How to search
Enter the search word or phrase in the box. The browser doesn't distinguish between lower- and upper-case letters, and it searches only for complete words. If you enter only part of a word, you won't therefore find citations containing the complete word. But you can use operators: an asterisk * can replace any string of letters, and a question mark ? can replace one letter.
If you wish to find two words or phrases simultaneously, they must be separated by a slash / in the search box.
Type in Polish characters from the keyboard. If you don't have a Polish keyboard, you can use the left apostrophe, ` (in most keyboard layouts this is located below the Esc key) - this is used after the Latin letter corresponding to the Polish character you want: e.g., you get ź by entering x`.
The browser has a built-in mechanism for dealing with Polish inflection. When this is activated, the search concordance will show citations with the word form or string as entered, as well as those with inflected forms. To deactivate the mechanism, put your query in quotation marks.
In the full online corpus you can search for a word or string within the context of another word. The context word can also include inflected forms. Length of left- and right-side context can be set in the appropriate boxes.
In the boxes further down you can specify the number of citations required, the length of concordance lines on each side (up to 30 words) and the method of alphabetical sorting: by the search word, by the first word to the right or by the first word to the left. The number of citations is limited to 10,000 in the full version and 300 in the demonstration version. The program selects the required number of hits from various corpus texts. E.g., if someone is searching for 1,000 examples of the use of the most common word in Polish, w, they will get them not just from the first book searched, but on the basis of a random selection of all occurrences of the word within the corpus. Additional information shows the total number of citations in the demonstration version and in the full online corpus. (In the diversified corpus there are just under 500,000 occurrences of the word w, while in the entire online corpus the word occurs over one million times.)
The citations will be shown on successive pages consisting of 300 lines. Larger pages would take too long to open. An additional window allows users with greater throughput capacity to increase the number of citations on one page to 500. The samples will be sorted throughout the concordance set, and not just on the current page. So that if we sort 1,000 hits for w, on the first page we will see the examples w 1945 roku, w akcji, and w Ameryce, while results from the same search with right-side context for further letters in the alphabet will only appear on subsequent pages.
Concordance lines can be re-sorted by clicking on the appropriate column heading. If the context you specify turns out to be too short, you will need to repeat the concordance process.
To check the source of a given citation, click on the query word within the citation. You will then see a larger portion of text - up to 50 words on both sides. Above it a window will appear with details identifying the text. Books are supplied with details of author, title, year of publication and of origin of the text or year of the first edition (if the corpus contains the first edition, then only one year is given). Descriptions of press texts contain: title, number of periodical or date of newspaper edition, year and place of publication. Spoken texts are provided with details of place and year of recording, type, and subject matter. In the same way we have information on subject matter, type, place and year of origin of ephemera and web site material.
More extensive citations can be viewed in tagged form.
The corpus browser doesn't distinguish between homonyms, or different meanings of the same word form. Citations for mamy for instance will include the noun form from mama and the verb form from mieć.
