Thursday 30 September 2010

The Logic Book, M. Bergmann, J. Moor, J. Nelson, McGraw Hill 2004, 10.4E, 12(b)

We are required to symbolise the following argument and show that it is valid. Despite the name used in the agrument, here abbreviated to lower case 'g', there is nothing that is cock-eyed about the argument.
Groucho Marx doesn't stay in any hotel that is willing to have him as a guest. Any hotel not willing to have Groucho Marx as a guest doesn't. Therefore, Groucho Marx doesn't stay in any hotel.
  1. (x)[(Hx • Wxg) ⊃¬ Sgx]
  2. (x)[(Hx • ¬ Wxg) ] ⊃¬ Sgx]
  3. ∴(x)(Hx ⊃¬ Sgx]
  4. * Hx ......... ACP
  5. * (Hx • Wxg) ⊃¬ Sgx ......... 1UI x/x
  6. * Hx ⊃(Wxg ⊃¬ Sgx) ......... 5Exp.
  7. * Wxg ⊃¬ Sgx ......... 4,6MP
  8. * (Hx • ¬ Wxg) ] ⊃¬ Sgx ......... 2UI x/x
  9. * Hx ⊃ (¬ Wxg ⊃¬ Sgx) ......... 8Exp.
  10. * ¬ Wxg ⊃¬ Sgx ......... 4,9MP
  11. * Sgx ⊃Wxg ......... 10Contrap.
  12. * Sgx ⊃¬ Sgx ......... 7,11HS
  13. * ¬ Sgx ∨¬ Sgx ......... 12MI
  14. * ¬ Sgx ......... 13Taut.
  15. Hx ⊃¬ Sgx ......... 4-14CP
  16. (x)(Hx ⊃¬ Sgx) ......... 15UG

Thursday 23 September 2010

Fog in many areas

I don’t like M.C. Escher’s art that much, but I can’t resist comparisons between his spirals and staircases to the English in some business documents I often have to read. It is amazing that business is done, money changes hands, buildings and bridges do not collapse (usually) despite the fact that what people say on paper is either vague, ambiguous or simply just drivel. Does this makes sense to you:

“The location of the investment shall be determined pursuant to the local zoning plan. If the local zoning plan has not been adopted for a specific area, the location of the public investment shall be determined by a decision on determining the location of a public investment, whereas other investments by a decision on development conditions.”

It always cracks me up when the recipient of a letter, a lawyer or consultant with whom I work closely, says they understand what the author meant to say even though they agree that what he actually says is impenetrable, and then draft a letter in reply to match or go one better in the impenetrability stakes. One wonders what the other side thinks of it!

And yet, they all understand one another! This must be where Grice’s conversational implicature comes in, and specifically the principle of cooperation. We do not, willingly, seek to misunderstand.

But if language of the stated kind has me tearing my hair out and thinking unprintable thoughts, I actually purposefully seek out ambiguity and test the limits of understanding for pleasure. And, interestingly, many of such verbal ambiguities resemble Escher’s art. The ones relying on lexical ambiguity are merely fun; the ones exploiting syntactic ambiguity are one level up on the former; while the ones hinging on both are simply brilliant. Here are a few:

A: Where are you going?
B: I am going down to the bank to get some money.
A: Who do you bank with?
B: I’m sorry, I don’t understand?
A: You said you were going down to the bank to get some money.
B: And so I am; I keep my money buried in a chest down by the river.
(Descriptions, Stephen Neale)

Humphrey Lyttelton used to ask on I’m Sorry I Haven’t a Clue: - How many legs have donkeys? The panelists wrestled with the question for a while whereupon Humph would deadpan: - Legs don’t have donkeys.

P.G. Wodehouse, Mark Twain, B.J. Priestly, Ambrose Bierce and Groucho Marx have delivered in the last category, to quote but one example: ‘Time flies like an arrow. Fruit flies like a banana.’

The quote resembles a cube in the oblique plane (hence the Escher connection), with the front and rear faces flipping back and forth.

