Summary
This lesson opens the second unit of the course, devoted to deduction. It presents its five lessons and the thread that joins them, from the definition of a formal deductive system to De Morgan's laws and the distributive laws; it places the unit on the path of the course, between language, already built, and semantics and metatheory, which are to come, and anticipates what these will say about it; it justifies the choice of an axiomatic system, which is built in two stages (Łukasiewicz's system, with three axiom schemas and modus ponens, and the course system, which adds to it the rule of replacement of double negation), and explains why that second rule is needed and why it is postulated instead of proved; it shows with a short deduction how much work an elementary piece of reasoning hides; it specifies what is taken from the previous unit, namely unique readability, induction (strong induction on the natural numbers and structural induction) and the unfolding of abbreviations, and gives advice for studying it.
Learning objectives
- Describe the sequence of the five lessons of the unit and explain which results of each one the next one uses.
- Place deduction on the path of the course, distinguish derivability from semantic consequence and explain what soundness and completeness will assert.
- Explain why the course system adds to Łukasiewicz's a second primitive rule, the replacement of double negation, starting from what the three schemas can and cannot see in a joint denial.
- Read a short deduction by numbered lines, distinguish the deduction of formulas from deduction about deductions, recognize what the unit takes from the previous one and organize its study starting from the entry assessment.
Entry assessment
Before working through the unit, it is worth measuring the starting point. This assessment is a self-assessment: it is not graded, and its only purpose is to guide your study. Its thirteen questions are of two kinds. The first three concern what the unit takes for granted from the previous one: unique readability, strong induction and the unfolding of abbreviations; anyone who misses one of them should review, before going on, the lesson that the question names. The remaining ten cover, in order, the five lessons of this unit: three the first, three the second, two the third, one the fourth and one the fifth. Anyone who has never studied a deductive system should not expect to get them right, and that does not matter: each answer, right or wrong, explains the point and names the lesson that deals with it. The same skills are measured again at the end, in the synthesis lesson.
Before starting the unit, measure your starting point. It is a short self-assessment: it is not graded and its result is stored only in your browser.
Question 1From earlier unitsThe string is a compound formula, of the form . What are its components and ?
is not a formula: a connective disk requires two formulas after it, and here it is followed only by the variable ; its accumulated weight is , not . Nor is the rest, two variables in a row, a formula.The unique readability theorem rules this out: if and the four components are formulas, then and . A cut at another place leaves at least one part that is not a formula.A compound formula has two components, and the empty string is not a formula. Moreover, the string that follows the initial disk has three variables and a single connective disk, and every formula has exactly one variable more than connective disks.After the initial disk, the weights of the words are accumulated: the connective disk subtracts , and the variables and add each; the accumulated weight reaches for the first time at the end of , and there ends. The formula is , and by unique readability it admits no other decomposition: this is what makes it possible to recognize without ambiguity the antecedent and the consequent of an implication when applying modus ponens.Question 2From earlier unitsIn a proof by strong induction that every natural number has a property , what may be assumed when proving ?
That is the hypothesis of simple induction. It suffices when case rests only on the previous one; when it rests on arbitrary smaller cases, like modus ponens on the lines of a deduction, all of them must be assumed.It includes , which is what is to be proved: assuming it makes the argument circular.Strong induction has no separate base case: it contains it, because for there are no smaller natural numbers and the hypothesis offers nothing. Only when the argument uses, for example, must the first cases be treated separately.It is the hypothesis of strong induction. The unit uses it to prove the deduction theorem: the last line of a deduction can be obtained by modus ponens from any two earlier lines, not necessarily from the immediately preceding one, and the hypothesis must reach both.Question 3From earlier unitsWith the official abbreviations of the course, which formula does designate, written with joint denial and negation?
The outer negation is missing: disjunction is the negation of joint denial, not joint denial itself, which is read “neither… nor”.That is the disjunction : in implication, the inner negation falls on the antecedent, not on the consequent.Each abbreviation is defined only once and designates a single string, which the unfolding determines. What does happen is the converse: one and the same string can receive several names.Implication abbreviates , and disjunction abbreviates the negation of a joint denial; hence . The unit unfolds the abbreviations in this way whenever it needs to know whether two expressions designate the same formula.Question 4Schema A1 is . Is the formula an instance of A1?
That substitution gives , not the proposed formula: the last occurrence of must also be .The outer form is not enough: the schema also requires that the last component repeat the first.Arbitrary formulas, variables included, can be substituted for metavariables: is an instance of A1.An instance substitutes formulas for the metavariables uniformly: the same formula for all the occurrences of . Distinct metavariables may receive the same formula, but one and the same metavariable cannot receive two.Review: Formal deductive systems
Question 5A deduction already contains the lines and , where and are distinct formulas. Which formula does modus ponens applied to those two lines allow one to write?
