About Paul Baker

CASS co-investigator Paul Baker is a Professor of Linguistics and English Language at Lancaster University. His research interests corpus linguistics, language and gender/sexual identities and critical discourse analysis.

Two approaches to keywords

On July 4th, 2013, I gave a presentation on keywords at a meeting of the Keywords Project at Jesus College, Cambridge University. The Keywords Project uses Raymond Williams’ concept of keywords as being socially prominent words (e.g. art, industry, media or society) that are capable of bearing interlocking, yet sometimes contradictory contemporary meanings, and the group meets a couple of times each year to discuss new keywords that have emerged in society. The group carry out analysis using a variety of different methods, involving deriving etymologies from the Oxford English Dictionary, making use of Google n-grams, referring to academic research on particular concepts and investigating corpora.

I was invited to give an alternative (or rather, complementary) perspective that was more focussed on around corpus linguistics. I discussed how the concept of keywords differs greatly in CL, and how keyness can be extended to include tagged words, semantic or grammatical groups of words, multi-word units or even punctuation marks. Using various reference corpora, I showed how keyness techniques could be used to aid the identification of potential emerging keywords, while concordancing and collocational analysis could help to to identify the range of meanings around a word at a given point in time.

For more information, see http://keywords.pitt.edu/index.html

In the press: Google and the perpetuation of stereotypes

The findings of a paper published by myself and Amanda Potts on the implications of Google’s auto-complete search function have been reported in Mail Online and The Telegraph (18 May 2013).

The paper examined what happens when the beginnings of questions about different identity groups are entered into Google’s search form. For example, typing “why do black”, “do gay people” and “should jews” results in Google offering auto-complete suggestions which could be considered offensive or perpetuating stereotypes.


The paper’s aims were to raise questions about the appropriacy of such auto-completes but also to investigate which sorts of stereotyping questions tend to be associated with different identities. We categorised 2,690 such questions as they occurred across 12 social groups, finding that the groups with the most negative stereotypes associated with them were male, black and gay people.


Our paper does not argue that people reading such questions will automatically internalise such stereotypes (although younger or uncritical users of Google may do so, and people who hold those stereotypes may feel that they are validated) but we believe that there should be an option for certain suggestions to be flagged as offensive and removed or hidden if they reach a certain level of complaints, similar to YouTube’s commenting system.


Baker, P. & Potts, A. (2013) “Why do white people have thin lips?”: Google and the perpetuation of stereotypes via auto-complete search forms. Critical Discourse Studies. 10:2, p. 187-204.

Available here.