Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Tue, 2 Jun 2020 at 23:27, Kerim Aydin via agora-discussion wrote: > On 6/2/2020 4:15 PM, Kerim Aydin wrote: > > On 6/2/2020 10:53 AM, James Cook via agora-discussion wrote: > >> This topic has been bugging me ever since some recent CFJ > >> discussion(s) (don't remember which) in which authorial intent was > >> dismissed as irrelevant to the meaning of the rules, because it's not > >> explicitly mentioned in R217. I think that when R217 says "the text of > >> the rules takes precedence", it is requiring us to interpret the > >> meaning of the rules text, and it's possible for intention to enter > >> the picture there. > > > > Can't remember the discussion, but IMO it's not because of R217 per se - > > the standard reason to ignore intent for legislation (in general, not just > > Agora) is that the rules are decided on by all the voters. If, in voting, > > I read something the author wrote, and interpret it a different way, and > > that matters for my vote choice, then there's no consensus intent that can > > be directly inferred from rules text - the author's intent is no more > > privileged than any other voter's intent at that point. > > > > Sometimes judges will go back to the conversations surrounding a proposal, > > and if there are conversations that provide evidence of intent then we can > > use that to a certain extent, as long as it shows "consensus voter intent" > > and not just "author intent". Thanks, that's interesting to hear. > Also remembering the game-nature of Agora here should make us > doubly-suspicious of legislative intent. After all, the "intent" of an > author may be to slip something past the voters that most voters would > read innocently but the author intends to be a scam. Or, as a voter, I > might read something and say "there's a scam there that the author > probably didn't intend, my intent in voting is to exploit that".. These are reasonable reasons to be suspicious of the author's intent. I'm just glad I'm not the only one who thinks R217 isn't necessarily the reason to ignore intent. - Falsifian
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On 6/2/2020 4:15 PM, Kerim Aydin wrote: > > On 6/2/2020 10:53 AM, James Cook via agora-discussion wrote: >> This topic has been bugging me ever since some recent CFJ >> discussion(s) (don't remember which) in which authorial intent was >> dismissed as irrelevant to the meaning of the rules, because it's not >> explicitly mentioned in R217. I think that when R217 says "the text of >> the rules takes precedence", it is requiring us to interpret the >> meaning of the rules text, and it's possible for intention to enter >> the picture there. > > Can't remember the discussion, but IMO it's not because of R217 per se - > the standard reason to ignore intent for legislation (in general, not just > Agora) is that the rules are decided on by all the voters. If, in voting, > I read something the author wrote, and interpret it a different way, and > that matters for my vote choice, then there's no consensus intent that can > be directly inferred from rules text - the author's intent is no more > privileged than any other voter's intent at that point. > > Sometimes judges will go back to the conversations surrounding a proposal, > and if there are conversations that provide evidence of intent then we can > use that to a certain extent, as long as it shows "consensus voter intent" > and not just "author intent". Also remembering the game-nature of Agora here should make us doubly-suspicious of legislative intent. After all, the "intent" of an author may be to slip something past the voters that most voters would read innocently but the author intends to be a scam. Or, as a voter, I might read something and say "there's a scam there that the author probably didn't intend, my intent in voting is to exploit that". -G.
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On 6/2/2020 10:53 AM, James Cook via agora-discussion wrote: > This topic has been bugging me ever since some recent CFJ > discussion(s) (don't remember which) in which authorial intent was > dismissed as irrelevant to the meaning of the rules, because it's not > explicitly mentioned in R217. I think that when R217 says "the text of > the rules takes precedence", it is requiring us to interpret the > meaning of the rules text, and it's possible for intention to enter > the picture there. Can't remember the discussion, but IMO it's not because of R217 per se - the standard reason to ignore intent for legislation (in general, not just Agora) is that the rules are decided on by all the voters. If, in voting, I read something the author wrote, and interpret it a different way, and that matters for my vote choice, then there's no consensus intent that can be directly inferred from rules text - the author's intent is no more privileged than any other voter's intent at that point. Sometimes judges will go back to the conversations surrounding a proposal, and if there are conversations that provide evidence of intent then we can use that to a certain extent, as long as it shows "consensus voter intent" and not just "author intent". -G.
