A tree representing a sentence contains all word-forms of the sentence including symbols of punctuation. Besides it contains only the root, whose attributes can take on the following values:
| Attribute name | Value |
|---|---|
| Lemma | # |
| tag | ZSB |
| form | #<serial number of the sentence in the framework of the file given in mstag |
| afun | AuxS |
| lemid | <identification of the sentence in the framework of ČNK> |
| mstag | <identification of the file for manual annotation at analytical level> |
| origf | #<serial number of the sentence in the file given in mstag |
| origap | space for annotator's notes |
| gap1 | |
| gap2 | |
| gap3 | |
| ord | 0 |
The number of nodes of a sentence tree equals the number of word-forms at the morphological level (including punctuation marks) plus one. In manual annotation no node is allowed to be cancelled or added.