Slashdot Mirror


Computer Program Learns Baby Talk in Any Language

athloi writes "Researchers have made a computer program that learns to decode sounds from different languages in the same way that a baby does. The program will help to shed new light on how people learn to talk. It has already raised questions as to how much specific information about language is hard-wired into the brain."

9 of 170 comments (clear)

  1. Re:Baby talk? I swear at my computer! by ShieldW0lf · · Score: 4, Insightful

    A computer learns something that a baby can learn, and this supports the extension that it is "Learning like a baby does"?!? What a load of crap.

    And what about this "hard wired vs soft wired" stuff? What is this supposed to prove? If I build a virtual machine, does this "prove" that the machine was made of software?

    Researchers examined the hardware of a babys brain, mimic it, and argue that it proves the baby learning language is in software.

    None of which is to say that I think language is hardwired, but this is such ridiculous logic it makes me feel stupider for having read it.

    --
    -1 Uncomfortable Truth
  2. Two speed bumps by DynaSoar · · Score: 2, Insightful

    > A computer program that learns to decode sounds from different languages ... is not the same as learning "talk". Talk is to sounds as molecules are to atoms. You can't predict the behavior of the former just from knowing the individual behaviors of the latter.

    > in the same way that a baby does

    McClelland's program only models it. The map is not the terrain. I haven't read his PNAS paper, but I'm definitely going to. I doubt it makes the kind of claims Reuters does.

    --
    "I may be synthetic, but I'm not stupid." -- Bishop 341-B
  3. Re:Baby talk? I swear at my computer! by Vegeta99 · · Score: 2, Insightful

    Well, some of the rules in language are pretty universal. I hesitate to say hard-wired because I can't cite it, but think about it. Every language consists of syllables that add up to words that add up to a complete thought.
    That's how an infant learns it. At first, they just babble as they figure out what sounds they can make - naturally, what sounds human language will have in them. Try and think of a language that doesn't have a soft A vowel as English does.

    And deaf babies babble too! It is, however, less complex than a non-hearing impaired infant's. That makes for some interesting theories. If a deaf infant can figure out some consonants, something's probably hard-wired.

    Pay attention to it! That babbling is nonsense, but eventually nonsense syllables.

    What are some first words you can think of? nana. dada. papi. That indicates that they've gotten far enough in development to know that syllables make up words.

    Now here's why I think you are right about the computer not "learning like a baby does". A baby can easily pick up what "dada" and "mama" are quick, what is mama gonna do every time the baby says it? This computer won't learn like an infant will, it will learn language as a blind infant would.

  4. Genetics IS a form of memory. by TapeCutter · · Score: 3, Insightful

    "I get so annoyed when people talk about "hardwired" like we have some kind of genetic memory."

    Genetics IS "memory", your DNA "remembers" what traits your parents passed on. It's in a baby's genes to "discover" their hands and practice moving them until the hands learn how to look after themselves (eg:touch typing).

    Same with language, a baby's genes will make them pick up on the phonetic sounds made by it's parents and try to copy them. It is more difficult for an adult to learn a radically different language (eg Asian vs European) because the adult brain refuses to hear the different phonetics, the adult brain long ago rejected those sounds as irrelevant to language and no longer even hears them in speech. This is why you get almost universal mistakes such as "engrish".

    --
    And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
    1. Re:Genetics IS a form of memory. by Anonymous Coward · · Score: 2, Insightful

      You are missing out something. Languages are not part of our DNA. Sure our brain has trouble adapting, but that doesn't go and re-write our DNA.

      Lets say I go and cut off your legs, and you in turn become adept at walking on your hands. If you went and had children they aren't going to get fantastic hand-walking dna.

      Same goes for languages. You can still take an Asian baby at birth and raise it in a European environment, and it won't speak "engrish".

    2. Re:Genetics IS a form of memory. by ChameleonDave · · Score: 2, Insightful

      PurpleBob accused you of the wrong fallacy. It was not ad hominem but straw man. The AC said that "Languages are not part of our DNA". Note the plural on "languages", which makes it clear that individual systems of encoding (e.g. English, French, Hindi) were the topic, rather than language as a capability, otherwise known as "speech". Your rhetorical question unfairly accused him of not understanding that humans have innate linguistic ability.

      The problem stems from the fact that your mention of children's "genes" (inherited from their parents) picking up on their parents' phonemes made it sound as though you thought that phonemes were inheritable. You almost certainly don't believe anything so silly, and merely worded the sentence a tad clumsily.

  5. Re:Baby talk? I swear at my computer! by muridae · · Score: 5, Insightful

    On the other hand, my laugh sounds exactly like my dad's. Not surprising until you find that I didn't live with my dad and didn't really spend much time with him at all. Many of our mannerisms are also the same. Like the way we walk, with a one hand in my pocket. The resemblence between our personalities is uncanny considering we didn't live together. So I have to ask, how much is based on what we see, and how much is based on our genes. The old nature vs. nurture question.

    You don't say if you knew your dad at all growing up, or if you looked at him as a father figure. If either or both of those fit, then even the child behavior of mimicking the mannerisms of adults could explain a lot of those traits.

    On the nature side of the argument, how much of your gate and posture is controlled by your muscle structure? Same goes for your voice.

    My opinion, you start with the genetic and add the environment later. It is hard for the environment to over come strong traits presented by genetic predisposition, but easy for it to mold how minor traits present.

  6. Re:Snootchie what? by Anonymous Coward · · Score: 1, Insightful

    Wait, flamebait? What? Is there really a lot of people who flame you for making Jay and Silent Bob references on slashdot, or are the mods on crack?

  7. I like the theory... by Quiet_Desperation · · Score: 3, Insightful

    ...that babies talk in baby talk because that's how everyone talks *too* them.