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
> Because what makes a judgement "appropriate" as per R591 is its specific > relationship to the statement, not the intended statement. > > -G. I was wondering whether evidence of the author's intention can affect the meaning of a statement. I'm pretty sure the meaning of a statement can depend on context. E.g. the statement of CFJ 3735 is "There was only one valid bid, namely for 1 coin by CuddleBeam.", which is only meaningful if you know what auction it's referring to. I don't think it's much of a stretch to say evidence of intention could form part of that context. I'd even go so far as to say that in some situations, there's no distinction between intended and actual meaning. E.g. if I get a personal email from a friend and have trouble understanding one of the sentences, I will try to figure out what they meant; there's no reason for me to imagine the statement has some kind of "pure" meaning divorced from intention. The way I see it, CFJ statements (and rule texts) are a different from that example because as public statements it's more important that we come to a common agreement on what they mean than it is to correctly infer the author's intended meaning. But I don't think it automatically follows that evidence of intention plays no role. This topic has been bugging me ever since some recent CFJ discussion(s) (don't remember which) in which authorial intent was dismissed as irrelevant to the meaning of the rules, because it's not explicitly mentioned in R217. I think that when R217 says "the text of the rules takes precedence", it is requiring us to interpret the meaning of the rules text, and it's possible for intention to enter the picture there. Of course, in the particular CFJ being discussed, I presented the evidence of my intention after making the statement. Maybe that means it shouldn't be taken into consideration. - Falsifian
Re: Fw: Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
> > Keep in mind that word choice, even where it's synonymous in a vacuum, > > conveys > > meaning. This is a feature of language, we unconsciously assume speakers are > > efficient and pragmatic and when expectations are broken it conveys > > additional > > meaning. You used "if and only if" where I would normally use "and". So my > > first thought was that you were focusing on something special about the > > biconditional. In this case, its implication that the two facts are > > meaningfully connected to each other. Ah, maybe I get it. "If and only if" is a strange thing to write, so you inferred a meaning that would require me to use such a strange wording? I'm not sure what you mean about "and", though; that would be another meaning entirely. I think my statement is fundamentally pretty strange, and I don't know a way to phrase it in a way that sounds natural, except maybe by spelling it out with a much longer sentence. - Falsifian
Fw: Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
(oops sent this straight to falsifian the first time.) ‐‐‐ Original Message ‐‐‐ On Monday, June 1, 2020 10:05 PM, nch wrote: > On Monday, June 1, 2020 9:44:27 PM CDT you wrote: > > > On Tue, 2 Jun 2020 at 01:57, nch via agora-discussion > > agora-discussion@agoranomic.org wrote: > > > > > On Monday, June 1, 2020 8:46:09 PM CDT James Cook via agora-discussion > > wrote: > > > > > Isn't that still a difference in intended meaning? Maybe I didn't > > > > phrase it clearly enough the first time, but my intended meaning was > > > > "Right now at the moment I'm calling this CFJ, the truth value > > > > (true/false) of 'Falsifian owns at least one blot' equals the truth > > > > value of 'English Wikipedia has an article titled "Sponge"'". > > > > > > If you had used "right now" or "currently" I'd agree with your reading, > > > see > > > below. > > > > > > > Also, CFJ statements about things like "Alice owns a blot" are usually > > > > assumed to be about the current situation at the time the statement > > > > was called. Are you saying the words "if and only if" override that > > > > default, and lead you interpret my statement as encompassing other > > > > times and/or situations other than the current one? Or am I > > > > misunderstanding your argument? > > > > > > There's no "override". In "Alice owns a blot" there's no ambiguity about > > > whether that statement is present progressive. When you