www.richmond-philosophy.net - © COPYRIGHT 2002-10 RUTC - Designed and maintained by Mat Carmody.

Fate and Fatalism

Mat Carmody

1. Introduction

You have just started to read this essay. Thank you for choosing to do so. Conceivably, you did not choose to do because you were told to do so. I hope that that is not the case. Yet even if you were told to do so, it seems you still had the choice. Even the highwayman who pointed his pistols at his victims and said, 'Your money or your life!' likewise offered them a choice. Of course, he was assuming that one of the two options was so undesirable that nobody would take it but it was a choice nonetheless.


Let us suppose you chose freely. Might it in fact be true that you simply felt free but were not really so? That you had no real choice in the matter because your reading this essay was inevitable? On hearing these words, you might already be resolving to put this article down just to prove that you are not compelled to read. But what if your doing so was likewise inevitable? What if everything you do and everything that happens to you occurs not through free choices you make but because it was always ever fated so to occur?


Fate has long been a topic of philosophical interest. In this essay, I shall introduce the following different varieties of fatalism:

  • Interventionist Theological Fatalism
  • Omniscience Fatalism
  • Physical Fatalism (Determinism)
  • Logical or Conceptual Fatalism

The majority of the paper will be devoted to the final type of fatalism. I shall consider three arguments of increasing strength that attempt to prove that, as a matter of sheer logic, there is only one way that the universe could be. These arguments are, in order, the Mortality Argument, the Idle Argument and the Necessity Of The Past Argument. I shall suggest a way that each argument can be resisted. By itself, this is not enough of course to disprove fatalism. There is no way to disprove fatalism, as we shall see. The anti-fatalist must therefore spell out what the preferable alternative to fatalism is.

2. The Idea of Fate

The earliest form of fatalism is what we shall call theological fatalism. It is the view that our lives are under the control of more powerful beings or spirits that we may collect together as gods. The gods may intervene to help or to harm us and thus prevent us from acting as we otherwise might. They may shape our lives by determining our destinies and the paths we are bound to follow.[1] The Greek poet Hesiod has bequeathed to us the idea of the Moirai - the three goddesses of fate. Life was pictured as a thread. Clotho spun the thread, Lachesis determined its length and Atropos cut it.[2] It was believed that the Moirai had power over the lives of mortals but there is evidence to suggest that it was also believed that they could exert control over the other gods too. The earlier Greek poet Homer spoke simply of moira - fate - as an impersonal power that influences even the lives of the Gods and which not even they can countermand.


The idea of trying and failing to outwit fate is also an old one. In Sophocles' great play Oedipus Rex King Laius of Thebes is told by the Oracle at Delphi that he must not have a child with his wife Jocasta or he will die at the child's hands and the child will marry his wife.[3] Laius has the child taken away and left on a mountain to die. The child is rescued by a shepherd and given to King Polybus of Corinth. Polybus raised the child, Oedipus, as his own. Later Oedipus hears rumours that he is not the son of Polybus and goes to the same oracle to find out who they really are. The Oracle now tells him that he will marry his mother and kill his father. Horrified, he tries to escape his fate but in vain. He marries Jocasta and kills Laius, admittedly without realising in either case who they are beforehand. Cicero recounts two tales of the apparent inescapability of fate.[4] In the first, Daphitas tries to outwit the oracle at Delphi by asking it whether he would find his horse when in fact he had no horse. The oracle said that he would and Daphitas took this as proof that they could not be taken seriously. Yet later in his life, he was sentenced to death by Attalus, King of Pergamum, by being thrown from a rock called Hippos - or The Horse. In the second, Philip of Macedon, father of Alexander the Great, was warned by an oracle to beware of a chariot and thus never rode in one but the sword with which he was assassinated had a chariot-race carved in ivory on its hilt. In the 20th century, we have the following short story related by the figure of Death in a play by W. Somerset Maugham:[5]


There was a merchant in Baghdad who sent his servant to market to buy provisions and in a little while the servant came back, white and trembling, and said, Master, just now when I was in the marketplace I was jostled by a woman in the crowd and when I turned I saw it was Death that jostled me. She looked at me and made a threatening gesture; now, lend me your horse, and I will ride away from this city and avoid my fate. I will go to Samarra and there Death will not find me. The merchant lent him his horse, and the servant mounted it, and he dug his spurs in its flanks and as fast as the horse could gallop he went. Then the merchant went down to the market-place and he saw me standing in the crowd and he came to me and said, Why did you make a threatening gesture to my servant when you saw him this morning? That was not a threatening gesture, I said, it was only a start of surprise. I was astonished to see him in Baghdad, for I had an appointment with him tonight in Samarra.

Both the servant and the figure of Death seem in the hands of fate: the one fated to take the life of the other.


The generic plan of these stories is that an agent A learns that an event E will happen and, despite taking precautions to avoid E happening, E happens nevertheless. From a literary point of view, there are (at least) four different ways we may understand stories of this kind. First, it is simply an accident that E happens. The oracle has no genuine powers of prediction. It made a lucky guess. Oedipus' freely-chosen actions in part contributed to his making the guess come true. The play is an exercise in tragic irony. Looked at another way, coincidences do happen and the unwary will attribute them to the hand of fate. This is what Cicero thought was the proper thing to say about the two stories above.


Second, the story presents a self-fulfilling prophesy. By trying to avoid E, the agent's own actions make E unavoidable. Here's an example of such a story. Bernard has an interview tomorrow and wants to avoid a sleepless night thinking about the interview. He goes to bed telling himself not to think about the interview which causes him to think about the interview and to stay awake all night. As with the first case, there is nothing but the appearance of inevitability.


To distinguish the third and fourth interpretations, we shall borrow an oft-borrowed idea from the writer Borges.[6] Imagine yourself entering a garden along a path. The path symbolises your life. You come to a fork with a path leading off to the left and a path leading off to the right. The fork symbolises a decision you have to make. You choose a path, walk farther and then another fork appears. And so on: life is full of decisions. From a bird's-eye view, we can see the single path you took out of the very many possible paths.


The third interpretation is the full-strength fatalist interpretation. Every event in your life - every event in the universe perhaps is inevitable through being part of a pre-ordained divine plan. If full-strength fatalism is true, then that single path was the only one you could have taken. No matter how free your choices felt, free choice was always an illusion. One certainly could read Oedipus Rex or the tale of the servant's appointment in Samarra this way. Yet a more interesting way to read them is according to the fourth interpretation of weaker fatalism. On a weaker fatalism, the network of paths is laid out by fate in such a way that (a) you can freely choose your route at least some of the time but (b) all paths lead to the same point or points. Oedipus' life has two 'fixed points', for example. If we view it in this light, we see his tale as genuinely tragic. He was free - but not free enough.


One way to distinguish the strength of fatalism is thus by the number of 'free forks' there are. Another is by the time at which one's fate is determined. At one extreme, Fate has laid out the path-way of your life before you were born. You are fated to take a certain route, despite feeling you have freely chosen it. If we move away from that extreme, you can make free choices but Fate has the power to be one step ahead and quickly construct the path ahead. So, Fate may have taken no interest in your life up to now but now decides that in five minutes' time, you will fall over. This will mean that she has to do a lot of work, for she must prevent you from being able to decide to lie down for ten minutes and successfully doing so. She may also have to interfere in the lives of others. For example, she must prevent someone from getting you into the passenger seat of car in the next few minutes, from which position falling over is impossible.


