The structure of the data has some peculiarities which need to be kept in mind to effectively use CorDon and especially its search function. This page lists some typical problems together with potential solutions.
Ensure that the search is set to "word form" and not "lemma".
Ensure that you are searching the right layer. The "original" layer contains the text including all special characters and metric diacritics (sŭgryʒ́o). The "transliteration" layer transliterates this text into the modern Lithuanian alphabet (sugryžo). The "modernized form" layer contains the form's modern Lithuanian equivalent in modern standard orthography (sugrįžo).
Pay attention to clitics, leaving them out if necessary. Clitics (ne-, be-, -gi) are detached in the "modernized form" layer, so that irgi appears there as ir + gi, nežinai as ne + žinai etc. The exception are the negated reflexives, where detaching the clitic would have left an ungrammatical "trunk" form (nesigėdi > †sigėdi). For these words the clitic is left attached and the non-negated form is given in addition (the lemma entry for nesigėdi is thus "nesigėdėti/gėdėtis").
Ensure that the search is set to "lemma" and not "word form".
Search separately for the reflexive form of the lemma. Reflexive verbs are lemmatized in their reflexive form (imasi has the lemma imtis, not imti).
Try again using "contains" instead of "whole word". Lemma entries may contain two forms separated by a slash (for double lemmata like dosnai/dosniai or nesigėdėti/gėdėtis) or a comma (for alternative lemmata, e.g. budinti,budyti). Because the search sees such entries as one "word", the individual forms listed there can only be found using "contains".
Try a separate search with the word capitalized. The search function is case-sensitive. Lemmata (except proper names), transliteration and modernized form are always lower-case, but the "original" layer preserves the original capitalization.
Write `I instead of Ì. Due to typesetting limitations the Nesselmann (1869) print edition puts the accent of a capital I in front of the letter rather than on it. The "original" layer preserves this quirk.
Search using "contains" instead of "whole word". The indeclinable possessive pronouns are lemmatized in a special format: as mano/aš, tavo/tu and savo/savęs.
Do a "word form" search in the "original" or "transliteration" layer. In all other layers, punctuation marks are invariably annotated with „-“.
Keep in mind that, for technical reasons, the categories offered there are exclusive, not inclusive. Meaning: A search for "noun > common noun" finds exactly those forms tagged as "common noun" (NA), but not those tagged as "common noun, adverbial" (NAA) or "common noun, predicative (NAP). You can use the "or" button to search for multiple types at once.
Keep in mind that punctuation marks are treated as tokens. For the purposes of the search, the sequence „Sveiks , dieve duok“ consists of four, not three, tokens. Thus you would need to search for the sequence adjective–[blank]–noun–verb in order to find this phrase.
Keep in mind that the search does not consider syntax. Any noun–adjective sequence found is not necessarily a noun–adjective syntagm.
These are catchwords indicating the first word of the following manuscript page. These lines can also be identified by the letter R (for "Reklamante") in their line ID.
Couldn't find your problem? Please contact Jolanta Gelumbeckaitė (gelumbeckaite at em.uni-frankfurt).
| 
| 