Yes: There are certain sentences which are true solely by virtue of the meanings of the words involved, so these sentences are not subject to empirical falsification. Example: “All bachelors are unmarried.” It is impossible for this sentence to be false, provided the words retain their ordinary meaning.
No: Every sentence is potentially open to empirical falsification. [EDIT: I guess the “No” answer would also be appropriate for those who believe that no sentence is open to empirical falsification, although I would be very surprised if anyone on this site fits that description.]
[VOTE BEFORE READING THIS COMMENT TO AVOID PRIMING.]
The most prominent critic of the distinction is Quine. You can read about the reasons for his opposition here. A quote:
Quine… offers a diagnosis of the persistence of the concept of analyticity. Philosophers find the idea plausible because they tend to assume, sometimes unwittingly, that there is a clear notion of cognitive meaning which relates each sentence to the experiences which count for it or against it and which can be applied to sentences taken one-by-one. Given that sort of notion of meaning, we could say: the synthetic sentences are precisely those to whose truth or falsehood experience is relevant; the analytic ones are those whose truth or falsehood is wholly independent of experience (and which can therefore be known a priori).
Quine criticizes this idea of atomistic (sentence-by-sentence) cognitive meaning.… Quine invokes holism, the idea that most of our sentences do not have implications for experience when they are taken one-by-one, each in isolation from the others. What has experiential implication is, in most cases, not an individual sentence but larger chunks of theory. Holism, Quine claims, undermines the atomism of atomistic cognitive meaning.
There is also Chalmers 2009 I guess paper about this, which breifly reviews the history of what happened after Quine polarized the topic.
Revisability and Conceptual Change. Chalmers attemps (in my view succesfully) to rescue 80% of what matters in the distinction, avoiding Quinean and post Quinean traps.
Is “some sentences of first-order logic are tautologies” a sufficient reason to vote yes? If so, clearly we should be talking about that rather than complicated human-language examples like bachelors. If not, what’s the difference?
The answer to your first question will be controverted for pretty much the same reasons that the analytic-synthetic distinction itself is, I think. Quineans will claim that insofar as those sentences of first order logic are actually used in science, they become enmeshed in the holism of cognitive meaning.
Other: Unique meanings are a useful approximation, but if you stretch an approximation thin enough it develops holes. For example, “bachelor” in middle english refers to a squire, and squires can be married.
I probably should have voted for “Other,” but I voted for “Lean toward: yes” because I still outright agree in certain contexts.
Quine’s Two Dogmas is certainly enough to make me doubt the usefulness of the analytic/synthetic distinction as regards ordinary language, but for formal languages, this is not the case. It’s also not clear to me whether it’s impossible to construct a language (for communication) clear enough to make sense of analytic/synthetic distinctions.
This is one of those wonderfully agnostic positions that philosophy often leaves me with.
The idea is that the meaning of an utterance isn’t just one thing; it’s kind of the overlap between two distinct propositions: the sense (that is, the concept or idea by which we find the referent), and the referent, the actual entity to which it refers. The standard textbook example is the word “water”; the sense of “water” would basically correspond to a descriptive, conceptually encoded “water”-iness, and the referent would be the substance itself. Basically, you’ve got your ideas about water, you’ve got the abstract entity you recognize or impute as a member of the reference class (the substance H2O considered in abstract, or a pond, or a glass of clear odorless transparent liquid on the table in front of you), and both are relevant to determining the actual semantic content.
So, OK, sticking with pragmatist’s example, can you summarize the conditions under which “All bachelors are unmarried” becomes false while the words retain their ordinary meaning? (I recognize that we might just turn out to disagree on what their ordinary meaning is, which I think would be uninteresting, but I’m hoping it won’t come to that.)
According to Quine’s meaning holism, explained by pragmatist here, the concepts of “bachelor” and “marriage” are embedded in a wider network of concepts like “human”, society”, “legal relation”, etc, and their use presupposes an amount of “truisms” such like that there exists humans, that humans can get involved in socially-endorsed legal relations, etc.
