«Abstract: Practical methods are introduced for the construction of definitions, both for philosophical purposes and for uses in other disciplines. ...»
we can choose between a vagueness-preserving and a vaguenessresolving definition. The following are examples of vaguenesspreserving definitions:
In the second of these definitions, the phrase “to the extent that” serves as the defining connective.
The following is a vagueness-resolving definition:
A substance is water-soluble if and only if at least 1 weight-unit of the substance can be dissolved in 10 weight-units of water at 20 °C.
No vagueness-resolving definition of baldness seems to be available, and indeed, none is needed. Generally speaking, vagueness-resolving definitions are primarily used when they are How to define – a tutorial 25 needed for scientific or technical standard-setting or for legal purposes. We have legal rules for how to handle inflammable substances, and therefore we have vagueness-resolving definitions of inflammability. Since we have no special rules for dealing with bald persons, we do not need any vagueness-resolving definition of baldness.
Multi-dimensional vagueness is exemplified by the concept of a “safe car”. There are many criteria that a car should satisfy in order to be counted as safe: it should have proper safety belts and airbags, proper seating for children, a well-constructed crumple zone, an anti-lock breaking system, a pedestrian protection system, etc. It is difficult, arguably impossible, to reduce all these properties of a vehicle into a single, one-dimensional measure. In cases like
this, a definition will have to be vagueness-preserving, such as:
A motor vehicle is safe to the extent that it has features that reduce either the probability of accidents or the effects on human health of any accident in which it may be involved.
9. Defining value-laden concepts Many of the words that we may want to define are strongly valueladen. Their value-ladenness can be uncontroversially positive, hence we all consider “justice” to denote something positive, or uncontroversially negative as in the case of “bureaucracy”. It can also be contested, as in the case of “religion” and “liberal”. These are words that some consider to have positive, others negative connotations.
Value-ladenness of any of these three types tends to be strongly connected to the word and almost impossible to remove.
This is important to observe both in lexical and stipulative definition work. A lexical definition of a value-laden term should have a definiens that is value-laden in the same way, since otherwise the definiens and the definiendum will not have the same meaning.
Hence, a value-neutral definition of “betrayal”, “pseudoscience”, or “accident” would be misleading.
Sven Ove Hansson In particular in the social sciences, attempts are often made to produce value-neutral versions of concepts that are value-laden in everyday language. This is sometimes done by constructing stipulative definitions that assign a value-neutral definiens to a word that is normally conceived as value-laden. Hence, some social scientists have wanted to define “bureaucracy” in a value-neutral way. A value-neutral concept of this nature is certainly useful for social science, but it turns out to be difficult to use the word “bureaucracy” for it. This word tends to retain its negative connotations however much one tries to define it as value-neutral. It is better to use some other word, in this case perhaps “administration”, for the value-neutral concept.
10. Defining controversial concepts Many words and concepts, perhaps in particular those used in politics and religion, are difficult or impossible to define in an uncontroversial way since there are widely divergent views on how they should be used. WB Gallie (1956) proposed that some of these concepts should be seen as essentially contested (essentially contestable). By this he meant that it is in their very nature to be interpreted differently according to the ideology of the interpreter.
The exact delineation of these concepts cannot be unequivocally determined by rational argument. Gallie provided four examples of essentially contested concepts, namely democracy, social justice, art, and Christian faith.
Not all contested concepts are essentially so. The touchstone of essentiality is that if the concept is made uncontested through some modification, then this modification also distorts it so seriously that it is hardly the same concept any more.
Even if a term is essentially contested, it may be possible to conduct a rational discussion on what it means. HLA Hart’s (1961, p. 156) analysis of the concept of justice is a classic example of how this can be done. He maintained that this concept consists of two parts: (1) a constant part, namely the injunction to treat equal cases equally, and (2) a variable part that consists of (competing) criteria How to define – a tutorial 27 for what it means that two cases are equal. John Rawls (1972, p. 5), who makes a similar distinction, calls the common part of the idea of justice the “concept” of justice and refers to variants of the variable part as different “conceptions” of justice. A libertarian and a left-winger can agree that equal cases should be treated equally, but they will have quite different views on what makes cases equal in the relevant sense. The left-winger will consider a rich and a poor person with the same medical condition as equally positioned in relation to a just distribution of medical resources, whereas the libertarian will tend to regard their cases as different in a relevant respect.
This distinction between concept and conception is often highly useful also in the definition of other contested terms. A definition that clarifies exactly what is common, and what differs, between different views on the meaning of a term, can be an important contribution to conceptual clarity.
11. Defining inconsistent concepts Some concepts are said to be even worse than contested: they are claimed to be inconsistent. Age-old examples can be found in the philosophy of religion. According to a common atheist approach to theodicy, the concept of God includes the properties of being omnipotent, omniscient, and perfectly good. Since these properties cannot be combined in one and the same being, it is claimed, the concept of God is inconsistent. Some authors have gone further and maintained that one or other of these three essential properties of God is in itself inconsistent. Most commonly, this has been said of omnipotence. Arguably, an omnipotent being cannot make a stone that it cannot itself lift, thus it cannot be omnipotent after all.
