Using conversational data to determine lexical frequency in British Sign Language: The influence of text type

Abstract This paper presents findings from an objective lexical frequency study in British Sign Language (BSL) based on 24,823 tokens collected as part of the BSL Corpus Project. The BSL study is only the fourth objective frequency study involving sign languages to be reported and is also the first study for any sign language to be based on entirely on spontaneous conversational data. When compared to previous frequency studies (both spoken and signed), some similarities can be observed although differences that may be attributed to text type are also recorded. When compared with subjective frequency ratings collected for BSL, a positive relationship is reported (similar to what has been observed for spoken languages). This is in contrast to a previous study which suggested a much weaker relationship between the two; however, this conclusion was based on a frequency count derived from narratives. These differences highlight the importance of using frequency measures derived from natural and spontaneous data, an opinion that has been emphasised in the spoken language literature.