If Fate is a capricious agent who only ever fixed the future a little before it happened, we should perhaps feel a little freer. We often make decisions with far-reaching consequences in mind. We believe therefore that our future will be radically different from what it would otherwise have been. If it were true that all paths will ultimately lead to the same place many years hence, we would understandably feel frustrated. If, however, all Fate will ever do is constrain us for a short time, then we may still feel as if we have a lot of control. No doubt we will think this is not enough. We want no constraints on our freedom. But is this realistic?


3. Being Free

Let us say that a maximally unconstrained being is one that can make anything at all happen: that is, make any proposition p true. Such a being would be free to satisfy every desire. Could there be such a being?


The obvious candidate is a god. Yet there is reason to think that the gods are limited at least by the laws of logic and mathematics. No god could make it true that 2+2=5 or create a three-sided square.[7] A god perhaps could make it true that copper no longer conducts electricity but no mortal being has the power to break the laws of nature. In virtue of those laws of nature, I am unable to fly unaided and in some sense perhaps 'unfree' but not in any sense that is important. There are also many thinks I am unable to do because I lack the skills and the means. I am not able to converse freely with a monolingual speaker of Latvian because I know no Latvian. I am not able to fly first-class to Australia tomorrow morning as I lack the financial means. Yet, once again, I am not interestingly 'unfree' even if I really want to fly to Australia first-class tomorrow. What if a prisoner were to say that he lacks the means to leave his cell, this being something he is rather keen to do, and quite certainly is unfree? We should say that I am not being deprived of anything that I could reasonably expect to have whereas he is being so deprived. This raises a complicated question about what a person is by default entitled to and can expect from life, his society and others and it is not one we shall pursue.


Let us now turn to crime. It is physically possible for someone to steal a car but not legally possible. What is possible in a legal sense is determined by what is permitted by the laws of the land. The same goes for a game: if I am abiding by the rules of chess, I cannot make the pieces move however I wish. Of course, I am free to move them however I wish so long as I am not playing chess. Now, do the laws of the land constrain my freedom? In some sense, they do. Do I care? It depends on what I think of the laws. If I am someone who thinks stealing is wrong and who would, as a result, never find themselves wanting to steal a car, then it seems I should not care. If I think stealing is wrong but am easily tempted by unlocked cars, I might be thankful that the law - or, rather, its consequences - exist to stop my giving in to temptation. If I think the law is wrong, I might still endorse it because it has practical value. For example, I might think that in principle, people should be allowed to take whatever drugs they choose but also believe that the people who happen to be in my society would abuse this freedom and turn to crime and so, on balance, I will be content with this restriction on my freedom. If I think the law is plain wrong, then I will be unhappy, as people often are in countries with repressive legal systems.


My possible actions are further limited by a variety of factors that have made me the person I am. Alongside my moral beliefs, I have various tastes and preferences that I did not consciously choose. I enjoy eating tapas in the civilised company of my colleagues and do not enjoy playing rugby. Yet I do not think I am unfree because I cannot happily play rugby. Perhaps my life would be the better for liking rugby but there is a limit to the activities we can indulge in and there is no reason to believe that everyone should enjoy the same things. Similarly, I will not willingly listen to acid jazz, eat tripe, or read celebrity magazines but I do not feel any the less free.


There are thus ever so many closed paths that I am happily indifferent to. Let us therefore consider a situation in which I am not indifferent. It is lunchtime and I am in the sandwich shop thinking about whether to have the ham sandwich (H) or the cheese sandwich (C). I like both. After some deliberation, I choose C. What made my choice free? According to some, it is this:


Alternative Possibility Principle: Where S chooses to ɸ, then S's choice is free just in case S could have chosen not to ɸ. [8]

This has some immediate intuitive plausibility. If I could not have chosen otherwise, then there was really only one option and my choice was forced. Further support comes from what we say about addictions. If I am addicted to cheese, then I would not be able not to choose the cheese sandwich and we would naturally say that my choice was not free as I am in effect a prisoner of my desires. Finally, there is an obvious link between this, freedom, responsibility and desert. If you are responsible for the crime, and deserving of punishment, you must have been capable of not performing the criminal action. You are being punished for choosing - freely - to do it neverthelesx

4. Some Types of Fatalism

In section 2, we introduced the following version of theological fatalism:


Interventionist theological fatalism: the future is under the control of divine beings. Understood in the most extreme way, these divine beings have planned out the entire history of the universe in advance and it is now merely 'unfolding'.

If it is true, then you couldn't have acted otherwise and you are not free. A similar worry arises in the context of the classical Christian conception of god (the so-called god of classical theism). God is omniscient. So, if God knows everything, then God knows what I will do tomorrow. It is a feature of knowledge that it is factive: if S knows that p, then p is true. So, there are truths about what I will do tomorrow known to God. But if there are such truths about my future, then my future is already fixed. If, for example, it is true tomorrow that I will get caught in a storm, then there is nothing I can do to make it false.[9]


We therefore also have:


Omniscience fatalism: if there is an omniscient being, he knows everything and so knows the entire future state of the universe. There is therefore no way the future can be other than it is.

Although they look similar, they are not quite the same. If there are beings who have a limited ability to intervene to fix the future, they may not know the consequences of their actions and hence not be omniscient. In the other direction, to say that God is omniscient is not to say that he has created the history he is omniscient about. Conceivably, an omniscient God could have not created reality: both could be uncreated and eternal entities. This said, people certainly do think of the God of classical theism as the designer and creator and hence as a being which would have knowledge of what will be.


Let us now introduce a non-theological fatalism:


Physical fatalism (determinism): the future is uniquely fixed in virtue of the physical nature of our universe: the strict laws of nature and their governance of all things.

The world in which we live is a very well-behaved world. Things happen, we say, in accordance with inviolable laws of nature. If I hold a book above the floor and release it, it will fall to earth. It will do so at a certain fixed rate of acceleration. Ordinarily, we cannot predict exactly how long it will take to land and where it will end up but we think that the problem is an epistemic not metaphysical one. It is not that it is fundamentally indeterminate how long it will take but that we can't work it out: we simply are not able to gather the data or perform or the calculation.


Let us therefore introduce the dramatic device of a super-being. A super-being, however, who knows the position and properties of every last speck of matter in my environment could work it out. Indeed, this super-being is able to know the total state of the universe at any one time t. This total state will contain data about the position and properties of every bit of matter and whatever else the physicists tell us is essential and fundamental. Let this super being also know all the laws of nature. This being could then calculate the total future state of the universe at any time t. It would be omniscient.


The super-being is a dramatic device because our concern is with the reality it knows. It is a deterministic reality. Each total state of the universe is determined by any past state plus the laws of nature. So, if we accept that the universe came into existence some 13.7 billion years ago, we are left with the unpleasant conclusion that the entire future of the universe was fixed at that very moment long ago.[10] If so, then everything that happens was bound to happen and nothing could have ever been otherwise than it is. If so, then I couldn't have chosen otherwise in the sandwich shop - nor at any other time in my life. If so, I am always unfree.


5. Logical Fatalism Introduced

Having introduced these forms of fatalism, we shall say no more about them. They make some very heavy presuppositions about the nature of reality, both natural and supernatural. Instead, we shall turn to three arguments for a slimmer version of fatalism that assumes as little as possible. We shall call this version of fatalism logical fatalism:


Logical fatalism: logic alone can prove that the sequence of events that is the history of the universe is unique: things could not have been otherwise.

We shall begin with the easiest argument to defuse, the Mortality Argument and work through the Idle Argument to the hardest of all, the Necessity Of The Past Argument.

