TECHNICAL FUNCTIONS

One common way to describe artifacts is in terms of how they technically function. In a telephone sound is transformed into electronic signals that are then transmitted over some distance and transformed back into sound by another telephone. Such technical functions are strongly related to human uses. Telephones are designed and built so that they can be used for transmitting the human voice over distances well beyond its normal range. Because references to technical functions are often the basis for assessing human uses of artifacts, and insofar as such assessments express certain values, the relation between technical functions and uses is an issue for any ethics of technology.

Judging Actions and Artifacts

All intentional human behaviors or actions are subject to normative judgments. These judgments are of two sorts: deontic and evaluative. Deontic judgments express what one ought and ought not to do or what one has reasons for doing. Evaluative judgments describe something as good or bad. Using an artifact is subject to these types of judgments, in the first place because it is a form of action. It is generally wrong, for example, to hurt another person with a knife, which is merely a specification of the judgment that one ought, generally, not to hurt someone.

Additionally, however, the use of artifacts is subject to judgments that relate directly to the particular function of the artifact. For instance, one may say that it is wrong to use a Phillips screwdriver to open a paint can. Assuming that the attempt to open the can is itself perfectly in order, the wrong here is not morally wrong but instrumentally or functionally wrong: Using the Phillips screwdriver will not smoothly lead to the desired outcome. Typical for artifact use, such judgments may be translated, so to speak, to the artifacts themselves. An artifact is said to perform its function well or to function poorly or to malfunction. One can also say that a particular artifact, in the prevailing circumstances, ought to do such-and-such a thing. Even natural objects can, in a context of use, be subject to such judgments, for instance when one says that a particular stone is a good stone to use as a hammer.

Functions

The use of the term function in the previous paragraphs sets aside a considerable philosophical debate about the meaning of functions, one that has taken place largely in relation to the analysis of functions in biology (the function of the heart is to pump blood) and the social sciences (the function of religion is to create social cohesion).

Briefly there are two major competing concepts of functions: system functions as first stated by Robert Cummins, and proper functions as first stated by Larry Wright (1973) and further analyzed by Ruth Millikan, Karen Neander, and others. According to Cummins (1975), who is primarily concerned with biological systems, something has a function insofar as it contributes to the capacity of some system. According to Millikan, by contrast, the proper function of an organ or system is what helps to account for the survival and proliferation of its ancestors (1993). Millikan aims for a theory of functions that applies to artifacts as well as organisms.

Against these attempts to bring all uses of the notion of function under a single theory, Beth Preston (1998) argues for a pluralistic theory of functions that includes Cummins's system functions and Millikan's proper functions. Wybo Houkes and Pieter Vermaas (2004) hold that theories of artifacts are overly function-oriented and that a theory of artifact functions can be derived from a theory of artifact actions. For Preston, as well as for Houkes and Vermaas and for many others, functions often become the locus in both science and technology for the uniting of deontic and evaluative judgments.

Uniting Deontic and Evaluative Judgments

There is, however, no consensus of what precisely unites deontic and evaluative judgments insofar as they jointly comprise the realm of the normative. One account proposed by Joseph Raz (1999) and Jonathan Dancy (2005) holds that normative facts are facts expressing how other facts—natural or positive facts—matter to the question how to act. The deontic judgment that To do X is right then expresses the normative fact that there is the positive fact of X possessing certain features, and that these features are such that, in the circumstances at hand, the balance of reasons points toward the doing of X. The evaluative judgment that X is good similarly expresses the normative fact that the features of X are such that one has reason, perhaps even a compelling reason, to adopt a certain positive attitude toward X. What this positive attitude could be depends on the nature of X.

In contrast to a lack of clarity concerning the normative in general, there is wide agreement among philosophers that instrumental value should be sharply distinguished from moral or ethical value. To see the difference between these two forms of value, consider the statement: This is a good knife to kill Mrs. Robinson with. The knife is instrumentally good as a means to an end, but the end, the killing of Mrs. Robinson, is morally bad. One has reason to disapprove of Mrs. Robinson's violent death, and ought to prevent it. But given that the killing of Mrs. Robinson is sought, to do it with this knife may be considered a good and recommendable choice. Instrumental value is therefore in a sense conditional: It concerns the fitness of a particular means to the realization of an end once that end is given, whereas it is not concerned with any pros or cons regarding the end itself.

