Modal Logic:

Note:

(1) The following elaboration is demonstrative rather than literary or accurate, meaning
that it is only supposed to provide an idea or understanding about the topic concerned as
articulated by me. There may be certain points that may not be as accurate, you can
refer to the book or online guides for a non-discrepant overview.

(2) The above point does not entail that the content here is useless, the content showcases
my attempt to provide an understanding of the relevant topic. It is also written in a style
that may loosely follow my stream of thought, so you may either find it repetitive at
times or maybe it might just fit your need. (Since I am writing for a general audience, I
have tried to elaborate most points)

(3) I have focused on intuitive understanding because I think that you can easily derive the
formulae and equations after it. I have not yet read the book because perhaps I would
have found it redundant to write this document after reading it (and I kinda didn’t feel
any need to read it yet), so whatever follows is a crude understanding, nevertheless, it
should enable you to solve the questions, and get a better grasp of the concepts
(whatever was taught till Jan last second week). Please read and provide suggestions
accordingly.

Part 1: An intuitive explanation for modal logic

Let us first consider the modal segment of modal logic; what exactly is it that we are doing
with logic to qualify it as modal logic? The Latin root of the word ‘modal’ is ‘modus’ which
implies ‘manner’ or ‘way’ [check any dictionary for reference]. Now, this ‘manner’ or ‘way’
does not pertain to the manner of doing logic, but instead, a consideration of something
that may qualify as ‘manner’ or ‘way’ is added as a component to classical logic. [do not
stress too much on this para; this is merely a reflection on the phrase ‘modal logic’ for
self-satisfaction. Simply speaking, although it likely is not accurate, you can consider modal
to be concerned with modes of being (so in what scenario could something be true or false).
Hence, we are generally concerned with possibility and necessity rather than mere truth or
falsity]. Find further elaboration below:

In classical logic (particularly, propositional logic), we simply take up propositions as our

basic units and deal with the relationships they have with each other, where these
propositions have binary values of truth or falsity. Propositions in classical logic are always
only necessarily true or false, making contingency in any form redundant. Now, must
propositions always be true or false in this manner?

Let us talk of an example: We want to enquire whether I shall get out of my room tomorrow
or not. Classical logic shall answer this question with a strict yes or no once we have set the
relevant premises, however, what about the scenario where I simply am not sure whether I’ll
get out of the room the coming day or not? This third scenario implies that there would be
certain cases under which I shall go out and certain ones under which I shall not. Classical
logic would need the required information about the premises to provide us with the
answer. However, what if you want to know whether it even is possible for me to come out
the next day or not?

Let us set our premises for the above case:

P1: It is possible for Ashish to get out of his room the next day only if his legs are not
broken

P2: His legs are not broken

C: It is possible for Ashish to get out of his room the next day

P2`: His legs are broken

C`: It is not possible for Ashish to get out of his room the next day

Now, in classical logic, we would consider the above-mentioned possibility as part of the
proposition itself, and not as something that is a quantifier or as something separate from
the proposition. The mention of possibility does not provide us with any new information
(meaning that we are not calling ‘Ashish’ getting out of his room as a proposition, say ‘p’, and
separating possibility from it; we are instead considering the entire statement as the
proposition. However, in modal logic, we’ll specifically perform this segmentation, and
consider possibility as an operator acting on the rest of the proposition). Things might be
slightly different in predicate logic, however, it is irrelevant to mention that here and
unnecessarily complicate things, not to mention that I have not read those things in detail
either [in case you are curious then you may follow this forum: link].

Now, if we add more premises, like the meaning of possibility in the above case, then we
may derive more information even through classical logic. For instance, the well-functioning
of my legs is only one of the factors that contribute to my getting out of my room. Hence,
despite my not having any injury to my legs, there is still no information available that
confirms my getting out of the room. However, say, if you have to take a decision that is
incumbent on my getting out of the room the next day, you may say that there still are
chances that I may get out of the room, hence, you can plan your moves accordingly after it.
If we talk about the other scenario, where my legs were broken, then it was certain that I
would not get out of the room, so once, again, you would know what decision to make.

