6 ms·

I’d be interested in hearing more about this. Don’t know much about formal semantics (and in fact am somewhat skeptical of the whole field), but I do have a bac

by bradrn 3y ago

I’d be interested in hearing more about this. Don’t know much about formal semantics (and in fact am somewhat skeptical of the whole field), but I do have a background in language construction, and would say I have a fairly good knowledge of linguistics more generally.

adastra22 3y ago
My email is in my profile. Formal semantics is a broad field full of some very interesting and practical theories, and a lot of quackery. I don't blame you for being skeptical. Specifically though I think that Lucas Champollion has done a great job of marrying Davidsonian event semantics with more modern work on qualifiers and counterfactuals that finally creates--I believe--an universal framework for capturing the full meaning of any speech utterance: https://champollion.hosting.nyu.edu https://champollion.hosting.nyu.edu If I am correct in this assessment, then the next logical step from a conlang/auxlang perspective is: can we build a direct, unambiguous spoken grammar for this universal semantics framework? If so, then we will have achieved what Lojban failed to even attempt: unambiguous meaning.
- bradrn 3y ago
  > My email is in my profile. OK, I’ll email you if you want to continue this discussion elsewhere. > https://champollion.hosting.nyu.edu https://champollion.hosting.nyu.edu Looks interesting… where should I start reading? > If I am correct in this assessment, then the next logical step from a conlang/auxlang perspective is: can we build a direct, unambiguous spoken grammar for this universal semantics framework? If so, then we will have achieved what Lojban failed to even attempt: unambiguous meaning. Personally, I’m not sure if ‘unambiguous meaning’ is possible or even desirable. Even if formal semantics allows us to achieve complete unambiguity when it comes to grammatical categories (quantifiers, temporal and spatial setting, etc.), I think it would be exceedingly difficult to make every single lexeme completely unambiguous. And even if this were achieved, I’m not sure if such a language would even be usable by humans. Still, it would be an interesting experiment!
  - adastra22 3y ago
    Oh I'm happy to continue this conversation here, although it's late where I am and I'll be going to bed soon, and HN threads are ephemeral. Feel free to email me if you'd like to actually collaborate on this. > Looks interesting… where should I start reading? IIRC he did some graduate summer schools that were recorded online and have attached lecture notes. Those plus the referenced papers would be a good place to start. The one in particular that I'm thinking of is, I believe, the Summer 2014 topic of "Integrating Montague Semantics and Event Semantics." That marriage of Davidsonian and Montague semantics into a single framework is, I believe, a universal theory of semantics. [It should be noted that he makes no such grandiose claims.] > Personally, I’m not sure if ‘unambiguous meaning’ is possible or even desirable By 'unambiguous meaning' I mean something more specific than the words alone imply. [I am aware of the irony.] In computational semantics, the goal is to transform a parsing of a sentence into a logical statement (or in the case of a question, a query) about the world. For example, the sentence "every dog barks" would be translated into a logical statement of the form "for every dog there exists at least one barking event with that dog as the acting subject." Or rather a mathematical statement to that effect that I don't know how to render on HN. But unfortunately in natural language even when a language is correctly parsed in terms of syntax--which is all Lojban seeks to accomplish--there are still many different possible semantic implications. For example, the sentence "Spot didn't bark" could mean 1. There is no event in which Spot has ever barked, or 2. In the case of a specific event, Spot did not bark. English doesn't differentiate without clarification, e.g. "Spot never barks" or "Spot didn't bark that time." Lojban has universal and existential quantifiers, but their use is not mandated. When I say I want unambiguous meaning, I mean something very specific: monoparsing of sentences to logical statements/semantic formalisms. There ought to be a one-to-one mapping between sentences in the language, and logical sentences in this universal formalism. The claims of those sentences could be as broad or narrow as desired, but there shouldn't be any ambiguity about what is being claimed.
    - bradrn 3y ago
      > Feel free to email me if you'd like to actually collaborate on this. Emailed, though I’m not sure how much time I’d actually have available to collaborate. > The one in particular that I'm thinking of is, I believe, the Summer 2014 topic of "Integrating Montague Semantics and Event Semantics." I’ll have a look at it, thanks! > When I say I want unambiguous meaning, I mean something very specific: monoparsing of sentences to logical statements/semantic formalisms. There ought to be a one-to-one mapping between sentences in the language, and logical sentences in this universal formalism. The claims of those sentences could be as broad or narrow as desired, but there shouldn't be any ambiguity about what is being claimed. OK, this makes far more sense. I still maintain there will be a lot of ambiguity, purely because the individual words themselves are ambiguous, but it’s an interesting goal nonetheless.