I have been playing around with my nemesis ‘the mole’ and have come up with this: There are different ways in which you can remove moles? A/ with laser, B/ with explosives, C/ by blowing their cover, but coming up with a piece of dialogue where the ambiguities are cancelled one by one, but not until they’ve had their run, has proved elusive so far. [mole: a/ a growth on the human skin; b/ a breakwater running out into the sea, c/ a spy, d/ animal that burrows underground (not intended by the question yet temptingly close)].

Understanding Symbolic Logic, Virginia Klenk, Pearson Prentice Hall, 2008, Unit 18, Ex.1, problem (r), p. 354

We have to derive the conclusion from the premises given. Hint: It is tempting to use indirect proof because the conclusion is a simple existential statement, but this would not work very effectively here. It is easier to instantiate the premises and work through them until we isolate, by Modus Ponens, Wz on the first line.
  1. (x){[Ax • ¬ (y)(Dxy ⊃Rxy)] ⊃(z)(Tzx ⊃Wz)}
  2. ¬ (∃x)(∃y)(Tyx • ¬ Dxy)
  3. ¬ (x)[Ax ⊃(y)(Tyx ⊃Rxy)
  4. ∴(∃x)Wx
  5. (∃x)[Ax • (∃y)(Tyx • ¬ Rxy) ......... 3CQ
  6. (x)(y)(Tyx ⊃Dxy) ......... 2CQ
  7. Aa • (∃y)(Tya • ¬ Ray) ......... 5EI x/a
  8. Aa ......... 7Simp.
  9. (∃y)(Tya • ¬ Ray) ......... 7Simp.
  10. Tma • ¬ Ram ......... 9EI y/m
  11. Tma ......... 10Simp.
  12. ¬ Ram ......... 10Simp.
  13. (y)(Tya ⊃Day) ......... 6UI x/a
  14. Tma ⊃Dam ......... 13UI y/m
  15. Dam ......... 11,14MP
  16. Dam • ¬ Ram ......... 15,12Conj.
  17. (∃y)(Day • ¬ Ray) ......... 16EG
  18. (∃y)¬ ( ¬ Day ∨ Ray) ......... 17DeM
  19. (∃y)¬ (Day ⊃ Ray) ......... 18MI
  20. ¬ (y)(Day ⊃Ray) ......... 19CQ
  21. [Aa • ¬ (y)(Day ⊃Ray)] ⊃(z)(Tza ⊃Wz)} ......... 1UI x/a
  22. Aa • ¬ (y)(Day ⊃Ray) ......... 8,20Conj.
  23. (z)(Tza ⊃Wz) ......... 22,21MP
  24. Tma ⊃Wm ......... 23UI z/m
  25. Wm ......... 11,24MP
  26. (∃x)Wx ......... 25UG

Saturday 18 September 2010

Deductive reasoning and common sense

The idea that it is impossible for an argument to be valid if the premises are true and the conclusion is false is central to all logic and mathematics. So if a sequence takes us from truth to falsehood, we can suspect an error in reasoning. All other combinations are perfectly acceptable, including from falsehood to truth, from falsehood to falsehood, and from truth to truth, of course.

The rules of deduction place three restrictions on us. The first is easy to state and quick to dispatch: an assumption must not be extended to all items of a certain kind until it is discharged. Suppose we assume that a car picked at random is of the colour brown. At this stage we can’t say that every car is brown – this would be making an unjustified leap in our reasoning. However, once we get to a stage, through valid reasoning, where we can say: ‘If a car picked at random is brown, then it is made in the US,’ by which we have discharged our assumption, we can say that ‘All cars which are brown are made in the US.’

The second restriction says that if, in the course of our reasoning, we introduce an individual specified by name next to an individual which we have earlier picked at random (strictly in this order), then, likewise, we can’t extend the relation between the two to a relation between all individuals and the one we have specified by name. Thus, by proceeding from ‘Every car in the car park belongs to someone,’ (true) through ‘A car in the car park picked at random belongs to Jamie,’ (possibly true), we come to ‘All cars in the car park belong to Jamie,’ (possibly false) and eventually ‘Someone is the owner of all cars in the car park,’ (possibly false), we have made an error in reasoning. We’ve gone from a true premise to a false conclusion.

