Another angle on this is that it isn’t about humans in general, it’s about some of the most gullible humans.
That aspect of things became salient to me when I read a man complaining about the women who get involved with prisoners—even those who have been convicted for murder. It’s so unfair when some ordinary guys get no female attention at all. I suddenly realized that he was talking about a very small proportion of women. The vast majority of women aren’t chasing prisoners.
Tolkien’s Sindarin (and, IIRC, Welsh) have two plurals, one for more than one and another for everything in a class. I am very envious. My ideal language would have a bunch of plurals—the two from Sindarin, plus “a noticeable minority” and “the vast majority”. I might even split “everything in a class” into “all we have seen” and “all by nature”.
Another angle on this is that it isn’t about humans in general, it’s about some of the most gullible humans.
This seems to imply that there is something fundamentally different about these humans compared to other humans. I’m not convinced this is the case. I would be rather surprised if you couldn’t make the average human drink bleach by exposing them to specifically tailored situations/information.
That’s not just arguing from fictional evidence, it’s arguing from hypothetical evidence. I don’t know, but I’ve been told that you can get a room full of Jews shouting “Sieg Heil” with a sufficiently rousing version of this. [1] You could probably get people to drink bleach by mislabeling a bottle.
Getting back to the general point, it’s not just important to know that an aspect exists, it’s important to know how strong it is.
[1] I was told it either by or about Robert Aspirin/Yang the Nauseating. He was quite a good performer.
I think this is a rather uncharitable interpretation of my argument. There is a difference between how strong an effect is and how common it is. If you had said “Another angle on this is that it isn’t about humans in general, it’s about [a small subset of humans that happened to experience an unfortunate convergence of misinformation and subjective context]” then we would have no disagreement.
I’m not contending that these situations are representative of the human population, but rather that these situations do not require “some of the most gullible humans” to occur.
The thing both of us are leaving out is that deciding what can be trusted is a genuinely hard problem. You can get badly hurt trusting conventional medical advice, too.
Tolkien’s Sindarin (and, IIRC, Welsh) have two plurals, one for more than one and another for everything in a class. I am very envious. My ideal language would have a bunch of plurals—the two from Sindarin, plus “a noticeable minority” and “the vast majority”. I might even split “everything in a class” into “all we have seen” and “all by nature”.
Why do you need grammatical inflection when you can just use quantifiers (“several sheep”, “all sheep”, “many sheep”, “most sheep”)?
Because having rationalist features built into the language means that it’s harder to slip in irrational premises. I don’t know whether there’s be research on whether native speakers of languages with evidentials think more clearly about the sources of their information.
Why the same thing would count as “built into the language” if expressed in one word but not if expressed in two? After all, spelling conventions for separating words are partly arbitrary—a language might have stuff like “severalsheep”, “allsheep”, etc.
It’s a question of what’s obligatory. Admittedly, this doesn’t have to be handled by what’s a single word. In English, you have to introduce nouns with ‘a’ or ‘the’. I’m not sure how valuable this is—many, perhaps most, languages don’t have that feature, but in English, you’re stuck with indicating whether something you’re talking about is especially important.
In the same spirit, you have to make an effort to avoid indicating a person’s gender.
In English, you can say “men do x” or “women do y” without a built in obligation to indicate or to notice to yourself whether you mean all, all that you’ve noticed, all inevitably, most, or some. I think not having a requirement to be clear about such things leads to a lot of stereotyping and pontificating.
In English, you can say “men do x” or “women do y” without a built in obligation to indicate or to notice to yourself whether you mean all, all that you’ve noticed, all inevitably, most, or some. I think not having a requirement to be clear about such things leads to a lot of stereotyping and pontificating.
I don’t think it’s just a matter of language. In Italian it’s extremely rare to use a noun without an article as the subject of a sentence—you’d use the definite article (lit. ‘the men’) if you mean something like ‘typical men’ (as in ‘men have opposable thumbs’—male amputees do exist but are irrelevant to the point being made) and a ‘partitive article’ (or an indefinite pronoun such as ‘someone’, rewording the sentence such as ‘there are men who’, etc.) when you mean ‘certain men’—and yet people use the former all the time even when they have very little evidence that something applies to an entire reference class except a few irrelevant exceptions.
