A Fundamental Mistake
The purpose of this chapter is to provide a concise unambiguous explanation of the nature and meaning of induction, which is not an alternative method of reason to deduction.
This chapter, and it's companions, "Epistemology Mistakes," "Concepts," and, "Cause," are necessary because these ideas are fundamental to almost all the remainder of philosophy and no philosophy today correctly identifies them, and, in fact has so corrupted them, no further progress in the field of philosophy is possible.
Induction is described as a kind of reasoning and is frequently explained as the converse of deduction, such as: "Deduction is reasoning from the general to the specific," verses, "Induction is reasoning from the specific to the general."
Neither deduction or induction are kinds of reasoning, however. Reasoning (thinking) is the mental process of acquiring knowledge (concept and proposition formation) and using knowledge to ask and answer questions and make judgments.
Language, mathematics, logic, and geometry are methods or tools of reasoning, not reasoning itself. All knowledge and thinking is done using language, one uses mathematics to identify and think about those aspects of reality which can be counted or measured, geometry to identify and think about those aspects of reality with spatial attributes, and logic to ensure one's concepts, propositions, questions, answers, and judgments are correct, but they are not methods of thinking or reasoning.
Logic, in particular, identifies the principles to which thinking and reasoning must conform to be correct. It only identifies when a process of reasoning is valid to insure one's reasoning does not result in mistaken conclusions. Logic does not determine what is or is not true. Reality determines what is true. Logic only ensures one reasoning remains consistent with reality by avoiding contradictions.
[NOTE: Please see the chapter, "Truth," for an explanation of what truth is, the chapter "Thinking And Reason," for an explanation of what rational thinking is and the principles of correct thinking and the chapter, "Knowledge Methods: Logic," for a discussion of the formal principles of reason.]
The Mistaken View Of Deduction
The mistaken view of deductive reasoning is the view that correct reasoning is done by means of some formula called, "logic," consisting of specific arrangements of words (or symbols). The two basic "arrangements of words," in all modern views of logic are, "the proposition," and, "the syllogism," or other formal arguments.
There is nothing wrong with the concepts of "propositions," or, "syllogisms," so long as it is understood, the words by which propositions and syllogisms are constructed only represent, "concepts," and that propositions and syllogisms have no meaning at all except any meaning that is based on the meaning of the concepts the words (or symbols) indicate. All modern views of logic either ignore the fact that a word or symbol has no meaning (sans concepts) or wrongly assume that words or symbols, themselves, have meanings. [Please see the chapter, "Concepts."]
The mistaken view that words have meanings is based on the Kantian view that, "words mean their definition." Every wrong view of reason can be traced back to the view that logic is the mere manipulation of words or symbols (without reference to any concepts or meaning), e.g., logical positivism and linguistic analysis, or that words mean their definition, e.g. the entire Humean/Kantian destruction of epistemology. [Please see the chapter, "Epistemology Mistakes."]
These epistemological mistakes lead the modern, "logician," to declare: "Although deductive reasoning is logically certain, it cannot provide new information. In a syllogism, the conclusion is already contained in the premises. The conclusion is just another way of stating the premises."
The purpose of logic is not to, "provide new information." It's only purpose is to insure whatever reasoning one has done is correct. New knowledge (true information) is only about what exists and its nature. Only reasoning that identifies what has not been previously identified is new knowledge.
Logic can be used to ensure one's reasoning contains no contradictions and that new knowledge is non-contradictorally integrated with all one's previous knowledge, but logic is not the reasoning by which one learns or discovers new knowledge.
The Mistaken View Of Induction
According to modern logicians, where deduction is always valid but provides no new knowledge, induction provides new knowledge, but is never certain.
New knowledge, according to the modern view of induction, is based on observation or experience beginning with one's direct consciousness of existence, that is, perception, (seeing, hearing, feeling, smelling and tasting). A number of variations of induction are usually cited, like: Generalization, Causal Inference, Analogical, Statistics, Prediction, and Authoritarian.
The primary meaning of induction is supposedly based on observation of some phenomena, such as a kind of entity, relationship between entities or actions, or the similarity between things. The validity of, "knowledge," supposedly based on these observations is determined by how often or consistently the same observation is made. Some variations of induction are explained as follows:
Inductive Generalization—based on repeated observation of some kind of entity. For example: If one observes 100 (or a thousand or a million) swans, and all are white, one might infer, "All swans are white."
Causal Inference—based on observed repetition of specific events always following other specific events. This most common form of induction is defined either as "concluding something may be the cause of a subsequent effect if the same cause always precedes the same effect and no other explanation is available. For example: "There are more colds in winter then in summer therefor cold weather is a cause of colds."
Analogical Induction—common features in similar entities suggest other shared features. For example: All observed organisms' blood is red because of it's iron content, therefor all organisms have red blood. (Except horseshoe crabs, snails, lobsters, squid, and some others have copper based blue blood.)
Statistical Induction (general)—general statistical induction uses a smaller set of data to make a generalization about larger sets. For example: The average number of books each person in America reads over the course of a year is twelve therefore Americans read about twelve books a year. [Except 40 per cent of Americans read no books in a year.]
Statistical Induction (specific)—statistical induction uses a analysis of a collection to make predictions about individuals in that collection. For example: The most frequently reported number of books read by each individual in America is four books per year, therefore if you are an American you probably read four books per year. [Except for those individuals who never read a book (40 per cent) and those who read 10 or more books per year (20 per cent)].
Predictive Induction—draws a conclusion about the future based on past experience or records. For example: "As far back as anyone can remember the volcano has never erupted, therefore it's not going to erupt." (Mount Vesuvius Gazette, Pompeii, 78 A.D.)
Authoritative Induction—conclusions based on the opinions of supposed authorities or experts. The basis of, "truth by consensus," and supposed, "peer review." For example: "Heavier-than-air flying machines are impossible." (Lord Kelvin, British mathematician and physicist, president of the British Royal Society, 1895.) and "There is not the slightest indication that nuclear energy will ever be obtainable. It would mean that the atom would have to be shattered at will." (Albert Einstein, 1932.)
Induction Is Observation And Discovery
Induction is not a kind of reasoning or logic. Induction is entirely a method of observation and discovery. It is not, for example, how science is done, but the recognition of the entities and events of physical existence that the physical sciences study.
The observation of entities, their differences and similarities, their behavior and relationships is the beginning of science. Noticing that things have different attributes and characteristics, that rocks, apples, and fish are all different, but that all rocks have similar attributes, and all apples have similar attributes, and all fish have similar attributes is the beginning of science. It is those differences that lead to the discovery that how a thing behaves and relates to other things is determined by a thing's nature (its qualities, characteristics, properties and attributes). A rock is hard and heavier than an apple or fish and can be used to pound or break things, but a fish or apple cannot. One can eat an apple or a fish, but not a rock. An apple can be picked off a tree, but a fish has to be caught in water. Rocks and apples only move if something moves them, fish move on their own.
As primitive as such observations are, they are essentially scientific observations. The first step in all of science is the identification of existents and their nature, i.e. what they are. Science, in earnest, begins with the discovery of how the behavior of entities is determined by their nature and their relationships to other entities. How anything behaves relative to anything else (the true nature of cause) is determined by an entity's own nature and the nature of all other existents it is related to. Scientific principles identify and describe those relationships.
It is the observation of the behavior and relationships between existents that lead to the discovery of scientific principles. Real science begins with questions about what is observed: what is a thing? what are its qualities? how does it behave? how does it relate to other things? Scientific principles answer those questions. None of those questions can be answered by inductive observation.
Induction Provides No Knowledge
Knowledge is only knowledge if what is believed to be true is true. From the chapter, "Truth:"
"Truth, then, is essential to knowledge. If what one presumes to know is not "true," that is, if what is believed is not correct about any aspect of reality, it is not knowledge. It may be error, or superstition, or delusion, but it is false, and not knowledge at all.
If what is believed is not certain, that is, if it cannot be known to be true, it is not knowledge. All that induction can provide, at best, are suggestions, surmises, hints, guesses, conjectures, assumptions, possibilities, probabilities, suspicions or hypotheses of what might be true, but can never establish any kind of knowledge.
The attempt to formalize induction, to make it a valid method of science, actually corrupts the true value of induction as a source for true scientific investigation. Consider the so-called varieties of induction:
Generalization is the heart of induction. It is the observation of things which are similar, but its significance is wrongly assumed to be that what things are is established by how often or consistently they are observed. The argument for this view says that if every instance of an observed entity has a specific attribute, it can be generalized that every such entity will have that attribute. If every bird one ever observes has feathers, it can be generalized that all birds have feathers. The fact that all birds have feathers is established on the basis of how many birds observed have had feathers, however, it is established by an entirely different method—concept formation. [See How Truth Is Established below.]
It is interesting that philosophers who regard themselves as empiricists insist that all new knowledge (such as the sciences) must be inductive. It is those same philosopher who insist no certain knowledge can be established by induction. One typical example is the so-called, "black swan," problem. It is true that no certain knowledge can be established by induction, but it is untrue that induction is the method by which new knowledge is established.
[NOTE: The Black Swan Problem is a piece of sophism meant to repudiate knowledge based on the wrong view of induction, which philosophers have mistakenly assumed is the method of science. The so-called problem is explained, "even if all the swans ever observed are white, it cannot be said, "All swans are white." To know that, one would have to look at every swan on earth to be sure there were only white swans. The discovery of just one swan of another color disproves the proposition, "all swans are write." It is usually pointed out this actually happened.