6. The Mortality Argument

The first argument is the Mortality Argument. You are, as we all are, a mortal being. So you're necessarily only going to live for a certain number of days. But what's necessary can't be changed. So, there's a certain unchangeable number of days that is your lifespan. The thread of your life has already been measured out!


We can lay the argument out thus:


(1)
You are mortal.
(2)
So, you necessarily have a finite lifespan: x minutes.
(3)
So, since x is necessarily your lifespan, it cannot be other than it is.
(4)
So, you have a fixed allotted lifespan.
(5)
So, you are unable to perform any actions to shorten or lengthen your lifespan.


Although it may not be easy to spot where the argument goes wrong, it should at least look rather fishy. For we can run a parallel argument with an even more absurd conclusion.


(1')
You are a human being.
(2')
So, you necessarily have a finite height: x centimetres.
(3')
So, since x is necessarily your height, it cannot be other than it is.
(4')
So, you have a fixed height.
(5')
So, you are unable to grow or shrink.


You surely believe that you are taller than when you were six years old and so either you must conclude that it is an illusion that you have grown or that the argument is problematic.[11]


The argument exploits the fact that the 'natural' languages we speak, such as English and French, do not always make logical distinctions clear. Consider the following sentence:


(SONG)
Everyone loves a song.


There are two ways of interpreting this sentence:


(SONG1)
There is a certain song loved by everyone.
(SONG2)
Everyone loves some song or other (so not necessarily the same song.)


The same sort of ambiguity occurs with the sentence:


(HEIGHT)
You necessarily have a height of x centimetres.


It can be interpreted in two ways:


(HEIGHT1)
There is a certain height x that you necessarily have.
(HEIGHT2)
Necessarily, you have some height or other x.


The sensible way to interpret it is (HEIGHT2). The argument 'works' because it moves from the sensible interpretation (HEIGHT2) in (2) to the silly interpretation (HEIGHT1) in (3) without any justification. It is simply pulling a trick. So, we can reject it:


(1')
You are a human being.
(2')
So, you necessarily have a finite height: x centimetres.
(3')
So, since there is a height x you necessarily have, it cannot be other than it is.
(4')
So, you have a fixed height.
(5')
So, you are unable to grow or shrink.


In the same way, we can dispense with the original argument because it confuses:


(LIFE)
You necessarily have a lifespan of x minutes


(LIFE1)
There is a certain lifespan x that you necessarily have.
(LIFE2)
Necessarily, you have some lifespan or other x.


The argument is therefore as follows:


(1)
You are mortal.
(2)
So, you necessarily have a finite lifespan: x minutes.
(3)
So, since there is a lifespan x you necessarily have, it cannot be other than it is.
(4)
So, you have a fixed allotted lifespan.
(5)
So, you are unable to perform any actions to shorten or lengthen your lifespan.


7. The Idle Argument

The Idle Argument comes in many forms and has a long history. Here is a version of it that employs the original idea. Bill is ill and thinking of sending for the doctor. His friend the fatalist philosopher Fred argues that there is no point as follows.


(1)
Tomorrow, either you will be better or you will not be better.
(2)
If you are not better tomorrow, then sending for the doctor was a waste of money.
(3)
If you are better tomorrow, then sending for the doctor was unnecessary.
(4)
So, either way, sending for the doctor is pointless! What will be will be!


Let us start with (1). We could deny it. We could say that the future is undetermined. If so, it is indeterminate whether I will get better and indeterminate whether I won't. So, it is indeterminate whether I will or I won't. Such a strategy is unattractive. Even if the future is open, it still seems right to say that one of the other situations must obtain tomorrow.


Let us now consider (2). (2) seems a perfectly reasonable thing to say. So, the problem must lie with (3). And it seems obvious that it is. If Bill is better tomorrow, then it may be just because he sent for the doctor. So, the argument fails because one of the premises is false.


This is, ultimately, the right thing to say. We cannot dispense with the argument quite so quickly, though. The key words in (3) are 'if' and 'unnecessary' and they need to be examined further.


The presence of oxygen is a necessary condition for combustion. If there is a fire, therefore, we can deduce that there is oxygen. So, we say:


(7.1)
P is a necessary condition for Q means if Q, then P.


So if P is unnecessary for Q, then this means that the truth of P has no bearing on the truth of Q. P will be true regardless of whether Q is true.


(7.2)
P is unnecessary for Q means (i) if Q then P and (ii) if not Q then P.


So, to say that the doctor is unnecessary for Bill's getting better is to say:


(7.3)
If Bill calls the doctor, Bill will be better tomorrow.
(7.4)
If Bill doesn't call the doctor, Bill will be better tomorrow.


So, to say that if Bill is better tomorrow, then calling the doctor was unnecessary is to say:


(7.5)
If you are better tomorrow, then [{if you call the doctor you will be better tomorrow} and {if you don't call the doctor, you will be better tomorrow} ]


We can re-express this more clearly as:


(7.6)
If you are better tomorrow, then whether you call the doctor or not, you will be better tomorrow.


The fatalist will now argue that (7.6) is true. Let us think through it slowly. You will agree that if Bill is better tomorrow, then Bill is better tomorrow. This is trivially true. It expresses a basic logical truth:


(7.7)
If P, then P.


Here, and in what follows, capital letters 'P' and 'Q' are dummy letters for arbitrary sentences (There's a bit more on the logic of 'if' in the Appendix.) Now, suppose that Bill is better tomorrow and the doctor was called. Does it follow that Bill is better tomorrow. Of course it does - that is part of our supposition! Similarly, suppose that Bill is better tomorrow and the doctor was not called. It still follows logically that Bill is better tomorrow. In other words, this is trivially true too:


(7.8)
If P, then (whether Q or not-Q), P.


If so, the fatalist has won. Calling the doctor is unnecessary. Indeed, we can apply the same logic to (2) which we passed by quickly. Whether the doctor was called or not, if Bill is not better tomorrow, then Bill is not better tomorrow. (One might say this to Bill, perhaps, if he is worried that the doctor would make things worse.) So, perhaps what will be will indeed be.


We can agree with the fatalist about this last claim. What will be will indeed be. This is a trivial truth. It is simply a version of (7.7):


(7.9)
For any proposition P, if P then P.


So, where's the mistake? Let us suppose that Bill called the doctor and that Bill is better the following day. So, both P and Q are true. Now, in order to establish that calling the doctor was unnecessary, what do we need to establish? We said above that the following need to be true:


(7.3)
If Bill calls the doctor, Bill will be better tomorrow.
(7.4)
If Bill doesn't call the doctor, Bill will be better tomorrow.


We know (7.3) is true. Do we know (7.4) is true? Notice first of all how odd it would be to say (7.4) after learning that Bill did call the doctor and did get better. An 'if...then...' statement is called a conditional and is composed of two parts: (i) the antecedent, which is the bit that follows the 'if' and the consequent which is the bit that follows the 'then':


IF-THEN
If antecedent, then consequent


Now, you would naturally utter (7.4) if you didn't know whether Bill had or was going to call the doctor. (For example, you might say it if you think that the doctor was dangerous). If you knew that Bill had called the doctor and wanted to say that doing was unnecessary, you would naturally say:


(7.10)
If Bill hadn't called the doctor, he would (still) have been better.


There is an important difference between (7.4) and (7.10). (7.4) is an indicative conditional. You don't utter a conditional like (7.4) if you believe that the antecedent is false. You use it because you are ignorant of whether the antecedent is true or false. You suppose the antecedent is true rather than false. (7.10) is a subjunctive or counterfactual conditional. You use such conditionals when you do know the antecedent is false as you want to say something not about what actually happened but about a different possibility reality - the way the world would have been if so-and-so had been the case.