The last restriction is hardest to state in plain English, but it is quite straightforward. Suppose that, figuratively speaking, we put into the same box in our head the things that we pick at random and that bear a relationship to one another. We can’t then, in the course of our reasoning, break the bond that connects them and put one thing into one box and the other into another box. Thus, from ‘Every car is identical to itself,’ (true) we can go to ‘A car picked at random is identical to a car picked at random’ (possibly true), and back to ‘Every car is identical to itself,’ but not to ‘Every car is identical to every car,’ as the latter is definitely false.

We can of course rehearse the last argument in a language approaching real life situations, as done earlier, for example, ‘Every driver admires himself,’ where, again, it would be incorrect to conclude that ‘Every driver admires every driver (that is, every other driver and himself too), but with such sentences, of course, we can’t easily demonstrate the truthfulness of the premise(s) or the falsity of the conclusion.

I will run through this again using mathematical examples another time.

Thursday 16 September 2010

Symbolic Logic, Irving M.Copi, Prentice Hall, 5th edition, 1973, p. 150, problem 9

The argument is set as follows:

There is exactly one penny in my right hand. There is exactly one penny in my left hand. Nothing is in both my hands. Therefore, there are exactly two pennies in my hands.

To prove it is not a challenge, to translate it into logical notation is.
  1. (∃x){Px • Rx • (y)[(Py • Ry) ⊃y = x]}
  2. (∃x){Px • Lx • (y)[(Py • Ly) ⊃y = x]}
  3. ¬ (∃x)(Lx • Rx)
  4. ∴(∃x)(∃y){Px • Py • Rx • Ly • ¬ (x = y) • (z){[Pz • (Rz ∨Lz)] ⊃z = x ∨z = y}}
  5. * ¬ (∃x)(∃y){Px • Py • Rx • Ly • ¬ (x = y) • (z){[Pz • (Rz ∨Lz)] ⊃z = x ∨z = y}} ......... AIP
  6. * (x)(y){[Px • Py • Rx • Ly • ¬ (x = y)] ⊃(∃z)[Pz • (Rz ∨Lz) • ¬ (z = x) • ¬ (z = y)]} ......... 5CQ
  7. * Pa • Ra • (y)[(Py • Ry) ⊃y = a] ......... 1EI x/a
  8. * Pm • Lm • (y)[(Py • Ly) ⊃y = m] ......... 2EI x/m
  9. * (x)(Lx ⊃¬ Rx) ......... 3CQ
  10. * Lm ⊃¬ Rm ......... 9UI x/m
  11. * Lm ......... 8Simp.
  12. * ¬ Rm ......... 11,10MP
  13. * Ra ......... 7Simp.
  14. * ¬ (a = m) ......... 12,13Id.
  15. * Pa ......... 7Simp.
  16. * Pm ......... 8Simp.
  17. * Pa • Pm • Ra • Lm • ¬ (a = m) ......... 15,16,13,11Conj.
  18. * (y){[Pa • Py • Ra • Ly • ¬ (a = y)] ⊃(∃z)[Pz • (Rz ∨Lz) • ¬ (z = a) • ¬ (z = y)] ......... 6UI x/a
  19. * [Pa • Pm • Ra • Lm • ¬ (a = m)] ⊃(∃z)[Pz • (Rz ∨Lz) • ¬ (z = a) • ¬ (z = m)] ......... 18UI y/m
  20. * (∃z)[Pz • (Rz ∨Lz) • ¬ (z = a) • ¬ (z = m)] ......... 17,19MP
  21. * Pr • (Rr ∨Lr) • ¬ (r = a) • ¬ (r = m) ......... 20EI z/r
  22. * (y)[(Py • Ry) ⊃y = a] ......... 7Simp.
  23. * (Pr • Rr) ⊃r = a ......... 22UI y/r
  24. * ¬ (r = a) ......... 21Simp.
  25. * ¬ (Pr • Rr) ......... 23,24MT
  26. * ¬ Pr ∨¬ Rr ......... 25DeM
  27. * Pr ......... 21Simp.
  28. * ¬ Rr ......... 27,26DS
  29. * Rr ∨Lr ......... 21Simp.
  30. * Lr ......... 28,29DS
  31. * (y)[(Py • Ly) ⊃y = m] ......... 8Simp.
  32. * (Pr • Lr) ⊃r = m ......... 31UI y/r
  33. * Pr • Lr ......... 27,30Conj.
  34. * r = m ......... 33,32MP
  35. * ¬ (r = m) ......... 21Simp.
  36. * r = m • ¬ (r = m) ......... 34,35Conj.
  37. ¬ ¬ (∃x)(∃y){Px • Py • Rx • Ly • ¬ (x = y) • (z){[Pz • (Rz ∨Lz)] ⊃z = x ∨z = y}} ......... 5-36IP
  38. (∃x)(∃y){Px • Py • Rx • Ly • ¬ (x = y) • (z){[Pz • (Rz ∨Lz)] ⊃z = x ∨z = y}} ......... 37DN

