Introduction: The Myth of Human Language

49^ Here is the point where my view of modulation breaks from Recanati’s, for he does seem to take the absolutely sharp meaning as being privileged, and we modulate from that.

50^ In the context of dynamic semantics, an utterance of a sentence would be supertrue iff it is true on every sharpening that resulted in the proposition expressed being determinately either true or false. This is added to account for the S-admissibility of sentences like ‘if five feet is long then six feet is long’ which seem to be determinately true whether the meaning of ‘long’ is determinate or not.

51^ Alternatively you might think that in such a case we are morally on the same page but just in disagreement about the best way to achieve some moral end. Imagine two utilitarians arguing about which strategy yields the greatest number of utils. Is this a moral dispute? I would have thought not, but there is no point in making an issue out of this.

52^ I am indebted to John McFarlane for conversations here.

53^ The proviso that it “plays a direct role in the derivation of the conclusion” is designed to allow us to ignore terms that have multiple occurrences but that are inert in the structure of the argument. (I am indebted to an anonymous reviewer for pointing out the need for this proviso.)
So, for example, consider the following argument form:
(P v Q)  R

Q  S

P .

R
The terms in Q have multiple occurrences, but they do not play a role in the derivation of R. Matters are a bit more complicated than this example lets on, however, because sometimes there are multiple ways to derive the conclusion in a sound argument. For example:

(P v Q)  R

Q

P .

R
We can either define soundness relative to a derivation (so that derivations are sound, not arguments), or we can say that the form of the argument is sound if there is at least one derivation path to the conclusion such that the argument consisting of just that derivation path respects the Dynamic Lexicon Constraint on Soundness. Other solutions are, most likely, available. See van Deemter and Peters (1997) for a survey of articles on the general problem developing logics that tolerate ambiguity, indeterminacy, and by extension underdetermination.

54^ If you wish, you can (following recent work in dynamic logic) treat each of the conditional premises as instructions for updating the common ground. In this case the argument wouldn’t really be an argument as above, but rather a series of instructions on modulating the meaning of ‘bald’ in the common ground. I remain officially neutral on this way of executing the idea.

55^ But see Armstrong and Stanley (forthcoming) for criticism.

56^ This view, or something very much like it, is advocated by Stine (1999).

57^ For a proposal of this nature, see Barnett (2009).

58^ Other experiments involve the naming of tangrams. See Wilkes-Gibbs and Clark (1992), Metzing and Brennan (2003).

59^ See Richard (1990) for a similar point about translation. Kripke, it should be noted, takes his point to extend beyond cases where we translate to cases of disquotation, but as we saw in section 5.1, the theory of the dynamic lexicon also leads us to be cautious about disquotation. The Disquotational Principle does not hold in cases where meaning shift has taken place.

60^ What happens to these senses when the sentence is evaluated in other possible worlds? Presumably one needs a theory of modal discards here. In other words, counterfactual evaluation calls for modulations that thin out the sense content of the word.

61^ Notice that here I am understanding ‘meaning’ to be inclusive of sense content.

62^ Heck(2002; 3) allows that this may be understood in a weak way: “On the weakest interpretation of (3), it speaks of ‘determination’ only in a mathematical sense: it claims only that senses are related many-one to references.”

63^ When this happens – when the metaphor “dies” – we don’t fossilize the invitation to compare categories of things, we rather (as noted by Stern 2000) see the effect of metaphors qua demonstratives. To illustrate this, consider a dead metaphor like ‘mouth of the bottle’. The initial modulation of this phrase invites us to consider important underlying properties shared between the mouth of an animal and the opening in the bottle, but once the modulation is taken on as routine it becomes a predicate the range of which includes openings in typical bottles as canonical instances. It now merely denotes these objects. Is the invitation to compare part of the meaning of the metaphor or is it pragmatic? The dynamic lexicon approach is fundamentally neutral on this approach I believe.

64^ If you find it problematic to think of the simple predicate ‘is a midwife’ has this robust thematic structure, you can invent a verb which characterizes that which the midwife does – ‘midwifing’ for example. Thus a midwife is someone who midwifes, and midwifes midwife states of affairs in which an agent brings about a result using an instrument. Following considerations in Larson (1998) I happen to believe that nouns like ‘midwife’ have very complex internal predicate structure that encodes these relations.

65^ I am forever indebted to Barry C. Smith for finding this quote.