introduce a modal, > > > you also introduce an ambiguity: now the sentence could be present > > > progressive or it could be conditional, which can refer to an "always" > > > time frame or a "currently" time frame without clarity. My honest first > > > take of your CFJ was a conditional always time frame. > > > -- > > > nch > > > > I think I have some linguistics to learn. I think a quick web search > > has taught me what the "present progressive" tense is, but I'm not > > sure I've grokked what a modal is. > > Modals are basically conditionals. It's more complex than that of course but > that's the quick and dirty. They cover possibility of truth or permission of > action. Also see modal logic, which extends classical logic with possible > worlds. > > > I do see that the "if and only if" wording opens the door to another > > interpretation. Interpreting it that way feels a little odd to me, but > > maybe that's because my intended meaning is still stuck firmly in my > > mind. > > Keep in mind that word choice, even where it's synonymous in a vacuum, conveys > meaning. This is a feature of language, we unconsciously assume speakers are > efficient and pragmatic and when expectations are broken it conveys additional > meaning. You used "if and only if" where I would normally use "and". So my > first thought was that you were focusing on something special about the > biconditional. In this case, its implication that the two facts are > meaningfully connected to each other. > > -- > > nch
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Tue, 2 Jun 2020 at 01:57, nch via agora-discussion wrote: > On Monday, June 1, 2020 8:46:09 PM CDT James Cook via agora-discussion wrote: > > > > Isn't that still a difference in intended meaning? Maybe I didn't > > phrase it clearly enough the first time, but my intended meaning was > > "Right now at the moment I'm calling this CFJ, the truth value > > (true/false) of 'Falsifian owns at least one blot' equals the truth > > value of 'English Wikipedia has an article titled "Sponge"'". > > If you had used "right now" or "currently" I'd agree with your reading, see > below. > > > Also, CFJ statements about things like "Alice owns a blot" are usually > > assumed to be about the current situation at the time the statement > > was called. Are you saying the words "if and only if" override that > > default, and lead you interpret my statement as encompassing other > > times and/or situations other than the current one? Or am I > > misunderstanding your argument? > > There's no "override". In "Alice owns a blot" there's no ambiguity about > whether that statement is present progressive. When you introduce a modal, you > also introduce an ambiguity: now the sentence could be present progressive or > it could be conditional, which can refer to an "always" time frame or a > "currently" time frame without clarity. My honest first take of your CFJ was a > conditional always time frame. > > -- > nch I think I have some linguistics to learn. I think a quick web search has taught me what the "present progressive" tense is, but I'm not sure I've grokked what a modal is. I do see that the "if and only if" wording opens the door to another interpretation. Interpreting it that way feels a little odd to me, but maybe that's because my intended meaning is still stuck firmly in my mind. - Falsifian
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Monday, June 1, 2020 8:56:35 PM CDT nch via agora-discussion wrote: > On Monday, June 1, 2020 8:46:09 PM CDT James Cook via agora-discussion wrote: > > Isn't that still a difference in intended meaning? Maybe I didn't > > phrase it clearly enough the first time, but my intended meaning was > > "Right now at the moment I'm calling this CFJ, the truth value > > (true/false) of 'Falsifian owns at least one blot' equals the truth > > value of 'English Wikipedia has an article titled "Sponge"'". > > If you had used "right now" or "currently" I'd agree with your reading, see > below. > > > Also, CFJ statements about things like "Alice owns a blot" are usually > > assumed to be about the current situation at the time the statement > > was called. Are you saying the words "if and only if" override that > > default, and lead you interpret my statement as encompassing other > > times and/or situations other than the current one? Or am I > > misunderstanding your argument? > > There's no "override". In "Alice owns a blot" there's no ambiguity about > whether that statement is present progressive. When you introduce a modal, > you also introduce an ambiguity: now the sentence could be present > progressive or it could be conditional, which can refer to an "always" time > frame or a "currently" time frame without clarity. My honest first take of > your CFJ was a conditional always time frame. > > -- > nch I don't think I'm alone with this reading either. G's arguments seem to implicitly rely on the "always"/"all possibilities" interpretation too. -- nch
