Descartes began his search for certainty about his own being and about God with the proposition "I think, therefore I am." [Rene Descartes, Discourse on the Method of Rightly Conducting the Reason and Seeking for Truth in the Sciences, Part IV, 1637, and Meditations on the First Philosophy, Meditation 2, 1641.] Most philosophers conclude that he failed.
Descartes should have begun with "there is thinking or talking about truth, therefore there is a logical language." That is the one absolute certainty. It is impossible to think or say anything about truth, let alone assert a true or false proposition, without a language that allows our logic to work. Here we can distinguish language that permits the making of true and false propositions from the language we use to express emotions such as fear, warning, sexual attraction, and just plain joie de vivre like a dolphin or a nightingale.
Where logical language is being expressed you can deduce that there is at least one language user. That does not prove the language user is human. After the destruction of all humans a lone computer could still be processing information. Beyond the basic certainty that there is at least one language user, deductive logic can only proceed by setting out alternative possibilities.
Either you are the only language user, or there are others. If you are the only language user, there is no point in trying to argue with others. If there are other language users, and we can understand each other, then there are various possibilities as to how we can relate to one another. If humans are not free to evaluate arguments then discussion of any question is meaningless. If humans are free to evaluate, their choice will either be arbitrary or based on reasons. If their reasons are organized systematically, we should be able to picture them in a logical model.
This is a book about the explanatory models used by followers of various religions and ideologies to explain their faith. And there is no way to prove logically that one system of explanation is certainly correct and the others wrong.
There can only be certainties within a logical system or model. This is easy to see in a very simple a two dimensional geometry of straight lines and circles. For that we need the language to define straight lines as the shortest distance between two points in a horizontal plane. We also need to agree that circles are lines which are equidistant from an imaginary point. In such a geometry we can be absolutely certain that it is impossible for a straight line to form a circle.
In Euclid's geometry all sorts of certainties can be deduced within the model because the certainties are built into the model by definition. It is also possible to discover some unexpected certainties as for example in Pythagoras' Theorem. Philosophers call such a system of certainties analytic truth. Logic is merely the sum of certainties within our model, and the number of models we could invent is infinite.
By adding the vertical to a two dimensional geometry we can make a three dimensional model. Then by including the dimension of time the model becomes four dimensional. In his Tractatus Logico-Philosophicus, 1921, first English translation 1922, Wittgenstein showed that a four dimensional science is a logical system with a theoretically limited number of true or false propositions. By increasing the scope of the model we multiply the number of possible certainties within it. For example in Newton's physics it is impossible for objects to move against gravity unless there is a force greater than gravity to move them. In Einstein's system the old system of certainties is abandoned and a new certainty is that space has to be curved. And E can never be anything else but MC squared.
By the time Wittgenstein wrote the Philosophical Investigations he had seen that there are thousands of certainties not only in scientific language, but also in many other kinds of language. In base 10 mathematics it is impossible for 2 and 2 to make five. Any child knows that if an object is above another, then the lower object cannot be above the first. If something is green, it cannot be red. Similarly in every kind of model there will be a number of built-in certainties. [See Ludwig Wittgenstein, On Certainty, Oxford: 1969; New York: Harper & Row, 1969]
Similarly there are certainties within each religious or ideological system. Within Islam there can be no possible doubt that we will face the last judgment. Within one form of Calvinism there can be no doubt that God chooses who will come to believe and be saved. In most forms of Condemnatory Theism it is certain that any of the heathen who have not heard about Christ deserve to go to hell. If we doubt that certainty, and say "but surely God knows the heart, and he can welcome anyone from any nation whose heart turns towards him" then we have already abandoned the condemnatory model and have moved into another which may not yet be clear to us.
In the theology of John's Gospel for example it is absolutely certain that no one will be in heaven except through Christ (John 14:6). But as C.S. Lewis pointed out that does not deny that some who claim to be atheists and others who have valiantly served the god Tash against the Lion of the Tribe of Judah, will be astonished to find themselves in heaven through Christ, the Son of God. [This is pictured in C.S. Lewis' The Last Battle, and more fully in his The Great Divorce]
So whatever religious or ideological system we adopt the model will inevitably generate its own certainties. It then becomes obvious that the model certainties of one religious system might agree in some points with another, but we must begin with recognizing that they are different as systems. That means that a proof which is perfectly valid in one system might seem total nonsense in another.
Since Wittgenstein it has become clear that systems of certainty can be invented for all sorts of purposes. In science fiction an internally consistent logical world is constructed by the writer. Very soon the reader catches on to the system of certainties and falsehoods within that model. In some cases the reader might say "that would be a wonderful kind of world to live in." Other writers can construct a world, as in Aldous Huxley's Brave New World or George Orwell's Animal Farm or Nineteen Eighty-four, which fills us with horror, and makes us care about quite different values. The story has moral power even though we know that no such world has so far existed.
We should also distinguish the certainties that belong to a model from the observations that can be made using the model. We know that two and two can never make five, and we know that waving your arms means you are not still, but we can disagree about whether there are two or five persons waving to us on a raft. The propositions concerning which the language does not give us certainty are called synthetic truth. Thus "all tigers are felines" is analytic by zoological definition, but "there is a tiger in my garden" could be proved mistaken, and so it is a synthetic proposition. Scientists often fail to make the distinction between the analytic truths of their system, and the truths they have discovered by collecting and analyzing data.
It took many years before social scientists accepted the fact that many of their certainties were merely a product of their model. In his psychiatric model Freud had managed to convince therapists that his model was somehow the only objective psychiatric truth. Within his model there were numerous certainties : repressions have harmful effects; in a dream fantasy a top hat is certainly a phallic symbol. There was an astonishing model shift with William Glasser's Reality Therapy, New York: Harper & Row, 1965. Suddenly the massive system of Freudian certainties became only one model among others. Social scientists of all kinds are now more comfortable with the idea of working with alternative models of sociology or psychology. Theologians still find that difficult. But this book has suggested that model theology is essential if we want to be honest and crystal clear about our theological differences. [See the appendix on Models]
Philosophers have concluded that none of the proofs for the existence of God are convincing. Now we should add that in each religious or ideological system there are certainties about God or the denial of God. It therefore makes no sense to discuss a proof about God's existence except within a system. And in a Theistic system God is the Creator by definition.
Does that reduce faith in God to relativism? We must certainly admit that God has put us in a world in which it is impossible to arrive at logical certainty about God, or even arrive at any kind of religious faith based on logical proof. We have seen that no proof is valid apart from the system in which it makes sense. That is why Paul tells us that "in the wisdom of God the world through its wisdom did not know him." [1 Corinthians 1:21] So we can only say another religion or ideology is wrong by adopting a standard of judgment. In that sense there is relativity of proof between religions and ideologies. But chapters 7 and 8 show that there is no relativism in a person's own moral commitment. Which suggests that it is moral commitment, not logical proof, that gives us the only certainty that is important for living our lives freely and creatively in this world.