What is not a corpus?
As we move towards a definition of a corpus, we remind ourselves of some of the things that a corpus might be confused with, because there are many collections of language text that are nothing like corpora.
The World Wide Web is not a corpus, because its dimensions are unknown and constantly changing, and because it has not been designed from a linguistic perspective. At present it is quite mysterious, because the search engines, through which the retrieval programs operate, are all different, none of them are comprehensive, and it is not at all clear what population is being sampled. Nevertheless, the WWW is a remarkable new resource for any worker in language (see Appendix), and we will come to understand how to make best use of it.
An archive is not a corpus. Here the main difference is the reason for gathering the texts, which leads to quite different priorities in the gathering of information about the individual texts.
A collection of citations is not a corpus. A citation is a short quotation which contains a word or phrase that is the reason for its selection. Hence it is obviously the result of applying internal criteria. Citations also because lack the textual continuity and anonymity that characterise instances taken from a corpus; the precise location of a quotation is not important information for a corpus researcher.
A collection of quotations is not a corpus for much the same reasons as a collection of citations; a quotation is a short selection from a text, chosen on internal criteria and chosen by human beings and not machines.
These last two collections correspond more closely to a concordance than a corpus. A concordance also consists of short extracts from a corpus, but the extracts are chosen by a computer program, and are not subject to human intervention in the first instance. Also the constituents of a corpus are known, and searches are comprehensive and unbiased. Some collections of citations or quotations may share some or all of these criteria, but there is no requirement for them to adopt such constraints. A corpus researcher has no choice, because he or she is committed to acquire information by indirectly searching the corpus, large or small.
A text is not a corpus. The main difference (Tognini Bonelli 2001 p.3) is the dimensional one explained above. Considering a short stretch of language as part of a text is to examine its particular contribution to the meaning of the text, including its position in the text and the details of meaning that come from this unique event. If the same stretch of language is considered as part of a corpus, the focus is on its contribution to the generalisations that illuminate the nature and structure of the language as a whole, far removed from the individuality of utterance.
As we move towards a definition of a corpus, we remind ourselves of some of the things that a corpus might be confused with, because there are many collections of language text that are nothing like corpora.
The World Wide Web is not a corpus, because its dimensions are unknown and constantly changing, and because it has not been designed from a linguistic perspective. At present it is quite mysterious, because the search engines, through which the retrieval programs operate, are all different, none of them are comprehensive, and it is not at all clear what population is being sampled. Nevertheless, the WWW is a remarkable new resource for any worker in language (see Appendix), and we will come to understand how to make best use of it.
An archive is not a corpus. Here the main difference is the reason for gathering the texts, which leads to quite different priorities in the gathering of information about the individual texts.
A collection of citations is not a corpus. A citation is a short quotation which contains a word or phrase that is the reason for its selection. Hence it is obviously the result of applying internal criteria. Citations also because lack the textual continuity and anonymity that characterise instances taken from a corpus; the precise location of a quotation is not important information for a corpus researcher.
A collection of quotations is not a corpus for much the same reasons as a collection of citations; a quotation is a short selection from a text, chosen on internal criteria and chosen by human beings and not machines.
These last two collections correspond more closely to a concordance than a corpus. A concordance also consists of short extracts from a corpus, but the extracts are chosen by a computer program, and are not subject to human intervention in the first instance. Also the constituents of a corpus are known, and searches are comprehensive and unbiased. Some collections of citations or quotations may share some or all of these criteria, but there is no requirement for them to adopt such constraints. A corpus researcher has no choice, because he or she is committed to acquire information by indirectly searching the corpus, large or small.
A text is not a corpus. The main difference (Tognini Bonelli 2001 p.3) is the dimensional one explained above. Considering a short stretch of language as part of a text is to examine its particular contribution to the meaning of the text, including its position in the text and the details of meaning that come from this unique event. If the same stretch of language is considered as part of a corpus, the focus is on its contribution to the generalisations that illuminate the nature and structure of the language as a whole, far removed from the individuality of utterance.
