Key ideas

The following are the key ideas of this chapter.

• Apart from anything else that they may be texts are a publicly available resource for doing science about language. Especially true of electronic text.
• There are good mathematical tools for studying codes and cyphers, and some of these are useful in linguistics.
• Linguistics could be seen as a branch of telecommunications engineering, if you wanted to.
• Linguists have to decide whether and how to exploit the availability of electronic textual resources.
• Actually having the data can be a challenge to the cherished pre-conceptions of current linguistics. Arguably this is no more than an artefact of the very recent history of linguistics.
• Statistics is a general method for handling finite samples of potentially infinite (or at least unmanageably large) datasets. It applies directly to data-intensive linguistics, addressing the central question of whether the finite samples available to us are in any appropriate sense representative of the language as a whole. All our arguments from data to general principles of language and language behaviour hinge on the assumptions which we make about this crucial issue.