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Monday, June 1, 2020 8:46:09 PM CDT James Cook via agora-discussion wrote: > > Isn't that still a difference in intended meaning? Maybe I didn't > phrase it clearly enough the first time, but my intended meaning was > "Right now at the moment I'm calling this CFJ, the truth value > (true/false) of 'Falsifian owns at least one blot' equals the truth > value of 'English Wikipedia has an article titled "Sponge"'". If you had used "right now" or "currently" I'd agree with your reading, see below. > Also, CFJ statements about things like "Alice owns a blot" are usually > assumed to be about the current situation at the time the statement > was called. Are you saying the words "if and only if" override that > default, and lead you interpret my statement as encompassing other > times and/or situations other than the current one? Or am I > misunderstanding your argument? There's no "override". In "Alice owns a blot" there's no ambiguity about whether that statement is present progressive. When you introduce a modal, you also introduce an ambiguity: now the sentence could be present progressive or it could be conditional, which can refer to an "always" time frame or a "currently" time frame without clarity. My honest first take of your CFJ was a conditional always time frame. -- nch
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Tue, 2 Jun 2020 at 01:35, nch via agora-discussion wrote: > On Monday, June 1, 2020 8:23:42 PM CDT James Cook via agora-discussion wrote: > > On Sun, 31 May 2020 at 19:29, nch via agora-business > > > > wrote: > > > On Sunday, May 31, 2020 2:06:51 PM CDT Kerim Aydin via agora-official > > > wrote: > > > > The below CFJ is 3837. I assign it to grok. > > > > > > > > status: https://faculty.washington.edu/kerim/nomic/cases/#3837 > > > > > > > > === CFJ 3837 > > > > === > > > > > > > > Falsifian owns at least one blot if and only if English Wikipedia > > > > has an article titled "Sponge". > > > > > > > > > > > > == > > > > > > Gratuitous: This CFJ should be found FALSE because the rules do not define > > > a biconditional relationship between these facts, regardless of whether > > > either individual fact is TRUE or FALSE. > > > > > > -- > > > nch > > > > Gratuitous response: > > > > When I published the statement, I intended "if and only if" to have > > the classical logic meaning, i.e. (I own at least one blot and English > > Wikipedia has an article titled "Sponge") or (I do not own at least > > one blot and English Wikipedia does not have an article titled > > "Sponge"). > > > > I suppose it could be interpreted differently. However, I think my > > intent is important here, since interpreting natural language is > > fundamentally an act of figuring out what someone was trying to > > communicate. I don't know whether there are past judgements on the > > subject of whether intent matters in a CFJ statement. > > > > - Falsifian > > We agree on what the biconditional means, I think. The difference between my > argument and your argument is at the pragmatics level not at the syntactic > level. > > Your argument is a de re interpretation: If one of these two sides is > currently true, the whole statement is true. > > My argument is a de dicto interpretation: If this statement is true under all > conditions for either side, then the whole statement is true. That's why my > interpretation thinks a iff relationship needs to be pre-established. > > -- > nch Isn't that still a difference in intended meaning? Maybe I didn't phrase it clearly enough the first time, but my intended meaning was "Right now at the moment I'm calling this CFJ, the truth value (true/false) of 'Falsifian owns at least one blot' equals the truth value of 'English Wikipedia has an article titled "Sponge"'". Also, CFJ statements about things like "Alice owns a blot" are usually assumed to be about the current situation at the time the statement was called. Are you saying the words "if and only if" override that default, and lead you interpret my statement as encompassing other times and/or situations other than the current one? Or am I misunderstanding your argument? - Falsifian