I find it conceivable that some of this truisms turn out to be false (e.g. imagine you are a brain in a vat) and that the entities you think of as humans are better described with a vastly different network of concepts, inexpressible with our currently existent ones. It may be that after you become aware of this and you acquire the better set of concepts, you will find your old concepts of “bachelor” and “married” confused to several degrees, and in a way such that the best way to make them survive implies that not all bachelors are unmarried.
One motivating example often used by Quineans is the law of excluded middle (“all meaningful propositions are either true or false”). It might seem analytic based on the meaning of “proposition”, “true” and “false”, but there are interpretations of quantum mechanics in which it is false. Whether those interpretations are the best ones is beside the point; the point is that there are several philosophers of physics who find the negation of the law of excluded middle conceivable in a new conceptual structure more appropriate to describe the new facts of quantum mechanics. The Quinean claim is that all propositions are revisable in this way.
It may be that after you become aware of this and you acquire the better set of concepts, you will find your old concepts of “bachelor” and “married” confused to several degrees, and in a way such that the best way to make them survive implies that not all bachelors are unmarried.
Sure, I can believe that. I mean, I can’t imagine it, but I believe it’s possible in some way that I can’t imagine. I certainly agree that meaning is holistic in the sense pragmatist explains, such that the meaning of such a sentence can change based on systemic effects.
But to believe that, and believe that “bachelor” and “marriage” have their ordinary meanings at the same time, is beyond me.
But to believe that, and believe that “bachelor” and “marriage” have their ordinary meanings at the same time, is beyond me.
The contention is that “the ordinary meaning” of a word is a fuzzy and ill-defined concept, once holism creeps in.
Consider another example (used by Putnam, I think). Physicists first introduced the concept of “momentum” as the product of mass and velocity. A central fact that made the concept useful was that momentum is conserved in an isolated system. Later, with relativity, it became clear that the conserved quantity is not really the product of mass and velocity, but includes a speed-of-light dependent factor as well. Physicists started then calling this quantity “momentum”.
Now, was this a change in the meaning of the word “momentum”, or a new fact discovered about the same physical entity momentum? This would seem to depend on whether the first early modern physicist who used the word intended “momentum” to have a fixed meaning as the product of mass and velocity, or as the quantity conserved in an isolated system. But he probably didn’t make his intention clear, and in any case his private intention does not matter if meaning is social and holistic. Even if most pre-Einstein physicists would have (if questioned, which they weren’t) agreed that “momentum” meant definitionally mass times velocity, post-Einstein physicists may be perfectly justified in saying they would have been wrong, that Einstein made a new physical discovery about the same quantity they were trying to talk about and not merely changed the meaning of words. What is “the ordinary meaning” (pre-Einstein) of “momentum” is not a question with a well-defined answer; the relevant unit of meaning was the whole physical theory, which was replaced by a new one, and we cannot make a clean distinction between which were changes in meaning and which were changes in factual beliefs.
It is more difficult to imagine something like this happening for “bachelor”, but according to Quineans, the difference is only of degree.
It is more difficult to imagine something like this happening for “bachelor”, but according to Quineans, the difference is only of degree.
Considering the present controversy over “the definition of marriage”, I think we can imagine many such cases.
Is a man who has lived with the same woman for ten years — but has never had a wedding — a “bachelor”? How about a man who has had a commitment ceremony with another man? (Does it matter if the invitations said “marriage” or “commitment ceremony”?) A man who has a marriage of convenience to a woman he has never slept with, for purposes of immigration? A man from a culture where he was, as a young boy, married by his family to a young girl, but who has left that setting and never seen her since? A man who believes he is married to a particular woman, but subsequent careful inspection of family history reveals that she is his long-lost sister and thus the marriage is invalid for incest? A Catholic priest?
