Sunday, August 13, 2006

The Meaning of 'Atheist'

This blog entry might be somewhat off of my normally well-beaten path, but it addresses an issue that has arisen more than once in this blog. I addressed it in passing, and I think that it warrants some more direct attention. It concerns the question, “What is an atheist?”

In an article in American Chronicle, David Glesson decides to address some “Common Misconceptions about Atheists and Atheism.” He starts off with the “misconception that, “Atheism is the belief that no Gods exist.”

I deny that this is a misconception.

A Few Words about Words

Language is an invention. It is a tool – like shovels and computers - that we design to do work. There is no “natural law” of language that dictates what a word must or must not mean. There is, instead, the purpose for which the invention of language is created and the question of what design will best serve that purpose.

The main purpose of language is communication. It is not the only purpose, but (particularly in this context) all of those other purposes are insignificant compared to the purpose of causing a particular idea to emerge in the mind of the listener or reader. As I write, the foremost questions that I have when I choose a word is, “When the reader sees this word, what will pop into his head? Will it be the same thing that I want to have pop into his head?” If the answer to the second question is, “No,” then I have to pick a new word.

In other words, the 'correct' definition of a word is the prediction of what thoughts pop into the mind of the reader or listener when they encounter the word.

What pops into the minds of almost all English speakers when they read or hear the word “Atheist?” It is, “Somebody who has the belief that no gods exist.”

We find some evidence for this in the fact that, if Gleeson were correct, then it should be instantly clear to every person who reads or hears the words ‘atheist’ that infants are atheists. Instead, what pops up in the minds of most English speakers is that those who would call an infant an ‘atheist’ is speaking gibberish. It is gibberish precisely because an atheist is one who has a belief that no god exists, and there is no way that an infant can have such a belief. Gleeson accepts that his definition of atheism means that all infants are atheists. He does not see that this conclusion is a reductio ad absurdum of his thesis.

Indeed, think about how absurd it is to argue that there is this word that is a normal part of the language. People use it all the time. People who read or hear the word almost never misunderstand those who write or speak the word. However, all these people who are using the word and not misunderstanding each other do not understand what the word really means.

This is a very strange claim to try to defend.

Note: In some cases, a word has a technical definition that does not correspond with the common usage. The technical definition of ‘argument’ among logicians is ‘two or more propositions where one proposition (the conclusion) is said to follow from the other propositions (the premises)'. However, when the kid comes to us and says, ‘My parents are having an argument,’ we do not say that he does not understand what the word means. We simply recognize that he is not using it in its technical sense.

The root of ‘a’ – ‘theist’

Gleeson defends his position as follows:

The word 'atheism' comes from the Greek prefix 'a', meaning without, and 'theist', meaning having a belief in a supernatural deity. Atheism, therefore, literally means "without theistic belief". Atheism does not positively assert anything; rather, it is a statement of withheld belief.

Now, consider this argument:

The word ‘atom’ comes from the Greek prefix ‘a’, meaning without, and ‘tomos’ meaning ‘to cut’. ‘Atom’ literally means ‘that which cannot be cut’. Yet, we have people today who assert that atoms can be split. Nothing can be more absurd than to say that it is possible to split something that, by definition, cannot be split.

The fact is, we cannot defend a definition on the basis of what the parts of a word may have meant to the ancient Greeks - not unless we are actually talking to an ancient Greek. We defend a definition on the basis of what ideas it causes in the mind of the readers and listeners who are competent users of the native language that contains the term.

I also want to note the condescension that Gleeson gives to those who do not share his opinion. He writes, “This statement's ubiquity is exceeded only by its utter falseness; not only is it misleading, but it is the complete opposite of the truth.”

Imagine that you are talking about splitting an atom, or about the parts of an atom, when somebody comes up to you, looks down his nose at you, and sneers, “You talk about splitting the atom. Your statement's ubiquity is exceeded only by its utter falseness; not only is it misleading, but it is the complete opposite of the truth.” He then goes on to say that the ancient Greek meaning of ‘a – tomos’ proves that you have no idea what you are talking about.

