Jump to content
Objectivism Online Forum

Announcing Study-group on Inductive Method of Thinking, Its Foundation

Rate this topic


Recommended Posts

Just posted related text in SGO, from the works of Burgess. Posting one part

 

Induction in POWER AND GLORY

Q. How Burgess inductively presented concept faith?
Ans:
The units subsumed by concept faith just prior to Jesus were

1. Metaphysical unit : A belief that God exists.
2. Epistemological unit : Assent to what God says as truth.
3. Ethical unit : Personal commitment to follow God's instructions for how to act.
4. Psychological unit : This element was taken from Judaism, while previous 3 were of greek origin. It was steadfast trust. That is trust in god that never changed.

Symbols were "pistis" in greek and "emunah" in Hebrew.

The "Conceptual Common Denominator" for the above concept is "fundamental ideas". The foil for Burgess being reason, another fundamental idea. Jesus however contrasted faith from doubt, considering steadfast as the primary unit of faith. Actually however, metaphysical unit is the starting point for other units of faith.

 

How Burgess defined reason?

Here is the definition from page 3 of POWER AND GLORY.

Reason is the ability of the mind to
1. Integrate ideas - concepts, principles, and theories - from a multitude of discrete sense-perceptions of reality.
2. Use various techniques of double-checking those ideas to make sure the mind has formed them logically.
3. Apply those ideas to understanding and solving problems of life, whether they are narrow technical problems or universal ethical issues.

He then gives application of reason, where person uses it to understand his sickness by integrating symptoms to medical and scientific knowledge.

Edited by RohinGupta
Link to comment
Share on other sites

One week from now main phase of the study group will start.

 

So today I requested participants to take some time to define their purpose of joining the study-group. Identifying this I think can lead to better focus while formulating the content, and also help others to collaborate better with the person.

Link to comment
Share on other sites

Here are some of my motivations in joining this study group

 

1. The need for a more systematic thinking method arose when I was writing the following set of posts on "Software Engineering".

http://forums.4aynrandfans.com/index.php?showtopic=12227#entry107444

 

2. As a cultural activist, I think lack of "abstract thinking connected to reality" is a major reason for many of the problems we see today. I plan to promote a culture where individuals think in terms of Principles derived from facts, and act purposefully after thinking through.

I realized the gravity of situation after reading about the following activist, who contributed to the devaluing of Epistemology in Western culture.

http://reasonversusmysticism.blogspot.in/2012/04/richard-rorty-postmodernist-mystic.html

Link to comment
Share on other sites

Today main phase of the study-group has started.

 

I posted following questions for - Introductory content "prior" to section on "The Nature of concepts"

 

Q1. Elaborate on the statement - "As our knowledge of the physical world has advanced, our understanding of knowledge itself has lagged behind."?

 

Q2. Why does this situation[described on Q1] persists in universities around the world? How did we arrive at this bizzare contradiction - with scientists developing technology that exploits our detailed knowledge of atomic structure, while philosophers revel in the alleged impotence of reason to grasp even relatively simple facts?

 

Q3. What is Induction? How can it be contrasted from Deduction?

 

Q4. What are "particular instances" and generalizations that induction deals with?

 

Q5. How is Induction related to the Nature of man?

 

Q6. Define generalization ? What are the implications of their correctness and falsehood?

 

Q7. What is meant by the statement - "The problem is to identify the method of induction, not to seek its justification".

 

Q8. What are the problems that one may encounter if formulating of generalizations is not started from particulars?

 

Q9. What is simple enumeration? Why can't induction be just simple enumeration?

 

Q10. If not simple enumeration, what approach can be used in induction to provide certainity?

 

Q11. What are the prerequisites for the "theory of induction"?

Link to comment
Share on other sites

Today I answered questions from Introductory content "prior" to section on "The Nature of concepts"

 

Sharing sample answers for wider audience

 

Q3. What is Induction? How can it be contrasted from Deduction?

 

Ans: From TLL
Induction is the process of inferring generalizations from particular instances. The complementary process of inferring[existing] generalizations to new instances is deduction. The theory of deductive reasoning was developed by Aristotle more than two millenia ago. This crucial achievement was a start toward understanding and validating knowledge, but it was only a start. Deduction presuposes induction; one cannot apply what one does not know or cannot conceive. The primary process of gaining knowledge that goes beyond perceptual data is induction. Generalization - the inference from some members of a class to all - is the essence of Human cognition.

 

MY FURTHER ANALYSIS

 

While the TLL text in answer covers the scope of question, there are two things I would like to go deeper into.

 

1. If induction is going from particulars to universals, does this include concept formation or not. While the text, taken in isolation, may indicate that concept formation is part of induction. But if we study rest of the book, we can conclude that induction is separate from concept formation. Concept formation involves organizing various entities, attributes, actions, lower level concepts, and as we saw in the example of furniture that higher level concept can also subsume lower level generalizations etc. by grouping their similar types. Wheras induction involves seeking relationship between various characteristics of entities.[1]

For e.g. the law of gravitation gives an inductive inference that (initially) describes relationship of "Force between different entities like Planets, sun" and "distance between these entities". Distance between planets and sun was calculated by Kepler or Copernicus(not sure if it was somebody else also), and it was the extension of concept length applied to very long distances. And concept Force, as described by Newton, involves change of state of rest or uniform motion of the entity.
What we are interested here is to get the difference between inductive generalization and concept. So concepts of "Planetary Force" and "Planetary distances" will be formed prior to the inductive generalization that establishes their relationship. Concept of "distances between planet and sun"(Planetary distance) will include characteristics like "locus of distances between planets and sun form elliptical orbits, Kepler's law that line connecting sun and Planets sweep equal areas in equal time periods, or Newton's observation that any body moving with constant speed will sweep equal area etc.". Or concept of Force will have essential characteristic that it is directly proportional to the change in velocity before and after applying it.
After the inductive generalization described by "law of gravitation" is established, the fact that "force on planet" is inversely propotional to the "square of distance with sun" will be added to the characteristic of concept "Gravitational Force". And the fact that "gravitational force on planets" depends on it will be added to characteristics of "Planetary distance with sun".