by bradrn 3y ago

adastra22 3y ago

My email is in my profile. Formal semantics is a broad field full of some very interesting and practical theories, and a lot of quackery. I don't blame you for being skeptical. Specifically though I think that Lucas Champollion has done a great job of marrying Davidsonian event semantics with more modern work on qualifiers and counterfactuals that finally creates--I believe--an universal framework for capturing the full meaning of any speech utterance: https://champollion.hosting.nyu.edu https://champollion.hosting.nyu.edu If I am correct in this assessment, then the next logical step from a conlang/auxlang perspective is: can we build a direct, unambiguous spoken grammar for this universal semantics framework? If so, then we will have achieved what Lojban failed to even attempt: unambiguous meaning.

bradrn 3y ago
> My email is in my profile. OK, I’ll email you if you want to continue this discussion elsewhere. > https://champollion.hosting.nyu.edu https://champollion.hosting.nyu.edu Looks interesting… where should I start reading? > If I am correct in this assessment, then the next logical step from a conlang/auxlang perspective is: can we build a direct, unambiguous spoken grammar for this universal semantics framework? If so, then we will have achieved what Lojban failed to even attempt: unambiguous meaning. Personally, I’m not sure if ‘unambiguous meaning’ is possible or even desirable. Even if formal semantics allows us to achieve complete unambiguity when it comes to grammatical categories (quantifiers, temporal and spatial setting, etc.), I think it would be exceedingly difficult to make every single lexeme completely unambiguous. And even if this were achieved, I’m not sure if such a language would even be usable by humans. Still, it would be an interesting experiment!
- adastra22 3y ago
  Oh I'm happy to continue this conversation here, although it's late where I am and I'll be going to bed soon, and HN threads are ephemeral. Feel free to email me if you'd like to actually collaborate on this. > Looks interesting… where should I start reading? IIRC he did some graduate summer schools that were recorded online and have attached lecture notes. Those plus the referenced papers would be a good place to start. The one in particular that I'm thinking of is, I believe, the Summer 2014 topic of "Integrating Montague Semantics and Event Semantics." That marriage of Davidsonian and Montague semantics into a single framework is, I believe, a universal theory of semantics. [It should be noted that he makes no such grandiose claims.] > Personally, I’m not sure if ‘unambiguous meaning’ is possible or even desirable By 'unambiguous meaning' I mean something more specific than the words alone imply. [I am aware of the irony.] In computational semantics, the goal is to transform a parsing of a sentence into a logical statement (or in the case of a question, a query) about the world. For example, the sentence "every dog barks" would be translated into a logical statement of the form "for every dog there exists at least one barking event with that dog as the acting subject." Or rather a mathematical statement to that effect that I don't know how to render on HN. But unfortunately in natural language even when a language is correctly parsed in terms of syntax--which is all Lojban seeks to accomplish--there are still many different possible semantic implications. For example, the sentence "Spot didn't bark" could mean 1. There is no event in which Spot has ever barked, or 2. In the case of a specific event, Spot did not bark. English doesn't differentiate without clarification, e.g. "Spot never barks" or "Spot didn't bark that time." Lojban has universal and existential quantifiers, but their use is not mandated. When I say I want unambiguous meaning, I mean something very specific: monoparsing of sentences to logical statements/semantic formalisms. There ought to be a one-to-one mapping between sentences in the language, and logical sentences in this universal formalism. The claims of those sentences could be as broad or narrow as desired, but there shouldn't be any ambiguity about what is being claimed.
  - bradrn 3y ago
    > Feel free to email me if you'd like to actually collaborate on this. Emailed, though I’m not sure how much time I’d actually have available to collaborate. > The one in particular that I'm thinking of is, I believe, the Summer 2014 topic of "Integrating Montague Semantics and Event Semantics." I’ll have a look at it, thanks! > When I say I want unambiguous meaning, I mean something very specific: monoparsing of sentences to logical statements/semantic formalisms. There ought to be a one-to-one mapping between sentences in the language, and logical sentences in this universal formalism. The claims of those sentences could be as broad or narrow as desired, but there shouldn't be any ambiguity about what is being claimed. OK, this makes far more sense. I still maintain there will be a lot of ambiguity, purely because the individual words themselves are ambiguous, but it’s an interesting goal nonetheless.