Page 355 - Encyclopedia of Philosophy of Language

(14)
In 1963, the President of the United States was assassinated in Dallas, Texas. (15)
In 1963, George Bush was assassinated in
Dallas, Texas. (16)
Modality. It is a necessary truth that nine exceeds seven, and it is a fact that the number of planets is nine. Yet (17) is not necessarily true:
The number of the planets exceeds seven. (17)
The truth or otherwise of (17) is not to be determined solely on the basis of the expressions it contains. In fact (17) expresses a contingent astronomical fact. That there are more than seven planets is something which has been discovered through intensive obser- vations and inference. So (20) does not follow from (18) and (19):
Nine necessarily exceeds seven. (18) Nine is the number of the planets. (19) The number of the planets necessarily exceeds
(20)
Besides these, there are many more constructions giving rise to opaque contexts. Just about every cat- egory of expressions contains elements which can cre- ate opaque contexts, for example, adjectives such as 'suspected' and 'alleged,' adverbs such as 'apparently,' and so on.
2. Modalitiesdedictoanddere
Philosophers react differently to the principle of exten- sionality in the case of opaque contexts. It might be argued that there is a reading for (20) in which this sentence does indeed follow from (18) and (19). This reading can be paraphrased as follows: that number which is in fact the number of the planets is necessarily greater than seven. Formally, this can be expressed in a predicate-logical language with a necessity operator D and a possibility operator O added (see Gamut 1991: ch. 3). These operators make formulas out of formulas. The resulting D<p should be read as: necess- arily <p, and O<p as: possibly (p. In this language, the reading of (20) in which it does follow from (18) and (19) translates as (21), whereas the reading in which it does not may be rendered as (22):
George Bush is President of the United States.
Consider the somewhat simpler examples (23) and (24), and their translations (25) and (26):
Necessarily there is something which is greater
than seven. (23)
There is something which is necessarily greater
than seven. (24)
Q3x(x>7) (25)
3xD(x>7) (26)
In (25) the scope of D contains 3x(x > 7), and in (26) it contains x>7. The scope of an occurrence of D may be considered to be the opaque context created by this operator. If the formula within the scope of D is a sentence, i.e., a formula with no free variables, then D is said to be a modality de dicto. As examples, then, one has (22) and (25). If on the other hand there is a free variable within the scope of D, that is to say a variable which may be bound by a quantifier outside the scope of D, then D is said to be a modality de re. Sentences (21) and (26) are examples of this modality. Traditionally, a modality de dicto was seen as an attribution of necessary (or possible) truth to a prop- osition (dictum) and a modality de re was seen as an attribution of a necessary (or possible) property to an entity (res). The traditional distinction does cor- respond to the formal one. In asserting the truth of (25) one does indeed assert that the proposition 3x(x > 7) is necessarily true; while in asserting the truth of (26) one asserts the existence of an entity which necessarily has the property of being greater than seven.
Some philosophers have objected to de re modalit- ies. For them, recognition of such modalities amounts to a revival of 'essentialism,' a philosophical position which distinguishes between accidental and essential properties of things. They have their objections to any such position and therefore reject modalities de re as meaningless and thus useless, at best suggesting to reduce modalities de re to modalities de dicto. One such vigorous opponent of modalities de re has been the philosopher and logician Quine.
Even leaving aside the question of whether recog- nizing modalities de re does indeed lead to essen- tialism, it would seem that a position like this is particularly unsuited to the present purposes. It can be argued that philosophical objections may never be allowed to weigh heavily if the aim is the description of natural language. For the aim is to give descriptions of how in fact we speak, not of how we would have to speak in order to carry the approval of philosophers. It is quite possible that speakers of natural languages do indeed make philosophically dubious assumptions, but that is a fact of life which should not be swept under the carpet of some philosophically more soph- isticated reformulation. Now, that modalities de re occur in natural language seems indisputable. One example is (27):
3x(x = the number of planets A Qx > 7)
(21)
D3x(x=thenumberofplanets Ax>7) (22)
Reading (22) says that in every possible situation the number of the planets, whatever it happens to be, will exceed seven. These two readings (21) and (22) of (20) comply with a distinction traditionally drawn in modal logic between modalities de dicto and de re, a distinction which can be made precise in terms of the scope of D-
Intensionality
333