In short, while concept formation and induction are different, they are also closely related by being complementary. Concept formation is a pre-condition to inductive generalization, as we also saw in the prep section of OPAR. And as we saw here, once inductive generalization is reached, it can increase the characteristics of corresponding concepts involved.

One more thing that needs elaboration is that once we have started induction, it does not mean that concept formation has stopped. Some concepts require and even subsume lower level generalizations. For e.g. we already saw example furniture. Further, law of inertia, a generalization, which states that body continues to be in state of rest or uniform motion unless compelled by external agency to change, is precondition for concept "Gravitational Force".

[1] While in our discussions we will use induction for referring to relationships between entity and action, or between various attributes in entity. There is another valid, more general reference to induction. This reference includes concept formation along with forming and validating generalizations. This is what Aristotle refers to when he says induction is the process of forming universals. Or what Ayn Rand refers to at the end of chapter "Abstraction of Abstractions" in ITOE. Here the context differentiates induction from deduction. In later discussions however, induction is contrasted from concept formation.
This digression on induction also leads to another property of concept. That is same word can serve as a symbol for different concepts. The similarity can range from just linguistic as in "axes - the plural of ax or axe/the plural of axis". Or metaphorical, where there is some overlapping between two concepts the word refers to. Mostly between some characteristics of respective concepts. As in "key to the door" and "key to the solution". Common characteristic here being "small but important thing to gain entry". Similarity of symbol "induction" here is metaphorical. In fact, our usage of induction is the subset of Aristotle's usage of induction.

 

2. Another thing that needs to be elaborated is that induction also involves moving from universals to particulars. While it is true that starting point of induction are particulars, and end point is a generalization, it does not mean that induction is unidirectional or even unidimensional[explained later].

By not being unidirectional I mean that we also have to move from generalizations to particulars during induction. The process is called reduction. Once we reach a final or intermediate generalization, we need to validate if the chain of steps used to reach the state is valid. That is at critical milestones we retrace steps where we a) Recall what percepts, first level generalizations, concepts, and lower level generalizations we have integrated. That is what we had to know to reach the current state. b.)  Check that there are no contradictions among terms and sentences identified in (a.). c) Integrate new percepts, first level generalizations, etc. if the connection of generalization to facts is found lacking. For e.g. Galileo did inclined plane experiments after pendulum experiments to confirm the hypothesis that horizontal motion is unacclerated.

By not being unidimensional I mean that we don't consider just percepts, concepts and first level generalizations that will eventually be subsumed by the generalization. We consider related percepts, concepts and generalizations also. For e.g. in lectures on "Objectivism through Induction", Dr. Peikoff while demonstrating forming of generalizations that lead to the concept of "Rational Egoism", also considers generalizations that are claimed to be subsumed by concept altruism. This is known as Genus method. Generalizations like "One's own life is the standard of values" vs "Sacrifice is the standard of moral code" etc. This consideration can be compared to identity stage in concept formation, where for percepts or for lower level concepts, similar existents and their foils are grouped. Like percept of chair in identity stage of concept table.

 

3. Dr. Peikoff in his lectures on "Objectivism through Induction" also clarifies that while Deduction requires Induction, Induction itself is self-sufficient. What I infer from this is that "reduction", while moving from universals to particulars need not be deductive. Here we recall(if we are validating our own generalizations) AND chronologically list(particularly for other person's path to generalization) the relevant steps in the formation of generalization. And reorganize the chronological order obtained, if we realize(from experience and history) some more intermediate steps are needed to complete the chain. The example that comes to mind is that while validating generalization "Reason is man's means of survival", we might infer in retrospect just by looking at hunting using man made stone weapons and farming. But when we look at history, we will realize that it required Industrial Revolution[for Ayn Rand] to really validate how reason without any element of initiation of force against other men can ensure survival and prosperity throughout life. (Greeks did not regard reason as means for practical ends, but an end in itself).

What I need to further investigate is if "Genus method" can "efficiently replace" or complement deductive reasoning in enabling new concretes to be subsumed by existing concept or generalization. One more thing that can be further investigated, and can help in complementing is if the Deductive reasoning can be subsumed under the Genus of "Genus Method". In Genus method, after reaching a generalization we can apply that generalization to the lower levels of component concepts, and then extend it to their variants. For e.g. in the method which I used previously(in prep section of OPAR) for establishing that "causality is the corollary of identity" was a Genus method. I first understood concept of "corollary" from an example in geometry, and then applied same concept to philosophy.

 

Q4. What are particular instances and generalizations that induction deals with?

 

Ans: From TLL
When we reason from "Men in my experience are mortal" to "All men are mortal"; or from "These fires burn me when touched" to "Fire by its nature burns"; or from "This apple and the moon obey the law of Gravity" to "Every physical object in the universe obeys the law", in all such cases we are passing from one realm to another. From observed to unobserved; from the past behavior of nature to its future behavior; from what we discover in a narrow corner of a vast cosmos to what is true everywhere in that cosmos.

 

MY FURTHER ANALYSIS

The passage introduces one problem that I am not very clear about. What is the starting point of induction? For a time being if we consider induction a variation of concept formation, that is a more complex form of concept formation done at later stage. With that outlook starting point of induction are percepts like rolling of various objects to reach generalization "among all objects spherical balls roll longest distance[weight, material and surface being same]". But given the complexity of inductive reasoning, I think its better to consider induction separate from concept formation. Treating concept formation as a preliminary rather than a part of induction. So induction for the generalization will begin once concepts like ball, solid objects, rolling, ground, distance are formed; and then only observations which will be later subsumed under the generalization be useful. With this context "particular sentences" like "Men in my experience are mortal" OR "These fires burn me when touched" are particular instances of generalizations. And "All men are mortal" or "Fire by its nature burns" are universals of induction, just as men and tables are universals of concept formation.

Edited by RohinGupta
Link to comment
Share on other sites

Yawn.  All of this appears to be rather cut-N-paste Scholasticism.  Repeat what you heard, and you pass the course....