The fatalist, of course, believes that there is no such thing as the way the world might have been. There is just the way things are and will be. So, the fatalist has to say that the only way to understand 'if...then...' is as an indicative conditional. They have to deny that we can meaningfully talk about the way things might have (not) been. They cannot of course use the argument we have just looked at because that only works if we accept that 'if' is to be understood their way and with it 'necessary'. So, they need a new argument.


As it happens, analysing subjunctives is a tricky business. We know that the antecedent is false. Knowing that the consequent is true or false in the actual world does not give us a clear indication as the following examples will make clear:


(7.11)
If I had grown up in Latvia, I would speak Lativan.


'I speak Latvian' is false in the actual world - i.e. actually false, as I don't speak it. It seems plausible to say that I would speak Latvian (had I grown up in Latvia) and so that the conditional is true.


(7.12)
If I had grown up in Latvia, I would speak English.


'I speak English' is true in the actual world. But it is plausible to say that if I had grown up in Latvia I would speak English as well because all children learn it in school.


Here we must leave the topic of conditionals (as mentioned above, a little more detail can be found in the Appendix). Whereas we dismissed the Mortality Argument rather comprehensively, the dismissal of this argument rests ultimately on making sense of subjunctive conditionals. Our final word on it will be that it certainly seems as if we do make sense of them and also that we need to make sense of them because talk of possibility is present in a lot of places:

  • Causation: some say that to say that X causes Y is to say that had X not occurred, Y would not have occurred.
  • Mathematics: mathematical truths are necessary truths, that is truths that could not possibly be false.

8. The Necessity Of The Past

The final argument for fatalism, the Necessity Of The Past argument, likewise comes in many forms and is associated in particular with Aristotle and Diodorus Cronos.[12]


On the 1st of June 2008, Paul and Steve are wondering whether it will rain in exactly a year's time in the spot where they are standing. For fun, they decide to bet. Paul says that it will and Steve says that it won't. Exactly a year later, it is raining on the day in question and Paul wins his bet. Explaining this to Jana, Steve says that, a year ago, Paul said, 'It will rain on this day in a year' and that what he said was true. Jana points out that it was fated that Steve would lost. When the present becomes the past, what is true becomes necessarily true. Nothing, Jana points out, can change now the fact that Steve began telling her this story thirty seconds ago. It is a fixed and frozen fact. So, since Paul uttered something true last year, it too was, thereafter, necessarily true. So, it was always going to rain.


But, of course, we needn't suppose there was anything special about Paul or last year. Had anyone uttered the sentence 'it will rain on 1/6/2009' in the past, it would have been true and hence necessarily true. And had anyone uttered anything about any other future event that happened, it would be necessarily true. the same goes for falsehoods. Steve uttered something false in 2008 and necessarily false. Here is a formal representation of the argument:


(1)
For any proposition p, either p is true or not-p is true.
(2)
What's past is fixed: truths about the past are necessarily true.
(3)
On 1/6/2008, Paul says, 'On 1/1/2009, it will rain'
(4)
On 1/6/2008, Steve says, 'On 1/1/2009, it will not rain'
(5)
Either what Steve says on 1/6/2008 is true or what Paul says on 1/6/2008 is true.
(6)
If it true on 1/6/2008 that it will rain on 1/6/2009, then it is necessarily true that it will rain on 1/6/2009.
(7)
If it true on 1/6/2008 that it will not rain on 1/6/2009, then it is necessarily true that it will not rain on 1/6/2009.
(8)
Either it is necessarily true that it will rain on 1/6/2009 or it is necessarily true that it will not rain on 1/6/2009.


It is tempting to reply to this argument by saying that what Paul or Steve said became true at the appropriate time in the future: it was not true when they said it. Now, there is a sense in which this is true but one has to be careful. It is important to distinguish between the sentence that someone utters and the proposition that they use the sentence to express. A proposition cannot become true or false: it simply is true or false. A sentence can be used at one time to express a false proposition and at another to express a true proposition and in that sense 'become' true. But this is irrelevant. Let me explain.


A proposition is the 'picture of reality' that a sentence expresses. So, the following different sentences all express the same proposition:

  • Badgers are mammals. (English)
  • Borsuki są ssakami. (Polish)
  • Les blaireaux sont des mammifères. (French)
  • Meles mammalia sunt. (Latin)

The same sentence can also express different propositions. If Paul utters the sentence 'I like sushi' he expresses the true proposition that Paul likes sushi. If Steve utters the same sentence, he expresses the false proposition that Steve likes sushi.


A proposition is also often much richer than a sentence. If Steve says, 'I am hot' on a Monday in Twickenham and says it again in Streatham on the next day, Tuesday, he does not say the same thing. He expresses the propositions that Steve is hot on Monday 1/6/2009 in Twickenham, UK and that Steve is hot on Tuesday 2/6/2009 in Streatham, UK. The proposition that the sentence expresses 'fills in' the time, place and other information needed for us to get something that can be true or false.[13]


A true proposition is also called a truth or a fact. Now, a fact can't become true. A fact just is true. If a fact became true, it would have been false beforehand. Consider the fact that Mat is sitting in Twickenham on 1/6/2009 typing. If this fact became true, then the following is true:


(8.1)
It was not true that Mat was sitting in Twickenham on 1/6/2009 typing and then it was true Mat that was sitting in Twickenham on 1/6/2009 typing.


This is just a contradiction. Perhaps you want to say that the proposition was neither true nor false but indeterminate. It was only when the right moment happened that reality decided what truths at that time were indeed truths. Perhaps. But the price of this is to say that no-one can truthfully say anything about the future. For whatever is said in the past about the future must be indeterminate. This is not an impossible line to take but one we won't pursue.


So true propositions don't become true propositions. By the same logic, false propositions don't become false. A sentence can 'become' true in the following way. Suppose it isn't raining. Then 'it is raining' does not express a true proposition. Five minutes later, it starts raining. Now the same sentence does express a truth. But it is not expressing the same proposition as it did before - two different times are involved. So, there is no real change in truth here - only a change in what is being said.


With the difference between a sentence and a proposition cleared up, we can now generalise. We can drop the need for people actually to say sentences that express true or false propositions. A truth is a truth and a falsehood a falsehood regardless of whether anyone bothers to say a sentence that expresses it! So, finally, every truth about the future relative to any past moment of time is necessarily true or necessarily false. Supposing, for convenience, a beginning to the universe, all truths about the history of the universe from that moment on are necessarily true or necessarily false. The future is fully fated.


Where now? We are accepting that every proposition is true or false and that propositions do not become true or false. Surely the claim that what's past is fixed - that past truths are necessary truths - is safe. Not even God can change the fact that I was here five minutes ago. But in fact it is just here that the problem with the argument might lie.


9. An Ockhamist Approach To The Necessity Of The Past Argument - I

The approach I am about to outline has it roots in the work of William of Ockham, an early 14th century logician.[14] Ockham was trying to resolve the earlier theological problem of reconciling God's omniscience and our freedom. Ockham made a distinction between two types of statements about the present. One type of statement is really about the present whereas another is really about the future. Suppose we are watching the end of the London Marathon. Bernard is a long way in the lead. He has only 100 metres to run. Alf, a commentator, says:


(9.1)
Bernard is smiling to the crowds in the Mall.


Charlie, another commentator, says:


(9.2)
The winner of the 2009 London marathon is tired but happy.


