Procedures

Contents:

• Processing new words
• Writing an essay
• Parsing sentences using the phrase-structure grammar rules

Processing new words (i.e. words for which you don't know its meaning or its pronunciation):

1. Go to the English->Polish page of the Cambridge Dictionary.
2. Enter the word in the search box.
3. Listen to the pronunciation of the word by clicking on the UK pronunciation button on the result page.
4. Add the word to your Words collection on cram.com. On the back side of the flash card enter the Polish translation and the note how to pronounce the word. Use Polish spelling in the pronunciation note. For example, for the word "conscious" it would be "'konszes" (note the ' sign at the beginning to mark the accent).

Writing an essay

1. Before you start writing anything, plan what you are going to say and in what order. 
2. Write the first version of your essay in Polish. In this way your imagination won't be restricted by the language which you don't know well yet.
3. Once you are happy with the Polish version, translate its every sentence in the following 4 steps:

• A. Translate it as best as you can on your own, checking only for unknown words. Make sure that the translated sentence has a well-formed structure. Check the correctness of your sentence using the LanguageTool for proofreading.
• B. Translate the Polish sentence into English using Google Translator.
• C. Compare your own translation with the one produced by Google. Keep in mind that Google, although amazingly clever, is not perfect and makes mistakes. Write a short comment, in Polish, saying what Google did well and where it erred.
• D. Modify your translation, if you think it necessary, adopting some of Google's "ideas".

Each sentence of a submitted essay should have four parts, A, B, C, D, as described above. Each part should be written as a separate line.

If necessary, count the number of words in your text using the Word Counting Tool.

 Parsing sentences using the phrase-structure grammar rules

1. The produced parse tree should have one and only one root node.
2. For every node above the level of speech part nodes there should be a corresponding rule used to produce the node.
3. The branches of the parse tree should not cross.