Now, is it impossible for me to get out of my room if my legs are broken? In the situation
relevant to us, it may just be the case that it would be impossible. Hence, you are defining a
certain number of premises. You are saying that if P1, P2,…Pn are true, then it follows that
‘C’ is true. Working legs may be one of the criteria for someone to be able to get out of their
room, hence, once you have this information, you can say with certainty that once you do
not have working legs, you cannot get out of the room, or, you may say that it is not
‘possible’ to get out of the room. Possibility is mentioned here because working legs do not
necessitate getting out of the room, it only represents a possibility which is dependent upon
other factors.

Moving on, if we attack our premises themselves, and say that it may be possible to get out
of the room despite broken legs if you have a wheelchair, superpower or something like that,
then you have shifted your initial considerations (that is, obviously, the premises). So, what
this point tells you is that ‘possibility’ and ‘necessity’ work only under a certain framework,
that you need to define the conditions valid within that framework. This gives rise to the
relationship between worlds in modal logic.

What are ‘worlds’ in modal logic? We just described possibility above: there are necessary
conditions for a certain event; if a necessary condition is not false, then the event is possible.
The event is possible and not necessary because we have no information about the other
conditions. However, once we know that one of the necessary conditions is false, we would
know that event is not even possible.

So, if we know a necessary condition to be true, yet we lack any info on the other necessary
conditions, we can still hypothesize this scenario or world, where all the other necessary
conditions are true (having no info on these other necessary conditions imply that we do not
know how many of them exist, if they do at all, therefore it is not absurd to assume a
scenario where these conditions do not affect the occurrence of our event). Therefore,
possibility implies that there is at least one world that is ‘connected’ to our world, where the
concerned event is true.

Going back to our initial example, my not having broken legs implies that it is possible for me
to get out of the room. The axioms or theories in modal logic say that in at least one
connected world I must get out of the room. So, let us say that I did not get out of the room,
and when asked I said that I wasn’t in the mood, hence I did not get out. However, it is
possible for you to conceive a scenario and say that had he been in the mood, he would
have gotten out of the room. Therefore, mood was this factor that we had no information
about, and it did not seem relevant enough, so we can still say that functioning legs enable
me to get out of the room. We can say that these two worlds were connected. If, however, I
had said that my room was locked from outside, hence it was not even possible for me to
get out of the room, then well-functioning legs did not necessarily make it possible for me to
get out of the room. It is absolutely possible to argue and say that in a world where my room
was not locked, I could have gotten out, however, this is where the relationship between
worlds come into play.

When I say that it is necessarily true that all things fall downward on earth, I am only talking
about the worlds where only the gravitational pull of earth is a dominating force. So you are
defining the rules that your worlds are supposed to follow, and it is on the basis of these that
the connection between worlds is derived. However, take note that this is not the theory
behind modal logic, I am merely using this as an example to elaborate upon the abstract
conception of relations in modal logic.

Part 2: Bringing forth the notations

Modal operator rules:

(1) ∼□A,i → ◊∼A,i

(2) ∼◊A,i → □∼A,i

(3) □A,i ; irj → A,j

(4) ◊ A,i → irj ; A,j

□: is the necessity condition/operator

◊: is the possibility condition/operator

i,j: these subscripts denote the world number

r: r stands for relationship, that the world i is connected to the world j, or that ‘i’ accesses
the world ‘j’

Rule (2): This rule states the following - it is ‘not’ the case that ‘A’ is possible in world ‘i’

If ‘A’ is not possible in world ‘i’ then it follows that all the worlds ‘i’ is connected to will not
have ‘A’. Basically, it says that even if we change the allowed conditions in world ‘i’, ‘A’
will not remain possible in it. Therefore (from the example in the previous section), even
if I had the mood to go out, broken legs would make it impossible for me to do so; so all
 those worlds in which I have broken legs, I shall not be able to go out (and our world ‘i’ in
this scenario is that in which the fact about broken legs cannot be changed).

We just mentioned that all the worlds ‘i’ is connected to will not have ‘A’. Therefore, we can
say that all of them will have ‘~A’. ‘~A’ henceforth is necessarily true in world ‘i’. We can
say that my ‘not’ going out is necessarily true in all those worlds where I have broken
legs.

