Truth conditions

Higher levels of inference depend on the truth conditions of the first level antecedent conditions, and thus higher levels of inference involve pattern matching and chained-inference logic. Higher level inference is done in the LISP processor, using various expert system paradigms, while the first level antecedents, which are computationally intensive, are evaluated in the parallel 68010 processor. 

Step 2 Identify the truth conditions of these sentences with their justification conditions. 

Let me run through an example. Suppose the word we are concerned with is dog . Step 1 the sentence This is a dog will be one that plays a part in the fixation of reference. Step 2 the truth conditions of this sentence are identical with the justification conditions, so it is decidable whether it is true on a given occasion. Step 3 dog refers just to those things which are demonstrated by the tokens of this on those occasions when the sentence is true. 

We may now compare this account with the accounts of reference implicit in the major theories of truth. First of all, this account is verificationist. As far as the fixation of reference is concerned, it does not really differ from a verificationist view like Dummetf s. Dummett would probably require though that the reference-fixing justification conditions be fully operational. The major difference concerns what happens after the fixation of reference. Whereas I will use reference to construct genuine, i.e. possibly verification-transcendent truth conditions, a fully-fledged verificationist semantics would construct verification conditions, instead. 

In the previous chapter I described in some detail how reference is fixed. There are simple sentences associated with each word which are responsible for its reference. Once we know the truth conditions of these sentences, the reference of a word is given by disquotational schemas. The truth conditions of these sentences are identical with their justification conditions. In the final analysis, reference is fixed by justification conditions. However, some justification conditions are inadequate, they are not conducive to truth in other words, they fail as truth conditions. What these inadequate justification conditions determine is relative reference the sort of reference we attribute to users of inadequate conceptual schemes when we give rationalizing explanation of their behavior. As to adequate justification conditions, we cannot do better than to identify them with our current justification conditions. Our current justification conditions may turn out to be inadequate. But we must take what we have. 

In 2.2. I have already outlined how we do that. Once reference is fixed, the truth conditions of the sentences which do not participate in the fixation of reference are generated by Tarskian sort of rules. The account as it stands is incomplete in at least two respects. It needs to be shown that the mixture of verificationist account of reference and the Tarskian approach is defensible. I also have to give a much more... 

This is important, because it is in virtue of this that the Tarskian mechanism yields truth conditions which may go beyond verification conditions. Let us see an example of how this happens. Suppose a physicist derives from his theory that a certain elementary particle p has a certain property Q. Suppose further that the justification condition which fixes the reference of includes the description of an elaborate experimental set-up. Suppose that the reference-fixing justification condition for g includes the description of a completely different sort of... 

Step 4 Generate the truth conditions of all sentences by broadly Tarskian recursive rules. 

This may seem incoherent because Step 2 and 4 involve different conceptions of truth. And it may seem circular because the truth conditions, which should come out at the end of the story, occur already at Step 2 so it may seem that truth conditions are explained in terms of truth conditions. 

It becomes clear now why my account is neither circular nor incoherent. It would be circular only if the truth conditions used in Step 2 were the same as the ones which result from Step 4, if Step 2 presupposed Step 4. But this not the case. First, because Step 2 deals with the truth conditions of a small set of sentences, the reference-fixing sentences, whereas Step 4 applies to all sentences. Second, Step 2 identifies the truth conditions it mentions with justification conditions, which renders this use of truth condition independent of the one resulting from Step 4. 

Another reason why they are compatible is that there are cases when they coincide, in the sense that they can be converted into each other. Consider the sentence Joe is smart . Its Tarskian truth condition is... 

In order to show that the verificationist and the Tarskian ingredients present in the account are incompatible, one has to show that the two can differ in the sense that the verificationist and the Tarskian truth conditions depict different and incompatible situations. But this cannot happen, because the justification conditions which figure in the determination of reference are infallible, as it was said in 3.1. It does happen though that we know the Tarskian truth condition of a sentence, but do not know its justification condition. But this is no objection, for it is part of the account. As it was pointed out, Tarskian truth can go beyond what can be verified. This is why the present conception is only modestly verificationist. 

If the truth conditions of a theory consist in all sentences that can possibly be deduced from the theory plus specific assumptions, as a standard view maintains, that would amount to all its possible explanations (postdictions) and predictions. Yet, scientific realism wants a theory to be more, making truth an obscure notion. 

Consider, again, our standard example Water reduces to H2O. I propose the following reconstruction of this sentence s truth-conditions Water reduces to H2O iff for every x, if x is water then (x is water because x is H2O), and being water = being H2O. We should add that no conceptual explanation is involved otherwise, all conceptual explanations, what was briefly gestured at in Sect. 2.2.3, such as this is a drake because this is a male duck , would turn out to be reductive explanations. Mere identity of properties does not suffice for conceptual constitution or analysis, or whatever conceptual dependence consists in - we cannot account for the concept of water in terms of the concept of H2O. Note that one may want to introduce a broader notion of reduction that encompasses both sorts of reduction, conceptual as well as non-conceptual reduction. These kinds of explanation may even bear striking similarities. Here, we are concerned with the non-conceptual case of reductive explanation only. This yields a first explication of the notion of reduction, here restricted to the reduction of types or kinds. 

So, prima facie, some by-explanations figure in truth conditions for at least some reduction statements. We thus arrive at a characterization of the reduction of events (here, again, construed as types) ... 

Possession of the notion of levels of constitutive structures enables us to give truth conditions for reduction statements of the form a reduces to b , where the terms substituted for a and b do not designate theories. 

---