While neither of the descriptions apparently make reference to future events, (9.2) is not a genuine statement about the present. By describing Bernard as the winner, it is really talking about the future as Bernard has not in fact won yet. By contrast, (9.1) is a genuine statement about the present.


When the present moment passes, present truths become fixed and necessary - but only those genuine present truths, those expressed by genuine present-tense statements. Once the precise moment at which (9.1) has been uttered has passed, that Bernard was smiling at that moment is now part of the unchangeable past. It is what has come to be called a hard fact or hard truth. But once the moment of utterance of (9.2) has passed and supposing Bernard does in fact win, the truth expressed by (9.2) is not part of the unchangeable past. That truth is that the winner of the 2009 London Marathon was tired but happy. It is what we call a soft fact or soft truth. It is not necessarily true once the moment has passed. For its truth depends on what happens in the future. It remains soft until Bernard has in fact won. And, crucially, in remaining soft, it remains possible that Bernard does not win. He could be struck down by lightning, for example, a metre from the finish line.


Now, why does this matter? Recall our puzzle. Imagine it is 1/6/2008. Paul says, 'it will rain on 1/6/2009'. We suppose it does. So, it is true on 1/6/2008 that it will rain on 1/6/2009. But, from the point of view of 1/6/2009, that truth is a past truth and a necessary truth.


We can now say that the truth that on 1/6/2008 it will rain on 1/6/2009 is a soft fact. The truth is clearly one that is looking to the future. It is not about the present moment at all. What is true about 1/6/2008 and true in a hard way was that Paul said, 'it will rain on 1/6/2008', that Paul was wearing a white shirt, that Steve ate lasagne for lunch, and so on.


In general, a statement of the form 'at t1, it was true at t2 that p' where t1 is earlier than t2 and p is indeed true at t2 expresses a soft truth: a truth seemingly about the present moment t1 but which is really about a future moment t2. We must resist the thought that once t1 has passed, these truths become necessary truths. They remain soft. We do believe that once the moment has passed, the past is fixed. But all we need to say to justify this intuition is that the truths that are genuinely about that present moment become fixed when the moment has passed. These truths are those expressed by sentences that describe the world in 'blinkered' terms that don't look forward to where things are heading. It is not easy to say just how we are going to do this. We naturally do describe things in the past in terms that sneak in a reference to the future, sometimes overtly, sometimes covertly. I shall end this section with some examples of how easy it is to do. In the next section, we finish off the solution by thinking about what makes true sentences true.

  • In 2001, before I met her, my girlfriend was studying classics at university.

    • Not a present-tense statement relative to 2001 because it describes a woman as my girlfriend, something she would only later become.

  • In 1997, Prime Minister Brown did not realise how long he'd have to wait stop being chancellor.

    • Not a present-tense statement relative to 1997 as it describes Brown Prime Minister, something he'd only become in 2008.

  • In 1564, Britain's most famous playwright William Shakespeare was born.

    • Not a present-tense statement relative to 1564 as he only became the most famous playwright thanks to future events.

10. An Ockhamist Approach To The Necessity Of The Past Argument - II

The solution we have before us runs as follows. Every statement is true or false. At a given moment t, a statement made about the future is true or false but it is not necessarily so. Such statements express soft facts. So, when Paul said truly on 1/6/2008 that it would rain, it was not necessarily true that it would rain. It could have turned out not to rain. Had Paul said on 1/6/2008 that Mat would be typing on 1/6/2009 he would have said something true. But for all we've said, I could have decided not to. So, I could have made it true this morning that what Paul said on 1/6/2008 was false. This is not some magical power I have to reach back into the past and change things. Remember that the truth Paul said something true is a soft truth: right up until the moment I decided to type, it could have been false.


It is natural to feel a nagging doubt that all of the above dissolves the puzzle by defining special terms - 'hard fact', 'soft fact' - and thus, to borrow the oft-used words of the philosopher Russell, has all the advantages of theft over honest toil. There are two concerns you might have which I shall take in turn.


Consider Paul again on 1/6/2008. He said something about 1/6/2009. It was true. Now, I said earlier that a truth can't become true. It simply is true. So, since it can't be become true, it must have always been true, in which case it is necessary after all!


It is true that a truth can't become true. This is just to say that if p is true, then p is true, as opposed to (e.g.) if p is true, then it might have once been false. A truth can't change its spots, so to speak. But it is quite consistent with this to say that p could have been false: that the truth might not have existed, to put it a little dangerously. I exist as a human and hence it is false that I might have earlier been a snail: our understanding of biology and physics rules that out. But I could have not existed: my parents may never have met. So, if it does indeed rain on 1/6/2009, then that truth is an unchangeable truth but there could have not been such truth at all.


The second worry runs thus. If Paul says something true in on 1/6/2008 about 1/6/2009, then even if what he says about 1/6/2009 isn't then necessarily true, it is still true. Paul said something true on 1/6/2008 and this is because on 1/6/2009 it is raining. So, the future does in some sense have an effect on the past.


Yes and no. We need to be careful here with the words 'because' and 'effect'. For they encourage us to think in causal terms: the future rain (on 1/6/2009) causes the past utterance to be true. But we understand causation to work in the other direction: causes come before their effects. Indeed, this seems connected with our very understanding of time itself. We determine the temporal order of things via the causal order of things. Imagine yourself sorting out a jumbled-up pile of photos of a fire in a house: you would try to construct the time sequence by thinking about the order in which things happened.


Now, the word 'because' doesn't have to suggest causation. For example, I could say that 17 is a prime number because it isn't divisible by any integers other than 1 and itself. Yet 1 and 17 aren't causing it to be prime in the way that two violins are causing the noise in the concert hall. The use of 'because' here indicates a logical relation. Here are two more examples:

  • Steve is at home because it is a Monday and if it is a Monday, Steve works at home.
  • The triangle is equilateral because the internal angles are each 60o.

One clear indication that we're not dealing with genuine causation here comes from Hume's observation that causes and effects are 'distinct existences'. If X is the cause of Y, then this is not something we can deduce as a matter of logic because it is always conceivable that X could happen without Y. It is conceivable that I could jump out of the window and fly, for example, or that I could throw a brick at my window and it would pass through. Unlikely, perhaps, but conceivable. By contrast, it is not conceivable that 17 is not prime or that the internal angles of an equilateral triangle are each not 60o.[15]


Consider the relationship between the true sentence 'Mat is typing on 1/6/2009' and the reality that makes it true. This is not a case of cause and effect. If so, it is conceivable that the reality could be what it is and the sentence could be false. Or, alternatively, that the sentence could be true and reality otherwise. This is nonsense. The connection between reality and true sentences is much closer and will not allow that. What is the connection if it is not causal? This is just another way to ask the question, 'What is truth?' and it is far too big a question to be dealt with here.


We have seen that Paul's utterance said something true but the truth was not necessary. The future could have been such as to make what Paul said false. As it happens it is true and, yes, the future does make Paul's past utterance true. But there is no mysterious causation here. We have a puzzle of an importantly different kind. For the puzzle concerns how reality makes any sentence true. It is just as much of a puzzle how reality is making the genuine present-tense statement 'Mat is typing true' as it makes Paul's past utterance true. But this is a problem quite different from the problem of logical fatalism.


11. Conclusion

As I said at the start, to dismantle the fatalist's arguments is not to disprove fatalism. There is no way to disprove fatalism. For the fatalist can always say, once the moment has passed and the past is fixed - something we all agree on - that history as it was is the only way it could have been. We can't go back in time and try to do different things to prove him wrong.