 

It seems to me that it's like any kind of self-motivated study. It is what you make of it. You can copy and paste answers or you can think about the concepts, concretize them in your own words, and chew them. It's hard to retain abstract ideas by simply reading them.

Link to comment
Share on other sites

Today I posted questions from section on "The Nature of Concepts" from "Chapter 1 - The Foundation"

 

Here are questions

 

Q1. Why is discussing Nature of concepts necessary? What aspects of the nature of concepts are discussed here?

 

Q2. Explain the role of measurement omission in forming the concept pendulum?

 

Q3. How does the chapter explain concept?

 

Q4. How does chapter explain similarities between "mathematics" and "concept formation"?

 

Q5. What does term "first level concept" refer to?

 

Q6. What are the similarities and differences between higher level concepts and first level concepts?

 

Q7. How does the chapter explain the role of definitions? Why are definitions necessary for conceptual level consciousness?

 

Q8. Why is it necessary to emphasize the cognitive role of definitions?

 

Q9. In this chapter, generalizations are identified as true and false, while concepts are identified as valid and invalid. What is the difference between true-false assesment and valid-invalid assesment?

Link to comment
Share on other sites

Today I answered questions from section on "The Nature of concepts"

 

Sharing sample answers for wider audience

 

Q1. Why is discussing Nature of concepts necessary? What aspects of the nature of concepts are discussed here?

 

Ans: From TLL

 

....I briefly indicate a few points that are particularly relevant to validating inductive generalizations. Reason is our faculty of acquiring knowledge by means of concepts. We form concepts by grasping similarities among existents, similarities that make a group of existents stand out against a background of different existents.
Similarity, of course, does not mean identity; Rand recognized that the existents united by a valid concept differ in every aspect. But she also recognized the nature of these differences. They are quantitative, i.e., they are differences in the measurement of characteristics. When we form a concept, our mental process consists in retaining the characteristics but omitting their differing measurements.

 

MY FURTHER ANALYSIS

Induction is used for forming a generalization, and later establishing it as true or false. Concepts are part of generalization. So without validating concepts as valid or invalid, generalization cannot be validated as true or false(or arbitrary).
Further, both involve moving from particulars to universals, so there are many similarities in the process also. As Dr. Peikoff mentioned in Introduction
"Every major aspect of the Objectivist view of concepts - including the role of similarities and differences, of integration, of hierarchy, of context - has a counterpart in the theory of generalizations. Indeed, generalization, Mr. Harriman explains, "is nothing more(or less) than an essential form of the method of concept formation."

Generalizations are built on top of concepts, the way concepts are built on top of percepts or on top of lower level concepts(like furniture) or also on top of lower level generalizations(e.g. concept of mass and Force). So understanding component concepts is a pre-requisite to understanding a generalization. One relevant aspect of concepts mentioned here is that we form concepts by grasping similarity. At perceptual level it can be grasped by observing together similar tables with a foil like a chair. Other extreme, particularly for forming higher level concepts like government and economics can be to recall characteristics of various units the concept may subsume. It is difficult to observe all varieties of units together with a foil in such cases. Intermediate methods can be closer to perceptual level like observing pictures or videos. Or closer to recollection where various characteristics and units of concept are listed in notes as outline.

 

Further, measurement omission while forming concepts can be of different types, and done in different ways :

 

1. Characteristic itself is omitted from essential characteristics - like color of table when forming concept table, or fabric of thread in pendulum.

 

2. The exact measurement of characteristic is omitted, but the range of characteristic concept subsumes is retained. For e.g. for concept table the exact size is not specified, but the range of size dictated by its function is specified. Though omitting characteristic, as in (1.) is a subset of (2.). While omitting color we are implicitly acknowledging that table can have any color(if omission is done consciously). But still this separation(between 1. and 2.) enables us to separate out fundamental characteristics from accidental characteristics. That is characteristics that enabled mental separation of the object.

 

Different ways of forming essentials by omitting measurements can be as follows :

 

1. Purely perceptual : When two tables are observed together with the chair as a foil, the grouping of tables percepts together in mind will require little effort. And the characteristic "table shape" will not be "abstracted" at this point(not even lined diagram of table). Instead any of the percept of two tables will be used as a "standard of measurement".

 

2. From percept as standard of measurement in (1), over a period of time the pattern of shape will be abstracted in the form of lined diagram, and will replace it as "standard of measurement". For e.g. symbol of table.

 

3. After forming concept, observing other characteristics like how it is used and range of its size, new characteristics will be appended while specifics of these characteristics will be omitted. E.g. use of table for keeping items useful to humans(This might require conceptualizing these items).

 

4. Some characteristics while forming concept like oscillation, like color and texture of the bob of pendulum can be omitted based on prior knowledge, where it is inferred that these characteristics

do not have any major impact on motion. This inference might be drawn by retrospecting on the formation of concept rolling, and recalling the inference that the change in color and material of the ball did not impact nature of rolling much. (This would also require subsuming oscillation and rolling under single concept motion).

 

5. In case of higher level concepts like government, (1),(2),(3),(4) are applied differently. Here too foils like private competitive enterprises are used, but rather than observing perceptually,‌ characteristics of each is listed and compared. Then differentiating characteristics like legal use of physical force is separated from non-differentiating characteristics like buildings, group of people working in these buildings, and reporting in hierachial manner. So CCD here is "Legal use of physical force".

 

6. In mathematics omission can be done by i) Retaining count but omitting the shape, material etc. Using "category of quantity" as CCD that is.
ii) Applying limits like "(delta x) tends to 0" in calculus. And thus form relationship between acceleration and time from relationship between velocity and time.
iii) Rounding off the length to form concepts like meter, millimeter etc.

Further, while similarity of group of objects is not identity, I think it is "part of identity". Identity of the object separates it from all other objects. So similarity mentioned in the concept that is subsuming the object will separate it from many of the objects that will not be subsumed by the concept. Combining similarity of concept with other differences with objects subsumed can establish the identity of the object.

 

Q9. In this chapter, generalizations are identified as true and false, while concepts are identified as valid and invalid. What is the difference between true-false assesment and valid-invalid assesment?