The distinction between moral value and instrumental value is closely related to a distinction among the sorts of reasons that back up an act or an attitude or a belief. If means M is fit to end E such that one ought to choose M or choose to do M, this concerns an ought on rational grounds. By contrast, if M is morally good or bad, such that one ought to approve or disapprove of M, this concerns an ought on moral grounds. This way of distinguishing rational grounds from moral grounds sees the notion of rationality exclusively as instrumental rationality. Not all philosophers will agree, however, that rationality should be viewed thus.

Designing and Using Artifacts

The design and use of artifacts is involved with both kinds of grounds for normative judgments, but in particular cases it is not always obvious whether one or the other kind is at issue. Malfunction judgments and judgments of poor or proper functioning certainly have a special relation to considerations of rationality. A statement such as Artifact A malfunctions expresses the positive fact that A does not or will not show the behavior it was designed to show. However, this positive fact does not exhaust the meaning of the statement. It also seems that when an artifact malfunctions or functions poorly, human beings by definition have a reason not to use it, or at least not to use it as designed, on rational grounds. One cannot go as far as saying that the notion of malfunction or of poor functioning semantically implies that the item ought not be used. There may be reasons such that, on balance, it is rational to use the thing anyway. But if one applies the judgment of malfunction prior to any considerations of use, as a mere factual statement of the artifact's failure to show a certain behavior, it makes no sense to then ask whether that fact means anything about what one will do with the artifact. When an artifact is said to malfunction, one necessarily has at least a reason not to use it as designed.

Similarly to say that a particular artifact functions well is not just to say that it shows a certain behavior, as a positive fact, regardless of anything that one might do with it. This judgment implies that the item shows a particular behavior and that one has a reason to use it as designed. In this case, however, the conditionality of instrumental reason really has a bite: One has an overall reason to use something to produce the result that using the artifact in question produces. If one does not have a reason to use a car in the first place, because one is not going anywhere, then neither does one have a reason to use this particular car, which happens to be a very good car.

Whether one also has a reason on moral grounds not to use a malfunctioning or poorly functioning artifact, or even ought not to use such an artifact on moral grounds, is a question that raises different issues. The judgment might be motivated, for instance, by fear that the artifact's use would pose a hazard for other people. But such judgment often depends on the particular case at hand and thus is not covered in the meaning of malfunction. It is hardly worthwhile to discourage someone from using a Phillips screwdriver to open a paint can on moral grounds.

The rationality of artifact use depends critically on knowledge. To judge that the use of a particular object is the best means to achieve a certain goal requires an adequate knowledge of the object's properties and the effects of manipulating it in the prevailing circumstances. The use of the object can be rational only to the extent that the user's beliefs about the object are rationally justified. Rationally, in this sense, refers to epistemic rationality, and not practical rationality, which was the form of rationality relevant in the preceding considerations. In practical rationality, the issue is what it is best to do, or what one has a reason to do, given one's end of realizing a particular situation. For epistemic rationality the issue is what it is best to believe, or what beliefs one has a reason to adopt, given the end of holding as many true beliefs as possible or holding only true beliefs.

Proper Use and Good Design

When someone uses an artifact in disregard of its designed function, that is, according to some privately conceived use plan, reasons of epistemic rationality seem all that matter. (The concept of function as a use plan is developed at length by Houkes, Vermaas, Dorst and de Vries, 2002.) When the artifact's use fails to have the desired result, there is no one to blame. This is no longer true when an artifact is used for its designed function, in circumstances that are consistent with the artifact's use plan as explicated in the instructions for use. When handing over an artifact to a client who ordered it, or to the market, the designer/producer is committed to the veracity of the predictions made about the artifact's behavior. These predictions have the force of a promise, and the commitment accordingly has the character of a moral obligation. One could say that a designer ought, on moral grounds, to be epistemically rational. In practice the extent of this obligation is articulated in the form of standards that say how much research and testing is sufficient to vindicate the claims that are to be made about the artifact's performance.

It is part of the human condition that neither the criteria of epistemic and practical rationality nor the criteria of moral obligations can guarantee the realization of plans. One may be disappointed by fellow human beings as well as, metaphorically speaking, by nature. The ubiquity of uncertainty shows in the use of language when one says that a particular artifact ought to do something when handled in a certain way. This may express the idea that one is epistemically justified in one's belief that the artifact will perform as expected, given the amount of research and testing adopted in designing or in repairing the artifact, but at the same time there is a recognition that there is always the possibility that something was overlooked. The statement may also express the idea that one has a right to the artifact's performance, on the basis of a promise by a designer/producer, retailer, or repair service person, while there is at the same time the awareness that such promises are occasionally broken.

