Multimodal parallel natural language learning

From Lukasgirtanner
Jump to: navigation, search

sorry, I have a lack of time at the moment and will need several days in order to write everything in complete sentences. The sentences are therefore now not complete(d) yet.

and don't forget UTSARA, even if it is "only" language

see preventing overstrain

The main goal of this page here consists in the prevention of "foreign language" learning in middle and late childhood and in youth because all time should be dedicated to mathematics, robotics and physical activity and no time should be lost with natural language learning that is perceived as "foreign" language learning because the child is already so old that it perceives a language already as "foreign". In order to still enable the adult person to know several languages, the MPNLL enables to learn several languages in a seamless, embedded and natural way directly from birth. Another goal is to convey the importance of mathematics directly by avoiding the focus on one single mother tongue.

What is better: A very good mathematician or a person with many natural language skills with little idea about mathematics? And if mathematics knowledge precedes the knowledge of natural languages, why shouldn't natural languages also be taught, but in a seamless and embedded way so that the much more important time for mathematics and robotics is not reduced?

It is not possible to develop sustainable technology and ultimately the most sustainable science - mathematics - with the knowledge of many foreign languages. Only as much as possible mathematics knowledge guarantees sustainability.

Possible titles:

  • include formal languages too as some way of keeping everything together mathematically/logically (and also in terms of a perfect language) and therefore not writing natural?

Acronyms: MPNLL

MPLL

"natural" is actually "natural natural": natural in terms of "natural language" but "natural" also in terms of a natural way of learning a language: casually, like a mother tongue, but as part of several mother tongues => actually MPNNLL instead of just MPNLL

or is it actually "parallel serial language learning"? or even "serial language learning"? The USB (Universal serial bus) or SATA (Serial ATA) also transmits data serially, so in that way, it would be more "serial" than "parallel" because only one language is being spoken at the same time (no neural interfaces yet and even in case of ultrasound, there would probably still be only one language at the same time, just significantly faster). So, on a small scale, it is actually serial learning. But when the process as a whole is labeled or named, the world "parallel" might be more suitable or appropriate because the process as a whole is perceived that languages are learned parallelly at least as far as every day is concerned (compared to conventional "foreign language" learning where a child has to wait years even for its first foreign language).

or is it just PLL?

Sapir-Whorf hypothesis: http://en.wikipedia.org/wiki/Linguistic_relativity

if only one sign language is used, the same signing will be repeated several times. use different sign languages instead? or would there at least be the advantage that at least the one sign language is repeated often and therefore the child can at least master one language very well from an early onset: the sign language?

singing language has several benefits:

  • more early onset. babies can already communicate with their mother before they are able to talk. this is also the reason why there are a lot of signing babies already being taught by their mother
  • stronger bond with their mother
  • multimodal (see also the points below)
  • body abilities increase
  • prioprosensing
  • movement is healhy
  • great additional potential for robotics (movement instead of just sound) and therefore indirectly for mathematics
  • better general body language because of constant training
  • acting

language relativism crucial in order to increase sense for:

  • mathematics
  • formal languages
  • a "perfect formal or natural language"
  • multimodality of languages (especially signing)
  • the Sapir-Whorf hypothesis: natural languages are relative and if one has to base one's own means of expression on something relative, learn at least several ways of it and do not have a preference for one particular (apart from maybe preserving an idiolect)

late stabilization? when? to which degree? at which age?

avoiding hybridization of language

avoiding mother tongue

avoiding a sense of matter of course for a natural language

communication with animals? and robots? help by robots? help by software?

role of the mother?

generational increase? mother->child->mother->child what about generational decrease?

time-related bandwidth efficiency: high-frequency / ultrasound / increased bandwidth

speeding up not in the ultrasound range by simply talking faster?

priority mathematics <-> robotics <-> MPNLL?

direct input by neural interfaces? (risky, only when deemed absolutely safe, no experiments with that)

most important: sign languages, signing babies which sign language to choose? several ones?

"The imp" What was meant with that?

Multimodal:

  • smell
  • lighting
  • specific rooms?
  • fixed order?

how many for a default start? related: analyzing the genetic potential of the parents and based on that pregnancy learning and learning immediately after birth? help of statistics?

how quickly go up and/or down?

individualization (down, up) for every individual baby? constantly learning at the edge

detection/assessment of cognitive overload? under-challenge?

specific/standardized announcement words?

learning during pregnancy: how long, how many, which ones, duration daily exposure, frequency, how loud, own speech of mother? degree of international standardization for pregnancy learning?

role of common precursor languages? speed up efficiency vs. hardly any speakers? (probably no)

fixed order? or changed order in order to increase flexibility even more? and if yes, from which age / cognitive readiness?

everyday adaptability and priority rules.

duration of everyday talk: 12 hours? continuous adaptation to constant talk while adhering to situational adaptivity and priority

Order of languages:

