Sportive Thoughts

I'm taking a special course in reading German this winter; it is special because the aim of the course is not to learn German. The aim is rather to learn how to read German and, with the aid of a dictionary, translate accurately into English. Thus, we are taught the many parts of speech without having to actually speak.

This is difficult but not impossible, and although a full knowledge of the German language is preferable to this more tedious methodology, learning how to read and translate German can be quite helpful.

As a supplement to the coursework, I've tried a few exercises of my own. My favorite so far is an experiment using the syntactic knowledge I've learned in class to smooth over the roughshod transcriptions of a translation "engine" (Google Translator). Here are my notes for the experiment. D(n) denotes the original German of a given sentence n; J(n) denotes my syntactic translation of n; G(n) denotes the engine transcription of n; and A(n) is the adjustment of J(n) and G(n) into idiomatic English.

D(1): Da der Begriff der Eironeia schon mehrfach behandelt worden ist, mtissen wir in den folgenden Bemerkungen, zustimmend oder ablehnend, die altere Literatur zu dem Themahiufig heranziehen.

J(1): The concept of irony has been treated frequently, we must increasingly extract from the old literature on the theme, for better or for worse, in the course of commenting.

G(1): As the notion of multiple irony already been dealt with, we must in the following remarks, approval or rejection, the older literature on the theme increasingly used.

A(1): As the concept of irony has already been dealt with quite frequently, we must, for better or for worse, increasingly extract from the older literature.

Notes: Google can perform a "spot translation" or transcription much faster than I. This can be an asset, because with the engine I have a quick gloss of most (but surely not all!) of the words used. My translation of the first sentence of Leif Bergson's "Eiron und Eironeia" follows the syntactic meaning of the text. Whatever the task of translation entails, it must include both a fast decipher of individual words and a critical examination of syntax. The mystery is how the two can be separated: how can the recognition of words lie outside the mechanism for parsing syntax? Efficiency is obviously lost. On the other hand, it is obvious that the agent who can decipher words and syntax simultaneously is said to know the language.

Neither I nor the engine know German, but I know English and the engine does not. Therefore, the engine is a tool for allowing me to know what the German says in English, provided that I already know how to construct English sentences and parse German ones.

To be continued...