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On 6/1/2020 6:23 PM, James Cook via agora-discussion wrote: > On Sun, 31 May 2020 at 19:29, nch via agora-business > wrote: >> On Sunday, May 31, 2020 2:06:51 PM CDT Kerim Aydin via agora-official wrote: >>> The below CFJ is 3837. I assign it to grok. >>> >>> status: https://faculty.washington.edu/kerim/nomic/cases/#3837 >>> >>> === CFJ 3837 === >>> >>> Falsifian owns at least one blot if and only if English Wikipedia >>> has an article titled "Sponge". >>> >>> == >> >> Gratuitous: This CFJ should be found FALSE because the rules do not define a >> biconditional relationship between these facts, regardless of whether either >> individual fact is TRUE or FALSE. >> >> -- >> nch > > Gratuitous response: > > When I published the statement, I intended "if and only if" to have > the classical logic meaning, i.e. (I own at least one blot and English > Wikipedia has an article titled "Sponge") or (I do not own at least > one blot and English Wikipedia does not have an article titled > "Sponge"). > > I suppose it could be interpreted differently. However, I think my > intent is important here, since interpreting natural language is > fundamentally an act of figuring out what someone was trying to > communicate. I don't know whether there are past judgements on the > subject of whether intent matters in a CFJ statement. It's ok to use intent in the arguments, but not so much in the actual act of judgement. In other words, it's ok to say "because the statement was worded weirdly, it's technically FALSE. So I judge FALSE. But the question the caller was obviously trying to get at is TRUE." Then, future judges respect it as a precedent of TRUE for the intended question. But FALSE is still the official judgement. Because what makes a judgement "appropriate" as per R591 is its specific relationship to the statement, not the intended statement. -G.
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Monday, June 1, 2020 8:23:42 PM CDT James Cook via agora-discussion wrote: > On Sun, 31 May 2020 at 19:29, nch via agora-business > > wrote: > > On Sunday, May 31, 2020 2:06:51 PM CDT Kerim Aydin via agora-official wrote: > > > The below CFJ is 3837. I assign it to grok. > > > > > > status: https://faculty.washington.edu/kerim/nomic/cases/#3837 > > > > > > === CFJ 3837 > > > === > > > > > > Falsifian owns at least one blot if and only if English Wikipedia > > > has an article titled "Sponge". > > > > > > > > > == > > > > Gratuitous: This CFJ should be found FALSE because the rules do not define > > a biconditional relationship between these facts, regardless of whether > > either individual fact is TRUE or FALSE. > > > > -- > > nch > > Gratuitous response: > > When I published the statement, I intended "if and only if" to have > the classical logic meaning, i.e. (I own at least one blot and English > Wikipedia has an article titled "Sponge") or (I do not own at least > one blot and English Wikipedia does not have an article titled > "Sponge"). > > I suppose it could be interpreted differently. However, I think my > intent is important here, since interpreting natural language is > fundamentally an act of figuring out what someone was trying to > communicate. I don't know whether there are past judgements on the > subject of whether intent matters in a CFJ statement. > > - Falsifian We agree on what the biconditional means, I think. The difference between my argument and your argument is at the pragmatics level not at the syntactic level. Your argument is a de re interpretation: If one of these two sides is currently true, the whole statement is true. My argument is a de dicto interpretation: If this statement is true under all conditions for either side, then the whole statement is true. That's why my interpretation thinks a iff relationship needs to be pre-established. -- nch