The anti-fatalist needs to supply an alternative. One strategy is to say that we have good evidence that the world is organised not by the hand of fate but by impersonal laws of nature. If fate could make the future turn out any particular way, we could not predict the future. Yet we take it that we can. We live our lives relying on the assumption that metals will still conduct electricity tomorrow, that water will still be essential for life, and so on. The danger with this, however, is that we end up as physical fatalists, or determinists. As it happens, there are people who think that this scientific picture is correct and yet that it allows us to be free. They think that it a mistake to think that freedom consists in being able to do otherwise. Freedom is essentially a matter of being able to do what you want and if your desires are determined, so what? You do not need to be a philosopher to realise that much of what you like and dislike has been determined by factors in your past over which you have had not control.


The second strategy is to defend the view that we could have acted otherwise. Common-sense and practice supports this, as we saw earlier. The chief problem with this strategy is to make good sense of the idea that we could have acted otherwise in exactly the same situation. It is perfectly sensible to think that, if I choose the ham over the cheese sandwich that I did not have to choose it in the sense that I am not always someone who chooses ham over cheese. I am not a ham addict. This is compatible with saying that my desires sometimes push me one way, sometimes the other. It is much less appealing to say that had I been in the exact same situation, I could have chosen differently. For if I had all the same desires and deliberations, wouldn't I be irrational - or worse, in no control of my mind - if I could just as easily act one way or the other?


Developing and defending one of these strategies is what the free will debate is all about once the problem of logical fatalism has been shown to be something we needn't accept. But this is a topic for another occasion.




Matthew Carmody
Richmond upon Thames College
matthew.carmody@rutc.ac.uk



References and Further Reading

For more introductory material on Fatalism, see:

  • Ayer, A.J. (1963) 'Fatalism.' In his Concept of a Person and Other Essays. Macmillan: St. Martin's Press, 3-20.
  • Bernstein, M. (2002) 'Fatalism.' In Kane (ed.) (2002) Oxford Handbook of Free Will. Oxford: Oxford University Press., pp. 65-81.
  • Stanford Encyclopaedia of Philosophy: Fatalism
  • van Inwagen, P. (1983) An Essay on Free Will, Oxford: Clarendon Press, pp. 23-54.

Ayer and van Inwagen briefly discuss the Idle Argument. For a different approach to it, see the opening pages of:

  • Dummett, M. (1964) 'Bringing About the Past', Philosophical Review 73, 338-359.

For an extremely detailed analysis of various types of the argument and discussion of fatalism and determinism in antiquity, the following is excellent:

  • Bobzien, S. (1998) Determinism and Freedom in Stoic Philosophy. Oxford: Oxford University Press.

Cicero talks about the argument in a general discussion of fate:

  • Cicero, De Fato, trans. H. Rackham. Cambridge, Mass: Harvard University Press, 1982.

For more on The Necessity of the Past Argument and the Hard Fact/Soft Fact approach, see the papers in Part 2B of the collection below:

  • Fischer, J. M. (ed.) (2005) Free Will: Critical Concepts in Philosophy. London: Routledge.

In particular, consider the following, especially the Freddoso, which presents a lucid and detailed version of the response I outlined to the argument.

  • Pike, N. 'Divine Omniscience and Voluntary Action'. Philosophical Review 74 (1965): 27-45. Reprinted in Fischer (2005), pp. 254-269.
  • Hoffman, J. & Rosenkrantz, G. (1984) 'Hard and Soft Facts' Philosophical Review 93: 419-434. Reprinted in Fischer (2005), pp. 291-303.
  • Freddoso, A. J. 'Accidental Necessity and Logical Determinism.' Journal of Philosophy 80 (1983): 257-278. Reprinted in Fischer (2005), pp. 304-324.

For more on Theological Determinism, see

Notes

[1]  The word 'destiny' comes from the Latin verb destinare meaning to bind.  [back] 


[2]  See Hesiod, Theogony, 215. (Read the text at the Perseus Project.) As so often happened, the goddesses reappear in Roman mythology as the Parcae: Nona (the spinner), Decuma (the determiner) and Morta (the cutter.)  [back] 


[3]  The oracle at Delphi was a priestess who channelled the god Apollo. An interesting but controversial idea is that the predictions of the priestess came from visions caused by gases that emerged from the ground on which the temple was built - read more here.  [back] 


[4]  Cicero, On Fate, III.  [back] 


[5]  W. Somerset Maugham, Sheppey (1933)  [back] 


[6]  Borges, The Garden of Forking Paths.  [back] 


[7]  For discussion, see Paul Sperring's RJP article Descartes, God and the Eternal Truths (PDF).  [back] 


[8]  The Greek letter phi - ɸ - here stands for any action.  [back] 


[9]  There's a relatively simple puzzle in this area we can dismiss easily. Knowledge is factive: if you know something, what you know is true. This is easy to show. Consider how odd it sounds to say, 'I know I left my keys at home but it turned out I took them with me.' By contrast, it is fine to say, 'I thought I left my keys at home but it turned out I took them with me.' Since knowledge is factive, if God knows what I will do X tomorrow, then it is true that I will do X tomorrow. But, equally, if you know that I will do X tomorrow, then it is true that I will do X tomorrow. This doesn't show you have the power of foresight! If I will do X tomorrow, whether freely or otherwise, and you have evidence that I will, then you can know that I will do X tomorrow. I make your knowledge happen, so to speak - your knowledge doesn't make me do what you know. How can you have sufficient evidence what I will do tomorrow? That's another question. Perhaps you can't in this case. But some knowledge of the future seems possible. If you and a friend are watching a film you both enjoy, you can know at any time that your friend will still be watching the film a second later, even though he could in principle freely decide to stop.  [back] 


[10]  Two short notes on two big topics. First, quantum physics. The idea that the universe runs along perfectly strict and in-principle predictable lines that characterises the idea of determinism belongs in the 18th century when it was first propounded. It is most famously associated with the philosopher and mathematician Pierre Simon de Laplace:


We ought to regard the present state of the universe as the effect of its antecedent state and as the cause of the state that is to follow. An intelligence knowing all the forces acting in nature at a given instant, as well as the momentary positions of all things in the universe, would be able to comprehend in one single formula the motions of the largest bodies as well as the lightest atoms in the world, provided that its intellect were sufficiently powerful to subject all data to analysis; to it nothing would be uncertain, the future as well as the past would be present to its eyes. The perfection that the human mind has been able to give to astronomy affords but a feeble outline of such an intelligence. (Laplace (1820), Essai Philosophique sur les Probabilités forming the introduction to his Théorie Analytique des Probabilités, Paris: V Courcier; reprinted in F.W. Truscott and F.L. Emory (trans.), A Philosophical Essay on Probabilities, New York: Dover, 1951.)


The 20th century has seen the scientific revolution that is the discovery of the quantum physical nature of reality. It turns out that at the supermicroscopic level, reality is not determinate but indeterminate or random. It is not possible to know everything about a fundamental particle such as an electron: in particular, its location and velocity (more accurately, still: its momentum). This is not because we're not smart enough to work it out. Nor is it the case that we can't measure the properties of the electron because we don't have accurate enough measuring equipment. Nor still is it the case that what we mean is that our attempts to measure it disturb the electron in the same way (for example) that sticking a thermometer into a liquid will affect the temperature of liquid slightly and prevent a perfect measurement of the temperature as it was. This does happen as well, but it is not what is of central importance. It seems that nature really is unsettled in certain respects independently of us. Now, if this is true, then it seems clear that the very idea of determinism must go. In fact, it's a little more complicated than that. For even though nature is apparently indeterminate, it is indeterminately determinate! This isn't as paradoxical as I've made it sound. If you roll a die, you can't know what face will land upright but you know that any face has a 1 in 6 chance of appearing. Similarly, although one cannot know precisely everything about an electron, it turns out that one can know the probability that it has certain properties. So, nature may unfold along determinate lines after all, albeit in a slightly different sense.


