The properties of something may be extrinsic or intrinsic. All objects have extrinsic properties except everything, and all objects have intrinsic properties except atoms.
There are several different ways to say what an apple is, or to give a description of it. One alternative is to define it functionally: it is something to eat, or something that grows on trees. These are extrinsic properties of an apple, because they depend on the apple's relationship with other things. The apple can also be defined intrinsically, by defining some characteristic property of apple-matter. That property is attributed to the apple itself; it is relatively independent of the apple's relationship with other things.
The distinction between intrinsic and extrinsic properties is important to keep in mind when describing the conditions for identity between two objects. Twins, for example, may be intrinsically identical if they have the same physical appearance and parts (under the assumption that different material of the same type is identical). They are not extrinsically identical, however: at the very least, their spatial positions are distinguishable.
A close approximation of this distinction between intrinsic/extrinsic properties is the distinction between interior/exterior properties.[17] For example, a property of apples such as “being eaten by people” is an extrinsic property. Properties such as “being eaten by worms” are also probably extrinsic properties, even if the worm is inside of the apple (which makes it clear that the matter is not always clear-cut). One might argue that the boundaries of the apple do not include the worm; in any case, the essential piece of information that makes a property extrinsic is its dependence on some other thing.
Given the relatively holistic presentation in this book, extrinsic properties are emphasized. A thing has extrinsic properties in virtue of its relation to other things. The intrinsic properties of a thing, therefore, can be defined as the extrinsic properties of that thing's parts. Accordingly, it is clear that a thing which is not a part of something else does not have extrinsic properties, and a thing with no parts does not have intrinsic properties. The two things satisfying these criteria are known as the universe and the atom, respectively. In terms of location (instead of properties), since there are no objects outside of the universe, there are no objects in terms of which to define it: hence, it cannot have an extrinsic definition. At the other end of the continuum of size is the atom: a thing with no parts cannot have an extrinsic definition between its parts, therefore it does not have an intrinsic definition.[18]
Intrinsic properties characterize the parts of a thing.
Intrinsic properties describe or define an object in terms of its parts. This method of definition is exploited by reductionism. For example, to understand the behavior of an individual reduces to the sciences of physiology and psychology (which describe the activities of the brain). Psychological understanding reduces to the science of biology, which studies the activities of the neurons that constitute the brain. Biology, in turn, reduces to chemistry or physics, which studies the molecules that make up the neurons.
Ultimately, this reduction results in a very detailed explanation, but not necessarily an increase of explanatory power. The big is not caused by the small, just as the parts are not caused by the whole. They are simply different levels of description, both of which are a valid description of reality (albeit descriptions of reality that deal with differently sized or shaped parts). While a description that uses small parts may be more detailed, it is also more complicated. So, while it is possible to describe a person by using a physical description that corresponds to the movement of their molecular parts, it is not necessarily of great benefit. In fact, the description of an individual in terms of various neurotransmitters might be substantially less useful than a physiological description, since we know how to affect physiological change more easily (which of course has an effect on chemicals in the brain).
Extrinsic properties characterize the whole of which a thing is a part.
Extrinsic properties can be investigated in the linguistic domain by examining the meanings of words (or more specifically, morphemes, lexemes, and phrases). The symbolic equivalent of the extrinsic definition of an object is the definition of one symbol (or phrase) using other symbols. The intrinsic description of the corresponding concept would be an analysis of its constituent words or morphemes. Both of these definitions occur in most dictionaries, which provide both the etymology of a word and the definition of that word using other words.
An interesting test for the extrinsic identity of symbols is known as linguistic substitutability. If two different words or phrases are able to be used in the same context (i.e. the same position in a given sentence), then they are linguistically substitutable, which most often implies that they are of the same type, or that they can play the same role. For example, ball and boy are substitutable in the following sentences (in that they do not change the meaning of the larger context), but bucket is not:
Kick the ball.
Kick the boy.
Kick the bucket. ( understood as a synonym for "to die" )
The last example is interesting because it demonstrates that “bucket” in this context is not a semantically complete thing. The idiomatic expression “kick the bucket” means to die, which is not a compound that involves the meaning of the word “bucket” . The semantics of “bucket” is irrelevant in this context: the word “bucket” acts like a phoneme instead of a morpheme.
The context of a word can determine one of several definitions of that word. More precisely, the single word is called a homonym , and the multiple words (i.e. those with different meanings) are called lexemes . To illustrate this, the preceding examples can be altered as follows:
We had a ball.
We had a boy.
We had a bucket.
“Having a ball” might connote either having a fun time or having a round toy in this context, which illustrates that the homonym “ball” contains at least two lexemes. In the second phrase, having a boy probably connotes that we have given birth to a child, which illustrates that the verb “to have” is a homonym. Finally, the word “bucket” in this context, as opposed to its context in the previous example, is once again a complete noun, meaningful on its own (or at least meaningful to a greater degree).
There are at least two different ways to understand homonyms. Under one understanding, a single word may contain multiple definitions, each of which is complete. Under another, the single word contains an incomplete definition, which can only be completed in context. And just as the definition of a word may be intrinsic or extrinsic, there are numerous things which are similarly incomplete or ambiguous without a larger (clarifying) context.
Properties characterize the relations of a thing.
