Note added 19 Apr 2014
I have discovered a freely available paper by John Shand, which overlaps in some
ways with this paper, though the details are very different.
John Shand, (The Open University) A Refutation of the Existence of God Think, Number 26, Volume 9, Autumn 2010 www.academia.edu/183655/A_Refutation_of_the_Existence_of_GodAnother is a book chapter on "Analytical Atheism" by Charles Pigden originally published as
Charles Pigden, 'Analytic Philosophy' in Bullivant and Ruse eds. 2013 The Oxford Handbook of Atheism, pp 307-319Google now provides far more entries for "analytical atheism" than it did in 2008 when the original version of this document was written.
Another one discovered: 10 Mar 2016 -- no longer exists 28 Aug 2016:
https://www.facebook.com/anatheism/
This document attempts to make clear that there are several forms of atheism, of varying 'strength' in their opposition to theism, and that there are several types of theism to which each type of atheism can be opposed. This means that instead of atheism being just a single viewpoint it covers a multitude of cases, each definable by a type of theism and a type of objection to that form of theism.
A slightly curious consequence of this is that a person who accepts one type of theism and is an atheist about another type of theism can be both a theist and an atheist. This is no more inconsistent than admiring some novelists and not others.
I first present some background to this discussion note: an exchange that occurred on the PHILOS-L email forum. I then mention a web site on intelligent design in which I described myself as an atheist, and then present the analysis of types of atheism. This helps me to be clearer about the type of atheist I am. Perhaps it will help others also.
I responded making three points:
1. As a D.Phil student in Oxford in the 1960s, I had met and interacted with the individuals named, and had generally not found that their catholicism had interfered with their quality as analytical philosophers. 2. I did not think the proportion of catholics amongst prominent analytic philosophers was particularly high. 3. My impression was that most of the well known analytic philosophers were atheists.I also wrote:
I was then and remain an atheist (of the kind who doesn't claim that God does not exist, but that the usual concept when analysed closely is incoherent, so that existence of an instance makes no sense -- is there a standard name for that sort of atheism?).That question provoked some comments, most on the list, and available here, in the thread "A popish plot?"
After a few comments had been posted, the moderator of PHILOS-L requested that further discussion be transferred to the Chora list, available here:
http://listserv.liv.ac.uk/cgi-bin/wa?A1=ind0804&L=choraI don't subscribe to Chora, as I already get far too much email.
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/teaching-intelligent-design.htmlI recommended allowing scientists to teach ID alongside evolutionary theory as part of an education in how to evaluate theories with competing explanatory claims.
In order to make my own position clear I stated in that document:
I don't recall exactly what convinced me that there was no god -- whether it was the news of the horrors on all sides of the war, whether it was the fairly obvious emptiness of the occasional perfunctory religious observances of my (Jewish) family or the hymn-singing and fatuous praying that started each day at school, whether it was the immorality of a god described as instructing a man to sacrifice his son on an altar (even if it was only a cruel hoax, to test his faith), whether it was something I read, or just the complete lack of any evidence for anything remotely like the good, all powerful, all knowing, loving god alleged to be listening to prayers of people on both sides of every conflict.
If one side could be wrong, why not both, since neither had better arguments than the other for the existence of their god?"
What does 'incoherent' mean in this context? There are different sorts of incoherence, but the types I am concerned with all involve producing a statement or claim that for some reason cannot be true and cannot be false either. Here is an example:
The universe is moving in a north-easterly direction at three miles per hour, but the motion cannot be detected because all measuring devices take part in the same motion.Why is this incoherent? Because
(a) the notion of 'the universe' includes everything that exists, including everything that can move and space itself, (b) the notion of something moving (in any direction at any speed) presupposes some fixed entities relative to which the motion can be measured.Consequently if everything is moving in a certain direction there cannot be anything relative to which things are moving. So the notion of everything moving undermines its own presuppositions and is therefore incoherent. (This is part of the core of Einstein's special theory of relativity.)
