FALSE? [N] = YES if N, seen as a TNT-string, IS a false statement of
number theory; otherwise NO.
(e.g., FALSE? [666111666] = NO,
FALSE? [223666111666] = YES,
FALSE? [7014] = NO)
The last seven examples are particularly relevant to our future
metamathematical explorations, so they highly merit your scrutiny.
Expressibility and Representability
Now before we go on to some interesting questions about BlooP and are led
to its relative, FlooP, let us return to the reason for introducing BlooP in
the first place, and connect it to TNT. Earlier, I stated that the critical mass
for Godel's method to be applicable to a formal system is attained when all
primitive recursive notions are representable in that system. Exactly what
does this mean? First of all, we must distinguish between the notions of
representability and expressibility. Expressing a predicate is a mere matter
of translation from English into a strict formalism. It has nothing to do with
theoremhood. For a predicate to be represented, on the other hand, is a
much stronger notion. It means that
(1) All true instances of the predicate are theorems;
(2) All false instances are nontheorems.
By "instance", I mean the string produced when you replace all free
variables by numerals. For example, the predicate m + n = k is represented
in the pq-system, because each true instance of the predicate is a theorem,
each false instance is a nontheorem. Thus any specific addition, whether
true or false, translates into a decidable string of the pq-system. However, the
pq-system is unable to express-let alone represent-any other properties
of natural numbers. Therefore it would be a weak candidate indeed in a
competition of systems which can do number theory.
Now TNT has the virtue of being able to express virtually any number-
theoretical predicate; for example, it is easy to write a TNT -string which
expresses the predicate "b has the Tortoise property". Thus, in terms of
expressive power, TNT is all we want.
However, the question "Which properties are represented in TNT?" is
precisely the question "How powerful an axiomatic system is TNT?" Are all
possible predicates represented in TNT? If so, then TNT can answer any
question of number theory; it is complete.
Primitive Recursive Predicates Are Represented in TNT
Now although completeness will turn out to be a chimera, TNT is at least
complete with respect to primitive recursive predicates. In other words, any
statement of number theory whose truth or falsity can be decided by a
BlooP and F100P and GlooP 417