(Anderson 1984. Cf. Puccetti 1963.) Theologians have dealt with these problems by refining the concepts involved. Hence, omnipotence can be restricted at least so that logically impossible feats are not required, and other restrictions can be applied to the other two properties that give rise to the theodicy problem.
Sven Ove Hansson However, the type of blatant inconsistency that threatens in the theodicy problem is relatively uncommon. Most cases of conceptual inconsistency are more sophisticated than that. In particular, some concepts will emerge as inconsistent only in the sense that there are logically possible situations in which they will be inconsistent. (Hansson 2000) The concepts that we use have been tailored to deal with the world that we live in, not with every logically possible world. Many such concepts are “overdetermined” with respect to empirical conditions. When, in real life, x and y always come together, we tend to incorporate this combination into a common concept. Such a practice has obvious advantages in terms of simplicity, but it also makes us conceptually unprepared for analyzing hypothetical situations in which x and y do not come together.
One of the best examples of a potentially inconsistent concept is that of a person. This concept has the two properties that (1) one person can never branch into two, and (2) continued consciousness constitutes identity of person. This combination is unproblematic in real life since both properties hold without exceptions for human beings as we know and conceive them.
However, it gives rise to inconsistency in certain science fiction contexts where persons can be duplicated, with continuity of consciousness preserved in both replicas.
What is the appropriate philosophical reaction to the potential inconsistency of concepts such as that of a person? A radical approach is to discontinue any serious use of them and in particular to block any argument that depends on them. This approach is exemplified by Parfit's (1987 ) appeal to the (potential) inconsistency of the concept of a person when arguing against the moral relevance of persons, and hence in favour of an impersonal ethical theory. The problematic nature of this way of
conducting philosophy was pointed out by Quine:
It would be overzealous to purge the language of potential inconsistencies. We can use a concept in our deliberations about the human condition and the world that we live in even if it would be inadequate in discussions on hypothetical worlds that differ radically from the one we live in. This is an application of the general principle that definitions should be adjusted to the intended usage of the terms we are defining.
References ANDERSON, C Anthony. “Divine omnipotence and impossible tasks: an intensional analysis”, International Journal for Philosophy of Religion, 1984, 15, p. 109-124.
BERLIN, Isaiah. Four Essays on Liberty, Oxford, Oxford University Press, 1969.
GALLIE, W.B. “Essentially Contested Concepts”, Proceedings of the Aristotelian Society, 1956, 56, p. 167-198.
GOULD, James. “Freedom: Triadic or Tripartite?”, Modern Schoolman, 1980, 58, p. 47-52.
HANSSON, Sven Ove. “Essentially inconsistent concepts”, Conceptus, 2000, 33, p. 57-66.
_______ The Structure of Values and Norms, Cambridge University Press, 2001.
_______ “Privacy, Discrimination, and Inequality in the Workplace”, p. 119-135 in Sven Ove Hansson and Elin Palm (eds.) The Ethics of Workplace Privacy, Peter Lang, Brussels, 2005.
HART, H.L.A.The Concept of Law, Oxford, Clarendon Press. 1961.
HURKA, Thomas. “Normative Ethics: Back to the Future”, p. 246in Brian Leiter (ed.) The Future for Philosophy, Oxford:
Clarendon Press, 2004.
LUCAS, J.R. “Discrimination”, Aristotelian Society, Supplement, 1985, 59, p. 307-324.
Sven Ove Hansson MAcCALLUM, Gerald. “Negative and Positive Freedom”, Philosophical Review, 1967, 76, p. 312-334.
MAKINSON, David. “On the Formal Representation of Rights Relations”, Journal of Philosophical Logic, 1986, 15:403–425.
MÖLLER, Niklas, Sven Ove Hansson, and Martin Peterson. “Safety is More Than the Antonym of Risk”, Journal of Applied Philosophy, 2006, 23(4), p. 419-432.
PARENT, William. “Recent Work on the Concept of Liberty”, p.
247-275 in KG Lucey and TR Machan (eds.), Recent Work in Philosophy, Totowa, N.J, Rowman and Allanheld, 1983.
PARFIT, Derek. Reasons and Persons, Clarendon Press, Oxford, 1987 .
PUCCETTI, Roland. “Is omniscience possible?”, Australasian Journal of Philosophy, 1963, 41, p. 92-93.
QUINE, W.V.Review, Journal of Philosophy, 1972, 69, p. 488-497.
RAWLS, John. A Theory of Justice, Oxford, Clarendon Press, 1972.
SUITS, Bernard. The Grasshopper: Games, Life and Utopia, University of Toronto Press, 1978.
WITTGENSTEIN, Ludwig. Philosophical Investigations (transl.
G.E.M. Anscombe), Oxford, Blackwell, 1953.