4 ms·

> Having fewer keywords does make Go simpler to learn and simpler to read other peoples code. that sounds like saying that latin-alphabet languages are easier

> Having fewer keywords does make Go simpler to learn and simpler to read other peoples code. that sounds like saying that latin-alphabet languages are easier to read than chinese or japanese because they have fewer characters. the complexity just goes in the way those characters are assembled together.

hnlmorg 6y ago
Funny you should use that as an example because one of my lockdown projects has been to learn to read simplified Chinese and I'd say your point about Latin-alphabet languages was also true despite it being intended as a falsehood. Learning all the different Hanzi is immensely harder because of the number of characters and how intricate they often are. I'd also say that last point means it's also easier to write Latin characters too.
- jcelerier 6y ago
  > Funny you should use that as an example because one of my lockdown projects has been to learn to read simplified Chinese and I'd say your point about Latin-alphabet languages was also true despite it being intended as a falsehood. but that's because you come from a latin background. I lived with a japanese girl for a long time and she was adamant that kanji were much easier and faster to parse for her than western text
  - hnlmorg 6y ago
    > but that's because you come from a latin background. I lived with a japanese girl for a long time and she was adamant that kanji were much easier and faster to parse for her than western text You could make your same argument here and say "that's because she comes from a Japanese background" ;)
    - jcelerier 6y ago
      but that is precisely my original argument. in the end everyone finds "simple" what they're accustomed to, and you end up with more-or-less the same amount of complexity overall, no matter the system used to express logical thoughts, if you let humans try to optimize it over decades (but systematically, for programming languages) or generations (but less systematically, for spoken languages). I'm mostly programming in C++ and I find understanding what happens in some Python scripts much harder than some template metaprograms I've worked on for instance.
      - hnlmorg 6y ago
        You're conflating different things now: simplicity vs familiarity. If someone is already a C++ guru then writing something in C++ for them will be easier than learning Go. But that doesn't mean that Go is easier to learn than C++. When talking about simplicity people generally compare all languages as if one has little or no familiarity with them. Otherwise it's not a fair comparison.
        jcelerier 6y ago
        > You're conflating different things now: simplicity vs familiarity. we're talking about humans, you cannot not conflate them. > When talking about simplicity people generally compare all languages as if one has little or no familiarity with them. that's a philosophically pointless exercice. Things have to be assessed in their actual, real-world context to bear any relevance !
        hnlmorg 6y ago
        ...and you asses something by comparison. Having different starting points isn't a comparison, it's a bias. I don't have an issue with people having a bias (we all have our own preferences, our own strengths and weaknesses, and stuff we just enjoy more than other stuff). But lets not pretend it's anything more than a bias if you're not willing to compare things from an even starting point.
      - flippinburgers 6y ago
  - blacksmith_tb 6y ago
    I took a year of Mandarin as an undergrad, and learned ~1,500 characters (at my best). The two approaches are just radically different (pun intended), with an alphabetic writing system (not just Latin, but Arabic, Cyrillic, etc.) if you know the system, you can read texts in other languages that share that system aloud - without knowing what you're saying. With logographic systems like Hanzi, you may be able to guess at the meaning of a character you don't know by looking at its radical, but you have no idea how to pronounce it. I do think broadly speaking alphabets are easier to teach and to learn, though still not easy (ask any small child).
  - turtlebits 6y ago
    Part of the hard part in learning Chinese and Japanese is that there is a separate phoenetic "alphabet" you have to learn to properly pronounce characters. (at least in Taiwan, kids learn bopofomo)
    - fomine3 6y ago
      Native Japanese never learn alphabets for pronounce.
  - alisonatwork 6y ago
    I can see some kind of argument that logograms are easier to parse than an alphabet because you can communicate more information in a smaller space, but that's only true if you already committed the thousands (!) of logograms to memory in the first place. This is no trivial task! The question is if it's worth putting a 5+ year barrier in front of people in order to become fluent in a written language that maybe, possibly, might be "faster" once you are fluent. For native speakers, perhaps that's worth the effort (although there are plenty who argue otherwise). For second-language speakers, I think the answer is almost certainly no. So bringing it back to Go. Perhaps Go seems more difficult and slow to people who already invested the years required to become fluent with something like C++. I doubt that, but I guess it's possible. But Go isn't designed for that case. It's designed as a second language that any programmer can learn quickly and become reasonably productive in. If you have a company with 1000 programmers who all know different languages, and you want to start creating good software right now, would you standardize on the language that all 1000 programmers can learn in a couple of weeks, or the language that 100 programmers know inside out and the other 900 are going to spend a year figuring out?
