Natural Languages

  Home

Table of Contents

• 1. Personal Context
• 2. Learning languages
  • 2.1. English
  • 2.2. German
  • 2.3. French
  • 2.4. Spanish
  • 2.5. Latin
  • 2.6. Greek
  • 2.7. Japanese
  • 2.8. Chinese
  • 2.9. Esperanto
  • 2.10. Sanskrit
  • 2.11. Italian 2004-04...
  • 2.12. Korean 2007-03...
  • 2.13. What's next?
• 3. Linguistics
• 4. Natural Language Programming

1. Personal Context

I've never been particularly good at learning languages, but I try. Over the years I've tackled a few in school, and quite a few on my own. In the process I've looked into formal linguistics and read a few texts in semantics and general sematics.

In addition, I have had brief exposure to Natural Language Programming (NLP), using grammar-based and statistic-based approaches.

2. Learning languages

My approach is:

1. Get a Teach Yourself ... book. It is a great series; cheap and effective. (I'll abbreviate that TY.)

2. Work through it, reading all the exercises aloud. Reading aloud lets your voice and your ear learn the language. It is nothing like the oral aural experience a kid gets growing up in the language, but it is better than nothing.

3. Get a bilingual dictionary and a graded reader (collection of gradually more difficult reading pieces). You can get these cheap at a used book store near a university. The stories begin to teach the culture and history. Continue reading aloud.

4. Get one or two classics in the language. E.g., German: Faust; Spanish: El Cid. Again, used copies can be cheap. Continue to read aloud. This is tough going. Use the dictionary and the grammar from the TY to ferret out the meanings. Try to nail down the tenses and odd turns of phrase. Somewhere along in here the flavor of the language and culture begins to come alive. You begin to shift from decoding to reading for content. I remember part way through El Cid, I began to lisp -- my mouth just felt right doing it.

5. If you happen to have a book with the translation on the opposing page (e.g., as in Loeb Library), don't look . You will find the translation is roughly the same meaning as the original but native turns of phrase are entirely lost. Those phrasings are not just accidental -- they are the heart and soul of the culture.

6. For contemporary languages, look for magazines and popular novels. Read for content. Make sure it is interesting material in its own right.

2.1. English

My native tongue. Not that I approve. I consider myself an aboriginal pre-Celtic Briton, whose land was overrun by assorted IndoEuropeans (Celts, Angles, Saxons, Norse, etc). My people were pushed into mountainous Wales (from Celtic "foreigner"), then to North America, then to the West coast of North America.

Be that as it may, I've read fairly broadly in English. Have spent many hours browsing libraries and bookstores. Tend to read the entire works of an author in a burst (1-2 books a day).

2.2. German

Studied it in highschool. Read Faust and a few other odds and ends. My grandmother had some "blackletter" books which were fascinating but pretty hard to read. With apologies to Germans everywhere, neither the language nor the culture does much for me.

2.3. French

Studied it in college. Followed by graded readers and then collections of pieces from 1700-1800's. I never caught on to the nuances of the multiple french languages. My instructor said she knew 5 languages, and 4 were french.

2.4. Spanish

Used TY . Ended with several contemporary novels and then El Cid. I liked the sound of El Cid.

2.5. Latin

Used Wheelock's Latin (c. 1963) text and the Using_Latin (c. 1961) texts. Ended with Gallic_Wars , Argicola , and assorted smaller readings. I tried to use the hard "c" sound (Tacitus as tack-i-tus), but I can hear how the soft "c" evolved.

2.6. Greek

Used Homeric_Greek and Reading_Greek . The long term goal is The Oddessey. It's been on on my bedside table for years.

2.7. Japanese

Used Japanese in 10 Minutes a Day and worked on Japanese for Busy People . Also have several readers.

I stopped when I realized I don't have a subtle bone in my body, and that the Japanese culture might as well be Martian as far as I'm concerned. Learning an entirely different language family and 3 scripts was just too much. We'll have to meet halfway. Like in Hawaii. At a luau.

So I'll just stick with translations, like haiku, koans, and Shonagon's pillow book.

2.8. Chinese

Awesome culture and awesome language. I started backwards -- learning 1000 kangi over a summer from Reading_and_Writing_Chinese (c. 1979) Of course I forgot most of them immediately. Then restarted with texts prepared by mainland China. These were pretty drab. I think I'm looking for a spicier dialog. Something like Suzy Wong does Berlitz. Anyway, reading Li Bo even in translation is a winner.

