Korpusu un individuālā vākuma salīdzinājums: ģenitīva un nominatīva konkurence saistījumā ar adverbu

The article describes the advantages and disadvantages of corpus data and individual collection. The availability of various grammatically annotated corpora of the Latvian language ensures more and more extensive grammar studies based on corpus data. On the other hand, the individual collection played a major role in the development of linguistics, and it is an older way of obtaining practical material. However, in today’s technological age, the individual usefulness of the collection has come into question. For a practical comparison of the two data acquisition methods, a common phenomenon in modern Latvian language usage was chosen – genitive and nominative competition (in connection with an adverb), which was found both in the individual collection and in the corpora data. In this study, three adverbs are selected – daudz ‘many’ (wordform vairāk ‘more’) maz ‘few’, cik ‘how many’ – which are analysed in greater detail in the syntactic centre of the sentence in connection with the genitive or nominative of the noun. The individual collection consists of relatively spontaneous unedited use of the Latvian language in speech and writing – 100 sentences with each adverb. For corpus-driven data analysis, four corpora of the Latvian language were used: The Balanced Corpus of Modern Latvian (LVK2018), Latvian Treebank (LVTB), Latvian Speech Recognition Corpus (LRK2013), and Corpus of Latvian Pandemic Diaries (PanDi). The phrases with the genitive form dominate the material of both the corpora and the individual collection. According to the used sources, nominative is more frequent in the Latvian Speech Recognition Corpus (LVR2013), but in the group of three analysed adverbs – more often in connection with the adverb cik ‘how many’.