  • degree of international standardization
  • linguistic diversity in order to maximize early childhood flexibility: diversity precedes economic/technological importance
  • priority/order of the languages
    • number of speakers?
    • economic/technological significance?
    • mathematical/IQ/technological significance?
  • how many preceding idiolects/dialects?
  • relation idiolect <-> (standardized) regional dialect <-> standardized language

script learning? how? when? how many?

help by technology (pattern recognition devices, translators, eye displays)

phonetic learning? (good standardization and learning opportunity)

constantly adapting priorities/orders for every generation

increase over time? realistic goals? when 20? when 100? when 1000?

the more the higher the bandwidth

ultrasound animals: for example bats, mice, rats, dogs, cats

bandwidth: proximity matters. the higher the bandwidth the less the power (but probably not a real problem). neural interfaces?

ultrasound extending hearing capabilities might be much more safe and much more early available than direct neural interfaces. Strive for ultrasound and only expect (much) afterwards neural interfaces to be sufficiently safe and feasible

don't forget that the ultrasound also has to be produced. voice?

begin in pregnancy only when the fetus can hear, not before: and be very careful, gentle, careful and considerate. If a child is not born yet, it is difficult to assess and prevent overstrain, so in case of doubt, it is better to do too little instead of too much.

what is natural language stabilization? alternative titles: (natural) language stabilization, ... at which age? or never completely? to which degree at which age? 90 percent at 10? or 20? or 40? or never? or only 95 percent at 40? priorities in language stabilization? at least one 99.5 percent or even 100 percent stabilized language in order to be "on par" with the other people?

(how long) do the children remain in a signing and mixed-language babbling stage because of MPNNL?

distribution between several persons (as it is now bilingual trilingual) and/or robots (other older children?) might also be an option, but is it standardized enough?

(minor but still existing problem: keyboard use, keyboard layouts, typing abilities)

beginning is unfamiliar and maybe even sometimes ridiculous and difficult but as soon as you have been used to it from your own childhood, no problem, with every generation it becomes more easy and ordinary/conventional, just the technologies and speed/numbers/efficiency might continue to change or increase

advantages:

  • no "foreign language" learning necessary
  • much faster learning abilities for new languages
  • better understanding for formal languages and mathematics
  • a better global mutual understanding, less limits, more knowledge
  • and because of that maybe even more global peace?

is it possible? what is the most important priority? relationship between mother (parents) and baby. happiness of the baby. that it can rest and sleep when it wants. but there is still plenty of unused time in the day and nowadays, only a small fraction of that time is used for language learning of usually only one or much more rarely two languages. much more would be possible, on average probably 4 languages (with intermediately delayed stabilization or 5-6 languages with more pronouncedly delayed stabilization on average, not talking about especially talented children where 8-10 languages or even 15-30 languages would be possible depending on stabilization goal). and the sign language keeps everything together and ensures communication anyway from a very early onset.

In the long-term, the goal is to learn not only all (as of 2010-05-24) 7589 languages of the ISO 639-3 standard, but also an increasing number of dialects and sub-dialects until actually at least theoretically there would be the ability or capacity to learn every idiolect. Constructed and extinct or "classical" languages would also be increasingly learned, but probably only beginning after about 30 to 100 of the most important extant natural languages have already been learned sufficiently.

In order to also preserve linguistic diversity, one could assign especially rare languages prematurely in order to prevent that they are only learned last when the chances that they will already be extinct is high. Therefore, as soon as maybe the 50 - 200 most important languages would have been learned, languages that are close to extinction could also be included or promoted in their priority. If mothers or parents would have the preference to speed up the process and even earlier include languages close to extinction, it would be encouraged. If a small or almost extinct language would be the idiolect, (sub)dialect or language of a parent, it would be included anyway as part of the principle that the own idiolect should be passed to the next generation with the highest priority regardless of its global significance.

Connection to other abilities like (apart from mathematics, robotics, ee and comp sci) musical instrument learning, motor abilities, singing, acting and art.

The archetype to prevent: the under-challenged young child sitting in public transport and repeating station announcements out of boredom. I have really witnessed that.

In contrast, the stress of having to learn languages perceived as "foreign languages" during late childhood let alone adolescence or adulthood can even be prevented like that because all languages that would have had to be learned at an older age are already rooted in the child's mind because it was able to learn them from birth and some of them even already beginning in the womb of its mother (but gently and tentatively as soon as it could hear).

In an ideal case, MPNLL is done in a casual way, without ever overstraining a baby and without any artificial stress, totally embedded within the healthy relationship within its mother. So, the question is actually only which degree of language stabilization should be necessary at which age and at which age multi-language babbling increasingly progresses into a more correct and less hybrid speech.

another topic or advantage of MPNLL: the prevention of accents

don't forget to mention on the multimodal parallel natural language learning page that actually at least one formal language should also be taught parallely. how could it or how was it possible to forget that so long? so teach at least one formal language better several ones in the quest/struggle for a perfect language? and like that and maybe only like that, wouldn't MPNLL also be integrated or "subjected" or committed more clearly into mathematization?

not only the "N", but also the "M" could be doubled because "M" might mean "multimodal" and "multiple", so the acronym would actually be MMPNNLL