Skip to main content

Quine (from my book "The Great Philosophers")

"…for all its a priori reasonableness, a boundary between the analytic and synthetic statements simply has not been drawn. That there is such a distinction to be drawn at all is an unempirical dogma of empiricists, a metaphysical article of faith." Quine

Quine is one of the most influential philosophers of the Twentieth Century. The son of a schoolteacher mother and entrepreneur father, Quine studied mathematics and logic at Oberlin College before winning a scholarship to Harvard. He spent his entire teaching career at Harvard, holding the Edgar Pierce Chair of Philosophy at Harvard University from 1956 to 2000. During WWII, Quine worked for US military intelligence.


Two kinds of truth

Many philosophers have drawn a distinction between two kinds of truth. Take these two sentences:

All bachelors are unmarried males
All vixens are female foxes

Both are true. But why? It’s tempting to answer: because of what the words “bachelor” and “vixen” mean. These sentences are true in virtue of meaning. The expression “bachelor” has the same meaning as “unmarried male”. They are synonymous expressions. So “All bachelors are unmarried males” is guaranteed to be true simply by virtue of the meaning of the words that go to make it up. And the same, you might suggest, is true of “All vixens are female foxes”.

We might add that, because these sentences are true in virtue of meaning, so there is no possibility of them being falsified by experience. Their truth is guaranteed, irrespective of how the world turns out to be.

Philosophers call those truths that are true in virtue of meaning analytic truths. They are contrasted with non-analytic, or synthetic, truths, such as:

All bachelors live in the vicinity of the Earth

This sentence is also true. But it is not true in virtue of meaning. The truth of this sentence, its tempting to say, is of an entirely different sort. This sentence is true because it correctly describes how things stand – as a matter of fact all bachelors do live in the vicinity of the Earth. Had they not, the sentence would have been false.

It seems, then, that there are two kinds of truth: analytic truth and synthetic truth.

  • {{{{{TEXT BOX:Explaining a priori knowledge
  • Armed with the notion of analyticity, several empiricist philosophers have attempted to explain how a priori knowledge is possible.
  • A priori knowledge is knowledge that can be established independently of the evidence provided by our five senses. It seems we can know a priori, for example, that all bachelors are unmarried. We don’t need to go out into the world and examine lots of bachelors to check are all unmarried. We know, even before we even start to look, that they will be. It seems we can also know a priori that 2 + 2 = 4 and that every surface is extended. And, according to Anselm (chpt XX), we can know a priori that God exists.
  • You can see how we might try to explain how certain truths can be known a priori by appealing to the notion of analyticity. How can I know, without examining any bachelors, that they are all unmarried? Because “bachelor” just means “unmarried male”. So I need only understand the meaning of the words that make up the sentence to be in a position to know that the sentence is true.
  • Armed with the notion of analyticity, we can also explain why it is a necessary truth that ever bachelors is also an unmarried male. If “bachelor” just means “unmarried male”, then “All bachelors are unmarried” says that all unmarried males are unmarried, which is a logical truth (for to deny it would involve you in a logical contradiction).
  • Of course, it is debatable whether all a priori necessary truths are analytic. But if they were, that would certainly explain why they are a priori and necessary. Which is why many philosophers have been drawn to the idea that all a priori necessary truths are analytic.END OF TEXT BOX.}}}}}

Quine’s attack on the analytic/synthetic distinction

The analytic/synthetic distinction has been one of the bread and butter distinctions of philosophy ever since Kant, who introduced the terms (though something like the distinction can be found in the work of Hume and Leibniz). While not all philosophers were convinced that all a priori necessary truths are analytic (Kant wasn’t), there was, before the 1950’s, a broad consensus that the analytic/synthetic distinction was a real one, and that there were, indeed, two categorically different kinds of truth. So when Quine presented a powerful-looking argument that the analytic/synthetic distinction was, in effect, empty, and that no distinction could be made between those true statements that are true purely in virtue of meaning and those that are not, the effect on the philosophical community was electrifying. Indeed, as a result of Quine’s work, many philosophers (particularly in the U.S.) are now convinced that there is no cogent notion of “analyticity”.

How does Quine attack the distinction?

The web of belief

According to Quine, our beliefs about the world form a web. They face the test of experience not separately, but collectively:

our statements about the external world face the tribunal of sense experience not individually, but only as a corporate body

Suppose, for example, that I believe that the Earth is flat. But then I come to possess two pieces of evidence that suggest my belief is false. A sailor tells me that he has sailed around the world and come back to the same place, and someone else has drawn my attention to the fact that, when boats disappear over the horizon, they disappear from the bottom up, as if the surface of the sea curved downwards.