- CobrastanJorji 6y ago
  Can I derail this to ask how you've been going about your project? I had the same lockdown project but with Japanese. It's been going well. I've found a little company called Wanikani that does a great implementation of a spaced repetition system. I've never rigorously used one before, but based on my progress, I'm now a believer in the technique and am a bit mad at myself for not doing this for school subjects as a kid.
gilbetron 6y ago
It would be interesting to find someone that grew up knowing a third type of language (no idea what that is, though!) and that then learned latin alphabet and chinese (say) as an adult and hear what they say :)
- cratermoon 6y ago
  Arabic, Hebrew, or one of the Indian subcontinent languages.
  - autarch 6y ago
    Those all use alphabets or syllabaries, so it's just a matter of learning a small new set of symbols.
    - cratermoon 6y ago
      Yes and no. There are complex rules around joining letters depending on where they fall in the words. Also there are no vowels but sometimes there are vowel markers.
      - autarch 6y ago
        I meant that if you know Hebrew, Arabic, or Hindu, learning the Latin alphabet is a lot easier than learning Chinese characters. And Latin user will probably find Arabic easier than Chinese too, even though the Latin alphabet is a bit simpler than Arabic.
        oblio 6y ago
        I guess it depends on which Latin alphabet variation. Vietnamese or Polish speakers will probably not feel intimidated by Arabic :-p
        cratermoon 6y ago
        It also depends a lot on the language. There are dozens of languages that use the same core Latin characters plus a few decorators. Someone coming from a strongly inflected, gendered language trying to learn English is going to struggle compared to trying to learn Spanish or French.
derefr 6y ago
The argument may be wrong, but the conclusion is true: literacy rates are actually lower in countries with ideographic languages — likely because there are more properties of words in ideographic languages that need to be memorized "literally", and fewer properties that can be intuited, or easily "sketched" verbally in order to ask the real spelling/pronunciation/etc. One study I recall (on my phone now, will cite later) reported that fewer 20-year-olds have "fluent" levels of reading comprehension for what they read in the [ideographic] newspaper in Japan or China, than in neighbouring countries with alphabetic languages derived from those same historical ideographic roots (Korea, Vietnam, Thailand.) Another study (will cite later) reported that the reading-comprehension rate for ideograph-heavy works in Japan is also decreasing over time, as more people spend their time reading mostly explicit "[ideograph-]light" novels, rather than literature that would actually teach them kanji. (Which is a vicious cycle.) Given that ideograph radicals are kind of analogous to alphabetic morphemes, this is kind of surprising. I think it's mostly because it's self-evident how to pronounce alphabetic morphemes if you know the pronunciation rules of the language, while 1. the names for ideographic radicals need to be memorized, more like the names of alphabetic letters (but with there usually being far more radicals in common use per ideographic language, than there are letters in an average alphabet.) And 2. ideograms aren't usually linear, but rather structured (things above/below/beside things, things inside things, things bonded to other things and reduced in the process, etc); and there's no universal cultural fluency for speakers of these languages in an ordering system for spelling out these radicals. You usually have to see a character written before you can truly know how to "spell out" its radicals. Because of these two problems, you can't just look at an ideograph you don't know, and casually, conversationally spell it out to someone else, in order to ask how it's pronounced / whether you're pronouncing it correctly. Unless, that is, you both know the names of all the involved radicals, and have a good geometric guess for the ordering. In practice, most people don't try to ask this question verbally, but either need to show someone the character written; or need to use a dictionary. I've noticed that this actually results in distinct approaches to how people treat "literature with words they don't know", between people whose first language is an alphabetic one vs. people whose first language is an alphabetic one. Given an alphabetic-language work with word they don't know, alphabetic-native people seem to be much more confident in just jumping in and looking words up as they go; while ideographic-native people seem to be more reticent. I believe they're subconsciously expecting new [alphabetic] words to be a much larger barrier for them than they actually would be, as they're used to thinking in terms of the barrier that novel ideographs impose.

