.

First, why did Dr. Peikoff say "active human process?" Did he mean "volitional" or "conceptual?" If so, that conflicts with percepts being an automatic integration of sensations or emotions being "the automatic results of man's value judgments integrated by his subconscious." If Dr. Peikoff merely meant there is an action of some kind going on, then this is redundant. All processes are actions and is he really saying integration is "a human process of putting elements together to make a whole?"

Then the meaning of Dr. Peikoff's definition of "integration" totally depends on what he means by a "whole" and he did discuss this further. He cited several OED definitions of "whole" as "something made of parts in mutual combination or connection" and "an assemblage of things united so as to constitute one greater thing" and "a complex unity or system" (DIM 2-1 9:55). These are even more abstract than "whole" and further clarification is necessary.

To concretize, Dr. Peikoff gave, as examples, an airplane made of many different parts (DIM 2-1 4:00) and a mixed drink containing the proper proportions of rum, lemon juice, tobasco, and beef bouillon that wouldn't be a "whole" if you put in one drop too much tobasco (DIM 2-1 8:50). This is even more confusing because I can't related the airplane or the mixed drink to the kind of integration that goes on when we integrate percepts or previous concepts into a new concept.

For instance, in PhilO's example, when he puts his "seemingly disparate" pieces of information together -- the progessive income tax, socialized medicine, antitrust laws, environmental regulations -- he sees that they are not disparate at all. He sees what they all have in common -- government control of the economy -- which is what he abstracts and integrates into the concept "socialist government."

But when it comes to putting an airplane together, what do a landing gear, engine, and cockpit window have in common? What does rum have in common with just so much, but not too much, tobasco? Integrating mental entities into a concept is integration according to similarities. It is a very different process than building an airplane or mixing a drink.

Dr. Peikoff's concept of "integration," as defined and concretized in DIM Lecture 2, is not just difficult for me to understand. It is difficult to reconcile with the use of the term in Objectivist epistemology when it comes to perceptual integration, emotional integration, and conceptual integration.

[Betsy Speicher]

After listening to the development of his ideas and the many qualifications and exemptions that Dr. Peikoff addressed, I found it confusing to try and grasp how to apply his hypothesis.

I wasn't able to get that far. I'm still stuck on trying to understand what his hypothesis is and what it means.

What really concerns me is that some people are taking the DIM Hypothesis and applying it by condemning other Objectivists as "disintegrated" or as "rationalists" who don't understand Objectivism. Considering that DIM is not yet fully defined and developed and, being an hypothesis, is not yet fully proven, "applying" it like that is appalling.

I am going to wait until he comes out with his book so that I can read his final thinking on this subject, rather than grasp many of the less than fully formed formulations, what he really means, and how would I apply the ideas. My biggest difficulty was in integrating DIM with my understanding of Objectivist epistemology, which is an issue that he barely touches on.

Considering that the subtitle of the DIM course is "The Epistemological Mechanics by which Philosophy Shapes Society" you would think that there should be a clear connection to Objectivist epistemology, but right now I see more conflict than connection.

I understand that the DIM Hypothesis is still being developed and I hope that the issues I have will be clarified and the contradictions will be resolved, but they are on such a fundamental level that I wonder how it can be done.

One of the biggest issues for me was how the concept of truth applied to the DIM. According to Dr. Peikoff, DIM is not concerned with the truth of an idea, only with how well the idea has been integrated. A person can be integrated and false as well as misintegrated and true in some respects. So what is the point of arguing whether someone is misintegrated or integrated? Is not truth and understanding reality the fundamental focus of cognition?

It is for me. That is something I will be writing more about when it comes to my Issue #2 - Dichotomy or Trichotomy?

[Betsy Speicher]

I listened to the first couple of lectures in the course over a month ago, and admittedly didn't give them my full attention, but I didn't understand Peikoff to be discounting differentiation as a fundamental process in cognition.

I didn't either. He just said he would not be dealing with it in the DIM lectures.

Yet I see serious discounting of differentiation in saying that integration is "the essence of human cognition from start to finish" and that "Integration is the essence of human cognition. It is the basic activity of a conceptual consciousness."

I thought that he was saying differentiation is something that happens automatically, on the perceptual level, and that is the reason why epistemology is concerned mostly with integration.

I don't think that is true as far as conceptual differentiation -- the kind of differentiation Ayn Rand was discussing in the section I quoted. Also observe that much of integration -- perceptual and emotional integration -- is automatic.

Upon a superficial introspection, I see that this matches up with what I can observe. When I look out at the world, I see a whole bunch of different things, whether I want to or not. But integrating them requires work and a method.

Is that really what is happening? While it is true that our senses are essentially "difference detectors" observe that they are constantly bombarding us with the evidence of thousands, and perhaps millions, of differences every second. Those differences don't become differentiations until you selectively and volitionally focus on them and consciously identify them as differences.

[dondigitalia]

I don't think that is true as far as conceptual differentiation -- the kind of differentiation Ayn Rand was discussing in the section I quoted. Also observe that much of integration -- perceptual and emotional integration -- is automatic....

Could you give examples of what you would consider to be conceptual differentiations and perceptual integrations?

[dondigitalia]

As soon as I posted that, I think I grasped "conceptual differentation." Is an example of that (straight out of ITOE) the concept of "coffee table"?

I am still at a loss for an example of a perceptual integration? If what you mean is emotional responses, I can see where you're coming from, but in those cases the integrations, although not under one's immediate control, are not a primary, and rely on prior volitional integrations on the conceptual level.

[Betsy Speicher]

Could you give examples of what you would consider to be ... perceptual integrations?

1) The integration of sensations into percepts -- into the awareness of entities which is the beginning of cognition in man and in the higher animals.

2) Perceptual association and memory. We automatically remember aspects of reality perceived simultaneously without conscious effort. (In fact, try not to remember that jingle from the radio commercial if you hear the first few notes or words.) Perceptual association and memory is how we train animals to respond to our commands.

3) Automatized behaviors (like driving a car, riding a bike, tying a shoe lace).

[Paul's Here]

After listening to the development of his ideas and the many qualifications and exemptions that Dr. Peikoff addressed, I found it confusing to try and grasp how to apply his hypothesis.

I wasn't able to get that far. I'm still stuck on trying to understand what his hypothesis is and what it means.

From my memory, I believe he said Hobbes could be classified in two of the categories (I don't remember which ones). He said the he wouldn't include Rush Limbaugh in any category because he is so completely unintegrated, espousing religion and freedom, etc.

What really concerns me is that some people are taking the DIM Hypothesis and applying it by condemning other Objectivists as "disintegrated" or as "rationalists" who don't understand Objectivism. Considering that DIM is not yet fully defined and developed and, being an hypothesis, is not yet fully proven, "applying" it like that is appalling.

I agree 110% with that. As unclear as the DIM is to me as presented by Dr. Peikoff, I find it amazing to see people actually argue "He's 75% D and 25% I" or some such classification derived from I don't know what. I am going to continue to evaluate people based on the truth of their ideas, the values they hold and the virtues they practice. I'm not sure how I would actually act toward someone who could be classified as integrated, misintegrated or disintegrated, since truth is not the issue here.

Considering that the subtitle of the DIM course is "The Epistemological Mechanics by which Philosophy Shapes Society" you would think that there should be a clear connection to Objectivist epistemology, but right now I see more conflict than connection.

Which brings up a point that should be addressed: I don't think it is valid to use "mechanics" in connection with "epistemology" unless one means a loose association (or analogy) and not a technical description. Consciousness is not mechanical in nature.

------------