These two pieces of evidence might together lead me to have doubts about the Earth being flat. But of course they are far from conclusive. I might deal with the sailor’s testimony by supposing that he is either lying or deluded about having sailed round the world. And I might try to deal with the observation about distant ships by supposing that light rays travelling close to the surface of the sea are somehow bent over a distance.

Of course, as more evidence against my theory that the Earth is flat begins to pile up, I might be more inclined to abandon my theory. But the fact is that, as long as I am prepared to make increasingly far-reaching adjustments to my other beliefs – my beliefs about the reliability of the testimony of sailors, about light, and so on – any apparent recalcitrant evidence can always be dealt with.

Now according to Quine, those beliefs that philosophers are inclined to call “analytic” are simply those beliefs that are particularly well-entrenched in our web of belief. They are the beliefs we will be slowest to give up, given recalcitrant evidence.

But, insists Quine, no belief is absolutely immune to revision in the light of experience in the way that “analytic” truths are supposed to be. Given that enough recalcitrant evidence mounted up, we might decide that, rather than making ever more complex and dramatic adjustments to the rest of our web of belief, the most satisfactory solution would simply be to abandon one of the supposedly “analytic” and/or “necessary” truths (interestingly, some have suggested that developments in quantum mechanics require that we revise the laws of logic).

No satisfactory definition of “analytic”

The other key component to Quine’s attack on the notion of analyticity is that no satisfactory explanation of the notion can be given. We can define analyticity in terms of meaning, of course, by saying that the analytic truths are those that are “true in virtue of meaning”. But, says Quine, the philosophical notion of meaning is just as opaque as that of analyticity. We might also try to explain analyticity in terms of the notions of the notions of synonymy, definition, and necessity. Indeed, each of these terms might be defined in terms of the others. The problem is that together they form a tight little circle of notions none of which is ever properly explained.

Here’s an analogy. Suppose I define a “wibble” as a collection of “woobles”, a “wooble” as an adult “doofer”, and “doofers” as the things that, when fully-grown, go to make up “wibbles”. I have defined each of my terms. But I have left you none the wiser as to what, if anything, I mean by them.

Quine’s point is that we are in a similar situation with respect to “analytic”. Yes we can define the term using other terms, but not in such a way that we get any closer to understanding what it is supposed to mean.

Of course, if I were to point to a pod of whales and say “That is a wibble”, what I mean by “wibble” suddenly becomes much clearer. Can’t we similarly explain what “analytic truth” means? Can’t we explain what it means by just providing some examples, such as “All bachelors are unmarried”?

No, says Quine. For why suppose that, say, “All bachelors are unmarried” is unrevisable in light of experience? True, we might not be able easily to envisage circumstances under which we would consider it false. But just because we can’t easily envisage such circumstances doesn’t mean they don’t exist.

If Quine is right that our statements are answerable to experience not individually, but collectively, and that any statement is potentially revisable in the light of experience, then even those statements that philosophers have traditionally called “analytic” are revisable. But then to point to such statements as examples of what “analytic” means would really be to beg the question. For if Quine is right, they are not analytic.

Since Quine’s attack on the notion of analyticity, it has largely, though not totally, fallen out of philosophical favour.


Larry Hamelin said…
I think the notion of analyticity still has some value, although nowhere near the deep connection with truth that earlier philosophers attributed to it.

Quine's elaborate construction that analytic statements are just those statements about the world are those we are most reluctant to change seems like a red herring. There is a much simpler, more direct way of drawing a distinction that keeps some passing resemblance to our intuitive notions about analytic and synthetic statements.

The concept of the "web of belief" refers to something entirely real, physical and concrete: human brains and the arrangement of neurons in those brains. Abstracted many layers, to be sure, but an abstraction of something prosaically real.

So the analytical/synthetic distinction retains some utility: synthetic statements are statements about the world outside our minds; analytical statements are statements about our minds themselves, specifically about the arrangement of our language-interpreting neurons.
anticant said…
How do we know that all bachelors live in the vicinity of the Earth? There may be bachelors on other planets, or in other universes.

We don’t know “a priori” that all bachelors are unmarried. We only know it because “bachelor” is the term conventionally used for unmarried men.

Neither do we know “a priori” that 2 + 2 = 4. We only know this because “2” and”4” are the conventional symbols used to designate certain quantities.

Pace Anselm, there is no “a priori” evidence that God exists.

“So when Quine presented a powerful-looking argument that the analytic/synthetic distinction was, in effect, empty, and that no distinction could be made between those true statements that are true purely in virtue of meaning and those that are not, the effect on the philosophical community was electrifying.” They must have been rather dim not to have realised this before Quine pointed it out!