 

Ans: From dictionary
Valid - Having sound basis in logic and fact.
True - In accordance with fact or reality.

 

Something that is claimed true can be refuted as false with a single observation. For e.g. "all pigeons are blue" can be refuted as false with the observation of white pigeon[1]. But something that needs to be demonstrated as invalid requires some integration after plain observation to be demonstrated as invalid. So to invalidate a concept we have to look at what units it subsumes, what is its definition or essence and CCD, and I think sometimes also the propositions[2] in which it is used.

 

The book gives example of lightness as an invalid concept, which was first conceptualized by the Greeks(Page 67 and Page 123). Lightness distinguishes air and fire from heaviness of Earth and water. Former going up and latter going down. And respective up and down motion is considered as the essential feature of each. But during late Renaissance Torricelli's experiment demonstrated that with nearby vaccum, air too can can have downward motion, as in barometer. Since the essence of concept lightness conflicted with one of the unit it subsumed, therefore the concept was invalidated[3].

 

The proposition "air or light objects move up" could be falsified just by the observation of Torricelli's experiments. But to invalidate concept lightness, reduction to what the concept subsumes and what is its definition and CCD is also required along with the observation of Torricelli's experiment.

So while concept light and heavy, with CCD as force exerted while holding them or more accurately weighing them are valid. Metaphorically similar concepts light and heavy that have CCD as motion of corresponding units is an invalid one, since under some circumstances the motion of units was reversed.

 

[1]Strictly speaking, some reduction like reduction of concept pigeon and blue to one percept of each is still required. But this reduction is much less than reduction needed for validating a concept, where apart from 2-3 existents concept subsumes, we also need to reduce the attributes used in definition(and further the percepts these attributes subsume).
[2] In simpler terms, propositions are sentences.
[3] One might argue, why lightness is invalidated, and spherical ball not rolling in magnet is delimited and not invalidated. I think reason can be, spherical ball rolling still has many situations as in understanding friction(where there are no magnets). But once lightness is invalidated, all its situations were invalidated since it was replaced by a valid concept mass to explain motions like projectile.

Link to comment
Share on other sites

Today I posted following questions from section on "Generalizations as Hierarchial"

 

Q1. What is pre-science? What is its role in formation of science?

 

Q2. Demonstrate that Generalizations like concepts are hierarchial?

 

Q3. What is reduction? What is its role in understanding generalizations of physics?

 

Q4. Reduce Galileo's generalization "horizontal motion is unaccelerated"?

 

Q5. Reduce the generalization "Light travels in straight lines"?

 

Q6. What are first level generalizations? What is their significance in induction?

 

Q7. How do you reconcile the fact that first level generalizations are self-evident, and that they have component first level concepts that are universal?

 

Q8. Why is it necessary to qualify the context for any generalization or concept?

Link to comment
Share on other sites

There has been good discussion on valid-invalid nature of concepts, Posting the details here

 

Q9. In this chapter, generalizations are identified as true and false, while concepts are identified as valid and invalid. What is the difference between true-false assesment and valid-invalid assesment?

Ans:
From dictionary
Valid - Having sound basis in logic and fact.
True - In accordance with fact or reality.

Something that is claimed true can be refuted as false with a single observation. For e.g. "all pigeons are blue" can be refuted as false with the observation of white pigeon[1]. But something that needs to be demonstrated as invalid requires some integration after plain observation to be demonstrated as invalid. So to invalidate a concept we have to look at what units it subsumes, what is its definition or essence and CCD, and I think sometimes also the propositions[2] in which it is used.
The book gives example of lightness as an invalid concept, which was first conceptualized by the Greeks(Page 67 and Page 123). Lightness distinguishes air and fire from heaviness of Earth and water. Former going up and latter going down. And respective up and down motion is considered as the essential feature of each. But during late Renaissance Torricelli's experiment demonstrated that with nearby vaccum, air too can can have downward motion, as in barometer. Since the essence of concept lightness conflicted with one of the unit it subsumed, therefore the concept was invalidated[3].

The proposition "air or light objects move up" could be falsified just by the observation of Torricelli's experiments. But to invalidate concept lightness, reduction to what the concept subsumes and what is its definition and CCD is also required along with the observation of Torricelli's experiment.

So while concept light and heavy, with CCD as force exerted while holding them or more accurately weighing them are valid. Metaphorically similar concepts light and heavy that have CCD as motion of corresponding units is an invalid one, since under some circumstances the motion of units was reversed.

[1]Strictly speaking, some reduction like reduction of concept pigeon and blue to one percept of each is still required. But this reduction is much less than reduction needed for validating a concept, where apart from 2-3 existents concept subsumes, we also need to reduce the attributes used in definition(and further the percepts these attributes subsume).
[2] In simpler terms, propositions are sentences.
[3] One might argue, why lightness is invalidated, and spherical ball not rolling in magnet is delimited and not invalidated. I think reason can be, spherical ball rolling still has many situations as in understanding friction(where there are no magnets). But once lightness is invalidated, all its situations were invalidated since it was replaced by a valid concept mass to explain motions like projectile.

 

More thoughts on valid-invalid nature of concepts

So far I have written about epistemological aspect of validating concepts. That is it requires more reduction than reduction of generalizations. But more important than this I think is metaphysical aspect of validation. Let me elaborate here a bit on this.

Generalizations often refer to facts about reality. So checking them with respect to what exists establishes their truth or falsehood or arbitrariness. But concepts at their base have an entity, and consciousness that grasps that entity. More specifically concepts derive from senses, senses that grasp entity in the form of percepts. So if the nature of senses change, nature of concepts also change. For e.g. if there are conceptual organisms as small as atoms, they will form concepts involving atoms perceptually. But humans have to infer these from prior concepts and from experiments based on prior concepts. So if the small conceptual organisms refer to atoms as perceptually given, without giving any prior context, their concept of atom is valid. But for humans same validation will require referring to prior concepts and experiments. Without that reference, the concept is invalid.

Concept atom, prior to experimental discovery was also invalid concept. So whether or not the corresponding existent exists is necessary but not sufficient condition for the concept.