In the same spirit, you have to make an effort to avoid indicating a person’s gender.
English is a lot better in that respect than most languages in the Indo-European family (and Hebrew, just because I happen to know it). Many languages have mandatory gender for all nouns. Not just “he” vs. “she”, but “he-chair” vs “she-sun”. In those languages you can’t talk about something without knowing its gender—masculine, feminine or neuter. You can’t address a person in Hebrew without knowing their gender—not even to ask for their name. (Think how fun that makes replying to email.)
My two mother tongues are Russian and Hebrew, but I read and write more in English. Every time I need to write something in Hebrew, I become angry at how gender forces its way to my attention.
ETA: I’m sure you know this already, but it was worth saying clearly.
I am constructing an artificial language (just for fun; also it isn’t going to be finished any soon) with three grammatical numbers: singular, plural (can be any number, zero and one included) and class plural for everything in the class. I have invented this independently of Tolkien not for practical reasons, but just for sake of elegance, as it better suits my desire to have the grammar work in part like set algebra.
Having grammar rules that enforce expressing the number more precisely than we are used to (e.g. different grammatical numbers for one vs. two vs. more than two but few vs. noticeable minority vs. majority vs. all) has disadvantages when the speaker doesn’t know the number or just wishes to describe a general situation where more than one number is applicable. Similar problems we face when our grammars enforce gender expression when using some pronouns (if we aren’t Finns or Turks or Chinese or...). People try to find a way around, either by using more complex expressions (“he or she”) or by attempts to update the grammar / vocabulary (“ey”). I am not sure whether I’d wish to have the same problem with grammatical number.
As for Welsh, a peculiarity with plurals I know about is that some nouns, usually denoting animals, have the basic (shorter) form denoting plural and the singular is derived by a suffix (e.g. adar/aderyn for birds/bird, pysgod/pysgodyn for fish, plant/plentyn for children/child—children are apparently sort of animals too). I haven’t heard/read about different types of plural though.
Another angle on this is that it isn’t about humans in general, it’s about some of the most gullible humans.
That aspect of things became salient to me when I read a man complaining about the women who get involved with prisoners—even those who have been convicted for murder. It’s so unfair when some ordinary guys get no female attention at all. I suddenly realized that he was talking about a very small proportion of women. The vast majority of women aren’t chasing prisoners.
Tolkien’s Sindarin (and, IIRC, Welsh) have two plurals, one for more than one and another for everything in a class. I am very envious. My ideal language would have a bunch of plurals—the two from Sindarin, plus “a noticeable minority” and “the vast majority”. I might even split “everything in a class” into “all we have seen” and “all by nature”.
This seems to imply that there is something fundamentally different about these humans compared to other humans. I’m not convinced this is the case. I would be rather surprised if you couldn’t make the average human drink bleach by exposing them to specifically tailored situations/information.
That’s not just arguing from fictional evidence, it’s arguing from hypothetical evidence. I don’t know, but I’ve been told that you can get a room full of Jews shouting “Sieg Heil” with a sufficiently rousing version of this. [1] You could probably get people to drink bleach by mislabeling a bottle.
Getting back to the general point, it’s not just important to know that an aspect exists, it’s important to know how strong it is.
[1] I was told it either by or about Robert Aspirin/Yang the Nauseating. He was quite a good performer.
I think this is a rather uncharitable interpretation of my argument. There is a difference between how strong an effect is and how common it is. If you had said “Another angle on this is that it isn’t about humans in general, it’s about [a small subset of humans that happened to experience an unfortunate convergence of misinformation and subjective context]” then we would have no disagreement.
I’m not contending that these situations are representative of the human population, but rather that these situations do not require “some of the most gullible humans” to occur.
