CLARIFICATION
a. Design Functions
The basic idea is that %’s design function is what x was designed to do. But this is not yet quite right, since my stopwatch was designed to resist water. Yet that is not its function.
Design functions, as well as those of the other two types, involve the important idea of a means of doing something. The designer designs x to be or to serve as a means of doing y. This mousetrap was designed by its designer to be a means of catching mice. But my stopwatch was not designed to be a means of resisting water. (I shall have more to say about the concept of means in section 6.) Alternatively, x may be a part of or belong to or otherwise be present in a “system” S which was designed by its designer to do y by means of x. If so we say that the (design) function of x in S is to enable S to do y. The function of the gasoline in this engine is to enable the engine to run. So we can write(1) If x was designed to be or to serve as a means of doing y then xs function (at least one of them) is to do y.
(2) If S was designed to do y by means of x then x s function in (with respect to) S is to enable S to do y.
Should design functions be restricted only to cases in which x (or S) is designed? A chemical can be produced, a person appointed, a rock placed where it is, to be or to serve as a means of doing something. If so we can speak of the chemical's, person's, or rock's function, even though these items were not themselves designed. Nevertheless, design was present. Perhaps then (1) can be generalized as follows:
If x was produced (created, established, appointed, placed where it is, etc.) by design to be or to serve as a means of doing y then x’s function is to do y.
A similar generalization of (2) would be possible.
However, with the inclusion of these additional activities all done by design we encounter cases in which x might have been designed for one end, produced for another, and placed where it is for a third.
Suppose that the Universal Design Company designed a certain type of bolt to serve as a means of bolting a car engine to the frame. The Ford Motor Company purchases this design but produces the bolt in its own shops to serve as a means of bolting the torsion bar to the frame. The bolt is then sent unattached to the dealers who are to install it when they receive the car. The dealers, however, ignoring the previous designs, affix the bolt to the wheel. What is the function of this bolt? Does it have three functions? Each of the agents noted—the designer, the producer, and the placer of the bolt—might well claim that the bolt has only one (proper) function and disagree over what this is.In reply we might proliferate types of functions still further by talking about production functions, placement functions, and so forth. But the more natural way to settle this dispute is to recognize that function statements do not always need to take the simple form “the function of x is to do y,” which philosophers tend to focus upon. We also make statements of the form “the function x was designed to serve is to do y,” “the function x was produced to serve is to do y,” “the function x was placed where it was to serve is to do y,” and so forth. Each of the three protagonists in our bolt story can then be understood as making a different type of function claim. When an agent from the Universal Design Company indicates the function of that bolt he means to indicate the function it was designed to serve; the agent from the Ford Motor Company is talking about the function it was produced to serve; the dealer is speaking of the function it was placed where it was to serve.
My solution to this dispute, then, is to recognize not different types (or senses) of function but different types of function statements. Several of the latter can be grouped together under the heading “design function statements,” all of which involve the general idea of design in origin or placement. One of their basic forms is this:
(3) The function x was designed (produced, created, established, appointed, placed where it is, etc.—all by design) to serve is to do y,
which is true if and only if x was designed (produced, etc.) to be or to serve as a means of doing y.
If x does not itself do y but S does we can write(4) The function x was designed (produced, etc.) to serve is to enable S to do y, which will be true if and only if x was designed (produced, etc.) to be or to serve as a means of enabling S to do y.
My claim now comes to this. A sentence of the form
(5) The function of x is to do y (or to enable S to do y)
is ambiguous. Someone who utters it may be making one of a number of different claims about design in origin or placement some of which are given by (3) and (4). But not all. The second chapter of Ayer's Language, Truth, and Logic, entitled “The Function of Philosophy,” states that the function of philosophy is to provide analyses of certain sorts or “definitions in use.” When Ayer wrote this he was aware that this is not the function that philosophy is produced or written by most philosophers to serve. Nor is his claim to be taken simply as one about the function of analytic philosophy. Rather he is indicating the function that (all) philosophy ought to be produced or written to serve. Accordingly, another type of design function statement which someone uttering (5) might mean to be asserting is
The function x ought to be designed (produced, etc.) to serve is to do y (or to enable S to do y),
which is true if and only if x ought to be designed (produced, etc.) to be or to serve as a means of doing y (or of enabling S to do y).
One can, of course, answer a question of the form “What is x’s function?” by uttering sentences other than (5), for example, by saying that x is an A (a church, a mousetrap). In such a case x is not designed to be a means of being an A, although the notion of means is not irrelevant here. If that building is a church then its function is to enable Christians to worship together. And the latter claim, which can be understood as being of type (4), will be true if and only if a corresponding means statement is.
b. Use Functions
When someone utters a sentence of form (5) he may mean to be saying something not about x’s origin or placement but about its use, e.g.,
The function x is used to serve is to do y (or to enable S to do y),
which will be true if and only if x is used as a means of doing y (or as a means of enabling S to do y).
Someone who says that the function of the human hands is to grasp objects may be saying that this is the function they are used to serve. And this will be so if and only if the hands are used as a means of grasping objects.Although what function an item has been designed (or produced, etc.) to serve often coincides with the function it is used to serve, sometimes it does not. Something may have been designed or even placed where it is to serve as a means of doing y although it is never in fact used, or although it is used only as a means of doing z. The function that trough may have been designed and placed where it is to serve is to water the pigs, even though it is never used or the only function it is used to serve is to water the flowers.[104] Moreover, x may be used to serve a given function without this being so by design. The function a mosquito's wings are used to serve is to enable the mosquito to fly. We need not say that the mosquito uses them by design or that they were designed or created by design.