Link to comment
Share on other sites

Today I posted questions from the section on "Perceiving first level causal connections"

 

Here is the list

 

Q1. What is the purpose of this section?

Q2. What are generalizations(whether first level or higher level ones)?

Q3. What is the purpose of generalizations?

Q4. What aspect of causal connection we will learn in this section, and what later?

Q5. What is a precondition for grasping first level generalization?

Q6. What is the precondition to understanding law of causality?

Q7. What kind of perceptual causal connections are understood in the beginning?

Q8. What kind of causal connections are discovered later?

Q9. What is the significance of perceiving perceptual connections?

Q10. How will you differentiate observation of causal connection from the regularity of observation?

Q11. Based on evolution of man's thought process, demonstrate how significant is the difference between causes having conscious being like self, and impersonal causes as in fluttering of leaves?

Q12. Impersonal metaphysics states that law of identity and not consciousness is the cause of natural phenomenon(and indirectly also man-made phenomena, since volition is also the part of identity of man)? Elaborate on how this idea was originated and transmitted into civilizations across History?

Link to comment
Share on other sites

Today I shared answers to Questions from section on "Perceiving first level causal connections".

 

Sharing sample answer for wider audience.

 

Q9. What is the significance of perceiving causal connections?
Ans:
From TLL
The primary method of grasping causal connection, therefore, is to perceive it. Building on this foundation, scientists develop more sophisticated, experimental methods of discovering causality, in higher level cases where perception of the cause is not possible. The use of such methods requires an analysis of variables going far beyond the first level cognition. In regrard to first level generalizations, however, direct perception of cause and effect is essential - and sufficient.

MY FURTHER ANALYSIS
Continuing my disagreement further on first level generalizations being self-evident, though I accept these are "validated"[reduced after being formed] by direct perception of cause and effect. I see error in statement that "first level generalizations are self-evident" as extension of intrinsic theory of concepts, Aristotelian variant. There is nothing innately written on the percepts that says "pushing rolls the ball". We first have to isolate frames of percepts from the observed sequence of ball rolling after being pushed. In first percept its not rolling and it is not pushed, in second percept it is pushed and in third percept its rolling. We then subsume second percept under concept push formed before, and third percept under concept rolling formed before. The ball in all these percepts is subsumed under concept ball. And when we look into three percepts, ball is the common entity combining these percepts. So forming first level generalizations requires isolation of perceptual frames, identifying concepts of existents in these perceptual frames, integrating the concepts using common entity across the perceptual frames.[so "Subsuming as unit percept of static ball just prior to being pushed" + "Subsuming as unit percept of pushing" + "Subsuming as unit percept of rolling ball" = "First level generalization" "Pushing rolls the ball"].

[in above percepts, first percept of "ball not rolling when not pushed" immediately precedes other two percepts often but not always. But second and third percepts have to occur chronologically always, such that percept of pushing[cause] has to be seen immediately before percept of rolling[effect] for the first level generalization "pushing rolls the ball" to form. The first percept can act as foil, like perception of chair acts for concept table.
Further, like initial first level concepts, I think initial first level generalizations can also be formed pictorically].

So when scientist tries to design experiments, he is creating circumstances that will help him grasp "higher level generalizations" as if they are "first level generalizations". Like balls falling same distance from table after rolling on the table through inclined planes of same inclinations. Here while the perception is not very different from what first level generalization "pushing or inclination rolls the ball" can subsume as unit. The inference drawn is a much higher level generalization "Horizontal motion is unaccelerated", because the percepts are subsumed by higher concepts like "horizontal", "motion" and "unaccelerated" by the scientist.
So perceiving of perceptual connections, where effect happens immediately after cause, can also help in validating hypothesis formed elsewhere(pendulum experiments in this case).
 

Link to comment
Share on other sites

Questions from this week and sample answer

 

Here are questions from the section on "Conceptualizing first level causal connections"

Q1. Why is conceptualizing first level causal connections necessary?

Q2. Recapitulate stages in evolution and expansion of "concepts in general", which will help us in understanding inducing of first level generalizations?

Q3. Why should the statement referring to first level generalizations be in terms of first level concepts?

Q4. How are first level generalizations universal in nature?

Q5. Why is it not possible for animals to form generalizations, even though they also experience causation directly?

Q6. How can we say that "method of induction" is a special case for "method of concept formation"? That is how measurement-omission is taking place in induction?

Q7. Summarize the role of "first level generalization" and "perceiving of causal connections" in induction?
 

 

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

 

Q3. Why should the statement referring to first level generalizations be in terms of first level concepts?
Ans:
FROM TLL
Notice that when our first-level inducer identifies a perceived causal connection in words, he does not do it as a description of unique concretes, even though that is all he perceives; he at once states a universal truth. His first remark is not "I see shimmering, yellow orange smoke-emitting flames, about one foot high, turning that huge front page of newspaper with its big headlines into a small pile of blackened ashes." The child at this early stage does not have the conceptual apparatus necessary to distinguish one instance of fire-burning-paper from another by means of words; first he must grasp that what he sees is an instance of "fire", "burning", "paper", i.e., of his earlier concepts. Only much later, when the vocabulary identifying specific measurements of these existents has been developed, can he use words in sophisticated combinations to describe the action of a unique fire. Logically, the generalization must come first; it is the direct product of applying one's conceptual apparatus to the perceived connection.

MY FURTHER ANALYSIS
The section takes the order in which we know one step further. So far we have studied first level concepts, then first level generalizations. Further, we have second level concepts like "furniture" which are broader in scope, or "dining table" which are narrower in scope. And when such vocabulary is sufficiently formed, which will also include higher level concepts of actions like shimmering, or high level concepts like newspaper; will the description of phenomenon in more specific terms be possible. So first level concepts and generalizations are known much before generalizations which will describe the perceptual connections in specific terms.