From the above two points, it follows that “‘not’ the case that ‘A’ is possible in world ‘i’” is
equivalent to “it is necessary that not A is the case in world ‘i’”. The equivalence
relationship here proves both rule (1) & (2)

Note: Necessity and possibility as mentioned above do not talk about the truth values in the
world ‘i’; what they talk about will be mentioned in rule 3 and 4

Rule (3): If ‘A’ is necessary in world ‘i’, then all the worlds connected to ‘i’ will have ‘A’,
meaning that the truth value of ‘A’ shall be ‘1’ in all the worlds that ‘i’ accesses.

Now, what are the worlds that ‘i’ accesses? We do not know that by default, it needs to be
additional information. Therefore, if we are told that ‘A’ is necessary in ‘i’, and that ‘i’
accesses a certain world ‘j’, then the truth value of ‘A’ in ‘j’ shall be ‘1’. [truth value of ‘A’
in ‘j’ can be written as follows: v(A,j) = 1]. We cannot comment upon the truth value of
‘A’ in ‘i’ even if it is necessary in ‘i’; v(A,i) would be ‘1’, if we have the relation iri or iRi,
meaning that ‘i’ accesses itself. I’ll need to think if I have to provide a relationship with
the initial example here, that is, if I have to explain the meaning of a world accessing
itself, so I am just skipping it for the moment.

I hope the above para sheds some light on rule 3.

Rule (4): If ‘A’ is possible in world ‘i’, then there ought to be a certain configuration of world
‘i’ in which ‘A’ is true. Again, I am using the term configuration here to talk about the
world that ‘i’ may be connected to. Configuration gives a better intuitive sense. Perhaps,
‘A’ is not true in world ‘i’ (say, take our initial example - because of my mood, I did not go
out of the room despite having well-functioning legs, so ‘A’ is false in this world,
however, there ought to be at least one world where I had the mood to go out as well; or
say, I did not go out the next day, but there may be some other day when I’d be of the
mood, hence ‘A’ is still possible, whereas if I had broken legs, then even the possibility of
going out would have been taken out).
From the above para, we can note that if something is possible in ‘A’ then there must be a
certain configuration of this world where this is true, hence we can say that there is at
least one world, say ‘j’ which is accessed by ‘i’, and ‘A’ is true in world ‘j’.

Other Rules:

The above section completes the modal operator rules, these are the ones primarily applied
when solving the relevant problems, along with the basic rules of classical logic. Sir, had
described the rules of classical logic as alpha and beta rules. These rules essentially
convert all the operators into the notations of negation, conjunction and disjunction
only. This is done because the equations for modal logic are checked for their validity
using tableaux, where you branch out your equation to solve it further. This shall be
explained later on. The rules are as follows:

Alpha:

1. Conjunction: A^B - essentially, both A and B need to be true

2. Disjunction: AvB - so either of A or B
3. Conditional: A⊃B - if ‘A’ then ‘B’, meaning ~A or B (~AvB)
4. Biconditional/Equivalence: A≡B -’A’ if and only if ‘B’, therefore [(A^B)v(~A^~B)]

Beta: (applying the negation operator to the above four)

5. ~Conjunction: ~(A^B) : ~Av~B

6. ~Disjunction: ~(AvB) : ~A^~B
7. ~Conditional: ~(A⊃B) : A^~B
8. ~Biconditional/Equivalence: ~(A≡B) : [(A^~B)v(~A^B)]

Example:

These complete the purview of the rules, now what exactly is to be done to solve the
equation in modal logic? Let us take an example (‘⊢’ is the notation of tautology or ‘it entails
that’)

1. □(A ⊃ B) ∧ □(B ⊃ C) ⊢ □(A ⊃ C) (2.4.6, page 25 of non-classical logic book)

Basic notion behind solution:

So, we are given a premise (or two if you remove the conjunction) and one conclusion in the
above example. Also, we know that an argument is valid only if the premises entail the
conclusion, therefore, there is no way for the premises to be true yet the conclusion false.
Since, we have to ascertain the veracity of the above equation or say, the validity of the
argument given to us in 1, we can say that if the conclusion is not entailed from the premises
then the argument would be invalid. In the absence of any other information, we assume
that our argument is invalid. Why so? Because there is only one scenario in which the
conclusion shall be invalid. If we fail to prove the invalidity, then we get our answer (that the
conclusion is valid); if we prove the invalidity, even then we get our answer. Hence, all we
need to do is assume our equation to be invalid, and wait to see if we run into a
contradiction.