The Flat Earth Society is alive and well:
Andrew Louis said…
This comment has been removed by the author.
Andrew Louis said…
I haven't read Quine, but he seems to be leading into pragmatism here. In that, there are no analytic truths, there is only what works as truth. (But, I see that as I'm a bit of a pragmatist myself and I'm not familiar with Quine.)

If for example, Sye's argument works, so what, what use does it have?

aside from believing in God to avoid hell (which is laughable) what other practical purpose does it have? What does religion do for humanity that humanity cannot do for themselves?

it reminds me of a joke I posted
Anonymous said…
"How do we know that all bachelors live in the vicinity of the Earth? There may be bachelors on other planets, or in other universes."

I suppose in our typically racist way we could simply restrict the definition of "men" to human males, having Earth passports, and "vicinity" to mean "as far as Earthmen have ever traveled".

Admittedly the possibility of other Universes does rather throw a spanner in the works but I suppose we could stretch the definition of "vicinity".

Mind you, we rapidly approach the other kind of truth claim.

Stephen said: "interestingly, some have suggested that developments in quantum mechanics require that we revise the laws of logic"

i think this is entirely the wrong way of looking at it. Mathematicians have been used to playing around with all sort of non-intuitive algebras and logic systems for ages. What we call "the laws of logic" usually refers to the sort of stuff applicable to the world of everyday experience. When some of the whackier stuff becomes everyday currency we will refer to the expanded set as "the laws of logic" and probably give some other name to the logic we currently use. what ever we all it though it will still be the same thing and still be valid in its context. Future logicians will be able to do the same manipulations as we now do although they may be explicitly aware that some facets of the physical world do not actually behave in this way. In the same way Newtons Laws of Motion are still as they are when he wrote them down. They are not changed by the discovery that they do not apply to the real world as well as he thought they did.
anticant said…
I know this thread will become endless if I ask others to define "truth" - but it is one of those words that gets bandied around pretty loosely! As I've remarked on previous threads, the so-called "laws of logic" aren't laws in the sense of eternal unbreakable edicts [whatever Sye may think]; they are simply semantic working rules designed to produce consistency in thought and discussion, and - as anonymous says - will no doubt be modified when they are found not to do this adequately in the light of developing knowledge.
Andrew Louis said…
On a side note,
it's been brought to my attention that my spelling is horrible. Yes, this is true.

I'm an engineer you guys, not an english major. Words never made a whole lot of sense to me....
anticant said…
"Words never made a whole lot of sense to me...."

Ouch! Why are you posting on a philosophy blog, then? :)
Andrew Louis said…
Ha ha ha Anticant, you got me. Nice one.

How about I rephrase that to the arangement of letters in the english language. I pay it no mind.

Fortunately in my area of expertise (statistics), it's not critical that I can spell it.
Anonymous said…
andrew :

Statistics - that's even worse! Most engineers systematically mis-spell either by throwing out non-functional letters or adding extra ones to meet the safety margins but you don't usually get random variation!

Don't tell me.. Your keyboard has error bars as well as a space bar?
Andrew Louis said…
anonymouse, TOO FUNNY! I'm not certain I can muster a response here.

But, yes... I think if you pull a random 30 piece sample of the way I spell certain words, you'll find variation with no direct correlation to the context. You'll further find (with a 99% confidence interval) that I just don't know what the heck I'm doing and it has nothing to do with my keyboard.
anticant said…
Well, I've lived for almost half a century with a top UK statistician, and he's the most niggling verbal pedant you can imagine. There are at least two dozen dictionaries in this house, and at least a couple of them are being consulted every day.
Timmo said…

I've got a question for you about Quine's critique of the analytic/synthetic distinction. Why has the distinction fallen out of favor?

Quine's critique relies quite explicitly on behaviorism in psychology. If we abandon that behaviorism, as we must if we are to be consistent with contemporary science, then much of the critique loses its sting.

Also, the consequences of following Quine here are going to be rough: there is no such thing as meaning. Once you erase or blur the distinction, a whole range of notions, like necessity, synonymy, and meaning bite the dust. But, most contemporary philosophers don't dispose of these notions.

Quine starts from a position we shouldn't hold and ends in a place many don't follow him to. So, why should the distinction fall out of favor?
Stephen Law said…
Hi Timmo - I am no Quine expert unfortunately. Here in UK this view is much less popular, certainly. Criticisms include:

(i) the fact that terms are defined in a circle is not a problem - that happens in science too (Grice/Strawson)

(ii) the term necessity can be explained independently without appeal to these other notions (Hooker).