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On Sun, May 31, 2020, 2:52 PM Kerim Aydin via agora-discussion < agora-discussion@agoranomic.org> wrote: > > > On 5/31/2020 12:39 PM, grok wrote: > > On Sun, May 31, 2020, 2:35 PM Kerim Aydin wrote: > >> On 5/31/2020 12:29 PM, nch via agora-business wrote: > >>> On Sunday, May 31, 2020 2:06:51 PM CDT Kerim Aydin wrote: > The below CFJ is 3837. I assign it to grok. > > status: https://faculty.washington.edu/kerim/nomic/cases/#3837 > > === CFJ 3837 > >> === > > Falsifian owns at least one blot if and only if English > Wikipedia > has an article titled "Sponge". > > > >> > == > >>> > >>> Gratuitous: This CFJ should be found FALSE because the rules do not > >> define a > >>> biconditional relationship between these facts, regardless of whether > >> either > >>> individual fact is TRUE or FALSE. > >>> > >> > >> Gratuitous: A judgement of IRRELEVANT is also appropriate - to evaluate > >> this, we are required to consider a world in which a common subject like > >> "sponge" is not in Wikipedia. A world like this might be strange in > other > >> ways. This is, literally and directly, an "overly hypothetical > >> extrapolation of the game or its rules to conditions that don't actually > >> exist" as defined for IRRELEVANT in R591. > >> > > > > This CFJ will (unintentionally, I believe) test the decision options in > the > > CFJ system. FALSE/DISMISS/IRRELEVANT all have different implications on > the > > gamestate long term. > > > > So for context, it wasn't unintentional. Falsifian was purposefully seeing > if e could "entangle" an otherwise irrelevant fact into being relevant. I > can imagine reasonably cogent arguments for any of those three options, so > it's definitely within the judge's purview to opine on whether the > implications of one of them in particular is "better" for the long-term. > Right, the objective to entangle irrelevant facts is a test of the CFJ system, but I think even though any of those options forms a coherent decision they all have different implications on the continuing interpretation of the ruleset >
Re: DIS: Re: BUS: Re: OFF: [Arbitor] CFJ 3837 Assigned to grok
On 5/31/2020 12:39 PM, grok wrote: > On Sun, May 31, 2020, 2:35 PM Kerim Aydin wrote: >> On 5/31/2020 12:29 PM, nch via agora-business wrote: >>> On Sunday, May 31, 2020 2:06:51 PM CDT Kerim Aydin wrote: The below CFJ is 3837. I assign it to grok. status: https://faculty.washington.edu/kerim/nomic/cases/#3837 === CFJ 3837 >> === Falsifian owns at least one blot if and only if English Wikipedia has an article titled "Sponge". >> == >>> >>> Gratuitous: This CFJ should be found FALSE because the rules do not >> define a >>> biconditional relationship between these facts, regardless of whether >> either >>> individual fact is TRUE or FALSE. >>> >> >> Gratuitous: A judgement of IRRELEVANT is also appropriate - to evaluate >> this, we are required to consider a world in which a common subject like >> "sponge" is not in Wikipedia. A world like this might be strange in other >> ways. This is, literally and directly, an "overly hypothetical >> extrapolation of the game or its rules to conditions that don't actually >> exist" as defined for IRRELEVANT in R591. >> > > This CFJ will (unintentionally, I believe) test the decision options in the > CFJ system. FALSE/DISMISS/IRRELEVANT all have different implications on the > gamestate long term. > So for context, it wasn't unintentional. Falsifian was purposefully seeing if e could "entangle" an otherwise irrelevant fact into being relevant. I can imagine reasonably cogent arguments for any of those three options, so it's definitely within the judge's purview to opine on whether the implications of one of them in particular is "better" for the long-term. Here's the context: Falsifian wrote: > ais 523 wrote: >> I think the best direction in this regard would be to allow CFJs that >> are not relevant to Agora directly, with some payment to compensate >> the judge for their time. "Agora as a ruleset interpretation service", >> if you like. So Agora would act entirely in a fact-finding role, not in >> any sort of enforcement role. (The person who commissioned Agora to >> come to a judgement could then do what they wanted with the resulting >> judgement and its reasoning.) > > > I suspect it's already possible to use Agora's CFJ system for > questions not directly relevant to Agora, by entangling the statement > with a relevant question. > > > CFJ: Falsifian owns at least one blot if and only if English Wikipedia > has an article titled "Sponge".