Ultimately, once the intruder admits to the 'ubiquity' of the term, he has already conceded defeat. He has already admitted that the common definition of the term 'atom' has become something other than 'that which cannot be cut.'

Atheism and Faith

After ‘proving’ his definition of atheism, Gleeson goes on to discuss what he calls another ‘misconception’, that Atheism requires just as much faith as theism.

Against this, he writes:

This misconception arises because of the misunderstanding of the term 'atheism', as described above. If atheism were indeed a positive assertion that no gods exist, then this criticism would be valid. After all, it would take just as much faith to claim that no gods exist as it would to claim that one god or many gods exist. But atheism makes no such claim.

His statement, If atheism were indeed a positive assertion that no gods exist, then this criticism would be valid. is false.

I look back at human history, at the number of different gods that different people have created, at the stories with their contradictions and inconsistencies, at the fossil record that tells us of evolution, at the fact that even today in the 'information age' people are inclined to believe ‘stories’ that are not only poorly founded but easily proved false (e.g., Saddam Hussein had weapons of mass destruction and was involved with those who planned 9/11), and I conclude from all of this evidence that the ‘God’ concept refers to a fictional character.

This is to say that the term ‘God’, like the terms ‘chimera’ and ‘Poseidon’, do not refer to any real world entity that has any type of predictive or explanatory power.

Can things with no predictive or explanatory power exist?

Possibly. There might be a parallel universe with absolutely no interaction with our own fictional characters exist. We have no choice but to remain agnostic about the existence of entities that have no role to play in explaining and predicting real-world events.

At the same time, we can say nothing at all about them that is not as likely to be true as fase. Any discussion about such entities – what powers they possess, what they look like, anything at all that can be known about them – would be a very short discussion. There is nothing to say, other than the same types of things that fiction writers put into their books as a matter of course.

My belief that the term ‘God’ refers to a fictional character is as secure as my belief that no ghosts exist, no leprechauns exist, the Loch Ness Monster does not exist, phlogiston does not exist, the earth is not flat, and it is not the center of the solar system. If the proposition that the term ‘God’ refer to a fictitious character is a matter of faith, then all of these other propositions are also a matter of faith.

Can this type of argument prove beyond all possible doubt that no God exists?

No, it cannot.

However, there is no proposition in science that can be proved true beyond all possible doubt. It is a part of the very nature of science that every theory that one can think of is a set of 'possibly false' propositions. In fact, it is the nature of all scientific propositions that they must be ‘falsifiable’.

The proposition, “The earth is 4.55 billion years old” is possibly false. It is not very likely to be false. It is, in fact, practically certain. Yet it is still ‘possibly false’.

When a person makes the assertion, “The earth is 4.55 billion years old,” he can do no more than state that this proposition has the most and strongest connection to everything else that we know – the best connection to our understanding of atomic theory, the nature of light, plate tectonics, the fossil record, and even facts about human perception (e.g., how our eyes work so that we can observe the results of our experiments).

When we look at human history, at different cultures and their invention of fictitious creatures, their invention of gods, the different types of gods that they invented, the propositions that has the best fit with our best understanding of all of these facts include the propositions, “There never were any dragons. There never were any ghosts. There never were any tree spirits. And there never were any gods.”


As I said, language is an invention. The meaning of a word depends on what ideas pop into the minds of the reader or the listener when they encounter the word (in that particular context). This, at least, is the meaning that any clear writer or speaker needs to use.

The meaning that pops into the mind of almost all readers and listeners of common English when the word ‘atheist’ is used is ‘one who holds that the term ‘God’ refers to a fictitious character.’ Gleeson admits this. In admitting it, he admits that the alternative definition he provides is mistaken.

Given enough time and enough effort, it may be possible to change the meaning of the word ‘atheist’ to ‘one who lacks belief in God.’ Given enough time and enough effort it may be possible to change the meaning of the word ‘atom’ back to ‘that which cannot be cut (or split).’ Language is one of those areas where even those who wrong, if they are persistent enough and persuasive enough, can actually make their false claim true.

The day may come when the word ‘atheist’ actually will conjure in the mind of the common listener or reader the idea, ‘one who withholds belief in a God.’

But it is not this day.


As a consequence of an anonymous comment below, I would like to add the following.

