Having the speeches in digital format allows us to explore them in new way: below are a just few. If you’re interested in exploring further, we’ve made the data available both as raw text and via machine-readable APIs.
Looking at word frequency – how often a word is used in speeches for a given year – can give us some insights into the concerns of the time. Examples include: tariff, wool, coal, war, vietnam, communist, terrorism, internet, ethnic, me, boats.
The complexity of language, average sentence length and the length of the speeches themselves tell us about styles of speech, both of the individual speakers and the time in which they spoke. Interestingly, the last two elections have used some of the simplest language of all. Does this indicate that the electorate is becoming less literate – or is it that contemporary politicians are trying to present their ideas in simpler, more accessible ways? Is this a good or a bad thing for our democracy?
All of the data in this archive is also available through a realtime API using the JSON data format: