The Nature of Knowledge


The following is a brief overview of the essential elements of knowledge, which is followed by an extended explanation of the true nature of concepts.


A concept consists of two "components" with a specific "function." The components of a concept are a "perceivable existent," and a "specification." The function of all concepts is to identify things. The perceivable existent is a word, its specification is a definition which isolates or differentiates (from all other things) what the concept identifies. What a concept identifies is what it means. A concept's meaning, however, is not part of the concept.

[NOTE: What a concept means is what it identifies. A definition only indicates exactly what it is the concept identifies, that is, what the concept's meaning is. A concept does not mean its definition, it means whatever the definition indicates it means. The meaning of a concept is what it identifies, and therefore not part of the concept.]

The only function of a concept is to identify existents. Existents can be anything: entities, events, qualities, relationships, concepts or ideas. If what is being identified is a specific single existent, the concept is a particular concept; if what is being identified is a category or class of existents, the concept is a universal concept. Most knowledge is in terms of universal concepts.

[NOTE: The distinction between particular and universal concepts is less important than is often implied. In one sense, all concepts identify classes or categories of existents, even those for categories which identify only one existent or particular. Such categories are universal in the sense that what they identify is the same entity on all occasions.]


Though words can be written and read, spoken and heard, even signed and recognized, words, as components of concepts, are "mental"—words as we think them. Written and spoken words are only marks on paper or sounds that represent the words we think. We first have to think a word before we can say or write it and a written or spoken word only has meaning if someone reads or hears it and mentally recognizes it.

Written and spoken words are symbols for the words we think, but words, as we think them, (except when we are thinking about them as words, their spelling for example), are part of concepts. It is words that provide the part of a concept we can be conscious of, that is, since perception is the only kind of consciousness there is, we are conscious of a concept by perceiving the word part of it.


Since the function of a concept is to identify existents, the purpose of a definition is to indicate which existents a concept identifies. Since the identity of any existent is all its qualities, a definition specifies those qualities or attributes of an existent or class of existents which distinguishes (differentiates) them from all other existents.

Concepts As Identification

If asked what he would like, a young boy wanting an apple might point to the apples in a bowl and say, "I'd like one of those, please." If there are no apples in plain sight, however, he might say, "I'd like an apple, please." The pointing and using the word apple have performed the same function—they have identified the kind of thing the boy wants.

It might seem like a quibble to emphasize it is not the word apple, but "using the word" apple that is the identifying action, because a spoken or written word is only a symbol, and it is the concept (symbol plus definition) that does the identifying. When the boy uses the word apple, he has already identified apples mentally before saying the word, else he could not say it. It is also not the spoken word that identifies an apple for the listener but the concept, apple, which hearing the word recalls to the listener's mind.

The identity of the apple is independent of anyone's knowledge or understanding of it. The apple's identity is a metaphysical fact consisting of all an apple's qualities and attributes, known or unknown. To identify an apple by means of the concept "apple" does nothing more than pointing and saying, "one of those," does. The obvious advantage of a concept is that identifying an apple does not depend on the presence of a physical one, because we can identify an apple anytime by means of the concept, we do not have to have any apples before us to think about them.

Concepts and Meaning

What a concept identifies is what it means. The concept "apple" means any apple there has ever been, is now, or ever will be. As the identifier of apples, what it identifies is the entire metaphysical nature of apples in their entire metaphysical context, because that is what apples are. That nature and context are not part of or in any way contained in the concept, apple; the nature and context pertain only to what the concept identifies, that is, apples themselves. Therefore, what the concept apple means is apples with all their qualities and attributes and all that can ever be known about them.

The concept, apple, used by a child who knows little more about apples than what they look and taste like, or the same concept used by a botanist specializing in the study of apples, means exactly the same thing. Neither the child's limited knowledge or the botanist's extensive knowledge is about the concept apples; the knowledge is about that which the concept identifies for both the child and the botanist, actual apples.