How shall we run into a contradiction? We try to break down the equation, and see if
assuming it to be invalid leads to an outcome that says, for instance, both A and ~A. If it
does, then it contradicts, if not, then it remains open and invalid. Also, take note that
sometimes, if start with disjunction, disjunction implies ‘either or’, meaning that we get two
branches (shall demonstrate them later when solving the question, or perhaps, just look at
any example in the book), if we get multiple branches, then we obviously need to run into
contradictions into each one of them, else our equation shall remain invalid.
Once, the above points are cleared, we can see that we can write the given equation in this
format: P^~C, where, P is the premise = □(A ⊃ B) ∧ □(B ⊃ C); and C is the conclusion = □(A
⊃ C). P^~C merely is another way to write what I described in the above para. We assume
the premise to be true and the conclusion to be false.

Acting on modal operators:

Now, alpha, beta, and all those rules can be applied for the classical logic operators. For,
modal operators, the thing can be slightly different as people might not be familiar with
these notations.
So, if we say, p = (A ⊃ B), and □(A ⊃ B), then we have □p (I am only writing it this way to
make it look easier and less messy; it is not mandatory). Now, our first step would be to
define the world we shall be talking about (the book uses the subscripts ‘i’ and ‘j’, and sir in
class uses ‘0’ and ‘1’, you may choose any of them). We define these worlds because when
talking about necessity and possibility, the necessity at world ‘i’ shall be different from
necessity at world ‘j’. Similarly, the truth values ‘v(p)’ derived shall pertain to specific worlds.
We cannot say that if v(p,i) = 0, and v(p,j) = 1, then it is a contradiction. It will only be a
contradiction if v(p,i) is both 0 and 1 (which obviously is a contradiction after all we are not
going into quantum physics -okay, let’s leave that, it was a joke).
Now that we have understood the importance of defining the world we are dealing with, we
take it as a necessary step to ascribe world notion, so we can write □(A ⊃ B),i or □p,i
If somehow, in the future, we get an information that describes the relationship of i with
some world j, that is, iRj, then we would be able to remove the necessity operator using rule
(3), and write the value of p at j (trace back if you are lost or confused).
Note: I am merely describing how to solve and not necessarily writing down the solution, for
the solution, you may refer to the book, or if needed, we can add it later. I am writing this so
that an in-person explanation may not be mandatory.
Since, we are negating the conclusion, we shall get a possibility operator in the case of the
conclusion (by using rule 2 or 1 or whichever it is to convert ~□ to ◊~). Once, we get a
possibility operator, we’ll get the information iRj (the future I just described would have
finally arrived), once we get that information, we shall be able to remove the modal
operators of necessity and possibility. Once that is done, we shall get the truth values at
particular worlds, and look for ourselves if there are any contradictions. And then, bingo! We
would have solved the problem.

Tableaux or tree method is used to solve these problems because that way it is easier to
keep track of our solution. We just draw lines and arrows and quickly solve the stuff, so there
we go.

Alright, please attempt to look at the example before proceeding forward now. I am only
writing the procedure and not the solution, so don’t loose track; just look at the example and
you’ll get it, else, if needed, we can always post the solution here.

Quantum Model:
Okay, this is not quantum computing (at least not yet, I guess) or physics, it is just that when
our conclusion remains invalid, meaning the we do not find contradictions in all the
branches (again, look at the tableaux method if you are confused with regards to what
branches are), we write the values that make the equation invalid (I hope this is clear; we
are just writing the scenario where the premises are true yet the conclusion is not). These
values are obviously not the quantum model, however, they are what lead to the
development of this thing called as quantum model.

When we solve the equation, we would have opened a number of worlds, that is, if we
found a relation between i and j, then we would have opened two worlds, if we had found
more relations, we would have opened more worlds. We also would have found truth values
at certain world. Therefore, the set the world, the relationships between these worlds, and
the truth values are the three things that define our invalid conclusion and the quantum
model.
Once we have noted down these three things, we draw these different worlds, their
relationships (connections using arrows), and different truth values at the different worlds.
The diagram thus generated is the quantum model.

Alright, I have done my part of explaining, rest of the things may either be added later on, or
can simply be taken from someone’s notes or the book.