Thursday 9 September 2010

Simple is difficult

Would that people asked only for complex things to be explained. Making a living would be very easy if one was in the business of providing the answers. It is the simple things that are a problem and it’s the simplicity that often brings tuts of derision.

The laws of reasoning are supposed to be very simple. Three, in particular, sound almost trivial. Suppose I look out the window and see a fresh molehill in the garden. My sentence is:

There is a fresh molehill in the garden.

It seems like nothing, but it’s a start. What is the meaning of this sentence? The meaning is: ‘true’, because it is a fact, and ‘true’, because first order deductive logic does not recognize other meanings than ‘true’ or ‘false’.

What can I infer from it? I could repeat it, for example, but it would be no use. I would end up with a tautology. I can conjoin to it any sentence I like using the disjunctive word ‘or’, including its own negation: ‘It is not the case that there is a fresh molehill in the garden’, because the meaning of ‘or’ guarantees that, in the first case, at least one of the sentences will be true, and, in the second case, precisely one sentence will be true. Either way, the whole new sentence will be true. The latter represents the Law of the Excluded Middle: either something is of a certain kind or it isn’t. There is no third option.

Suppose I don’t look out of the window, but hold the thought in my head. What is this thought? It is an assumption, or hypothesis. What is its meaning? I do not know. I can repeat my hypothesis at will because, it being a hypothesis, I am not committing myself to anything, only probing and exploring. However, assumptions are not made to be just random thoughts floating freely in my head – that is a stream of consciousness. We make an assumption in order to prove something.

I can, of course, attach to my assumption another sentence via ‘or’:

There is a fresh molehill in the garden or the gate is open.

This in itself suffices as my conclusion – something I derived from the original sentence ‘There is a molehill in the garden’, but I’m not sure I know what to do with it. Far more convenient, if less illuminating, is to repeat the hypothesis and stop at that:

Assumption: There is a molehill in the garden.
Repetition: There is a molehill in the garden.
Conclusion: If there is a molehill in the garden, then there is a molehill in the garden.

This may not be much of a conclusion, but it is in fact an illustration of the Law of Identity, which, like the Law of the Excluded Middle, holds regardless of fact. When I look out and see a molehill, the conditional sentence is true. When I look out and see no molehill, the conditional sentence is also true. If I now choose to paraphrase my two sentences:

There is a fresh molehill in the garden or it is not the case that there is a fresh molehill in the garden.

If there is a fresh molehill in the garden, then there is a fresh molehill in the garden.

then it turns out that I can paraphrase each with:

It is not the case that both there is a fresh molehill in the garden and there isn’t.

The key word in this sentence is ‘and’. The sentence illustrates the Law of Non-contradiction. We can’t say that something both is and isn’t. A closer look reveals that all these laws are in fact one law. At any rate, all three hold regardless of the facts on the ground, so to speak.

These three laws were known to the ancient Greeks, and we use them so automatically that we never give them a second thought. When the Skeptics searched for answers as to what is and each avenue led them to deeper skepticism, they turned to these fundamental laws as something unchangeable throughout time, something that could be relied on. I rely on them alright but the more I try to explain them, including to myself, the more tortuous the explanations get.