FROM TLL
Similarly, a toddler sees a particular ball, but his identification of it is simply "ball". At this early age, the child does not and cannot know any other integrations or narrower subtypes or cross-classifications of "ball"; he cannot identify a ball as "a human artifact", or as "a yellow tennis ball", or as "a product of capitalist profit seeking". To him, at the start of the conceptual process, the verbalized object is "ball", pure and simple. The same applies to the child's experience of himself as the particular pushing agent. His identification must be of "pushing" as such, not of "voluntary human action", nor of "exerting force", nor of "his own individual act of pushing" - since he does not yet know any such relatively more abstract terms; the concept simply denotes any and every place of his observation, this is par excellence an irrelevancy to the child(and to anyone). Inherent in forming and applying a concept is the understanding that what counts cognitively is only the identity of referents. The mere passage of time or the change of location, assuming everything else remains the same, makes no difference to one's conclusions, because the concept of an existent subsumes all instances everywhere, past, present, and future.

MY FURTHER ANALYSIS
So summarizing the order of learning
1. Percepts
2. Forming first level concepts of entities from perceptual frame.
3. Forming first level concepts of attributes from percepts.
4. Forming first level concepts of actions from percepts.
5. Perception of phenomena.
6. Identifying first level concepts in the phenomena. Like Fire, paper, burns, hot, bright.
7. Forming first level generalization by connecting phenomena observed and first level concepts identified in (6.). "Fire is bright", "Fire is hot" being few examples of simpler first level generalizations. And "Fire burns paper", "fire cooks food" etc. being example of complex first level generalizations.
8. Adding simpler characteristics of attributes to concept of entity from generalizations. Like hot, bright, etc. for fire.
9. Adding some less simple characteristics to corresponding first level concepts based on first level generalizations formed from them. Like characteristic "burns paper" for fire, or "burned by fire" for paper, or "acts on paper", "caused by fire" for burns.
10. Forming second level concepts of entity, attributes, and actions from first level concepts and their characteristics identified in 8-9 using generalizations. Like lighted that subsumes fire, bulb, tube, computer, mobile, lamp etc.Or "glowing" that subsumes "burning", "bright", "shining" etc. Or "changing" that subsumes "burning of paper", "breaking of glass", "moving of chair" etc.
11. Forming second level generalizations from second level concepts and from new percepts, or instead of new percepts from recollection of percepts that were used for forming first level generalizations. Like "Night is lighted", "Laboratory is glowing" or "There are many changes in one day".

Note: While the text mentions that "Pushing rolls the ball" as first level generalization. I think prior to that generalization "Ball rolls" should be formed. But I think we can classify both of these as first level generalizations, since they are formed from first level concepts.

Edited by RohinGupta
Link to comment
Share on other sites

Questions from this week and sample answer

 

Here are questions from the section on "The structure of inductive reasoning"

 

Q1. How is this section similar and different from previous sections?

Q2. What is reasoning? What might be the need to introduce this concept at this point?

Q3. What role logic plays for reasoning in general, and deductive reasoning in particular?

Q4. How is inductive reasoning different from deductive reasoning, when it comes to integrating the premises to form conclusion?

Q5. What is the problem of induction?

Q6. How does study for reaching "first level generalizations" contribute to the solution of "problem of induction"?

Q7. What leads can we get to the solution by studying Benjamin Franklin's kite experiment and related facts?

Q8. Summarize similarity and difference between first level and higher level generalizations?

Q9. Elaborate on the conceptual framework that Franklin had to integrate when he did kite experiment?

Q10. What is the role of integration and logic in developing inductive generalizations?

Q11. What is the difference between premises of induction and deduction?

Q12. Why problem of induction has been elusive for so long?

 

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

 

Sample answer is as follows

 

Q7. What leads can we get to the solution by studying Benjamin Franklin's kite experiment and related facts?

Ans:
MY ANALYSIS
This part shows the process of induction for reaching higher generalizations. So far we have studied induction in detail only for first level generalizations.

FROM TLL
Let us examine a case typical of scientific induction, in which we do not directly perceive causal connections, and in which we are not restricted to a handful of first-level concepts. The example I have chosen is Benjamin Franklin's famous kite experiment. It will serve to illustrate the pattern that we will see repeatedly in the next four chapters.
Franklin set out to prove that lightening is essentially electricity. Thunder clouds, he hypothesized, are electrically charged, and lightening is merely an electric discharge, when this charge runs off. Scientists do not always have their explanation of events worked out as a hypothesis in advance of their experiments, but sometimes they do, and Franklin did in this case. He setup his apparatus accordingly.
......If Franklin was right, the electrically charged thunder clouds should cause the key to become charged, and this charge will flow into and accumulate in the Leyden Jar. Of course, everything Franklin predicted turned out as he expected, and he concluded - from his experience of one thunderstorm - with a sweeping and necessary generalization describing the nature of lightening as such, here, everywhere and always. Why did this prove his generalization?

First, what did Franklin actually see during this experiment? What concretes did he observe that could also have been observed by a child or a savage? Among other things, he saw sparks flying from the key to the wire in the Leyden jar. He saw pieces of the wet kite string become rigid and repel each other. And he found that if he held the jar with one hand while touching his finger to wire entering it, he felt an unpleasant shock.

These concrete observations are essential[but not sufficient] to the experiment; you could learn nothing from it[experiment] without them[concrete observations]. Yet these observations would be meaningless to an ignorant person.....
Besides his percepts, Franklin needed a series of sophisticated concepts; otherwise, he could neither have designed his experiment, nor interpreted its results....concepts as "electricity", "discharge", "conductor", "Leyden Jar"..... These concepts were made possible by and represent a wealth of earlier knowledge([some of]which was also discovered by means of experiments). Without this conceptual framework, Franklin could only have stared uncomprehendingly at sparks and shocks. Given such a framework, however, he can at once identify what he is seeing : The kite apparatus is a long conductor, and thus the electrically charged thundercloud causes the key to become charged, and then the key discharges into the insulated Leyden jar.
.....The conceptual framework enables him to identify with the concrete measurements omitted, the essential causal chain; nothing else explains the observations.
If one grasps the observations in this case, and learns that the conceptual framework is valid, then the generalization follows necessarily.