Just as subsequent physics discoveries can problematize the definition of “momentum”, subsequent social and personal-history discoveries can problematize the definition of “bachelor”. The “ordinary meaning” is only had by choosing to ignore problems.
For my own part, I would say that if we take the relevant unit of meaning to be the whole physical theory (a position I find compelling in principle, if unwieldy in practice), it follows that changing the physical theory does not preserve preexisting meanings. I would not say that the meaning of “momentum” changed, precisely, but that “momentum” acquired a new meaning in addition to its old one, and anyone talking about momentum in a relativistic context is using the new meaning, even though people talking about momentum in a non-relativistic context can go on using the old meaning. (I would also say that the intent of the first physicist to use the term is effectively irrelevant.)
This also implies that people who try to copy over assertions about momentum from non-relativistic contexts to relativistic ones are essentially confusing homophones… similar in principle to what happens if I try to copy over assertions about monarchs from lepidopterological contexts to governmental ones.
But, OK, I can understand how someone could sensibly argue that no, the meaning is preserved, because the meaning was always fuzzy in the first place, we just became aware of the fuzziness late in the game. (This seems to in turn depend on a strongly externalist account of meaning.)
Other: I think the distinction clearly exists (in the sense that it’s a human
concept that carves reality more or less at the joints, so that a person can
learn the definition and then come up with examples), but it’s not useful for
much apart from arguments-about-words-rather-than-things, and, because of the
fuzziness of definitions, in practice it’s more like a continuum than a
strictly binary distinction.
Unless you buy into Kant’s synthetic a priori arguments, that’s really all analytic means. Of course, in practice it’s far more interesting & complicated, and it even leads to the kind of applications that have made secure internet commerce possible, not to mention the computers we use to do that.
At least, on some days I think that’s what ‘analytic’ means. Maybe.
Crud. Misread the poll as about the viability of the a prior / a posteriori distinction. (True without empirical content {the usual example is first-order logic or mathematics} vs. truth-value only after empirical content).
History note: Kant wanted to assert the existence of a priori synthentic statements—a position I find nonsensical.
Analytic-synthetic distinction: yes or no?
[pollid:78]
Yes: There are certain sentences which are true solely by virtue of the meanings of the words involved, so these sentences are not subject to empirical falsification. Example: “All bachelors are unmarried.” It is impossible for this sentence to be false, provided the words retain their ordinary meaning.
No: Every sentence is potentially open to empirical falsification. [EDIT: I guess the “No” answer would also be appropriate for those who believe that no sentence is open to empirical falsification, although I would be very surprised if anyone on this site fits that description.]
The Yes answer seems obvious, is there some sort of gotcha?
[VOTE BEFORE READING THIS COMMENT TO AVOID PRIMING.]
The most prominent critic of the distinction is Quine. You can read about the reasons for his opposition here. A quote:
There is also Chalmers 2009 I guess paper about this, which breifly reviews the history of what happened after Quine polarized the topic. Revisability and Conceptual Change.
Chalmers attemps (in my view succesfully) to rescue 80% of what matters in the distinction, avoiding Quinean and post Quinean traps.
Is “some sentences of first-order logic are tautologies” a sufficient reason to vote yes? If so, clearly we should be talking about that rather than complicated human-language examples like bachelors. If not, what’s the difference?
The answer to your first question will be controverted for pretty much the same reasons that the analytic-synthetic distinction itself is, I think. Quineans will claim that insofar as those sentences of first order logic are actually used in science, they become enmeshed in the holism of cognitive meaning.
Good point to raise, nevertheless.
Other: Unique meanings are a useful approximation, but if you stretch an approximation thin enough it develops holes. For example, “bachelor” in middle english refers to a squire, and squires can be married.
I probably should have voted for “Other,” but I voted for “Lean toward: yes” because I still outright agree in certain contexts.