There is a more subtle incoherence because the notion of moving in a north-easterly direction makes sense only when applied to something moving on or relative to a location on the surface of the earth, where the four compass directions are defined. North-east will be different for different locations on earth. So it makes no sense to say that the whole earth or the solar system, or our galaxy, let alone the whole universe, is moving in a north-easterly direction. Further examples of incoherence are given below.)
The discovery of the incoherence of the theist position, and some variants of its denial is a special case of something that has happened several times in the history of philosophy, as follows:
-- Then an opposing thesis is presented also with apparent supporting arguments.
-- Later, it is gradually realised, at least by some of those involved in the debate, that it is an endless debate that can never be resolved -- because it involves propositions that appear to be capable of being true or false, but which can never be either because they are either too lacking in specificity or incoherent, sometimes in unobvious ways.
There are many examples of statements which appear to be intelligible when first encountered, but which one can learn to see are incoherent.
The statement about motion of the whole universe is an example.
Another example comes from Wittgenstein: we know what it means to say that is now 5pm in London, or 5pm in New York. So it may seem obvious that we understand what it means to say that it is 5pm on the sun (or at the centre of the Earth, or in another galaxy).
But closer analysis shows that a presupposition of talking about what time it is at location X, is that the sun (usually the sun of our solar system, though it could be the sun at the centre of some similar planetary system) has an angular elevation that changes at that location.
But the sun cannot have an angular elevation at the sun, or at the centre of the earth etc. So not all statements of the form "It is now Xpm at location Y" are capable of being either true or false. That is not immediately obvious until one has learnt to analyse preconditions for the applicability of concepts one uses.
There are many examples of such disguised nonsense in the history of philosophy.
I shall call the theory that theistic claims cannot be true and cannot be false because they are incoherent "analytical atheism", because it is arrived at by conceptual analysis.
For now I am more concerned with clarifying the difference between that kind of atheism and more common kinds of atheism and agnosticism, than with producing arguments for any of the positions.
When I recently tried, using google, to find examples of other people putting forward what I call analytical atheism, my search revealed little though that may be simply because I did not look hard enough. An example that comes close is this:
Note, however, that being self-contradictory implies being false. So this statement:
This table is made entirely of rubber and this table is not made entirely of rubberis logically contradictory and is therefore false, according to our standard understanding of 'and'. It has a truth value, which is independent of how things are in the world. That is not the same as being incoherent and incapable of having a truth value, as in this statement:
This table is made entirely of days of the week.
In the remainder of this paper I shall try to make more precise what the thesis is that I refer to as "analytical atheism" and what the alternatives to it are. Producing and evaluating all the best arguments for and against the thesis and showing that this atheistic thesis is correct, or at least better supported by arguments and evidence than any of its rivals, would require a much longer document.
Readers who are tempted by that view need to be aware that it can take a complex and unobvious argument to show that some set of propositions is logically inconsistent even if there are people who believe they are all true. Showing that a collection of propositions that appear to make sense cannot form a coherent whole, because some of them undermine the presuppositions of others, can also take a complex and unobvious argument. The existence of such a proof of incoherence is perfectly compatible with many people feeling they understand something coherent by that set of propositions.
Compare the five year old child who was asked where a fast racing car and a slow moving van would meet if they simultaneously started moving towards each other after starting some distance apart. His answer was to point at a location nearer the starting position of the racing car. When asked why he sincerely answered: "it is moving faster so it will get there sooner".
He was not ready to understand his own confusions. Adults can be in the same state. Training in analytical philosophy can help them to recognize and get over some of their own confusions.
Being able to discover that some of one's thoughts are nonsensical requires a kind of intellectual development, which for many people can be very difficult if they have not studied analytical philosophy.
Whereas some people fail to realise that thoughts that feel meaningful can include hidden contradictions or be incoherent, others who appreciate the need for constraints on meaningfulness propose the wrong constraints. For example "concept empiricists", and their modern re-inventors, defenders of "symbol grounding theory" propose overly strong criteria for thoughts to have meaning, e.g. that all the concepts used must have been derived from experience, or must be "grounded" in definitional links to sensorimotor relationships. For a discussion of this error see
The specific kind depends not only on the kind of rejection, but also on what is being rejected. There are different sorts of theism and to each one of them there can correspond several varieties of atheism. Not all varieties of theism are incoherent in the way I shall attempt to define.