I can understand why the Chinese people had to revolt against the crushing burden of the Mandarin heritage. But once they've made their own place in the world, they need to go back and rescue some of the earlier culture and writings.

2.9. Esperanto

Used Richardson's Esperanto (c. 1988) I like cultures that are willing to use romance to coax readers along another page. Also did the on-line email course. Reading is easy. Getting the idioms right for writing is tough.

For a while I lurked on soc.culture.esperanto, but it is (was) mostly about the language. Kind of like listening to shortwave and hearing about receivers and antennas.

2.10. Sanskrit

Installed ITRANS, so I could print texts. Did a few chapters in TY . I like the sound. It is fun seeing IndoEuropean terms and grammar showing up in an alternative form. Haven't done much more.

2.11. Italian 2004-04...

It started innocently enough with an interest in Italian movies and cooking. Then I read (in English) Barzini's "The Italians". Before I knew what was happening, I'd purchased another TY book, several dual-language readers, and a dictionary. Several months of intermittent study brings me to crude decoding-level reading.

I still can't understand any of the movie dialogs, but I care about the people more fully.

...Just started Umberto Eco's "Il nome della rosa".

2.12. Korean 2007-03...

Using TY plus several dictionaries and vocabularies. Just getting started.

2.13. What's next?

Mostly I've become too busy with other things to do languages. Here are some options:

1. Languages represented by my family's own heritage: German, Gaelic (Irish, Scot, Welsh), Swedish, Danish, Polish, Hawaiian, Cherokee.

2. Languages spoken within walking distance of my house prior to European invasion. Assorted Salish-based languages. Several local tribes offer courses: you learn a traditional story, perform it, and become personally responsible for it. True oral tradition.

3. Languages spoken currently within walking distance of my house. I'm personally aware of these: Swedish, Norwegian, Danish, Irish, Scot, Welsh, German, French, Mandarin Chinese, assorted other Chinese, Vietnamese, Thai, Korean, Cambodian, Spanish. My wife (a nurse) tells me she hears from every continent and culture.

3. Linguistics

chomsky98 (2007-09-28)

Noam Chomsky. "On Language". The New Press, 1998 ISBN 1-56584-475-0. Reprint of:

• "Language and Responsibility",1977
• "Reflections on Language", 1975.

Covers his seminal notions re generative grammars, genetically-determined brain structures which process language, and use of language for propaganda).

A phrase used regarding someone else applies here; THis guy's mind impresses yours.

aronoff2003

M. Aronoff, J. Rees-Miller, editors. "The Handbook of Linguistics". Blackwell Publishing, 2003. ISBN 0-631-20497.

Intended to be a reference work, used as needed. I (of course) read it cover to cover.

comrie1987

B. Comrie, editor. "The World's Languages". Oxford University Press, 1987. ISBN 0-19-520521-9.

Brief (e.g., 10 pages) summaries of language families and individual languages.

nirenburg2001

Sergei Nirenburg, Harold SOmers, Yorick Wilkes, eds. "Readings in Machine Translation". MIT Press, 2001. ISBN 0-262-14074-8.

Collection of key papers in the field; some historical, some fairly recent. Might safely be subtitled, "Ideas that didn't work". 50 years of very bright people funded by national intelligence services, and we still can't do as well as a reasonably well-informed bilingual native speaker.

4. Natural Language Programming

Also known as Computational Linguistics. I've read a few texts, and have done MS CompSci-level exercises, but haven't really dug in. I'm vaguely interested in doing human-assist Machine Translation, mostly for vocabulary. I find it is easy to recognize grammar in languages you have once known, but the vocabulary is easily lost.

I'm just beginning to build OSS tools for multi-lingual text editing, such as mined and scim.

allen1995

J. Allen. "Natural Language Understanding" Benjamin/Cummings, 1995. ISBN 0-8053-0334-0.

jurafsky2000

D. Juragsky, J. H. Martin. "Speech and Language Processing". Prentice-Hall, 2000. ISBN 0-13-095069-6.

manning1999

C. D. Manning, H. Schutze. "Foundations of Statistical Natural Language Processing". MIT, 1999. ISBN 0-262-13360-1.