Concepts, Propositions, and Truth

Except by implication, no concept is either true or false. Concepts can be good or bad, that is, they may identify confused ideas, or be vague and poorly defined, or even identify what does not materially exist, (as though it did). What those concepts identify are fictions, but the concepts are neither true nor false. A concept only identifies things, and is just as valid when identifying fictional things as when identifying actual things.

Only propositions can be true or false. A proposition is a statement (sentence) that asserts something about an existent or class of existents. For example, "Zeus is a god worshiped by the ancient Greeks," asserts something about Zeus. If what is being asserted is correct, the proposition is true; if what is being asserted is incorrect, the proposition is false. The assertion, in this case, and therefore the proposition, is true, even though the concept "Zeus" identifies a fiction. The same concept can be use in both true and false propositions. "The phoenix is a common bird found in the forests of Colorado," is false, but, "the phoenix is a mythical bird of ancient Egypt," is true. See the chapter, "Propositions."]

Since only propositions can be true or false, knowledge consists entirely of propositions; but all propositions are constructed of concepts, without which no knowledge would be possible. Concepts identify the existents all our knowledge is about. Technically, concepts are not knowledge, but a definition, if correct, is knowledge because it is stated as a proposition.

One might say, all correctly defined concepts constitute a kind of knowledge, but notice, it is really only the definitions that are the knowledge, not concepts as identifiers, which is their only function. Concepts imply knowledge, and most concepts would be impossible without knowledge, but ascribing knowledge to concepts themselves is an epistemological mistake. It is that mistake that is the source of such confused ideas as those that suggest knowledge somehow changes the meaning of concepts, so that what a child means by an apple, and what a botanist means by an apple would be different things.

The True Nature of Concepts

Ayn Rand, Dr. Peikoff, David Harriman, and Dr. Binswanger write a lot about how children form their earliest concepts but all they describe is extremely unlikely. Children seldom, if ever, form their own concepts (nor do most adults). All a child's concepts are learned from parents, other adults, teachers, friends, and what they hear, see, and read.

How concepts are formed is not very important to epistemology. What describes a correct concept is extremely important, because that is the means of judging whether the concepts we have are valid or not, and on those occasions when one must form a concept, understanding what a correct concept is must be the guide to that creation.

[NOTE: The emphasis on how things come to be is a kind of philosophical mistake. However something comes to be does not change the nature of that thing, and seldom elucidates that nature. It is that mistake, I believe, behind the view that our philosophical understanding of things requires appeals to evolution or history, for example.]

Baby's First Words

A baby's earliest words, which we recognize as words, are similar to (and are obviously attempts to mimic) our own. Babies obviously learn some words from observation, but most of them are taught. "Where's Tina's nose?" we say while touching Tina's nose. "Where's Tina's ear?" touching Tina's ear. "Kitty!" we say, guiding Tina's hand to touch the family pet.

When, later, Tina says, "kitty!" upon seeing the family cat, we conclude she has learned the meaning of the word "kitty." Has Tina now a concept "kitty" for a cat?

If the components of a concept are a word and a definition, what is the definition of "kitty" that Tina understands? It is obvious Tina would understand nothing about animals in general, or the meaning of felix domesticus. "Kitty," to Tina is the thing that looks and feels like the family cat.

A "definition" only has to indicate or isolate that which a concept means, that is, what it identifies, from all other things. At what point Tina goes from simply repeating the word "kitty" when she sees the family cat to making the connection that it is "kitty" we do not know, but it is that connection that changes the word "kitty" from a mere word, a sound she can repeat, into a spoken symbol for the concept "kitty."

[Note: Tina does not have to do anything, mentally, to "isolate" Kitty from other things. Tina recognizes Kitty because Kitty is already isolated from all other things by its its own attributes, both as a cat, as well as the specific cat that belongs to Tina's family. The fact that it is a thing's own attributes that differentiates it from all other things is an ontological fact not learned until much later.]