All of the technical complexities are of secondary importance in the issue of whether we are free or fated. Although we feel disturbed by the idea that the universe unfolds in a strict and determinate way, freedom is not restored by randomness. If nature is such that what will happen tomorrow is essentially indeterminate, then I still have no hand in shaping it. It would be a little like living one's life by the role of a die!


For more on determinism, see the Stanford Encyclopaedia of Philosophy entry. (It is rather challening for the beginner, though.) For an accessible book on quantum physics, try John Gribbin's Schrodinger's Kittens and the Search for Reality.


Second, a quick note on the laws of nature. It is not supposed by modern physicists that the laws of nature came into existence ready formed at the moment of creation but were shaped in the initial moments of the universe (and by 'moment' here, I mean a very, very thin sliver of a second.) A good book on this (despite being twenty years old, which is a long time in modern physics!) is Steven Weinberg's The First Three Minutes.  [back] 


[11]  The Eleatic philosophers in Ancient Greece, such as Parmenides and Zeno, actually denied that genuine change was ever possible: all change is merely apparent.  [back] 


[12]  Aristotle's famous argument is know as the Sea Battle argument and it features in De Intepretatione, Book IX. Diodorus Cronos' argument is known as the Master Argument.  [back] 


[13]  The traditional view of propositions is 'eternalist'. If I say now, "I am listening to the radio", I express a different proposition than if I repeat those words a minute later. The first proposition was Mat is listening to the radio at 23.14 on 6th June 2009 and the second is Mat is listening to the radio at 23.15 on 6th June 2009. A proposition typically includes reference to a time and so it is not possible to have the same proposition expressed at different times. (Not all propositions include reference to a time. Some propositions are timeless: e.g. 2+2=4, whales are mammals.) Some philosophers - 'temporalists' - think that if I say, "I am listening to the radio" at two different times, I do express the same proposition and it is true at one time and not at another. However, even if this is the right way to think, it won't provide an easy way out of the Necessity of the Past argument. It can be rephrased entirely in terms of eternal proppositions.  [back] 


[14]  For more on Ockham, read the Stanford Encyclopaedia of Philosophy entry.  [back] 


[15]  Briefly, Hume's point is the following. Suppose I hold a pen in my hand above the floor. What will happen if I let it go? We naturally think that it will fall to the ground. But the only reason we think that is because that's what has happened up to now. There's nothing logically incoherent in the idea that something else could happen. It doesn't help to say that we've proved that there's a law of nature that means that the gravitational force between Earth and pen will cause it move to Earth (and the Earth to the pen, remember.) Our knowledge of the laws of nature is based on past experience too. We have no guarantee that the laws of nature will hold the same way tomorrow. The past is what we use as a guide to the future but we can never treat it as logically faultless guide to the future.  [back] 






Appendix

The Idle Argument makes essential use of conditionals: sentences of the form 'if...then...' As we said above, a conditional is composed of two parts: the antecedent ('if it rains tomorrow...') and the consequent ('then I will stay indoors'). Some conditionals are indicative conditionals. Here are some examples:

  • If it rains tomorrow, I will stay indoors.
  • If it is Thursday, Jeeshan is teaching in room 2E7.
  • If your keys are on the table, then mine will be in my pocket.

Where have an indicative conditional, we do not know whether the antecedent is true. We assert conditionals to express what will happen if it turns out that the antecedent is true.


Under what conditions is a conditional true? Suppose I tell you that if is rains tomorrow, I will stay indoors. Suppose now it rains tomorrow - and so that the antecedent is true.

  • Case #1: I stay indoors. You should say that I spoke truly.
  • Case #2: I go out. You should say that I spoke falsely.

Logicians work out how complex sentences are true and false by building up 'truth-tables'. A truth-table contains all the possible combinations of truth and falsity for the basic sentences, from which we can work out the truth and falsity of more complex sentences. So far, we have built half of the truth-table for 'if...then...':



AntecedentConsequentIndicative Conditional
TrueTrueTrue
TrueFalseFalse



What if the antecedent is false? I say that if it rains tomorrow, I will stay in and then it turns out fine. Logicians argue that we should say that the conditional is true. The complete truth-table is therefore:



AntecedentConsequentIndicative Conditional
TrueTrueTrue
TrueFalseFalse
FalseTrueTrue
FalseFalseTrue



Now, you may find it very odd to say that 'if' should have this interpretation. You would not be alone. There are two ways to defend it.


First, a logical one. If I say, 'if it rains tomorrow, I will stay indoors', I am saying that the following will not happen: that it will rain and I will not be in doors.


Let us use some symbols to make things (hopefully) a little clearer. By the dummy letters 'P' and Q' we shall understand any (assertoric) sentence. So, I might say that 'P' is 'it is raining'. We now understand 'P & Q' to mean 'P and Q', '~P' to mean 'it is not the case that P', and 'P Q' to mean 'if P, then Q'. So, if 'P' is as above and 'Q' is 'I will stay in', then (e.g.) '~Q' is 'It is not the case I will stay in' (i.e. 'I will not stay in'), 'P & Q' is 'It is raining and I will stay in' and 'P Q' is 'If it is raining, I will stay in'. The logical defence can now be put thus:

  • P Q     means the same as    ~(P & Q).

Now, conditions under which ~P and P & Q are true are not controversial. ~P is true iff P is false; and vice versa. P & Q is true iff both P and Q are true and false otherwise:



PQ~QP & ~Q~(P & Q)P Q
TrueTrueFalseFalseTrueTrue
TrueFalseTrueTrueFalseFalse
FalseTrueFalseFalseTrueTrue
FalseFalseTrueFalseTrueTrue



As you can see, the conditions under which ~(P & ~Q) and (P Q) are true are the same.


This can be informally supported in the following way. When I assert a conditional, I am in effect concerned to rule out the possibility that the antecedent is true and the consequent false. If I say that if it rains, I will stay in and then, when it rains, I go out, I have made a mistake (be it through lying or through ignorance). I must retract what I said. By contrast, if I say that if it rains, I will stay in and it turns out sunny, I don't have to retract what I say. I can do what I like: stay in or go out. So, since what I said wasn't false, it was true. Its truth is harmless because we are (almost) never in fact in the situation of asserting a conditional whose antecedent we believe to be false. If it is sunny now, what point could there be in saying, 'if it is not sunny...'?


A second informal one. I just said that we are almost never in the situation of asserting a conditional whose antecedent we believe to be false. Sometimes, we do, for dramatic effect. For example, I might say, 'if he turns up on time, then I'll eat my own legs!' My assertion is honest. I am trying to say something true. So, a conditional with a false antecedent is a true conditional.


The conditional just described is referred to as the material conditional. Despite what I have just said, it is controversial to claim it is the correct interpretation of the English 'if'. But if we accept that it is, we can easily show that if P then (if Q then P) is a logical truth, as promised earlier in section 7:



PQQ PP (Q P)
TrueTrueTrueTrue
TrueFalseTrueTrue
FalseTrueFalseTrue
FalseFalseTrueTrue



You will recall that the Fatalist has to show that if I get better tomorrow having called the doctor, he would need to prove that I would have still got better had I not called the doctor.


