Background - The Prime Machine

The Prime Machine
Go to content



The Prime Machine is a user-friendly corpus tool for English language teaching, linguistic analysis and self-tutoring based on the Lexical Priming theory of language; the software was developed by Stephen Jeaco.

My doctoral thesis explains some of the pedagogical rational for the development of the original 32 Bit Windows version of The Prime Machine, as well as some technical details and an evaluation.
Jeaco, S. (2015). The Prime Machine: a user-friendly corpus tool for English language teaching and self-tutoring based on the Lexical Priming theory of language. Unpublished Ph.D. dissertation, University of Liverpool.
Available for immediate download from:

From the very start, The Prime Machine was developed with language learners and teachers in mind.  You can use it to find examples of real language use (also called naturally occurring language).  It has also been further developed to offer a good range of tools for English majors and students studying linguistics or TESOL to use for corpus research projects.  The patterns and summary information of features of Lexical Priming can be used to compare specific instances of English language use (perhaps an expression in an English learner text or a test item, or creative uses of language in speeches, articles or literary extracts) against patterns representing a norm in one of the ready-made online corpora.  For example, a combination of words found in a single text outside the corpus can be compared with patterns of collocation and co-text of these words in a corpus of similar text types.   The “art” of a sentence from a novel not in the corpus can be explored to see how some patterns in the sentence follow conventions found in similar text types, while other features of the patterns deviate from norms.  Projects may also involve the construction of new corpora.  Typical DIY corpus projects may focus on the exploration of differences between genres/registers, the exploration of different authors’ styles as well as differences in English translations (corpus stylistics) or the exploration of vocabulary and language patterns for English language teaching materials for a specific subject discipline.  

Most of my publications can be accessed on a secure part of my personal website, under the publishers fair use and/or access policies. Some are author accepted versions, others are the final publications. You can access this part of the website here.

Dr Stephen Jeaco, Senior Associate Professor, Department of Applied Linguistics, Xi’an Jiaotong-Liverpool University, China.

Background to the software
Note on key word analysis
If you are doing corpus research using key words (keywords) you may be interested to know that my 2020 article on the use of log-likelihood and other measures for key word analysis –  Jeaco,  S. (2020) Key words when text forms the unit of study: Sizing up the effects of different measures – is now open access and is freely available from

The article explores key word metrics, considering different aims of different kinds of corpus work using keyness.  It engages with ideas raised by Gabrielatos (2018) and Gabrielatos & Marchi (2012), providing a rationale and investigation of the use of log-likelihood key words when exploring words which are key in texts and collections of texts in corpora.

Researchers using key words (Scott 1997) in tools such as WordSmith Tools (Scott, 2016), LancsBox (Brezina, McEnery & Wattam 2015), AntConc (Anthony, 2019), Sketch Engine (Kilgarriff, Rychly, Smrz & Tugwell 2004), and The Prime Machine (Jeaco, 2017) may find it helpful as a guide to the usefulness of Log-likelihood in key word analyses and discussion of the nature of effect size for word frequency data.

See the article for the references.
Back to content