hnlmorg 6y ago

Funny you should use that as an example because one of my lockdown projects has been to learn to read simplified Chinese and I'd say your point about Latin-alphabet languages was also true despite it being intended as a falsehood. Learning all the different Hanzi is immensely harder because of the number of characters and how intricate they often are. I'd also say that last point means it's also easier to write Latin characters too.

jcelerier 6y ago
> Funny you should use that as an example because one of my lockdown projects has been to learn to read simplified Chinese and I'd say your point about Latin-alphabet languages was also true despite it being intended as a falsehood. but that's because you come from a latin background. I lived with a japanese girl for a long time and she was adamant that kanji were much easier and faster to parse for her than western text
- hnlmorg 6y ago
  > but that's because you come from a latin background. I lived with a japanese girl for a long time and she was adamant that kanji were much easier and faster to parse for her than western text You could make your same argument here and say "that's because she comes from a Japanese background" ;)
  - jcelerier 6y ago
    but that is precisely my original argument. in the end everyone finds "simple" what they're accustomed to, and you end up with more-or-less the same amount of complexity overall, no matter the system used to express logical thoughts, if you let humans try to optimize it over decades (but systematically, for programming languages) or generations (but less systematically, for spoken languages). I'm mostly programming in C++ and I find understanding what happens in some Python scripts much harder than some template metaprograms I've worked on for instance.
    - hnlmorg 6y ago
      You're conflating different things now: simplicity vs familiarity. If someone is already a C++ guru then writing something in C++ for them will be easier than learning Go. But that doesn't mean that Go is easier to learn than C++. When talking about simplicity people generally compare all languages as if one has little or no familiarity with them. Otherwise it's not a fair comparison.
      - jcelerier 6y ago
        > You're conflating different things now: simplicity vs familiarity. we're talking about humans, you cannot not conflate them. > When talking about simplicity people generally compare all languages as if one has little or no familiarity with them. that's a philosophically pointless exercice. Things have to be assessed in their actual, real-world context to bear any relevance !
        hnlmorg 6y ago
        ...and you asses something by comparison. Having different starting points isn't a comparison, it's a bias. I don't have an issue with people having a bias (we all have our own preferences, our own strengths and weaknesses, and stuff we just enjoy more than other stuff). But lets not pretend it's anything more than a bias if you're not willing to compare things from an even starting point.
    - flippinburgers 6y ago
- blacksmith_tb 6y ago
  I took a year of Mandarin as an undergrad, and learned ~1,500 characters (at my best). The two approaches are just radically different (pun intended), with an alphabetic writing system (not just Latin, but Arabic, Cyrillic, etc.) if you know the system, you can read texts in other languages that share that system aloud - without knowing what you're saying. With logographic systems like Hanzi, you may be able to guess at the meaning of a character you don't know by looking at its radical, but you have no idea how to pronounce it. I do think broadly speaking alphabets are easier to teach and to learn, though still not easy (ask any small child).
- turtlebits 6y ago
  Part of the hard part in learning Chinese and Japanese is that there is a separate phoenetic "alphabet" you have to learn to properly pronounce characters. (at least in Taiwan, kids learn bopofomo)
  - fomine3 6y ago
    Native Japanese never learn alphabets for pronounce.
- alisonatwork 6y ago
  I can see some kind of argument that logograms are easier to parse than an alphabet because you can communicate more information in a smaller space, but that's only true if you already committed the thousands (!) of logograms to memory in the first place. This is no trivial task! The question is if it's worth putting a 5+ year barrier in front of people in order to become fluent in a written language that maybe, possibly, might be "faster" once you are fluent. For native speakers, perhaps that's worth the effort (although there are plenty who argue otherwise). For second-language speakers, I think the answer is almost certainly no. So bringing it back to Go. Perhaps Go seems more difficult and slow to people who already invested the years required to become fluent with something like C++. I doubt that, but I guess it's possible. But Go isn't designed for that case. It's designed as a second language that any programmer can learn quickly and become reasonably productive in. If you have a company with 1000 programmers who all know different languages, and you want to start creating good software right now, would you standardize on the language that all 1000 programmers can learn in a couple of weeks, or the language that 100 programmers know inside out and the other 900 are going to spend a year figuring out?
CobrastanJorji 6y ago
Can I derail this to ask how you've been going about your project? I had the same lockdown project but with Japanese. It's been going well. I've found a little company called Wanikani that does a great implementation of a spaced repetition system. I've never rigorously used one before, but based on my progress, I'm now a believer in the technique and am a bit mad at myself for not doing this for school subjects as a kid.