At first, "kitty" is a particular concept, the identification of a single entity, the family cat. There is a sense in which even particular concepts have a universal aspect, however. If the neighbor's cat comes into Tina's view, if it is similar enough to her own family's cat, she will probably use "kitty" to identify it. Though Tina has not yet learned the concepts of similarity or multiplicity, her concept "kitty" means, "the kind of thing that looks like that," even though she does not know there is more than one of "that kind of thing." For Tina, the same cat on different occasions, or different cats (if similar enough) are the same thing.

It will not be until Tina happens to see two cats, perhaps her family cat and the neighbor's, at the same time, that she might first conclude that "kitty" identifies more than one thing. It will be on such an occasion that Tina will have a "universal" concept.

Two kinds of Concepts

A universal concept identifies things, not as particular individual things, but as members of a class or category of things. It is the difference between Tina's "kitty" identifying the family cat and her later more sophisticated use of the word "kitty" to mean "one of those kinds of things." Most of our concepts are universal.

The concept "dog" for example, means any dog there has ever been, is, or will be, real or imaginary. There is a difference, however, in a concept like "dog" and a concept like, "Athenian," meaning a citizen of Athens. The difference is something called "essence."

"Athenian," like, "dog," refers to any Athenian there has ever been, is, or will be, real or imaginary, but a dog is a dog because that is it's metaphysical nature; an Athenian is not an Athenian because it is his nature, but because Athens is where he lives. He would have the same nature (human) if he lived elsewhere.

The reason a dog is a dog, rather than something else, is inherent in the dog itself; it is the kind of being it is; the reason an Athenian is an Athenian is not inherent in Athenians, they would be the same kind of being if Londoners or Parisians. Philosophers call what is inherent in dogs that makes them dogs their essence. The problem for philosophy is to explain exactly what essence is. Most of the explanations have been disasters that have plagued philosophy with errors that make a rational epistemology impossible.

Another View of Concepts

The only function of a concept is to identify the existents that it is the concept for.

[Note: The meaning of a concept, which are the actual existents it identifies are also called a concept's referents, units, or particulars.]

If the concept is a particular concept, the identification is accomplished by any definition that succeeds in describing what a thing is, and isolating the specified existent from all others. Since the identity of any existent is all its qualities (attributes) it is those attributes that both identify the existent and isolate it from all others. The technical term for that isolation is differentiation. If the concept is a universal concept, it must not only identify what makes those existents different from all other existents, but must also identify what makes those existents members of the same class or category of existents, that is, what is the same about them.

For universal concepts like Athenian, cook, or trinket, the sameness is something external to the existents identified by the concepts and is determined by things such as where one lives, what one does for work, or what something is used for. I identify those kinds of universal concepts as extrinsic concepts.

For universal concepts of things that are members of the same category because of something inherent in the existents themselves, like dog, planet, rock, or elephant, it is their very nature as existents that is the same. I identify those kinds of concepts as intrinsic concepts.

Kinds of Things and Essence

What makes things the same kind of things? More importantly, what makes anything the kind of thing it is? The answer to the first question is their essence; everything with the same essence is the same kind of thing. It is the answer to the second question that is the explanation of what essence is. The answer to that question, is epistemological for extrinsic concepts, but metaphysical for intrinsic concepts. That difference is one of the most important concepts in epistemology.

The essence of existents identified by extrinsic concepts is something known about those existents which is why the essence is epistemological. The essence of existent identified by intrinsic concepts is their metaphysical nature, which is why the essence is metaphysical.

Essence and Qualities

Every existent identified by a universal concept has two kinds of qualities: necessary and possible.

Necessary qualities are all those qualities of an existent that it must have to be the kind of existent it is, and without which it would not be that kind of existent.

Possible qualities are all the qualities an existent may have, but will be the same kind of existent whether it has any of those qualities or not.

