Interesting post, which I would like to make some comments to.
First of all, I don’t think it’s necessarily a bad thing to be associated with the logical positivists. In their days, they were, in my view, one of the most interesting proponents of the scientific world-view. The fact that their program (which was unusually well specified, for being a philosophical program—somethinghat contributed to its demise, since it made it easier to falsify) ultimately was shown to be unviable does not show that their general outlook was mistaken. Verificationism and the notion that philosophy should construct a scientific world-view can still be good ideas (in fact are good ideas, in my view) even though the logical positivists’ more specific ideas were misguided.
Secondly, Yudowsky is right that unverifiable statements are not meaningless in the same sense as true nonsense is meaningless. In his essay “Positivism Against Hegelianism”, Ernest Gellner makes the same point:
“The logical positivist definition of meaning was inevitably somewhat confused. Clearly, though, it could not define ‘meaning’ in the sense used by working linguists as he classes of sound patterns which are emitted, recognised and socially accepted in a given speech community. By such a criterion, ‘metaphysical’ [i,e. unverifiable—my note] statements patently would be meaningful. The anti-Platonism of paradigmatic logical positivism equally prevents us from interpreting the delimitation of meaning as the characterisation of a given essence of ‘meaning’, as for them there are no such essences (thiugh in some semi-conscious manner, and in disharmony with their nominal anti-Platonism, I strongly suspect that this was precisely what many of them did mean).
The only thing which in effect they could mean, plausibly and in harmony with their other principles, was this: the definition circumscribed, not the de facto custom of any one or every linguistic community, but the limits of the kind of use of speech which deserves respect and commendation. It was a definition not of meaningful speech, but of commendable, good speech. Their verificationism was a covert piece of ethics. Meaningless was a condemnation, and meaning a commendation.” (Gellner, Relativism and the Social Sciences, pp. 30-31)
As I argue in my article “Ernest Gellner’s Use of the Social Sciences in Philosophy” (Phil of Soc Sci, 2014:1), this interpretation is not quite right, though. Gellner’s interpretation is over-charitable—the logical positivists really did saw their assertion that unverifiable statements are meaningless as descriptive (even though Gellner is right that they at some level intended it to be normative). More importantly, so did their chief critic Quine who dealt an important blow to logical positivism by showing that that the logical positivists’ conception of meaning failed to illuminate how language actually works (in “Two Dogmas of Empiricism”). He subsequently argued that it should be replaced by his own notion of stimulus meaning—a behaviouristic notion which he held to be empirically acceptable, unlike the logical positivists’ notion of meaning. (Word and Object)
The logical positivist notion of meaning was, in short, not empirically grounded in any way. They just asserted that some statements are meaningless, and some are not, while producing little argument for it—and certainly no empirical evidence. My guess is that there is an important lesson to be learnt here. For all their talk of a scientific world-view, the logical positivists were rather influenced by the German tradition of a priori philosophy (e.g. Carnap was influenced by neo-Kantianism). Also, there was a strong anti-psychologistic trend in early 20th century philosophy, inherited from the 19th century (especially Frege: for an excellent overview over why psychology was severed from philosophy in Germany, read Martin Kusch’s Psychologism: A Case-Study in the Sociology of Philosophical Knowledge, where it is convincingly argued that this happened for social, non-rational reasons.
Naturalistic philosophers have for centuries tried to make philosophy more empirical and more based on the sciences, but although some progress is done, it seems that it seldom goes far enough. E.g. the later Wittgenstein – in many ways a naturalistic philosopher—argued that philosophers should “not think, but look”, and that we should look upon language in an “anthropological way”, seeing how it really works (rather than constructing a priori models, as philosophers often had done). Still he did no empirical investigations himself. Likewise the logical positivists venerated science but used a non-empirical notion of meaning.
There might be several reasons for this, but the most important one seems to me be that philosophers are more or less exclusively trained at a priori-reasoning and don’t really have a lot of other useful knowledge—certainly not cutting-edge knowledge. In order to make philosophy thoroughly naturalistic, philosophers must—as has been argued on this site—be extensively trained especially in cognitive psychology (which I hold to be the empirical discipline most useful to philosophers) but also as far as possible (and depending on specialization) in other disciplines.
Lastly I would like to add that Karl Popper’s famous falsificationism was probably closer to Yudowsky’s thinking, since Popper did not see falsifiability as a criterion of meaning, but rather as a criterion of whether a theory should be seen as scientific. Popper was, though, much more positive to metaphysics (e.g. he was a realist) than the logical positivists, and I’m not sure if Yudowsky would like to follow him on that point.
