2
$\begingroup$

I'm interested in finding a solution for the following problem:

  • problem space: any language that has more than 1 article
  • let's take German language as example. So the articles in German are "der", "die", "das". - every word has an article as prefix and you must use the article. for example you can't say just "Auto"(car), you must say "das Auto".
  • If you are not a native German speaker, you have bad luck, because you have to memorize the article too, when learning a new word, since there are only 3-4 rules to know which word takes which article. And those rules are perhaps about max. 10% percent of the vocabulary. (for example if a word ends with "-ung", it takes "die")


So here comes the funny part: as a non-native-German-speaker, i wanted to analyze the language from an IT-point of view and asked some friends of mine random words with the following properties:

  • first I reduced the input set from "any German word" to "a German word", for which no known article rule exists".
  • then I extended the input set with "made up words", which do not exist.

Every candidate had the same answer for German words, which should be not surprise. But when I asked them per word "why do you think/feel that the article of the word is "der/die/das"?" they couldnt give an answer. They just know it, without knowing why.

Here comes the real hammer: every candidate had the same answer for any made-up non-German word. I say anything, make any meaningful voice, and they all give the same answer (ie. article).

I'm pretty sure that I'm not the first human being in the world who made thoughts about this topic. And I'm interested in any scientific papers/research about that topic. By the way, what I definitely do NOT need is, any method (like neural networks) which outputs a pure mathematical function. I thought about following possibilities:

  • focusing on feelings of subjects psychologically (I think it can be managed with colors somehow etc.)

  • analyzing the words statistically according to their syllable structure

Where can I begin? What are your suggestions? Are there any Machine Learning research about that topic?

$\endgroup$
5
  • 1
    $\begingroup$ I would guess that gender is determined by analogy, and that instead of 3-4 rules you really have much more, only they're too specific to be very useful for someone learning the language. Also, I would say this is off-topic here, more appropriate in a linguistics forum. $\endgroup$ Jan 20, 2013 at 5:53
  • 2
    $\begingroup$ Also, this is very much language dependent. You can't generalize this, and should focus on one language at a time. Moreover, many implicit rules and analogies are made subconsciously in native speakers. For example, I do not know too much about the formal grammar rules of my native language, but I speak it well. You might want to pose this question to people who have studied the grammar formally and deeply, and are not simply native speakers. For German, german.se might be a good place to start. $\endgroup$
    – Paresh
    Jan 20, 2013 at 7:32
  • 1
    $\begingroup$ Also there is a linguistics stackexchange: linguistics.stackexchange.com. $\endgroup$
    – usul
    Jan 20, 2013 at 21:02
  • $\begingroup$ As a German native, I can assure you that not everybody gives the same answers. Especially for words taken from English in recent years, all kinds of articles float around. But for non-abstract words, there is a "rule": biological gender resp. genderlessness. Obviously, that only gets you so far. Swedish is worse in this regard; but then, I'm no native of that language. $\endgroup$
    – Raphael
    Apr 21, 2013 at 15:14
  • $\begingroup$ Oh, and: "der Schwung", "der Dung", "der Sprung", ...; I guess you were thinking mainly of nominalization? ;) It's a good rule-of-thumb, anyway. $\endgroup$
    – Raphael
    Apr 21, 2013 at 15:40

1 Answer 1

1
$\begingroup$

A language is a result of a complex historical development. If the development were guarded by simple language generating rules it would've been different like some artificial languages such as the Esperanto language.

Many natural languages have nouns split into several categories. These categories are often referred to as the Gender. Though the link between the grammatical gender and the natural or biological gender is not always obvious.

In Slavic languages for instance the gender of a word can be guessed from the flection at the end of the word. E.g. in Russian "облако" (a cloud) is neutral, "-о" is the flection, and "зима" (the Winter) is femenine, "-а" is the flection.

In Old English the category gender did exist but was eliminated during the evolution towards the Modern English. In general all languages that stem from the Latin language inherited some form of grammatical gender (for more details see the discussion in Wikipedia).

When learning the language one has to memorize at least the words and there meaning. In learning languages that keep grammatical gender one has to memorize the gender as well. You can find the gender of a word in any German dictionary.

Look at the small letter f, m, or n, printed in italics next to the word.

E.g.

Bündel, n a bundle;

Bus, m a bus;

Tasse, f a cup;

That is "Das Bündel", and "Der Bus", and "Die Tasse".

When a comprehensive dictionary is available for a language with all the nouns annotated with their respective grammatical gender, I guess there is little space left for machine learning. It is much more easy to keep the complete dictionary in memory than to devise a set of rules that would always be prone to errors.

$\endgroup$
1
  • $\begingroup$ There are examples where machine learning may be useful to resolve ambiguities. For example, in Swedish the "same" word can have different meanings depending on whether it's the -en or -et variant. In the presence of typos, things get worse: compare "der Falter" and "die Falte", and their respective plurals "die Falter" and "die Falten". $\endgroup$
    – Raphael
    Apr 21, 2013 at 15:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.