[Note: Necessary qualities also exclude any qualities that would make an existent a different kind of thing as well as determining what qualities are possible to the existents.]

A dog is a physical entity, a living organism, an animal, a mammal, and a canine. Physical, living, animal, mammal, and canine are all necessary attributes of a dog. If something had all these attributes except one, it would not be a dog.

A dog can have short hair, long hair, or no hair, a tail, no tail, be very big, or very small and exhibit individual patterns of behavior. Hair, tails, a certain size and variations of behavior are all possible attributes of a dog, but none are necessary; a dog will be a dog with or without them.

From ontology we learn the identity of every existent is determined by three necessary conditions: 1. it must have some qualities, 2. it must be different in some way from all other existents; therefore, it must have some quality or qualities that are different from those of all other existents, and 3. every existent has some relationship to all other existents; therefore, it must have some quality or qualities it shares with all other existents.

Obviously it is an existent's necessary qualities that determine the kind of existent it is and equally obvious that every existent with the same necessary qualities is the same kind of thing.

The essence of a thing, therefore, consists of all its necessary qualities. A dog is a dog because it shares with all other dogs the same necessary qualities, that is, those that are a dog's nature or identity.

For there to be more than one of any kind of thing, everything that exists must have some quality or qualities that are different from all other existents.

The difference in existents of the same kind must be differences in possible qualities.

Every dog has some possible quality or qualities that are different from the possible qualities of every other dog. No two dogs can have exactly all the same possible qualities. (The differences can be pronounced or very slight.)

The essence of any existent is all the necessary qualities it shares with all existents of the same kind, that is, all those identified by the same concept. Every unit or referent of that concept will have some different possible quality or qualities from all other units of that concept.

"Referents" of Universal Concepts

A referent of a universal concept is also called a particular or unit of a concept.

The necessary qualities of any single referent of a concept are all the qualities it has at any moment. The identity of every individual thing, then, is all the necessary qualities it shares with all other existents identified by the same universal concept, and all possible qualities it has at any moment.

The possible qualities of any individual existent, are those that can change over time, or have at some times and not at others. The necessary qualities of an individual existent are those that cannot change without that existent becoming a different kind of existent. For any individual existent, at least one of its necessary qualities, as that existent, must be a possible quality of all other existents identified by the same concept, and it must be different from the possible qualities of all other units of that concept.

All of an individual existent's necessary qualities are its essence as an individual unit.

For all intrinsic concepts, the essence is metaphysical, or more specifically, ontological.

For all extrinsic concepts, the essence is epistemological (based on what is known about existents such as their purpose, use, function, membership in some collection, residence or relationship to other things). If the units of an extrinsic concept are material, the material aspects themselves have intrinsic essences, but as units of an extrinsic concept, the units intrinsic material qualities, are only necessaryqualities if they cannot be changed (e.g. a Musician's human attributes are necessary qualities) but are possible qualities if they can be different in different units or referents of the concept (e.g. a drinking cup may be glass, ceramic, plastic, or some other material).

For example, I own many pairs of chop sticks. They are all units or referents of the concept, "chop sticks." Their essence as chop sticks is, "two straight sticks used as implements for eating." The necessary qualities are, "sticks" that are "straight" with the exact quantity "two" with the function of being used "to eat with." I'll call this essence, "chop-stickness."

The essence,"chop-stickness," is strictly epistemological. While sticks are physical entities, and straight is a physical attribute and those qualities are necessary to chop sticks, "chop-stickness" is not inherent in chop sticks, because straight sticks can have many other functions (cue sticks or drum sticks, for example) and those qualities are attributes of many other kinds of existents. Notice it is the function, "to eat with," combined with the physical attributes (two straight sticks) that differentiates chop-sticks from all other things, including eating implements of other shapes.

