Here's an analysis of the 'fake news' frequency data:

1. Data: The normalized frequency of "fake news" was consistently low from 2010 to 2015. A dramatic spike occurred in 2016, followed by peak usage in 2017 and 2018. Subsequently, the frequency decreased steadily from 2019 to 2024, with a slight uptick in 2025, but still remaining substantially higher than pre-2016 levels.

2. Possible explanation: The surge in "fake news" usage from 2016 to 2018 strongly suggests a direct correlation with a specific event or period that brought the concept to the forefront of public discourse. The 2016 US Presidential election and its aftermath, characterized by heightened media scrutiny and accusations of misinformation, likely triggered the initial increase. The continued high usage in 2017 and 2018 indicates sustained public and media focus on the issue of fabricated or misleading news. The subsequent decline from 2019 onwards may reflect a gradual decrease in the novelty or perceived urgency of the term, potentially due to desensitization or the emergence of alternative framings of the same concept.

3. Competing constructions: Prior to 2016, when the frequency of 'fake news' was low, competing constructions likely included more general terms related to deception and inaccurate reporting. These could include "false information," "misinformation," "rumors," "propaganda," "biased reporting," or simply "lies." These terms cover the same semantic space, referring to untrue or misleading statements.