Language Unlocked

Our methods

Our methods

We use corpus linguistic and stylistic methods to produce objective analyses of how your organisation is represented in texts. To do this, we construct large databases of text (referred to as corpora) and subject these to rigorous computational analysis.
Our analyses are informed by the latest linguistic theories and insights into the construction of meaning.

We use a variety of analytical tools to unlock hidden meanings in texts. Some of these tools are explained below, along with some examples of how such analyses can be practically useful for organisations.

frequencyFrequency analysis

Frequency analysis offers an insight into how often particular words are used in a data set.

This simple statistical measure can be indicative of the overriding concerns expressed in a text. Frequency analysis can be used to investigate lexical change across time or differences between texts.
The results of frequency analysis can help clients see clearly how their organisations are presented and perceived.

For example, the graph below shows the changing frequency of the word choice in party political manifestos between 1945 and 2010:



The best way to carry out closer analysis of a keyword, or any word in a corpus, is to investigate its meaning in context. To do this we use software to generate concordance lines. A concordance is a list of all the sentences in which our target word occurs.

Concordances are helpful because they can allow us to see patterns of language use. For example, opposite is an extract from a concordance of the word slim (excluding metaphorical usages), taken from a corpus of British English:
It is apparent from reading the corpus lines that slim is an adjective used more often to describe women than men. It is easy to envisage that this information could be useful to an organisation aiming to encourage weight loss in men, and may consequently lead to alternative marketing strategies focusing on, for instance, fitness rather than appearance.

Following keyword analysis of your data, we focus on detailed analysis of concordances in order to provide you with a clear sense of common patterns of usage which you may wish either to emphasise or break away from.

Keyword analysis

KeywordingFrequency analysis is useful for investigating the occurrence of a particular word that we already think might be interpretatively revealing. However, if we have no prior expectations about which words are likely to be informative, then we can use keyword analysis to discover the statistically significant words in a corpus. Keywords are all the words that occur more (or less) frequently in one corpus than they do by comparison with another corpus.

For example, in a corpus of texts relating to the national day of action on November 30th 2011 collected from various unions (including press releases and information issued to members), the keywords are pensions, action, members, justice, ballot, government, consultation, support, campaign, negotiations and guidance. However, in a collection of news stories in the British press covering the day of action, the keywords are strike, strikes, walkout, chaos, anger, disrupted, protestors and taxpayers.

Read more about our methods by downloading our brochure or contact us for further information.