Use function, like design function, statements can be prescriptive. Someone who claims that the function of a college education is to arouse intellectual curiosity and not just to get a job might mean that this is the function it ought to be used to serve, not that it is in fact used to serve this. There is a related use of function sentences of form (5). Pointing to the wings on a particular mosquito I might say that the function of those wings is to enable the mosquito to fly, even though in this case the wings are broken or for some other reason the mosquito never flies. My claim, then, could be that
(6) The function those wings are supposed to be used to serve is to enable that mosquito to fly.
And I may make such a claim because I believe something about the function of wings in mosquitos generally, namely, that
(7) The function a mosquito’s wings are (generally) used to serve is to enable the mosquito to fly.
The general statement (7) provides a norm on the basis of which the function statement (6) about a particular x can be asserted.
On the other hand, pointing to those particular broken wings I might also say that they have no function at all, since they cannot be used.c. Service Functions
One who utters (5) might mean to be claiming something stronger than simply what function x was designed or is used to serve, namely,
(8) What x in fact does, the performance of which is a function that it serves, is y (or to enable S to do y).
Thus we might say that although the function the regal chair was designed to serve is to seat the king, and the function it is used to serve is to block a doorway, the only thing the chair actually does, the performance of which is a function it serves, is to draw crowds to the palace. We might say this because crowds are in fact drawn to the palace by means of the chair and this is of benefit, while the king is in fact not seated, nor is the doorway really blocked, by means of that chair.[105] More generally, a statement of form (8) is true if and only if
(9) y is in fact done by means of x (or S is in fact enabled to do y by means of x), and either x was designed (produced, etc.) to be or to serve as a means of doing y (enabling S to do y), or x is used as a means of doing y (enabling, etc.), or y’s being done confers a good.
A number of points must now be made about the last clause in (9). It would be a mistake, I think, to follow those writers who require only one specific type of good such as survival and reproduction. As Sorabji (1964, p. 293) notes, a certain creature might have an organ which shuts off sensations of pain when lethal damage has been done to its body. We might then identify this as the function of the organ even if shutting off the pain does not increase the creature's chances of surviving and reproducing. Analogous examples are possible against views committed to other specific goods.
Possibly, those who select one type of good such as survival and reproduction do so because they think of this as an ultimate good—as worthy of having for its own sake; other states of affairs can be beneficial only because they contribute to the ultimate good.
But even if we distinguish ultimate and intermediate goods it is doubtful that there is a unique ultimate good with which all functions can be associated. Another reason for selecting survival and reproduction as the only good is that the authors that do so are concerned almost exclusively with biology. The function of the nose, they will say, is to breathe, not to hold up eyeglasses, although both are done by means of the (human) nose and their being done confers a good. To this the correct response is that the nose serves or can serve a number of different functions not all of which are biological. Whether doing y is classifiable as a biological function will depend at least in part upon the type of good it confers, upon what, and how; however, I shall not here try to offer criteria for such functions. If holding up eyeglasses is not a biological function of the nose it is nevertheless an important function of the nose in the eyeglass “support system.”Doing y need not be of benefit to x itself or to its owner or user but to something else. The sickness and subsequent death of various animals in a species may be of no benefit to them, but to other members of the species, who will now have fewer competitors for food. So the sickness and death of these animals can have a function for the species, even if not for the animals themselves.
Doing y can confer a good upon a person S even if doing y is not something which S knows about or regards as beneficial even if he knows about it. One of the functions of basic training in the army is to teach recruits to obey orders, even if the recruits do not realize that basic training is teaching them to do this and even if they do not want to be taught or regard this as beneficial. One who believes that basic training is of value in this way can make such a function statement.
By contrast, one can also make a service function statement without committing oneself to values, if any, in that statement. Suppose that watching television dulls the mind to the problems of the world and people act as if so dulling the mind is beneficial for them. For the television viewer, one might say, what television does, the performance of which is a function it serves, is to dull the mind. One can make such a statement from the point of view of the television viewer with his set of values. On the other hand, employing one's own more intellectual values one might say that for the television viewer what television does, that is, dull the mind, serves no function at all, since it confers no benefits. The speaker can make explicit which point of view—which set of values— he is employing, or this can be clear from the context. This means that speakers with quite different values can still agree over the truth of those service function statements based on the conferring of a good, provided that they invoke the values of the party in the function statement, not their own.
Finally, it should be clear, I am not maintaining that all function statements, or even all service function statements, commit one to the claim that doing y confers a good. We can utter a true service function statement of form (8) by saying that what that button on the sewing machine in fact does, the performance of which is a function it serves, is to activate the exploding mechanism, even if activating this mechanism is of no benefit at all.
5.
More on the topic CLARIFICATION:
- References
- SALIENT FEATURES
- “The Pragmatic Character of Explanation” is reprintedby permission from Philosophy of Science Association Symposium Proceedings 1984 (1985), 275-292.
- PREFACE
- RuskintheRomantic
- Conclusion
- Is Chemical Research Ethically Neutral?
- CONCLUSION
- The Meaning of Welfarism and Non-welfarism
- Conclusion: Directions for Future Research on the Geographies of Palestinian Children