It seems natural that in what is summarily described as good design the grounds distinguished above play a role. An artifact that can be termed a good design must be instrumentally fit for its function in a range of plausible circumstances. However a well-designed artifact must also be one that it is morally vindicated to use. This can either mean that it is not likely to lead to outcomes of low moral value, for instance by being safe, or that it is likely to lead to outcomes of high value, which will often be a comparative matter.

Thus the features of a particular artifact may give rise to reasons, even compelling reasons, for its use in order to contribute to the realization of one's goals, by which such artifact is instrumentally good. It may also have features such that one has reasons, even compelling reasons, to approve and promote its use, by which it is morally or ethically good. Additionally, artifacts are often judged on the basis of a third criterion, previously not discussed, namely aesthetic appeal. Technical artifacts may have both instrumental and ethical value, or both instrumental and aesthetical value, or even all three. Some trash receptacle for public use may not only function perfectly as a trash receptacle, but it may also encourage people to use it to a larger extent than another type of trash receptacle, and on top of that be considered a beautiful object.

MAARTEN FRANSSEN

SEE ALSO Engineering Design Ethics;Engineering Ethics.

BIBLIOGRAPHY

Audi, Robert. (1989). Practical Reasoning. London: Routledge. An introductory text on practical rationality.

Cummins, Robert. (1975). "Functional Analysis." Journal of Philosophy 72(20): 741–765. The classical text for the theory that an object's function is its contribution to the overall behavior of a system of which it is a part.

Dancy, Joseph. (2005). "Ethical Non-Naturalism." In The Oxford Handbook of Ethical Theory, ed. David Copp. Oxford: Oxford University Press. Proposes, among other things, a general characterization of normativity.

Fransen, Maarten. (2005). "The Normativity of Artefacts." Studies in History and Philosophy of Science 36(x). Analyzes the interrelations between the various normative statements in which artifacts figure and their grounding on criteria of rationality and morality.

Grice, Paul. (1991). The Conception of Value. Oxford: Oxford University Press. A general text on the notion of value and evaluation.

Houkes, Wybo, and Pieter Vermaas. (2004). "Actions versus Functions: A Plea for an Alternative Metaphysics of Artifacts." The Monist 87(1): 52–71. Argues that use is a more basic notion to make sense of artifacts than function. Discusses (ir)rational and (im)proper use of artifacts.

Houkes,Wybo; Pieter Vermaas; Cees Dorst; and Marc de Vries. (2002). "Design and Use as Plans: An Action-Theoretical Account." Design Studies 23(3): 303–320. Develops the notion that the function of an artifact has to be understood from the role it plays in a use plan by which an individual seeks to produce a certain end.

Korsgaard, Christine. (1996). The Sources of Normativity. Cambridge, UK: Cambridge University Press. A general philosophical text discussing various conceptions of normativity.

Millikan, Ruth Garrett. (1993). White Queen Psychology and Other Essays for Alice. Cambridge, MA: MIT Press. This is a more accessible statement of a view first argued in Millikan's Language, Thought, and Other Biological Categories: New Foundations for Realism, Cambridge, MA: MIT Press, 1984, that all functional statements, including those concerning the function of language and the meaning of sentences, an be grounded in terms of evolutionary biology.

Preston, Beth. (1998). "Why is a Wing like a Spoon? A Pluralist Theory of Function." Journal of Philosophy 95(5): 215–254. Argues that only a pluralist theory of function is able to do justice to biological as well as artifact functions.

Raz, Joseph. (1999). Engaging Reason: On the Theory of Value and Action, 2nd edition. Oxford: Oxford University Press. An important text, first published in 1975, on the relation between value and reasons for acting.

Wright, Georg Henrik von. (1963). The Varieties of Goodness. London: Routledge & Kegan Paul; New York: The Humanities Press. Discusses the interrelations among a broad spectrum of evaluative terms.

Wright, Larry. (1973). "Functions." Philosophical Review 82(2): 139–168. First statement of the view that an object's function is what is causally responsible for its existence, which underlies later evolutionary accounts such as Millikan's.