For example, Einstein sometimes claimed to be religious and to
accept a kind of God, but, in saying that, he was not referring to
the usual kind of Judeo-Christian type of God, but to some deep
features of the natural world that make it awe-inspiring even to a
hard-headed scientist. He was not endorsing more common kinds of
theism, or religious belief. For more on Einstein's views and how
they are sometimes distorted by theists, see
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/einstein-religion.html
In order to define various kinds of atheism we need to specify various kinds of theistic thesis. Different subsets of those theses are accepted by different theists (and different religions).
It should be obvious therefore that there will be different atheistic (or anti-theistic) positions, denying the truth of different combinations of theistic claims.
For each theistic position, TS, defined by the conjunction of a set S of theistic propositions, we can define the following six anti-theistic views more or less strongly opposed to TS:
I am not claiming that there are always sharp distinctions between these cases -- there may be atheists whose views straddle two or more of these cases.
It should be clear from this that for every different set of theistic propositions, TS, there will be (at least) six different atheistic (or anti-theistic) positions that can be held, according to the schemata presented above,
A1(TS) A2(TS) A3(TS) A4(TS) A5(TS) A6(TS)four of which treat the theistic position as saying something that might conceivably be true and might be false, and the sixth one, A6(TS), analytical atheism about TS, stating that TS is incoherent and therefore cannot be either true or false. Note that, as explained above, this is a stronger criticism than the claim that TS is contradictory and therefore necessarily false.
I am an analytical atheist, in the sense of A6, about various sets of theistic propositions, discussed below.
This should not be read as a claim that I am dogmatically or irrevocably committed to A6.
Unlike most theists I always accept the possibility that I shall later find that my views are mistaken -- as has happened in the past, when I have changed my mind on some philosophical or scientific question.
E.g. I may one day find that theism can have additional flaws not listed above. Or I may learn that there is some variety of theism that I have never encountered (or have encountered but failed to understand) that comes very close to the types of theism I have summarised here but is neither self-contradictory nor incoherent.
This is one of the main differences between scientific or philosophical atheists and religious theists: the latter typically will not even admit that it is possible that they are mistaken, and in some cases try to make a virtue of having faith that is not backed up by evidence or reason.
Educating children to have closed minds of that sort is thoroughly unethical and can be compared with foot-binding. I have called it mind-minding, since it constrains intellectual development, leaving minds permanently immature -- stunted mental growth. This is a cause of much harm that people do to one another in the name of religion. It is for that reason that state-funding should never be used to support faith-schools.
being Aware of ... being Unaware of .... AW(A1(TS)) UW(A1(TS)) AW(A2(TS)) UW(A2(TS)) AW(A3(TS)) UW(A3(TS)) AW(A4(TS)) UW(A4(TS)) AW(A5(TS)) UW(A5(TS)) AW(A6(TS)) UW(A6(TS))Between the extremes of being fully aware of one's attitude to TS and being fully unaware of one's attitude (like an infatuated lover whose infatuation is obvious to all his friends but not to him).
It is also possible to distinguish differences between the reasons why people hold a position or whether they have reasons. I am fully aware of being an analytical atheist in relation to some common varieties of theism AW(A6(TS)), and I have reasons for that, indicated elsewhere in this paper.
The important point is that for each type of theism there is some set of propositions all of which or some subset of which are accepted as true by the adherents. For any such subset there will be six corresponding varieties of atheism represented by replacing TS with a label for that set of theistic propositions in each schema above.
What follows illustrates the sorts of theistic theses, various subsets of which, at least in the communities I have encountered, are commonly regarded as characterising religious beliefs.
There is something or someone, often referred to as God,
Different theists will accept different subsets of those assertions. Some may accept none of them: there are many kinds of religion and many kinds of theism. I suspect there are mystical kinds of theism for which all the above statements are far too simple and definite.
The main point here is that for each kind of theism, as long as its adherents put forward some set of statements TS that they regard as true, there will be in principle (at least) the six possible forms of atheism about TS.