The common-language 'meaning' of the terms atheist, theist, and agnostic depend on how one would answer this question.

Does God exist?

(Almost) Certainly Yes: Theist.

Probably: Weak theist.

I don't know: Agnostic.

Probably not: Weak atheist.

(Almost) Certainly No: Atheist.

This is the breakdown that makes the most sense of how people actually use the terms when they talk to each other. This is the breakdown that has the best explanatory and predictive power as the best theory of what the terms mean in English.


Nath said...

I think the meaning of the word 'atheist' depends on who I'm speaking to, and in what context. As Bertrand Russell put it:
I never know whether I should say "Agnostic" or whether I should say "Atheist". It is a very difficult question and I daresay that some of you have been troubled by it. As a philosopher, if I were speaking to a purely philosophic audience I should say that I ought to describe myself as an Agnostic, because I do not think that there is a conclusive argument by which one [can] prove that there is not a God.

On the other hand, if I am to convey the right impression to the ordinary man in the street I think I ought to say that I am an Atheist, because when I say that I cannot prove that there is not a God, I ought to add equally that I cannot prove that there are not the Homeric gods.

In day-to-day conversation, most people do take 'atheist' to mean 'one who believes that no gods exist'. However, day-to-day language is notoriously imprecise and hard to reason in. This whole issue arises because most people are simply not capable of properly applying the law of excluded middle.

That said, if someone tells me he's atheist, the only assumption I make is that he is not theist.

Alonzo Fyfe said...


It is true that day-to-day language is imprecise. As a result, it is sometimes necessary to stipulate a more precise definition while writing. These are cases where an author says, "There is no word in our language that means 'X', so I will use this word. In what follows, whenever I use this word, take it to mean 'X'."

I am doing this constantly -- such as when I say "If an agent has a desire that P, and P is true in S, then S fulfills that desire."

However, in making a stipulated definition, I am not saying that those who use the term in is traditional sense are wrong -- that they do not know what the word means.

Ultimately, this posting is not concerned with what you assume when somebody tells you is an atheist. It concerns what it is reasonable to think that others assume when you use the word 'atheist.' As I said, a clear writer must choose his words according to what he can reasonably predict will come to the mind of those who read it.

When you say that, In day-to-day conversation, most people do take 'atheist' to mean 'one who believes that no gods exist', it follows that, as a writer, if you use the term, you must either take this to be your definition, or you must stipulate that you are using a definition that deviates from this common understanding.

Anonymous said...

In other words, the 'correct' definition of a word is the prediction of what thoughts pop into the mind of the reader or listener when they encounter the word.

So, people are infallible when it comes to definitions? Whatever people imagine is the definition, is the definition - it's not possible to be wrong about the definition?

When people hear "the theory of evolution," most people seem to think that the origin of life is included. Therefore, the 'correct' definition of 'evolution' includes the origin of life - despite the fact that in science the origin of life is separate from biological evolution.

When people hear "scientific theory," most people seem to think that we are talking about some sort of guess or hunch. Therefore, the 'correct' definition of 'scientific theory' is a guess or hunch made by scientists - despite the fact that in science 'theory' isn't a guess or a hunch.

Obviously not all of this can be true. Why? Because it's possible for people to be mistaken about what words mean. A person can have a definition pop into their head which isn't correct. Their reasons for being mistaken might be understandable or reasonable, but it's still a mistake. Sometimes, people need to be told that the theory of evolution doesn't really include the origin of life. Sometimes, people need to be told that in science a 'theory' isn't a guess.

And, sometimes, people need to be told that people who call themselves 'atheists' includes those who simply aren't theists.

If the purpose of language is communication, then that purpose is undermined if I have to use 'atheist' to mean 'one who is in denial about God' because that's what pops into the mind of the theist I'm talking to but this same theist has to use 'atheist' to mean 'not a theist' because that's what pops into my mind. But of course, now what's popping into our minds is switched and we have to switch again... and so on into eternity, forever talking past one another.

If language is a tool for communication, what are we communicating about? Atheists. Who are these atheists - the narrow group of people the theist in question claims or the people who actually call themselves atheists and identify as atheists? Most likely the latter, in which case the theist in question is obviously wrong because they are using a definition which doesn't refer to those people.