The problem with the so-called, "black swan problem," is, defining swans as white birds was absolutely correct. The discovery of black birds that were similar to swans in every way except color did not disprove that all swans are white. No matter how similar the black birds were to swans, they were not swans, because the concept, "swan," only identified, "white birds."
When the black swan-like birds were discovered, there was no concept to identify them. At that point, based on the similarity of the black birds to swans, the concept swan could be expanded to include the black birds as a subset of swans, or a new concept could be formed (swanoirs, perhaps) to identify black swan-like birds.]
The value of observations of multiple instances of entities with similar specific attributes is that they suggest possible reasons for the similarities which ought to be investigated. It was just that kind of observation that inspired the discovery of genetics, for example.
Causal Inference has all the problems of generalization applied to the relationship between events. The argument for causal inference is that events that are always observed together where one of those events is always followed by the other may be generalized (assumed) to be a causal relationship, that is, that the first event is the, "cause," of the second event. Of course no causal relationship can be established on the basis of how often such relationships are observed, even if the observations were exhaustive. There is no way from observation alone to differentiate coincidence from cause. The bigger problem with so-called causal inference is that is it based on a totally wrong explanation of the nature of cause. [See the chapter, "Cause."]
Analogical induction assumes the observation of similarities in phenomena is a good suggestion for conjecture and hypothesis about other possible similarities in such phenomena, but such observation alone cannot establish such similarities. [Analogy is a method of reasoning outside the scope of induction and is based on the relative and hierarchical nature of existence. See Analogous Reasoning in the, "Thinking And Reason," chapter.]
Statistics of any kind are unable to establish the certain truth of anything. Statistics can only establish the likelihood or probability of things. While some statistic may be correct (the average life expectancy in country xyz is 65, for example), it is not knowledge of anything but the statistic. A statistic is not knowledge about any actual entities or events. No statistic determines when any death will occur or how long any individual will live.
The worst of statistical deceptions are those used to defend notions of cause. Not only are they based on the wrong meaning of Cause, they are based on the false idea of inductive generalization.
Prediction of any actual future phenomena is not possible by any method. Prediction is certainly not possible by induction based on how consistently an event has occurred in the past.
It is simply not possible to know the future for two reasons: 1. there are no isolated events and it is not possible to know every possible influence on any future event, and 2. no specific behavior of any organism can be predicted, especially the behavior of volitional human beings.
1. All strictly physical phenomena are determined by the nature of physical reality, as understood by the physical sciences, physics, chemistry, and biology. Every physical existent is related to every other physical existent and the context of all physical behavior is the entire physical universe. It is simply not possible to know everything there is know about the entire universe at any moment, and, no matter how isolated any event is, it is impossible to predict what in the entire universe might influence any specific event.
Science is not about predicting actual events. All scientific principles identify how an entity will behave in any particular physical context and can specify the limits of that behavior relative to that context. All physical laws imply, "within the limits of a specific context," an existent will have these properties and behavior." [Please see the chapter, "Cause," for the explanation of what scientific prediction means.]
2. What distinguishes living organism from merely physical entities is their behavior. The behavior of physical entities is determined entirely by their physical nature while the behavior of living organisms is determined by the life of the organism. Every organism is a physical entity and as a physical entity its behavior is determined physically, but its living behavior cannot be described in physical terms. [See the chapter "The Nature Of Life."]
Authority and consensus are never sources of knowledge. For every bad or wrong idea ever held or promoted in this world there will be some authority, expert, or leader espousing it. Nothing true can be established on the basis of how many people believe or except it. How this got put over as a form of reason is the work of academics and philosophers promoting some ideology for which there is otherwise not rational foundation.
What Is Induction Good For?
There is nothing wrong with the view of inductive reasoning being a means to good conclusions about what is possible, likely, or worth investigating. The concept of induction is only a problem when it is assumed to be the method by which truth in science, or philosophy, or any other discipline is established.
How Truth Is Established
All knowledge is held in the form of propositions, and all propositions are formed from concepts. [Please see the chapters, "Epistemology, Concepts," and, "Epistemology, Propositions."
Concepts identify existents—ontological, material, natural existents (physical entities, living organisms, conscious organisms, or human beings, or any of their attributes, behaviors, or relationships) and epistemological, psychological existents (all knowledge and knowledge methods, language, mathematics, logic, geometry, all science, history, geography, arts, literature, and inventions). All knowledge consists of propositions that state what is known about what exists. The beginning of all knowledge is the forming of (or learning) the concepts that identify the existents all one's knowledge is about.
