数学基础——数理逻辑

1. Introduction

本文无预备文章。同时，为了避免逻辑名词翻译的错误，本文用英文书写。

Mathematics is a human activity, driven by human impluses to understand esthetically and to model efficiently. Some objects explored in mathematics are subjective, which perhaps only exist in the pure rational world, such as Banach-Tarski paradox, the collection of all sets and so on. Pragmatists believe these objects are unvalued since the projection of them into the real world is vacuum. But who knows? We (if we are regarded as mathematicians) think we need to add the post-modifier into the sentence, that is at least in the range bounded by the researched real world.
This is a formal article, where we will learn the language and way of thinking of mathematicians, not just calcalation, but the whole system of basic foundations where proofs are built.
We set out to develop a language in which we can write proofs, that is, airtight arguments showing the validity of a statement. In order to do that, all terms of the statement (and of the proof) need of course to be unequivocally defined, which is not as easy to achieve as you may think at first.
If we want all terms involved in a mathematical statement and its proof to be well- defined, we are immediately faced with the problem that common language does not provide any unambiguous definition, that is, all definitions are context dependent and subject to interpretation. Indeed, if you look up a word in the dictionary, it is defined in terms of other words, which you may in turn look up in your dictionary, and this process will eventually turn circular, getting you back where you started. You will never get to foundational terms that are not defined in terms of other words.
It means if you are an alien with no priori intuition about human-kind, since there are no primitive word forming a foundation of a language, you can not learn anything from wikipedia! Hence, in common language meaning depends on ever changing relations among words and is heavily context dependent. The linguistic turn in philosophy during the 20th century has much to do with exploring the consequences of this simple observation. A precise static meaning would require a fixed collection of primitive foundational words that are not defined in terms of other words but “stand on their own.” There are no such notions in common language.
As a consequence, common language cannot be enough for writing proofs and thus we need to develop a deductive theory, that is, an alternative adequate language to formulate our arguments. Any such theory is necessarily founded on primitive notions, that is, undefined notions—for otherwise, a notion could be defined modulo an infinite string of other notions without ever being able to rest on a concept whose meaning stands on its own.
Primitive notions can be any symbol you like, for example, $\mathcal{O},>,\mathcal{Z}$ and so on. But we need to give independent meaning to these symbols. Unproven propositions called axioms declare true certain facts regarding the primitive concepts. The meaning of a primitive notion is thus delineated by the axioms that involve it, that is, the properties we ask of the object are what defines it. Nonprimitive notions of the theory are defined from primitive notions modulo specific syntactic rules. Propositions of the theory are obtained from axioms modulo logical inferential rules.
In Set Theory, which constitutes the foundation of all mathematics, the notion of set is primitive, and so is the membership relation $\in$.
$$x\in A$$
is read as "$x$ is a member of $A$" or "$x$ belongs to $A$". As we mentioned above, the actual meaning of the word set and of the expression “$x\in A$” is determined by the axioms of Set Theory, some of which we will examine later, after we introduce the language of logic necessary to formulate them.

2. Proposition

Although logic is independent to the commmon language you use, no matter 中文 or English, we have the same logical thinking method, but to express the meaning of your thinking, we need to use statement which is combined by our common language. A statement may be true or false, even cannot be clarified by any true value.

Definition 1. A proposition is a statement with exactly one well-defined true value, either true(T) or false(F).

Example 1.

• “She is taller than 5ft 8 inches” is a statement but not a proposition, since this sentence cannot be assigned true value unless who “she” refers to is clearly specified.
• “This sentence has F value./This sentence is false” is not a proposition, since if it is true, then it is false, if it is false then it is true. So this sentence cannot be assigned a true value.
• “This sentence has the opposite value to its true value.” can be assigned by F value, so it is a proposition, and it is false.
• “Life will be extinct on Earth by 3000.” is a proposition, it has a unique true value, although the true value is unknown until 3000.

We will often consider propositional variables $p,q,r,$ etc., that is, propositions that are not specified, beyond the fact that they are propositions, hence are either true or false (and not both). We use them as arguments (that is, as variables) of logical connectives to form new propositional forms, that is, formulas that include propositional variables and become propositions if the truth values of each propositional variable is assigned.

3. Propositional connectives

Definition 2(Negation). Suppose $p$ is a proposition, then "It is not the case that $p$" is another proposition, denoted by $┐p$.

Definition 3(Disjunction). Suppose $p,q$ are propositions, then "$p$ holds or $q$ holds" is another proposition, denoted by $p\vee q$.

Definition 3(Conjunction). Suppose $p,q$ are propositions, then "$p$ holds and $q$ holds" is another proposition, denoted by $p\wedge q$.

In terms of truth tables, we have

$p$	$q$	$┐p$	$p\vee q$	$p\wedge q$
T	T	F	T	T
T	F	F	T	F
F	T	T	T	F
F	F	T	F	F

If you are familar with the proof techniques of mathematics, you may ask how to prove the truth table, but please note that we do not have any proof for it, these constructed proposition and their assigned true value is just assigned artificially, they cannot be proved in the frame of logic, and what we can do is just believe it.

Exercise 1. Show that $(p\vee q)\vee$