Alonzo Fyfe said...


My whole article is written in the context of somebody who, I claim, made a mistake about what a word means. Obviously, I hold that people can make mistakes.

As I said, a writer must choose the term that causes the idea that he wants it to cause within his reader's mind. When he makes a prediction, he may be mistaken.

Another type of mistake is the translation error. For example, when I translate a passage from French into English I may not pick the English words that communicate the same ideas as the French word.

It is still the case that what it means to translate the French phrase into English is to find the English phrase that more closely causes the same ideas to arise in the mind of the reader/listener as the French phrase being translated.

Your examples of 'theory of evolution' and 'scientific theories' are translation errors. In this case, the translation is from a subgroup who are using a term in a specific way to communicate among each other. The translator substitutes some other meaning for their local meaning.

Yet, even here it is the case that the meaning of the word depends on what those using the word expect their intended audience (other members of the group) to understand them to mean.

Gleeson's article was written about people whose 'intended audience' is the general public. In this case, he is wrong to state that the general public expects the listener/reader to understand him as meaning anything other than, 'one who denies that God exists'.

The standard distinction used in public discourse is this:

One who holds that the proposition "God exists" is true is a theists.

One who holds that the proposition "God exists" is false is an atheist.

One who holds that he does not (cannot) know whether "God exists" is true or false is an agnostic.

If you go with these definitions, you can, with great reliability, accurately predict and explain the use of these terms.

vjack said...

We're going to have to agree to disagree on this one. My position can be found here .

"In other words, the 'correct' definition of a word is the prediction of what thoughts pop into the mind of the reader or listener when they encounter the word."

What implications might this statement have for definitions involving homosexuality? Wouldn't this mean that if someone thinks of homosexuals as immoral animals that their mistaken definitions would have to be correct?

***Dave said...

No, but it would mean that if you know that's how someone understands "homosexual" then to use the word without taking that interpretation into account would be foolish at best.

Another way of putting it is that language at one level is prescriptive -- we teach kids what concrete items associate with what basic words, like, "green," "ball," "three" -- but, the further we get from basic and concrete terms, the more language becomes descriptive -- it represents a consensus of what that word refers to, not a hard rule that can be dictated from a dictionary. Dictionaries, in fact, change with usage more than usage changes with dictionaries.

Alonzo Fyfe said...

If I use the phrase, "Jim's car," and the idea that comes to your mind is a candy-apple red mustang convertible, this does not imply that Jim's car is a candy-apple red mustang convertible. But, then, "candy-apple red mustang convertible" is not what the phrase 'Jim's car" means. It means, "That car which Jim owns," which may or may not be a candy-apple red mustang convertible.

This distinction calls upon the distinction between a priori and a posteriori truth -- which, I admit, I did not use in the article, because I did not want to write a 20,000 word essay. (My postings are too long as is.)

However, since you brought it up, meaning (or a priori truth) is what necessarily comes to mind when using a term, where as a posteriori truth is that which is only contingently connected to the term. Everybody recognizes that the "Jim's car" can also refer to a blue and white 1967 Chevy pickup. The candy-apple red mustang convertible is only contingently connected to Jim's car.

Wrongness is necessarily associated with murder. Murder means 'morally unjustified intentional killing.' Killing in self-defense is not murder. On the other hand, wrongness is only contingently associated with homosexual acts. An sex act would not cease to become a homosexual act if it were not wrong in the way that a killing would cease to be murder if it were not wrong.

With the term 'atheist' in common discussion, it is thought impossible for an infant to be an atheist in the same way that it is impossible for a justified killing to be murder. The capacity to attribute a possibility of being true to the proposition "god exists" is a necessary part of its meaning (the word does not apply to entities that lack this capacity). Those at or near 0% probability of being true are atheists. Those at or near 100% chance of its being true are theists. Those in between are categorized as various levels of agnosticism.

Dar said...

Thanks Alonzo. I will direct people to your site who also ignorantly confuse "atheist" with "hate monger".

Anonymous said...