gilbetron 6y ago

It would be interesting to find someone that grew up knowing a third type of language (no idea what that is, though!) and that then learned latin alphabet and chinese (say) as an adult and hear what they say :)

cratermoon 6y ago
Arabic, Hebrew, or one of the Indian subcontinent languages.
- autarch 6y ago
  Those all use alphabets or syllabaries, so it's just a matter of learning a small new set of symbols.
  - cratermoon 6y ago
    Yes and no. There are complex rules around joining letters depending on where they fall in the words. Also there are no vowels but sometimes there are vowel markers.
    - autarch 6y ago
      I meant that if you know Hebrew, Arabic, or Hindu, learning the Latin alphabet is a lot easier than learning Chinese characters. And Latin user will probably find Arabic easier than Chinese too, even though the Latin alphabet is a bit simpler than Arabic.
      - oblio 6y ago
        I guess it depends on which Latin alphabet variation. Vietnamese or Polish speakers will probably not feel intimidated by Arabic :-p
      - cratermoon 6y ago
        It also depends a lot on the language. There are dozens of languages that use the same core Latin characters plus a few decorators. Someone coming from a strongly inflected, gendered language trying to learn English is going to struggle compared to trying to learn Spanish or French.

derefr 6y ago

The argument may be wrong, but the conclusion is true: literacy rates are actually lower in countries with ideographic languages — likely because there are more properties of words in ideographic languages that need to be memorized "literally", and fewer properties that can be intuited, or easily "sketched" verbally in order to ask the real spelling/pronunciation/etc. One study I recall (on my phone now, will cite later) reported that fewer 20-year-olds have "fluent" levels of reading comprehension for what they read in the [ideographic] newspaper in Japan or China, than in neighbouring countries with alphabetic languages derived from those same historical ideographic roots (Korea, Vietnam, Thailand.) Another study (will cite later) reported that the reading-comprehension rate for ideograph-heavy works in Japan is also decreasing over time, as more people spend their time reading mostly explicit "[ideograph-]light" novels, rather than literature that would actually teach them kanji. (Which is a vicious cycle.) Given that ideograph radicals are kind of analogous to alphabetic morphemes, this is kind of surprising. I think it's mostly because it's self-evident how to pronounce alphabetic morphemes if you know the pronunciation rules of the language, while 1. the names for ideographic radicals need to be memorized, more like the names of alphabetic letters (but with there usually being far more radicals in common use per ideographic language, than there are letters in an average alphabet.) And 2. ideograms aren't usually linear, but rather structured (things above/below/beside things, things inside things, things bonded to other things and reduced in the process, etc); and there's no universal cultural fluency for speakers of these languages in an ordering system for spelling out these radicals. You usually have to see a character written before you can truly know how to "spell out" its radicals. Because of these two problems, you can't just look at an ideograph you don't know, and casually, conversationally spell it out to someone else, in order to ask how it's pronounced / whether you're pronouncing it correctly. Unless, that is, you both know the names of all the involved radicals, and have a good geometric guess for the ordering. In practice, most people don't try to ask this question verbally, but either need to show someone the character written; or need to use a dictionary. I've noticed that this actually results in distinct approaches to how people treat "literature with words they don't know", between people whose first language is an alphabetic one vs. people whose first language is an alphabetic one. Given an alphabetic-language work with word they don't know, alphabetic-native people seem to be much more confident in just jumping in and looking words up as they go; while ideographic-native people seem to be more reticent. I believe they're subconsciously expecting new [alphabetic] words to be a much larger barrier for them than they actually would be, as they're used to thinking in terms of the barrier that novel ideographs impose.