Passing from the consequent to the antecedent is the fallacy of affirming the consequent, not modus ponens. The first lesson proves, with a marking that the axioms and modus ponens preserve, that is not derivable from in any way.It is the converse of the implication written, and no rule produces it from that implication: an implication is not reversed.Nothing in the two lines speaks of the negation of ; modus ponens only detaches the consequent of an implication whose antecedent is already written.From and one obtains ; here is already written, and modus ponens does not authorize going back up the implication.Review: Formal deductive systems
Question 6It is known that , that is, that there is a deduction of whose only premise is . What does the deduction theorem then assert?
Dropping the premise changes what is asserted: from the premise one deduces , and is not a theorem. The premise does not disappear; it becomes the antecedent.It reverses the direction: the discharged premise is the antecedent, and the conclusion, the consequent.Deductions are not traversed backwards: from the premise one deduces, with A1 and modus ponens, for any , and from that implication is not recovered.The deduction theorem discharges a premise and turns it into an antecedent: if , then ; here is empty. It is proved by strong induction on the number of lines.Review: Formal deductive systems
Question 7What is derivable, with no further premises, from the set ?
Modus ponens needs the antecedent written, and is not a premise: from two implications one obtains another implication, not a loose formula.It chains in the reverse direction: the premises go from to and from to , and the chain is not traversed backwards.It would assert that and both hold, and the premises assert neither of the two: they only say what follows if holds or if holds.It is hypothetical syllogism: implications are chained. It is deduced in seven lines with A1, A2 and modus ponens, and the lesson turns it into a rule applicable under any set of premises.Review: Four deduction techniques
Question 8The rule of replacement of double negation (RDN) changes, in a line, a single occurrence of a subformula, and it acts on the string of the base language. How many occurrences of the subformula does the string designated by contain?
It reads the abbreviation as if it showed the whole string. The arrow hides the outer negation, , which writes its argument twice.Implication abbreviates ; the negation of is in the string even though the arrow does not show it, and the rule acts on the string.Four would be the copies of an argument under two consecutive negations, as in ; here there is a single outer negation, which duplicates it once.By the official definitions, is , that is, : the outer negation duplicates its argument. It follows that putting a double negation before the antecedent requires two applications of the rule, one in each copy; the second lesson uses this when it extends the deduction theorem.Review: Four deduction techniques
Question 9Which formula is obtained from by contraposition, without adding double negations?
It is the converse: it reverses the direction without changing the signs, and in general it is not derivable from the given formula.It changes the signs without reversing the direction; nor is it derivable, in general, from the given formula.It mixes two operations: it negates as it crosses over, but leaves as it was. The first form of contraposition would give , with one more negation.It is the second form of contraposition: the members change sides, and each one changes sign as it crosses over; loses its negation and gains one. The two formulas are provably equivalent.Review: Four deduction techniques
Question 10Do and designate the same formula?
The definition says that is , not . Taking one for the other is the mistake that the unit avoids: in other texts it is a definition; in this course, a result.The English reading does not decide the identity of two strings: it is decided by unfolding the abbreviations, and these two have different lengths.They are provably equivalent: the third lesson proves it with two applications of the rule RDN, with no need for any other tool.When the implication whose antecedent is is unfolded, appears in the first place of the joint denial. Passing from one formula to the other requires removing or adding that double negation inside the joint denial, and that is what the rule RDN allows; since the outer negation duplicates its argument, the rule is applied twice, once in each copy.Question 11It is known that , that is, that each one is derivable from the other. What justifies that also ?
Negating both sides is not a rule of the system, whose only rules are modus ponens and the replacement of double negation, which only interchanges a subformula with . That the operation is legitimate is a theorem, and it has to be proved.The definition says which string is, but it does not relate to : two equivalent formulas are not the same string, and neither are their negations.It is true, and the unit proves it; moreover, a formula can be replaced by an equivalent one in any place of any formula.From one obtains ; contraposition gives , and the converse of the deduction theorem, ; the other direction is the same. For joint denial, the fundamental lemma is needed as well.Question 12From the set of premises one deduces a formula and its negation . What can be concluded?
It confuses the assumption with the conclusion: the contradiction refutes the assumption , so that is concluded.That would require a contradiction deduced from alone. Here it was deduced from , and it may be due to the added assumption; from every formula is indeed derivable, by explosion.On the contrary: that an assumption leads to a contradiction allows it to be rejected. It is reasoning by contradiction, present throughout mathematics.It is reductio ad absurdum in its classical form: if assuming together with leads to a contradiction, is derivable from . The fourth lesson proves it with explosion, the deduction theorem, A3 and the theorem .Question 13To which formula is the negation of an implication, , equivalent?
It negates each member without touching the connective. The negation of an implication is not another implication: that of “if it rains, the street gets wet” is not “if it does not rain, the street does not get wet”, but “it rains and the street does not get wet”.It negates only the consequent. If does not hold, both and hold, and a formula cannot hold at the same time as its negation.It swaps the roles: an implication fails when the antecedent holds and the consequent does not, not the other way round.To negate “if , then ” is to assert that holds and does not. The fifth lesson proves it as a variant of De Morgan's laws: is ; De Morgan gives , and double negation, with the replacement of equivalents, .
Loading the content…
Completed
Sign in to save your progress