I have chop sticks that are made of wood, others made of bamboo, and still others made of plastic. Wood, bamboo, and plastic are all intrinsic concepts; the essences of wood, bamboo, and plastic are ontological, the necessary attributes of these substances are physical. As any of these essences is instantiated in chop-sticks, they are only possible qualities, however, because chop sticks can be made from many different substances and still be chop sticks.


An existent's identity is all it's qualities which both determined what that existent is and differentiates it from all other existents. Except for those things described abstractly, it is almost impossible to know all of an existents qualities, and attempting to "define" concepts by means of listing all the qualities of a concepts referents would be epistemologically impossible.

There are two practical limits to all definitions. 1. They must be made within the limits of one's knowledge, and 2. they must include such qualities of the referents that one can recognize as those referents. For a boy, 'red, shiny, round, sweet, fruit,' would suffice as a definition of an apple. (It will have to change when he learns about plums and pomegranates.) For a botanist, the technical taxonomic definition would identify the same thing.

[NOTE: It is not quite enough for a definition to only differentiate the referents of a concept from all other things. The main purpose of a definition is not to indicate what a thing is not, but what it is. I used the the generic phrase, "that one can recognize," to indicate what qualities of a thing must be included in a concept's definition, because only such qualities of an existent that are required for one to comprehend what existent or kind of existent is identified by a concept are required, because all other qualities will be implied by them.]

The Importance Of Intrinsic versus Extrinsic Concepts

The failure to distinguish between intrinsic and extrinsic concepts results in bad epistemology.

The following is an example of a wrong hypothesis of concepts, which is nevertheless, perhaps the best after Peter Abelard's. It is an example of Objectivism's failure to identify the true nature of concepts. It is the same as Dr. Binswanger's.

Ayn Rand did not use the word essence, nevertheless, that is what is being described. I'll summarize that description:

[NOTE: Nevertheless, in Introduction to Objectivist Epistemology, Rand does define the word essence: "Objectivism holds that the essence of a concept is that fundamental characteristic(s) of its units on which the greatest number of other characteristics depend, and which distinguishes these units from all other existents within the field of man's knowledge." She uses essence with regard to the concept, not the existents themselves and regards it as strictly epistemological.]

Measurement-Omission Theory of Concepts

The units or referents of a concept are the same kind of existents because they are similar. Similar means partly the same and partly different. What is different about existents of the same kind is more or less obvious, the difficulty is determining what is the same about them. Rand's explanation is that two or more things are similar when their qualities (or attributes) are the same; what is different about units of the same concept is the magnitude or "measurement" of those shared qualities. This is called, "measurement omission." Things are the same kind of things (have the same essence) when they share the same defining attributes with the measurement of those attributes omitted. (Omitted does not mean discarded, it means unspecified.)

This explanation differs from the correct one at two points. First, it makes no distinction between intrinsic and extrinsic concepts. Second, it accounts for only one kind of possible qualities, magnitude.

The first difference is important because most of our concepts are extrinsic. The only intrinsic concepts are those of physical substances, naturally occurring physical entities, organisms, naturally occurring physical events, relationships, and attributes, and all concepts derived from these, such as attributes of physics, biology, medicine, ethics, and similar concepts.

All other concepts are extrinsic, including all man-made physical existents, activities, and services. Almost all extrinsic concepts are defined by their purpose, function, relationship to things or places, or the use or purpose of the units subsumed by the concept, none of which have magnitude as an attribute. [A tire is a tire. Tires may have different sizes, and, within limits, different shapes and those differences might be measurable, but those shapes and sizes are not necessary qualities of tires, because tires are defined extrinsically. It is what they are used for that makes them tires, and there is no measurement of "tireness."]

Consider Rand's own example which is intended to describe the formation of the concept "table."