An existent is whatever all its qualities (attribute, characteristics, or properties) are, which is how the existent a concept identifies is specified by the definition. What is mistakenly called an, "inductive generalization," is actually concept formation. While a generalization cannot possibly be certain, a concept correctly formed, cannot be uncertain.
A defender of inductive reasoning would say a proposition like, "all birds have feathers," is generalized from the observation that every bird one sees has feathers. That idea is not only untrue, but impossible, because it assumes there is already a concept, "bird," before their attributes are observed and identified. The concept bird does not exist until at least one bird has been observed and the attributes that are unique to it as a bird that distinguishes it from all other things have been recognized. If one sees a bird and notices it has feathers, wings, two legs, a beak, and flies, a combination of attribute not observed in anything else, one may form a concept for that new kind of observed entity, and designates the word (symbol), "bird," for that concept.
When and if one observes other entities with the same attributes as the one identified as a, "bird," they are also birds. It is not necessary to see many birds to form the concept bird, (one is enough), but the concept will apply to every bird (entity with the same attributes) no matter how many there are, one or one billion. A concept is not a generalization, it is an identification by specification. [See the chapters, "Proof," and, "Truth."]
No Problem Of Induction
There is no so-called, "problem of induction." If one has formed the concept bird meaning any entity with the attributes," feathers, wings, two legs, a beak, and that flies," the first time one comes across a flightless bird like a penguin or ostrich, the certainty of their concept is supposedly invalidated. The concept bird, does not deny the possibility of the existence of entities similar to birds. A concept only identifies those entities that have the attributes of the entities it identifies. The concept bird formed before any similar but flightless entities only includes those that fly.
Language is an epistemological tool, a knowledge method invented by human beings. There is no ontological or scientific principle determining how concepts must be formed. If concepts are to provide knowledge they must be formed in the context of what is known and observed.
When the first organism with all the attributes of birds, except flight, is discovered, one must make a choice: include the flightless creature as a subset of the concept bird (flightless birds) or form a new concept for those flightless creatures. Neither choice cancels or invalidates any previous knowledge—birds are still birds. [The same principle applies to the so-called "black swan," problem, which is only a problem for ignorant philosophers.]
All Knowledge Of And About Existents Identified By Concepts
Except for those earliest concepts identified ostensively, all the concepts one knows are by means of the definition witch indicates the actual existents the concept refers to. That definition is a proposition that describes the existents in terms of those existents' qualities, often as a subset of some broader concept. The definitional proposition is the basic knowledge one has about any concept, that is, what the identified existents are and what their basic nature is.
All knowledge is about the existents concepts identify. The first knowledge about any existent is the knowledge of what that existent is, usually in the form of the definition of the concept that identifies that existent. A child who knows the concepts, "dog," and, "cat," knows what those existents are, and probably a little about them such, "dogs bark," and, "cats purr," and a little about what they eat and do. A veterinarian will a great deal about both dogs and cats the child cannot possibly know yet, but it is the very same entities the child's limited knowledge and the veterinarians extensive knowledge are about.
Correct Concepts Are Certain Knowledge
Every correctly formed concept is certain knowledge. Since a concept's only function is to identify existents, any actual existent identified by a concept cannot be anything other than that existent.
From the list of things we know for certain in the "Certain Knowledge" chapter:
—nose, ear, eye, mouth, arm, hand, leg, foot, bedroom, kitchen, bathroom, door, pants, shirts, socks, underwear, dress, slacks, blouse, sweater, water, bacon, eggs, tomatoes, potatoes, onions, grapes, apples, meat, flour, cake, cookie, cracker, hamburger, hot dog, taco, knife, fork, spoon, plate, bowl, glass, cup, saucer, refrigerator, stove, dishwasher, washing machine, sink, bathtub, toilet, shower, car, truck, boat, plane, train, motorcycle, book, television, table, chair, desk, lamp, computer, cat, dog, cow, chicken, pigeon, goat, sheep, pig, and horse.
Each of these words is a symbol for a concept that identifies a kind of entity, and it is certain each of those kinds of entities exist and are what they are. The concepts each word signifies are certain because the existents have the attributes that are the existents identified. The concept would be certain even if there were only one of each kind of entity identified, that is, a refrigerator would be a refrigerator even it it were the only one that ever existed, because it has that attributes of a refrigerator, and cow would be a cow, even if only one cow had ever been observed because it would have the attributes of a cow.
In fact, there are many of each kind of entity identified by the words (concepts) in that list, and every entity that has the attributes of a refrigerator is a refrigerator and every entity that has the attributes of a cow is a cow. The concepts that identify existents can never be wrong because the concepts are formed based on the extent's attributes (and defined accordingly). A potato has the attributes of a potato and every entity that has those attributes is a potato.
[Please see the chapters, "Epistemology, Concepts."]