Quine’s Two Dogmas is certainly enough to make me doubt the usefulness of the analytic/synthetic distinction as regards ordinary language, but for formal languages, this is not the case. It’s also not clear to me whether it’s impossible to construct a language (for communication) clear enough to make sense of analytic/synthetic distinctions.
This is one of those wonderfully agnostic positions that philosophy often leaves me with.
Other: Lean toward two-dimensionalism.
Can you elaborate on this? I haven’t heard of it in this context before.
The idea is that the meaning of an utterance isn’t just one thing; it’s kind of the overlap between two distinct propositions: the sense (that is, the concept or idea by which we find the referent), and the referent, the actual entity to which it refers. The standard textbook example is the word “water”; the sense of “water” would basically correspond to a descriptive, conceptually encoded “water”-iness, and the referent would be the substance itself. Basically, you’ve got your ideas about water, you’ve got the abstract entity you recognize or impute as a member of the reference class (the substance H2O considered in abstract, or a pond, or a glass of clear odorless transparent liquid on the table in front of you), and both are relevant to determining the actual semantic content.
This was the most surprising poll result, to me.
I assume you also expected far more of the (currently) less popular answer?
Huh, yeah, I’m pretty surprised too. Possibly even for the same reason :P
Huh. Well, I’m willing to be convinced.
So, OK, sticking with pragmatist’s example, can you summarize the conditions under which “All bachelors are unmarried” becomes false while the words retain their ordinary meaning? (I recognize that we might just turn out to disagree on what their ordinary meaning is, which I think would be uninteresting, but I’m hoping it won’t come to that.)
According to Quine’s meaning holism, explained by pragmatist here, the concepts of “bachelor” and “marriage” are embedded in a wider network of concepts like “human”, society”, “legal relation”, etc, and their use presupposes an amount of “truisms” such like that there exists humans, that humans can get involved in socially-endorsed legal relations, etc.
I find it conceivable that some of this truisms turn out to be false (e.g. imagine you are a brain in a vat) and that the entities you think of as humans are better described with a vastly different network of concepts, inexpressible with our currently existent ones. It may be that after you become aware of this and you acquire the better set of concepts, you will find your old concepts of “bachelor” and “married” confused to several degrees, and in a way such that the best way to make them survive implies that not all bachelors are unmarried.
One motivating example often used by Quineans is the law of excluded middle (“all meaningful propositions are either true or false”). It might seem analytic based on the meaning of “proposition”, “true” and “false”, but there are interpretations of quantum mechanics in which it is false. Whether those interpretations are the best ones is beside the point; the point is that there are several philosophers of physics who find the negation of the law of excluded middle conceivable in a new conceptual structure more appropriate to describe the new facts of quantum mechanics. The Quinean claim is that all propositions are revisable in this way.
Sure, I can believe that. I mean, I can’t imagine it, but I believe it’s possible in some way that I can’t imagine. I certainly agree that meaning is holistic in the sense pragmatist explains, such that the meaning of such a sentence can change based on systemic effects.
But to believe that, and believe that “bachelor” and “marriage” have their ordinary meanings at the same time, is beyond me.
The contention is that “the ordinary meaning” of a word is a fuzzy and ill-defined concept, once holism creeps in.
Consider another example (used by Putnam, I think). Physicists first introduced the concept of “momentum” as the product of mass and velocity. A central fact that made the concept useful was that momentum is conserved in an isolated system. Later, with relativity, it became clear that the conserved quantity is not really the product of mass and velocity, but includes a speed-of-light dependent factor as well. Physicists started then calling this quantity “momentum”.