The creation of parts is a process which is necessarily relativistic: a part depends on its counterpart, or the complement of that part, for its definition (in particular, its extrinsic definition). To state the matter slightly differently, when characterizing a thing with properties or attributes, the complement of that thing is also characterized. When everything is divided into something and not -something, something has a certain characteristic property in light of which the division is possible in the first place. The not -something, on the other hand, does not have that property: further, the not-something has the not-property.
For example, consider a table. Now, imagine a part of that table: a table-top thing, which is composed primarily of the surface of the table. In virtue of (conceptually) creating this part, a complementary thing has been created: the legs of the table (i.e. the remainder after the partition). The fact that it is possible to distinguish the table top from the rest of the table implies that it has some property that the rest of the table does not: let us say that it has the property that we can put drinks on it. Because the object has this characteristic property, the complementary object has a complementary property, i.e. the legs of the table have the property that drinks cannot be placed on them. If they did not have this complementary property, then the basis for creating the dichotomy in the first place would disappear (assuming that the table or legs do not have other characteristic properties, which in reality they certainly do).
Under this analysis, the creation of an object is analogous to naming one part of a divided thing: creation is a division in addition to a collection. Every time we create something, we implicitly create at least two somethings. Neither thing is ontologically prior to the other, although we often name only the object on one side of this boundary (which side to name is most often a pragmatic decision). For example, within the context (or superset) of fruits, some subset may be designated as “apple” : there is no (simple) designation for the object which is materially constituted by “all fruits that are not apples” . So the latter object must be referred to by a complex expression, by negating that which has been named: “non-apple” . Note that this negation (or complement-formation) requires that we know the whole from which the part was created: a non-apple in the context of “all food” means something other than a non-apple in the context of “all fruits” .
It is a mistake to see non-apple things as only lacking in something: possessing the property of being “not-apple” is every bit as characteristic as the concept of apple. Of course, “being a not-apple” may be a less useful piece of information compared to “being an apple” , because it is a characteristic of a comparatively large number of things. Still, having a property has no more reality than having the opposite of that property, just as the thing apple has no more reality than the thing not-apple (note that we are not talking about the symbols apple and “not-apple” , where one is a compound word and the other is not). The creation of a decision boundary results in two things, each of which has a characteristic property. Again, which object is named or labeled, as opposed to which object is referred to through negation, is a pragmatic concern. Similarly, the difference in formulation between having a given property and having the negation of that property is a feature of references, not of the things to which those references refer.
Additionally, whether an object possesses a property or not depends on the counterpart of that object. As a concrete example of this conceptual relativism, consider whether “strong people require weak people” . In particular, imagine a woman who is strong (e.g. someone in a gym who is lifting a heavy weight). Suppose that her ability remained roughly constant, while everybody else on the planet started weight training, and became capable of lifting weights that she could not lift. If we still called her strong, there would be nobody to call weak anymore. Perhaps we would call everyone else “super-strong” , but it seems more likely that we would not call her strong; we would call her weak (even though her ability did not change).
If we do not change the label which we assign to her, the semantics of that label have to be greatly altered. Although we may continue to call her strong, she was relatively strong, and she is now relatively weak. In either case, she is not in control of being weak or strong. Calling her strong depends on other people; it is a relative judgment that depends on the whole of which she is a part. Superman is not super compared to his friends from planet Krypton; he's Regularman.
Some people might maintain that certain attributes of a thing are not relativistic in this sense, or that some attributes have semantics which do not depend on that thing's complement. One example that the scientifically-minded might raise is the mass of an object: the mass of an apple does not depend on the mass of a banana, does it? The banana is not directly used to compute the mass of the apple, but the measure of a thing is always taken with respect to something else. For example, suppose that the mass of the apple is expressed in kilograms. A kilogram is defined as the mass of a certain volume of water at sea level.[19] It is by definition relative; the primary difference between this and the previous example about a given person's strength is that in this case the comparator is a single object (a certain volume of water), whereas in the last example the comparator is a number of objects (other people). Although some choices of measurement may allow consistent application to a wider range of phenomena, there is no a priori reason to use one comparator as opposed to another.
Some people may object that the strength of a person may change, but the mass of a specific volume of water does not. To know that the mass of an amount of water does not change, however, we have to weigh it (let us suppose that it weighs one kilogram). In other words, it weighs as much as some other object that weighs one kilogram. If the mass were to change, all we know is that the mass of other objects must have changed at the same time. So we cannot conclude that the mass does not change in an absolute sense, but only that it does not change with respect to something else (unless this is how we define absolute change in the first place).
This relativistic viewpoint is closely related to a conundrum proposed by Henri Poincare: if the size of the world doubled overnight, would you notice it? If you assume that all other masses and laws of physics were adjusted as necessary, it is not possible to tell the difference (whether such an undetectable difference is in fact a difference at all is left as an exercise for the reader).
[17] We should note that this is a relatively simplistic characterization: there is a large body of literature examining the difference between intrinsic/extrinsic properties.
[18] Since everything is all-inclusive, it can have no extrinsic properties: there is nothing outside of it with which to relate it. Neither does it have intrinsic properties, as it is considered as a whole (i.e. unless we are subdividing it, we cannot give an intrinsic description). So, we say that it is beyond description (and apologize for describing it by saying that it is beyond description).
[19] At least this was the original definition of the kilogram: the mass of a liter of water at sea level. Since that mass can vary, the kilogram was subsequently defined in terms of International Prototype Kilogram, a particular object located in France. Since the mass of that object has also been found to vary, the kilogram is currently being redefined.