You are wrong, Alonzo. The terms "scientific theory" and "theory of evolution" are not translated wrong in a manner analogous to how French words might be translated incorrectly. You would be right if I had simply used "evolution" as an example, because that's a term with a specific meaning in a sub-group like biologists but different meanings elsewhere. The terms I used have specific meanings, though, and false meanings are attributed to those terms by people who don't understand the concepts (or who are lying about them). When a person tries to discuss evolutionary theory, and defines "theory of evolution" wrong, then they are wrong. When a person tries to discuss atheists, and defines "atheism" wrong, then they are wrong.

The "distinctions" you list, for example, do describe how many people see the terms - and they also reveal that many people fail to understand that knowledge and belief are not mutually exclusive things. Thank you, then, for making my point: people can be objectively mistaken in what "pops" into their minds in connection with certain terms. It is my position that such mistakes should not be pandered to - they should be corrected. It would be foolish, as Dave says, not to keep such errors in mind, but it's equally foolish to just ignore the fact that they are errors and not make corrections.

Your position appears to be one to pander to others by simply using whatever "definition" you think they are using rather than what is actually correct in the relevant context. Knowingly perpetuating and encouraging error strikes me as morally wrong as stating a deliberate falsehood.

By the way, if you look at, they cite dictionaries which make room for atheism to be defined as not having a belief in gods.

Alonzo Fyfe said...


My claim is also that mistakes should not be pandered to, but that those who understand 'atheist' to mean 'one who holds that the proposition 'God exists' is (almost certainly) false' are not mistaken.

It is Glesson (and Smith) who are making the mistake, and it is their mistake that I am refusing to pander to. They are making a mistake because you do not determine the meaning of a word by looking at the sum of its ancient Greek parts. You determine the meaning of a word by looking at how the intended audience would typically use the word. If your claim about what the word 'means' does not match its intended usage, then your theory of meaning is mistaken.

An examimation of how the term 'atheist' is used shows that it means -- among 99% of the people who read, speak, write, or hear the term, 'One who holds that 'God exists' is (almost certainly) false.' This theory most reliably and explains and predicts how speakers use the term.

If our civilization were destroyed, and ancient archaeologists were to rummage through our books, they would not translate the word 'atheist' to mean 'one who lacks belief in God.' They would translate the phrase to mean 'one who holds that the proposition 'God exists' is (almost certainly) false.

They would, of course, find set of writings from people who asserted this alternative meaning. Yet, they would be able to conclude that this assertion clearly contradicts how the term is actually used, thus it does not describe the actual meaning of the term.

As a writer, when I use the word 'atheist', I theorize (correctly, I assert) that the vast majority of the readers will understand the term to mean 'One who (almost certainly) believes that the proposition 'God exists' is false.' It is certainly true of my use of the term. Whether I am communicating effectively is determined by whether my word-use theories such as this are correct. If they are mistaken, people will not understand me.

So, your charge of pandering is question-begging. I am, in fact, refusing to pander to those who I think have a less accurate theory of the meaning of the word 'atheist.'

Anonymous said...


I know we have disagreed on the details of a theory of ethics/morality, but on this subject I can only say I agree very much with the ideas you have expressed here, and you have expressed them very well.

Thank you, I found it a deeply satisfying read.

vjack said...

What about the origins of the word atheism? The closest meaning from the Greek origin appears to be "without theism," does it not? Just because additional meanings have been attached (i.e., active assertion that god does not exist) does not mean the definition should be expanded.

Having an accurate definition has many important implications, not the least of which is the burden of proof issue. If we define atheism as the conviction that god does not exist, we are no better off than the theist in that we are claiming something which cannot be verified.

Perhaps the crux of my argument is that we need to educate people so that the a priori meaning of atheism shifts to the more accurate definition I am proposing. I hadn't really thought of it that way, but it does appear that the concept has collected inappropriate baggage.

The "common discussion" to which you refer is quite variable. I am convinced that it is not only possible but absolutely necessary for an infant to be an atheist because the theistic concept has never been encountered. Similarly, I deny that agnosticism is actually a valid position. Rather, I see it as a subcategory of atheism.

One who answers anything other than "yes" to the question "Do you believe that a god or gods exist" is without theism - an atheist.