Now, was this a change in the meaning of the word “momentum”, or a new fact discovered about the same physical entity momentum? This would seem to depend on whether the first early modern physicist who used the word intended “momentum” to have a fixed meaning as the product of mass and velocity, or as the quantity conserved in an isolated system. But he probably didn’t make his intention clear, and in any case his private intention does not matter if meaning is social and holistic. Even if most pre-Einstein physicists would have (if questioned, which they weren’t) agreed that “momentum” meant definitionally mass times velocity, post-Einstein physicists may be perfectly justified in saying they would have been wrong, that Einstein made a new physical discovery about the same quantity they were trying to talk about and not merely changed the meaning of words. What is “the ordinary meaning” (pre-Einstein) of “momentum” is not a question with a well-defined answer; the relevant unit of meaning was the whole physical theory, which was replaced by a new one, and we cannot make a clean distinction between which were changes in meaning and which were changes in factual beliefs.
It is more difficult to imagine something like this happening for “bachelor”, but according to Quineans, the difference is only of degree.
Considering the present controversy over “the definition of marriage”, I think we can imagine many such cases.
Is a man who has lived with the same woman for ten years — but has never had a wedding — a “bachelor”? How about a man who has had a commitment ceremony with another man? (Does it matter if the invitations said “marriage” or “commitment ceremony”?) A man who has a marriage of convenience to a woman he has never slept with, for purposes of immigration? A man from a culture where he was, as a young boy, married by his family to a young girl, but who has left that setting and never seen her since? A man who believes he is married to a particular woman, but subsequent careful inspection of family history reveals that she is his long-lost sister and thus the marriage is invalid for incest? A Catholic priest?
Just as subsequent physics discoveries can problematize the definition of “momentum”, subsequent social and personal-history discoveries can problematize the definition of “bachelor”. The “ordinary meaning” is only had by choosing to ignore problems.
OK, I think I follow.
For my own part, I would say that if we take the relevant unit of meaning to be the whole physical theory (a position I find compelling in principle, if unwieldy in practice), it follows that changing the physical theory does not preserve preexisting meanings. I would not say that the meaning of “momentum” changed, precisely, but that “momentum” acquired a new meaning in addition to its old one, and anyone talking about momentum in a relativistic context is using the new meaning, even though people talking about momentum in a non-relativistic context can go on using the old meaning. (I would also say that the intent of the first physicist to use the term is effectively irrelevant.)
This also implies that people who try to copy over assertions about momentum from non-relativistic contexts to relativistic ones are essentially confusing homophones… similar in principle to what happens if I try to copy over assertions about monarchs from lepidopterological contexts to governmental ones.
But, OK, I can understand how someone could sensibly argue that no, the meaning is preserved, because the meaning was always fuzzy in the first place, we just became aware of the fuzziness late in the game. (This seems to in turn depend on a strongly externalist account of meaning.)
Not the source of my surprise.
Thanks for priming everyone who reads this thread before voting, though.
If you are interested in exploring this issue, Quine’s Two Dogmas is probably the best place to start.
http://www.ditext.com/quine/quine.html
Other: I think the distinction clearly exists (in the sense that it’s a human concept that carves reality more or less at the joints, so that a person can learn the definition and then come up with examples), but it’s not useful for much apart from arguments-about-words-rather-than-things, and, because of the fuzziness of definitions, in practice it’s more like a continuum than a strictly binary distinction.
Wow. I voted “No” because modus ponens and such are just procedures that produce evidence, not some kind of magical truth juice.
I guess if “Yes” means you can construct consistent definitions that are not empirically verifiable, then I’d vote Yes.
Unless you buy into Kant’s synthetic a priori arguments, that’s really all analytic means. Of course, in practice it’s far more interesting & complicated, and it even leads to the kind of applications that have made secure internet commerce possible, not to mention the computers we use to do that.
At least, on some days I think that’s what ‘analytic’ means. Maybe.
I’m not interested in this philosophy stuff, but our names have an amusing symmetry.
internet fist bump
Crud. Misread the poll as about the viability of the a prior / a posteriori distinction. (True without empirical content {the usual example is first-order logic or mathematics} vs. truth-value only after empirical content).
History note: Kant wanted to assert the existence of a priori synthentic statements—a position I find nonsensical.