According to the, "measurement-omission" theory of essence, the formation of the concept table begins with the observation that tables are differentiated from other things by their distinctive shape: a flat surface with supports, for example. The concept table leaves out the measurement of geometric characteristic of a table, but retains those geometric characteristics; for example: the concept leaves out the shape of the surface (round, square, octagonal), the number of legs, height, and weight, but retains the fact it must be some shape, have a leg or legs, be some height, and weigh something.

What actually distinguishes a table from all other things is not its shape, but its function. The possible number of different shapes a table can be is infinite, and for almost all those shapes, there are other kinds of furniture and objects with exactly the same or very similar shapes. In some parts of the world there are tables that look more like beds, and beds that look more like tables—if shape alone were what distinguished tables from beds they could not be distinguished. A table and a bed might have exactly the same shape (and measurements); what distinguishes them is not their shape but what they are used for.

The second difference is interesting, because, for a small subset of intrinsic concepts, Rand's explanation seems to work. The "measurement-omission" explanation does work for that subset of existents which are strictly defined by their geometric nature or quantitative attributes. A triangle is a triangle because it has three sides, and every triangle has those same qualities (attributes) and the difference between triangles is strictly the difference in the lengths of their sides, and of course the resulting difference in their corresponding angles. But notice, even for such a simple concept, there is not one defining attribute the measurement of which is omitted, but six, three sides and three angles, and though they are mathematically interdependent, they are each a separate measurable attribute.

[NOTE: Triangles are actually epistemological concepts from the field of geometry, but there are entities which are triangles such as certain drafting tools and musical percussion instruments. They happen to be identified by extrinsic concepts, however.]

It is, in fact, difficult to name something that is defined in terms of measurable attributes that can vary. From grains of sand to heavenly bodies, from water to the most complex chemical substances, the defining attributes are not variable measurable attributes. Water, for example, has many attributes and those attributes are necessary qualities. For example, "water is a clear, colorless, odorless, and tasteless liquid essential for most plant and animal life and the most widely used of all solvents." The attributes, "colorless, odorless, and tasteless," cannot be measured because they are absolute. A liquid either has those attributes or it doesn't. Definitions of water often include the fact water is a compound of hydrogen and oxygen (but so is hydrogen peroxide) so the formula h2o is included. None of these attributes is measurable. In fact, even the measurable attributes of water (weight, freezing and evaporation temperatures, coefficient of expansion, etc.) are identical for all water so cannot be the measurable attributes from which the measurements are omitted.

The fact that it is not variable measurable attributes that define concepts is true of all substances, with the possible exception of substances of which there is more than one variety, such as isotopes, in which case, each of the isotopes cannot be identified in terms of variable measurable attributes, because whatever measurable attributes they have must be identical for all examples of the isotope.

It is, according to the measurement omission theory, measurements that can vary between the referents of a concept that are a concept's defining attributes. It does not matter how many measurable attributes something has, if all the referents of a concept must have the same attributes with the same measurement, those attributes cannot be what distinguishes referents of a concept from each other.

This mistake is partly the result of the failure to distinguish between intrinsic and extrinsic concepts. Since "table" is an extrinsic concept, as are all concepts for man-made things, (See note), its essence will not be comprised of measurable necessary qualities, though it may have measurable possible qualities.

[NOTE: Some man-made things do have intrinsic essences. These are mostly substances, such as pigments, chemical compounds, drugs, metals, and specific kinds of plastic, though, generically, "plastic" is an extrinsic concept.]

[Note: An automobile is an extrinsic concept, determined primarily by an automobile's use. But an automobile is a real metaphysical object. Part of the essence of an automobile will be that it is a motorized physical object with wheels, but that cannot define an automobile. The defining "essence" is what an automobile is designed and used for.]

Rand was very close. The units or referents of a concept all share the same essence, that is, the same necessary qualities, with their "differentiating possible qualities left out." Rand's shortcoming was that she only identified one kind of differentiating possible qualities, magnitude (of measurable attributes), which might work for one small subset of existents, geometric objects, which are not entities however, having only epistemological existence.