Short answers, pulled from the story.
Margaret Masterman (the 4th of May 1910 - the 1st of April 1986) was a British linguist and philosopher known for pioneering work in computational linguistics and machine translation. She founded the Cambridge Language Research Unit in 1955 and is recognised as having anticipated many ideas that later became standard in artificial intelligence and natural language processing.
The Cambridge Language Research Unit (CLRU) was a research centre Masterman founded in 1955, growing from an informal discussion group started in 1953. Located in a former museum called Adie's Museum in Cambridge, it operated outside the official university structure for roughly twenty years, funded by US agencies including AFOSR, ONR, and NSF, as well as UK government bodies. Its staff never exceeded ten people.
Masterman attended Wittgenstein's lectures at Cambridge and was one of the few students selected to take notes during his 1933-34 course. Her notes and those of others were compiled as The Blue Book. She also co-authored the Yellow Book with Alice Ambrose, containing notes taken during pauses in the Blue Book dictation.
At the Fourth International Colloquium in the Philosophy of Science in London in 1965, Masterman argued that Kuhn used the concept of paradigm in at least twenty-one different senses, which she grouped into metaparadigms, sociological paradigms, and artefact or construct paradigms. Kuhn accepted this criticism, and it was a key factor in his shift from the term paradigm toward the concept of incommensurability.
When Masterman died on the 1st of April 1986, William Williams shut down the CLRU. Its library of early machine translation documents was thrown into a skip, despite two university bodies having offered to house it. Serious research at the unit had already effectively stopped around 1978.
Masterman proposed that sentences could be split into breath group segments, each carrying a distinct meaning. Translation would work by converting each segment independently into the target language and then reconstructing the sentence. She attempted to implement this approach after 1980 using two North Star Horizon computers programmed with the Forth language, contrasting with the dictionary and rule-based Systran system then dominant in the field.