MY FURTHER ANALYSIS
The most important inference for induction in this section is regarding higher level generalizations. First level generalizations require relevant first level concepts, and perception of causation in phenomena the generalization will refer to. Similarly higher level generalization requires higher level concepts, and like first level generalization referring to a perception depicting causal connection. That is, higher level generalization also refers to percept containing causal connection. Further, we also see, as in case of Galileo, the concepts, more specifically their characteristics contribute to the design of experiment. Further, for humanities subjects like history, while having experiment is not possible, as we saw in DIM, these characteristics can help select relevant subjects, existents, and enable their interpretations from various eras and fields(What is being integrated? How it is being integrated?)

Coming to Franklin's experiment, here too we follow similar path as in "First Level Generalization". While ultimately it established[connected units of] concept electricity, explicit scope of this experiment was to establish the generalization "clouds are electrically charged". The behavior of lightening, or characteristics of concept lightening led to the formulation of hypothesis. More specifically, its heat, light and equivalent shock characteristics are similar to sparks that came around charged particles[This part is not written in the text, but we can infer this from later text]. And the experiment established other aspects of clouds, like it is charged since the string of kite got charged. This being established by spark near Leyden jar and repulsion among the strings.

These are the steps that I think would be involved here(all are not necessarily chronological, but some can be) :-
1. Forming concepts of entities whose units exhibit electrical characteristics. Concepts like amber rod, spears etc. in this case.
2. Forming concepts, probably as phrases initially, for characteristics that subsume properties of charged materials or conductors. Like attracting-repulsing, shock producing, spark producing, heat producing etc. Or transferring shock, spark etc.(I also assume concepts of actions like attraction, shock etc. are formed before).
3. Forming causal connection between the characteristics in (2.), and entities in (1.), leading to higher level concepts like charge and charged materials.
4. Subsuming new units to materials exhibiting characteristics identified in (2.), like key, wet string, and Leyden jar.
5. Observing lightening and forming the concept that subsumes its various instances.
6. Adding characteristics to lightening using "first level generalizations". Like it produces shock, produces "heat and light" etc. These first level generalizations have concepts of attributes derived from (2.).
7. Formulating experiment that can be used to connect entities involved in lightening and entities involved in charge. The formulation is because of hypothesis that lightening and "sparks of charged materials" are same, and hypothesis is reached because of few similarities like shock producing, heat producing, and light producing between the two.
8. Doing and noting various observations in experimental setup.
9. Comparing new observations in the setup to observations of relevant phenomenon in materials identified in (1.).
10. If the actions are same, reaching the conclusion that concepts of attributes contributing to the actions are also same. And therefore concluding that entities involved in lightening have some attributes, that are same as that of charged materials. That is clouds are also electrically charged.
11. Forming / Expanding concept electricity based on the generalizations derived from observations. Generalizations involving flow of charges, sparks being produced, shock, heat and light.

Some concepts in the above experiment will come under pre-science knowledge. Like lightening, clouds, kite, spark, heat, light, etc. Others like charge, Leyden jar, conductor will be discovered by science. Even if one of the concept, its characteristics, or the observation is invalid or false, the generalization will collapse. But we know from physics, each of this is true or valid. Valid meaning consistent, and true meaning as is perceived.

 

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

Link to comment
Share on other sites

Other recent posts that I think are important are as follows

 

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

On certainty of "first level generalizations" and referrants of "first level concepts"

Certainty of "first level generalizations"

I have been using obvious to describe the status of "first level generalizations." Given its greater emphasis on completeness in understanding, I think certainty is a better term. Here I will clarify the context of certain.

I think certain is a genus that subsumes "self-evident", "obvious" and "proven" as units. Self-evident being perceptual. Obvious meaning easily reduced, like single percept as in "two tables with chair as a foil" for concept table. Or reduced to continuous percepts as in "pushing rolls the ball". Proven will also be ultimately reduced to perceptual, but the percepts will be distributed. Few of the many percepts of G=(m1)(m2)/(d^2) being motion of objects falling on earth, rotation of planets and moon, and experiments involving circular motion.

Here is relevant snippet from the excerpt of "How We Know"
------------------------------
p21-23

“Self-evident” is not a synonym for “obvious.” To one who has learned
arithmetic, it is obvious that two plus two is four, but that truth is not selfevident;
it is inferred by a process of comparison and counting. But that the
page you are reading exists is not an inference: it is self-evident.
The data of sensory perception are self-evident, but the conceptual
interpretation of that data, and inferences drawn from it, are not self-evident.
They must be validated by reducing them back to the self-evident.
-------------------------------

Referrants of "first level concepts"

Here is the relevant snippet from "How We Know"

-------------------------------
p151-153

To form a concept of an entity, we contrast two or more instances of
the entity with a foil — e.g., some tables vs. a chair. Likewise, to form the concept of an attribute, we contrast this and that instance of the attribute with a foil — e.g., two or more shades of blue vs. a shade of green. To form the concept of an action, we contrast this and that instance of an action with a foil — e.g., two or more instances of a thing moving vs. being at rest.
And, as with entity-concepts, concepts of characteristics are formed by
measurement-omission, on the “some but any” principle, and are integrated
into a new mental unit by means of a word.

We form higher-level concepts of characteristics just as we do in the case
of higher-level concepts of entities. “Blue” is first-level, within attributeconcepts; “color” is a widening; “indigo,” and “ultramarine” are narrowings. Narrowing by cross-classification is exemplified by “pastel blue,” if we allow two words to count as a concept, since “pastel blue” stands for those shades.

74 A technical issue arises regarding “levels,” because the term has two senses. In one sense, only concepts of entities are “first-level”: only entity-concepts presuppose no prior conceptualization. Since concepts of characteristics presuppose concepts of entities,
concepts like “blue” are not “first-level” in this sense. But, in another sense, “first-level” denotes concepts that do not integrate or subdivide any prior concepts, and in this second sense “blue” is first-level: it conceptualizes what is directly perceivable. Accordingly, these concepts need no validation or checking (there’s no such thing as getting “blue” wrong). Concepts like “blue,” “round,” and “moves” are part of the incontestable base to which more abstract concepts must be reduced and against which their validity is to be judged. As such, they could be called “reductively first-level.”