You could be right.
The thing both of us are leaving out is that deciding what can be trusted is a genuinely hard problem. You can get badly hurt trusting conventional medical advice, too.
Why do you need grammatical inflection when you can just use quantifiers (“several sheep”, “all sheep”, “many sheep”, “most sheep”)?
Because having rationalist features built into the language means that it’s harder to slip in irrational premises. I don’t know whether there’s be research on whether native speakers of languages with evidentials think more clearly about the sources of their information.
Why the same thing would count as “built into the language” if expressed in one word but not if expressed in two? After all, spelling conventions for separating words are partly arbitrary—a language might have stuff like “severalsheep”, “allsheep”, etc.
It’s a question of what’s obligatory. Admittedly, this doesn’t have to be handled by what’s a single word. In English, you have to introduce nouns with ‘a’ or ‘the’. I’m not sure how valuable this is—many, perhaps most, languages don’t have that feature, but in English, you’re stuck with indicating whether something you’re talking about is especially important.
In the same spirit, you have to make an effort to avoid indicating a person’s gender.
In English, you can say “men do x” or “women do y” without a built in obligation to indicate or to notice to yourself whether you mean all, all that you’ve noticed, all inevitably, most, or some. I think not having a requirement to be clear about such things leads to a lot of stereotyping and pontificating.
I don’t think it’s just a matter of language. In Italian it’s extremely rare to use a noun without an article as the subject of a sentence—you’d use the definite article (lit. ‘the men’) if you mean something like ‘typical men’ (as in ‘men have opposable thumbs’—male amputees do exist but are irrelevant to the point being made) and a ‘partitive article’ (or an indefinite pronoun such as ‘someone’, rewording the sentence such as ‘there are men who’, etc.) when you mean ‘certain men’—and yet people use the former all the time even when they have very little evidence that something applies to an entire reference class except a few irrelevant exceptions.
Thanks. That’s a good example of mental defaults pulling in one direction even though the language is pulling in the opposite direction.
English is a lot better in that respect than most languages in the Indo-European family (and Hebrew, just because I happen to know it). Many languages have mandatory gender for all nouns. Not just “he” vs. “she”, but “he-chair” vs “she-sun”. In those languages you can’t talk about something without knowing its gender—masculine, feminine or neuter. You can’t address a person in Hebrew without knowing their gender—not even to ask for their name. (Think how fun that makes replying to email.)
My two mother tongues are Russian and Hebrew, but I read and write more in English. Every time I need to write something in Hebrew, I become angry at how gender forces its way to my attention.
ETA: I’m sure you know this already, but it was worth saying clearly.
BTW, see this.
If it’s built into the language then, presumably, you can’t not use it.
That’s the main difference, I think.
I am constructing an artificial language (just for fun; also it isn’t going to be finished any soon) with three grammatical numbers: singular, plural (can be any number, zero and one included) and class plural for everything in the class. I have invented this independently of Tolkien not for practical reasons, but just for sake of elegance, as it better suits my desire to have the grammar work in part like set algebra.
Having grammar rules that enforce expressing the number more precisely than we are used to (e.g. different grammatical numbers for one vs. two vs. more than two but few vs. noticeable minority vs. majority vs. all) has disadvantages when the speaker doesn’t know the number or just wishes to describe a general situation where more than one number is applicable. Similar problems we face when our grammars enforce gender expression when using some pronouns (if we aren’t Finns or Turks or Chinese or...). People try to find a way around, either by using more complex expressions (“he or she”) or by attempts to update the grammar / vocabulary (“ey”). I am not sure whether I’d wish to have the same problem with grammatical number.
As for Welsh, a peculiarity with plurals I know about is that some nouns, usually denoting animals, have the basic (shorter) form denoting plural and the singular is derived by a suffix (e.g. adar/aderyn for birds/bird, pysgod/pysgodyn for fish, plant/plentyn for children/child—children are apparently sort of animals too). I haven’t heard/read about different types of plural though.