How might we represent this logically? The fatalist might suggest the following:

  • if [doctor called and get better] then [if doctor not called then get better]

Now, it turns out that this expresses a logical truth. It is an instance of:

  • (P & Q) (~P Q)

The truth table for this is given below:



PQP & Q~P~P Q(P & Q) (~P Q)
TrueTrueTrueFalseTrueTrue
TrueFalseFalseFalseTrueTrue
FalseTrueFalseTrueTrueTrue
FalseFalseFalseTrueFalseTrue



Now, if this is a logical truth, then so too are the following which capture the other possibilities:

  • if [doctor not called and get better] then [if doctor called then get better]
  • if [doctor called and not get better] then [if doctor not called then not get better]
  • if [doctor not called and not get better] then [if doctor not called then not get better]

(You can try proving that they are yourself by constructing truth-tables. Alternatively, you can find them laid out in the following footnote.[1]). So, has the fatalist proven after all that whatever I do is pointless? No. Suppose I call the doctor and get better. We are not concerned with the first of the following conditions but the second:

  • if I did not call the doctor then I will get better.
  • if I hadn't called the doctor, I would have got better.

The second conditional is not an indicative conditional but a subjunctive conditional. In a subjunctive conditional, the antecedent is presumed false.[2] Here are some examples:

  • If it had rained today, I would have stayed indoors.
  • If you were more helpful, you'd have got me some coffee.
  • If you'd grown up in Latvia, your would have spoken Latvian.

When we assert an indicative conditional, we are concerned with the actual world. The point of asserting an indicative conditional is to say that I will do something if a certain situation is true and I don't know whether it is true. If I did know that it was true, I would simply say what I am going to do. When we assert a subjunctive conditional, we are concerned with what is non-actual - what might have been.


Now, whether a subjunctive conditional is true cannot be determined by considering the truth-values of its parts. Since, by definition, the antecedent is false, there are just two possibilities.


1.  The consequent is true/false and so the conditional is false/true.
2.  The consequent is false/true and so the conditional is true/false.


Neither will work. Let's start with the first. Consider the following subjunctive conditionals:


A.  If I had been born in Latvia, I would have spoken Latvian:
     [I was born in Latvia: FALSE], [I speak Latvian FALSE]

B.  If I had been born in Latvia, I would have spoken English.
     [I was born in Latvia: FALSE], [I speak English: TRUE]


Is A true or false? It is tempting to say that it is true - but perhaps I might have been born in Latvia to an English family who never let me mix with native speakers. If we say it is true, then we should say that B is false (in line with 1 above). But this seems harsh. For I might well have gone to school and learned English very well.


Now, in line with 2, consider the following pair of conditionals:


C.  If Obama had had less money, then he would have still be president.
     [Obama had less money than he did: FALSE], [Obama is president TRUE]

D.  If Caesar had had tanks, he would have used them against the Gauls
     [Caesar had tanks: FALSE], [Caesar used tanks against the Gauls: FALSE]


Is C true? Arguably, yes. America wanted a change and Obama was the candidate for change. He would have made it with less money. But then we should say that D is false (in line with 2). But then D is arguably true.


Quite how we determine whether A, B, C or D is true - indeed, whether it makes sense to say that subjunctive conditionals are always true or false - is a complex matter that we cannot pursue here.


A strategy that the fatalist might pursue now is to say that this talk of what might have been is meaningless. There is just the one reality, our actual reality, and only one way it could be. The word 'if' can only be interpreted as the material conditional. This is certainly worth pursuing but the topic of conditionals is something we cannot enter into further here.


Appendix Notes

[1]  The three other conditionals are given below with their truth-tables

  • if I get better tomorrow having not called the doctor, I would have still got better had I called the doctor.
  • if [doctor not called and get better] then [if doctor called then get better]
  • (~P & Q) (P Q)
P Q ~P ~P & Q P Q (~P & Q) (P Q)
TrueTrueFalseFalseTrueTrue
TrueFalseFalseFalseFalseTrue
FalseTrueTrueTrueTrueTrue
FalseFalseTrueFalseTrueTrue


  • if I don't get better tomorrow having called the doctor, I wouldn't have got better had I not called the doctor.
  • if [doctor called and not get better] then [if doctor not called then not get better]
  • (P & ~Q) (~P ~Q)
P Q ~Q P & ~Q ~P ~P ~Q (P & ~Q) (~P ~Q)
TrueTrueFalseFalseFalseTrueTrue
TrueFalseTrueTrueFalseTrueTrue
FalseTrueFalseFalseTrueFalseTrue
FalseFalseTrueFalseTrueTrueTrue


  • if I don't get better tomorrow having not called the doctor, I wouldn't have got better had I not called the doctor.
  • if [doctor not called and not get better] then [if doctor not called then not get better]
  • (~P & ~Q) (~P ~Q)
P Q ~P ~Q ~P & ~Q ~P ~Q (~P & ~Q) (~P ~Q)
TrueTrueFalseFalseFalseTrueTrue
TrueFalseFalseTrueFalseTrueTrue
FalseTrueTrueFalseFalseFalseTrue
FalseFalseTrueTrueTrueTrueTrue


 [back] 


[2]  The word 'subjunctive' is an unfortunate one. It is most commonly a term used in linguistics to refer to a feature of verbs. For example, if you have learned some French, you will know that with some constructions you have to use indicative forms of the verb and sometimes subjunctive. If you want to say I know he is coming, you would say Je sais qu'il vient where the verb is the indicative. If you want to say I fear he is coming, you would have to put the verb in the (present) subjunctive and say J'ai peur qu'il vienne. The difference between the indicative and subjunctive is a difference of mood. The mood of the verb conveys something about what the speaker is trying to do with the verb (or the larger phrase or sentence in which it is embedded) or is trying to say about the relationship between what he is saying and reality. It is best conveyed by example. Here are three moods you will be familiar with:

  • 'Jim is coming here': indicative mood
  • 'Is Jim coming here?': interrogative mood
  • 'Jim - come here!': imperative mood

Depending on whether we are aiming to state a fact, raise a question or give an order, we use the same words differently, both by re-organising the sequence of words and adjusting the verb itself. A change in mood is not always signalled this way. In English (especially Australian English), by uttering 'Jim is coming here' so as to raise your voice towards the end, you make it into a question: 'Jim is coming here?' In English, if you want to walk about what must be the case as opposed to what is the case - 'Jim must come here!' - you use an extra or auxiliary verb 'must'. In other languages, you would change the verb itself.


The subjunctive is connected to the irrealis mood or the 'unreal' mood. We often talk not about how things are but how things could be or might be, about how the world should be, about how we wish it to be, and so on. In English and French, for example, you use the subjunctive in wishes: 'I wish he were here' (not 'I wish he was here') and 'Je souhaite qu'il soit ici' (not: 'Je souhaite qu'il est ici'). But it is not the case that all languages handle irrealis moods the same way. Some languages lack a subjunctive mood: that is, they lack a distinctive form of the verb. English almost lacks a subjunctive. Although we have just seen it in action, typically we use auxiliary verbs: 'I wish he would come', 'He must leave the building immediately'. Many languages likewise use other verbs or even other moods. Furthermore, the subjunctive is often demanded in cases where it is hard to see anything 'irrealis' going on. In French, for example, you must use the subjunctive after 'avant que' - 'before' - but not after 'après que' - after.


Philosophers often avoid talking about subjunctive conditionals and talk instead of counterfactual conditionals. A counterfactual condition is one that runs contrary-to-fact - one whose antecedent is presumed false. It is this feature that we are interested in.  [back]