Interesting post, which I would like to make some comments to.
First of all, I don’t think it’s necessarily a bad thing to be associated with the logical positivists. In their days, they were, in my view, one of the most interesting proponents of the scientific world-view. The fact that their program (which was unusually well specified, for being a philosophical program—somethinghat contributed to its demise, since it made it easier to falsify) ultimately was shown to be unviable does not show that their general outlook was mistaken. Verificationism and the notion that philosophy should construct a scientific world-view can still be good ideas (in fact are good ideas, in my view) even though the logical positivists’ more specific ideas were misguided.
Secondly, Yudowsky is right that unverifiable statements are not meaningless in the same sense as true nonsense is meaningless. In his essay “Positivism Against Hegelianism”, Ernest Gellner makes the same point:
“The logical positivist definition of meaning was inevitably somewhat confused. Clearly, though, it could not define ‘meaning’ in the sense used by working linguists as he classes of sound patterns which are emitted, recognised and socially accepted in a given speech community. By such a criterion, ‘metaphysical’ [i,e. unverifiable—my note] statements patently would be meaningful. The anti-Platonism of paradigmatic logical positivism equally prevents us from interpreting the delimitation of meaning as the characterisation of a given essence of ‘meaning’, as for them there are no such essences (thiugh in some semi-conscious manner, and in disharmony with their nominal anti-Platonism, I strongly suspect that this was precisely what many of them did mean).
The only thing which in effect they could mean, plausibly and in harmony with their other principles, was this: the definition circumscribed, not the de facto custom of any one or every linguistic community, but the limits of the kind of use of speech which deserves respect and commendation. It was a definition not of meaningful speech, but of commendable, good speech. Their verificationism was a covert piece of ethics. Meaningless was a condemnation, and meaning a commendation.” (Gellner, Relativism and the Social Sciences, pp. 30-31)
As I argue in my article “Ernest Gellner’s Use of the Social Sciences in Philosophy” (Phil of Soc Sci, 2014:1), this interpretation is not quite right, though. Gellner’s interpretation is over-charitable—the logical positivists really did saw their assertion that unverifiable statements are meaningless as descriptive (even though Gellner is right that they at some level intended it to be normative). More importantly, so did their chief critic Quine who dealt an important blow to logical positivism by showing that that the logical positivists’ conception of meaning failed to illuminate how language actually works (in “Two Dogmas of Empiricism”). He subsequently argued that it should be replaced by his own notion of stimulus meaning—a behaviouristic notion which he held to be empirically acceptable, unlike the logical positivists’ notion of meaning. (Word and Object)
The logical positivist notion of meaning was, in short, not empirically grounded in any way. They just asserted that some statements are meaningless, and some are not, while producing little argument for it—and certainly no empirical evidence. My guess is that there is an important lesson to be learnt here. For all their talk of a scientific world-view, the logical positivists were rather influenced by the German tradition of a priori philosophy (e.g. Carnap was influenced by neo-Kantianism). Also, there was a strong anti-psychologistic trend in early 20th century philosophy, inherited from the 19th century (especially Frege: for an excellent overview over why psychology was severed from philosophy in Germany, read Martin Kusch’s Psychologism: A Case-Study in the Sociology of Philosophical Knowledge, where it is convincingly argued that this happened for social, non-rational reasons.
Naturalistic philosophers have for centuries tried to make philosophy more empirical and more based on the sciences, but although some progress is done, it seems that it seldom goes far enough. E.g. the later Wittgenstein – in many ways a naturalistic philosopher—argued that philosophers should “not think, but look”, and that we should look upon language in an “anthropological way”, seeing how it really works (rather than constructing a priori models, as philosophers often had done). Still he did no empirical investigations himself. Likewise the logical positivists venerated science but used a non-empirical notion of meaning.
There might be several reasons for this, but the most important one seems to me be that philosophers are more or less exclusively trained at a priori-reasoning and don’t really have a lot of other useful knowledge—certainly not cutting-edge knowledge. In order to make philosophy thoroughly naturalistic, philosophers must—as has been argued on this site—be extensively trained especially in cognitive psychology (which I hold to be the empirical discipline most useful to philosophers) but also as far as possible (and depending on specialization) in other disciplines.
Lastly I would like to add that Karl Popper’s famous falsificationism was probably closer to Yudowsky’s thinking, since Popper did not see falsifiability as a criterion of meaning, but rather as a criterion of whether a theory should be seen as scientific. Popper was, though, much more positive to metaphysics (e.g. he was a realist) than the logical positivists, and I’m not sure if Yudowsky would like to follow him on that point.