(iii) Quine is operating with a behaviouristic, reductionistic model - which is too crude.

You might check e.g. the Stanford Internet Encylopedia of Philosophy for more....
Anonymous said…
"True, we might not be able easily to envisage circumstances under which we would consider it false. But just because we can’t easily envisage such circumstances doesn’t mean they don’t exist."

Is this not just a re-statement of the problem of induction, further highlighting the difficulties in applying logical/mathematical reasoning to the physical world?

if Quine asks us to accept "But just because we can’t easily envisage such circumstances doesn’t mean they don’t exist." does he not also have to accept the same for the notion of analytic truth?

Quine may have given us a bit of a jolt from complacency but I feel that we should not discard a useful practical distinction.
Anonymous said…
something something Quine, something something worldview, something something, how does that logic apply to mine, something something, how does Quine account for logic,.... something


Timmo said…
something something Quine, something something worldview, something something, how does that logic apply to mine, something something, how does Quine account for logic,.... something

Is this perhaps the most cogent line of reasoning we've seen from Sye?
In arguing against the analytic/synthetic distinction, is Quine arguing against the possibility of any analytic truths, for example, those simply stipulated when engaging in set theory or mathematics?

It would seem that nothing stops one from stipulating that the set X consists of elements a,b,and c. Some statements as to the elements of set X being elements of set X wil be true by definition or analytically.

Is not this possibility of stipulation an argument for an analytic/synthetic distintion, even if trivial?
I think that some of the significance of Quine's Two Dogmas of Empiricism is that it anticipated the work of Kripke and Hilary Putnam on natural kinds, wherein empirical discoveries of the properties of natural kinds such as gold can be incorporated into our knowledge of these kinds without it constituing a change in definition.

One supposed consequence of Identity and Necessity and Naming and Necessity is that if mental state A turns out to be brain state B, then this is necessariy true, and NOT because it is an analytic truth.

However, to dispense with the distinction itself and not merely to suggest that it is less applicable than believed is an extreme position.

Popular posts from this blog


(Published in Faith and Philosophy 2011. Volume 28, Issue 2, April 2011. Stephen Law. Pages 129-151) EVIDENCE, MIRACLES AND THE EXISTENCE OF JESUS Stephen Law Abstract The vast majority of Biblical historians believe there is evidence sufficient to place Jesus’ existence beyond reasonable doubt. Many believe the New Testament documents alone suffice firmly to establish Jesus as an actual, historical figure. I question these views. In particular, I argue (i) that the three most popular criteria by which various non-miraculous New Testament claims made about Jesus are supposedly corroborated are not sufficient, either singly or jointly, to place his existence beyond reasonable doubt, and (ii) that a prima facie plausible principle concerning how evidence should be assessed – a principle I call the contamination principle – entails that, given the large proportion of uncorroborated miracle claims made about Jesus in the New Testament documents, we should, in the absence of indepen

Plantinga's Evolutionary Argument Against Naturalism refuted

Here's my central criticism of Plantinga's Evolutionary Argument Against Naturalism (EAAN). It's novel and was published in Analysis last year. Here's the gist. Plantinga argues that if naturalism and evolution are true, then semantic epiphenomenalism is very probably true - that's to say, the content of our beliefs does not causally impinge on our behaviour. And if semantic properties such as having such-and-such content or being true cannot causally impinge on behaviour, then they cannot be selected for by unguided evolution. Plantinga's argument requires, crucially, that there be no conceptual links between belief content and behaviour of a sort that it's actually very plausible to suppose exist (note that to suppose there are such conceptual links is not necessarily to suppose that content can be exhaustively captured in terms of behaviour or functional role, etc. in the way logical behaviourists or functionalists suppose). It turns o

Suggesting a new named fallacy: the Non Post Hoc Fallacy (or David Cameron Fallacy)

Many of us are familiar with the Post Hoc Ergo Propter Hoc Fallacy (' after this, therefore because of this) - Post Hoc Fallacy for short). It's the fallacy of supposing that, because B occurred after A, A must be the cause of B. For example: My car stopped working after I changed the oil, so changing the oil caused it to stop working. Or:  I wore my red jumper to the exam and I passed, so that jumper is lucky: it caused me to pass. This fallacy is so common, it gets a latin name. However, there's a related common fallacy that I think also deserves a name. I am going to call it the Non Post Hoc Fallacy (' not after of this, therefore not because of this), or, perhaps more memorably, the David Cameron Fallacy. Every now and then someone desperate to ‘prove’ that X is not causally responsible for Y – e.g poverty is not a cause of crime, will commit the following fallacy. They will argue that as X has often occurred without Y following, therefore X was not the