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

Personally I picture levels in 2 dimensions. X axis being entities at origin, attributes at 1, actions at 2 etc. Higher levels of entities like Furniture, Household items, man-made objects etc. being at (0,1),(0,2),(0,3) etc. if table is at origin. Dining Table, desk etc. being at (0,-1). Rounded being at (1,0), as attribute for entity ball in origin. Rolling being at (2,0) as action for ball etc. Motion being at (2,1) as higher level concept subsuming rolling.

 

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

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

 

Foils while forming "first level generalization

We have foil chair while forming "first level concept" table. Like concept is a universal, so is generalization. Therefore forming "first level generalization" also requires foils. Further if we retrospect on formation of our own "first level generalizations", we will see that without foil, forming first level generalization would not be possible.

If for e.g. bulb always remained lighted because of some hidden connection, child will never be able to form the first level generalization "switching lights the bulb" by just playing with the switch. He has to see first hand that with switch down bulb is on, and when switch is up bulb is off.

Off state of bulb I will call as metaphysical foil. I think there needs to be an epistemological foil also, which in this case will be "wish that bulb is not switched on when I push the switch down". Let me elaborate how. Before child actually pushes the switch down, he will have to make the decision at some level, mostly really low conscious level that is bordering sub-conscious level. So he makes the decision to push the switch down, actually pushes the switch down, and then the bulb glows. At this point he might start thinking that it was his wish and not push down of switch that made the bulb glow. But at later point lets say he wishes the bulb to glow without pushing down the switch, and the bulb does not glow. At this point he will be certain that "pushing down switches the bulb".

The epistemological foil I gathered from examples Dr. Peikoff gave in OPAR section we studied. "Wishing that marbles dont rattle", or "pillow rattles", or "balloon goes down on release" and "stone goes up" etc.

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

Link to comment
Share on other sites

Here is final update from the study-group

The final week was for "Review and Summary" So first I posted the the scope of "Review and Summary", and then my own "Review and Summary".

 

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

Basic structure of "Summary and Review"

Now that we have finished the main study text, we enter the concluding week. Please try to post pending posts this week.

Purpose of this concluding week is to summarize what we have studied, what we have learnt, how we have learnt it, and highlight the important aspects of the study-text. More importantly, we mention how we will follow up this study for even deeper understanding.

 

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

 

SUMMARY AND REVIEW

To summarize the path of (formal)study, I first studied about concepts and concept formation. Existent, identity, and unit stages from ITOE gave me better understanding about things I had already learned, and how to better organize my knowledge for improved understanding and living. That is through reduction, and keeping snapshots of units and characteristics in various states of unit stage, identity stage or even existent stage of some concepts. Studying causality helped me to better organize knowledge related to various events I observe and those I know of. Examples from physics(and from chemistry, biology, history, I recollected) gave me even better grasp of causality. Co-relating my day to day thinking to that of Galileo and Newton I think will make me a better person by enabling me to solve problems better through systematic understanding.

We apply principles often(and laws get applied on us), but don't validate them as much as required. Now I think I will validate my goals, motivations, principles, and events impacting me more often. And fine-tune the hierarchy of my values, based on proper analysis of emotional triggers I get. Professionally, I will continue to effectively question key policy reversals in my company happening due to cultural bankruptcy of humanities. Specifically, removal of ratings system in my case. But primarily prioritize my work related tasks better, and organically improve my knowledge of software systems and their genesis. I can already see much better problem solving, partly due to experience, but also by better understanding similarities and differences between various systems and their versions.

While most of us know that we form concepts and generalizations from percepts(thanks to Aristotle), now many of the missing links have been filled. To further solidify my understanding, I plan to again study "the full book", "Abstraction from Abstractions", "Cognitive Role of concepts", and "Concepts of consciousness" from ITOE. Reading How we know excerpts itself improved quality of my posts. I will definitely follow it up with complete study. But more importantly I plan to cash in on the epistemological assets I have built through this study group. So expect my next study-group to be on Ethics, probably Virtue of Integrity.

My understanding of first level concepts and generalizations has been greatly enhanced though. I can now think of any first level concept like table, chair, red, rolling, etc. to have 6 perspectives. The existent, its percept(s), its characteristics, essence in concept, its CCD, and the category like enity, attribute, action it belongs to. I intend to build on this understanding and have better understanding of complex categories involving concepts-of-consciousness, abstractions-of-abstractions(-of-abstractions...), axiomatic concepts etc. However I am still thinking of one aspect. If induction is moving from particulars to universals, why are we spending so much time in moving from universals to particulars using reduction. I think reason can be that moving from universals to particulars for known concepts like tables, furnitures, gravity, etc. can provide us more (explicit) CCDs and better understanding of categories of concepts. Armed with this better understanding, we can then better classify new existents we encounter. And thus enabling us to form correct, and more abstract concepts and generalizations.

Representation through symbols, whether its concepts, mathematics, physics or deduction, it improves the understanding of finer aspects. Introducing symbolism in concept formation and induction(not to that extent in Induction though) I think is a major take away here. Whether its through sets in various stages of concept formation and induction, or setting and organization(prioritization through fundamentality) of characteristics, essence or CCD.

So to conclude, I will study epistemology privately in short term. Publicly I will try to connect epistemology studied so far to ethics, politics, economics and law. So Tara Smith's book on ethics, and Yaron Brook and Don Watkin's "Free Market Revolution" are what I am eyeing. Apart from these I will also rescan archives of SGO, and co-relate it to what I studied here.

Thanks to all for participating. Apart from academic, enhancing my project management and marketing skills was a major takeaway. This "study group" will always remain special. In totality it took minor part of my year, but gave me unprecedented joy in return....!

Amen, not in the name of supernatural, but "In the name of Best Within Us"(David Harriman, Leonard Peikoff, ARI, Brad, all the particpants including myself, and the fountainhead whose soul shall always reside in these study-groups and inside many of us - Burgess Laughlin!!).

 

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

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...