The Logic Book, M. Bergmann, J. Moor, J. Nelson, McGraw Hill, 2004, 10.5E, 3(d), p. 563

Is the following a theorem of predicate logic: (x)(Ax ⊃Bx) ∨(∃x)Ax ? Yes, it is provable without premises, and the proof is straightforward.
  1. (x)(Ax ⊃Bx) ∨(∃x)Ax
  2. * ¬ (x)(Ax ⊃Bx) ......... ACP
  3. * (∃x)(Ax • ¬ Bx) ......... 2CQ
  4. * An • ¬ Bn ......... 3EI x/n
  5. * An ......... 4Simpl.
  6. * (∃x)Ax ......... 5EG
  7. ¬ (x)(Ax ⊃Bx) ⊃(∃x)Ax ......... 2-6CP
  8. ¬ ¬ (x)(Ax ⊃Bx) ∨(∃x)Ax ......... 7MI
  9. (x)(Ax ⊃Bx) ∨(∃x)Ax ......... 8DN

Thursday 2 September 2010

The Logic Book, M. Bergmann, J. Moor, J. Nelson, McGraw Hill, 2004, 10.5E, problem 3f, p. 563

We are asked to show that the following statement is a theorem in predicate logic: (x)(∃y)(Ax ∨By) ≡ (∃y)(x)(Ax ∨By). Under Claim 1 we cannot just drop the quantifiers and restore them in reverse order because of the restriction saying that if every x is in a relation to some y, it doesn't follow that there is a y such that every x is in this relation to it. We are not bound by such a restriction under Claim 2.
  1. (x)(∃y)(Ax ∨By) ≡ (∃y)(x)(Ax ∨By)
  2. Claim 1: (x)(∃y)(Ax ∨By) ⊃ (∃y)(x)(Ax ∨By)
  3. * (x)(∃y)(Ax ∨By) ......... ACP
  4. * * ¬ (∃y)(x)(Ax ∨By) ......... AIP
  5. * * (y)(∃x)(¬ Ax • ¬ By) ......... 4CQ
  6. * * (∃y)(Ax ∨By) ......... 2UI x/x
  7. * * Ax ∨Bm ......... 6EI y/m
  8. * * (∃x)(¬ Ax • ¬ Bm) ......... 5UI y/m
  9. * * ¬ Aa • ¬ Bm ......... 8EI x/a
  10. * * ¬ Bm ......... 8Simp.
  11. * * Ax ......... 10,7DS
  12. * * ¬ Aa ......... 8Simp.
  13. * * (y)Ay ......... 11UG
  14. * * (∃y) ¬ Ay ......... 12EG
  15. * * ¬ Ar ......... 14EI y/r
  16. * * Ar ......... 13UI y/r
  17. * * Ar • ¬ Ar ......... 16,15Conj.
  18. * ¬ ¬ (∃y)(x)(Ax ∨By) ......... 4-17IP
  19. * (∃y)(x)(Ax ∨By) ......... 18DN
  20. (x)(∃y)(Ax ∨By) ⊃ (∃y)(x)(Ax ∨By) ......... 3-19CP
  21. Claim 2: (∃y)(x)(Ax ∨By) ⊃(x)(∃y)(Ax ∨By)
  22. * (∃y)(x)(Ax ∨By) ......... ACP
  23. * (x)(Ax ∨Bm) ......... 22EI y/m
  24. * Ax ∨Bm ......... 23UI x/x
  25. * (∃y)(Ax ∨By) ......... 24EG
  26. * (x)(∃y)(Ax ∨By) ......... 25UG
  27. (∃y)(x)(Ax ∨By) ⊃(x)(∃y)(Ax ∨By) ......... 22-26CP
  28. {(x)(∃y)(Ax ∨By) ⊃ (∃y)(x)(Ax ∨By)} • {(∃y)(x)(Ax ∨By) ⊃(x)(∃y)(Ax ∨By)} ......... 20,27Conj.
  29. (x)(∃y)(Ax ∨By) ≡ (∃y)(x)(Ax ∨By) ......... 28BE