If a theistic doctrine says nothing true or false, but merely consists of exhortations to act in certain ways, or to take up certain attitudes of awe or reverence or respect to the universe, or some portion of it (which is how R.M.Hare characterised his religious belief in the 1960s), then none of the above forms of atheism is relevant. Of course, other forms, opposing the exhortations, are possible.
(I shall extend the list of theistic propositions, and possibly the list of types of atheism, from time to time. Suggestions welcome.)
Originally there were only five types of atheism on my list, until searching literature on atheism made it clear that types 5 and 6 needed to be distinguished. Many people seem to have noticed type 5, without noticing type 6.
Different subsets of that list can be taken to define different sets of theistic propositions,
TS1, TS2, TS3, .....
For each subset, TS, there will be different kinds of Atheism, e.g.
A1(TS) A2(TS) A3(TS) A4(TS) A5(TS) A6(TS)I am not claiming that these are the only varieties of atheism. There may well be views or theories that are opposed to kinds of theism but do not fit any of the above formulae. My main concern here is to indicate how A6 differs from other forms of atheism.
The other theists make more or less vague statements about their attitudes to life, or their sense of mystery or awe about the universe, or they utter prescriptions about how people should live. But they don't claim to be saying something that is true. Sometimes they don't even try to convince others that they should believe or accept anything: i.e. they are not evangelical at all. We could call them 'non-assertive' theists. Nothing written here implies that what non-assertive theists say is incoherent. (It may, in some cases, be perfectly coherent, but ethically objectionable.)
My 'personal' tutor while I was a student at Balliol College in Oxford working for a DPhil on philosophy of mathematics was Richard Hare, who was a non-assertive theist.
I am not an analytical atheist about non-assertive theists. I just don't care to join them, in their claims, prescriptions, rituals, etc.
Unfortunately a normal school syllabus will not, as far as I know, teach students to do conceptual analysis, even though it is a most important skill. Usually, learning to do it requires a degree in philosophy, with a strong component of analytical philosophy.
However it is possible for people to teach themselves. In 1978 I published a tutorial on conceptual analysis in Chapter 4 of The Computer Revolution in Philosophy, now available online here. However, many readers will find that quite heavy going. Clear thinkers probably have a natural talent for learning how to do it if they have not been exposed to too much intellectual mush in their education. It possible to get the idea by working on a few relatively simple examples before starting on the really difficult conceptual analysis tasks, such as trying to clarify concepts like "consciousness", "mind", "emotion", "love", "faith", "free will", "good", "bad", "evil", "survival after death", "omnipotence", "omniscience", or "god".
One way to start is to consider the following questions. Each of them appears to be asking a question which requests an answer in the form of a proposition. The right answer will be a true proposition -- if there is a right answer. If the question is incoherent, no answers can be true or false.
The notion of conceptual analysis has hidden depths. I have tried to uncover some of them here:
I have not yet read it all.
Anyone who wishes to criticise the incoherent concept of God should not point the finger at me, but at the priests, theists and other believers who specify that that's the sort of God they believe in and worship. (BBC Radio 4's regular religious slots provide plenty of examples of this sort of incoherent theism -- along with unctuous self-righteousness and a recurring tendency to refer to what "we" need, or think, or feel, or want, without any evidence of research backing up the generalisations made!)
What I criticise as incoherent is not my invention but the type of concept of god that I have heard put forward by others, mainly in the Jewish/Christian and possibly (though I am not so sure) Muslim traditions.
As noted above, this criticism of incoherence need not apply to all religions and in particular it does not apply to the forms of Buddhism that I have encountered. There may be other forms that avoid the criticism. Good luck to the proposers: they may have some need for such a concept. I don't.
I have no interest in looking for an improved concept of god -- there are enough other tasks to keep me busy for a few more centuries (like trying to understand the evolutionary origins of human mathematical competences, and the forms of virtual machinery found in animals and artefacts).
If someone else finds a significantly improved god-concept, no doubt I'll hear about it. Though I don't expect it to be of any importance for the things that concern me.
Maintained by
Aaron Sloman
School of Computer Science
The University of Birmingham