The Geriatric Medicine Research Unit at Dalhousie University, is very interested in the way that spontaneous languages changes over the course of Alzheimer's disease , and how it responds to treatment. We have worked on this with a linguist, Dr. Elissa Asp, of St. Mary's University in Halifax. The team have found that for many patients, there are distinct patterns in how language is lost, and how it responds to treatment.
One pattern that I have seen is the person who has problems finding the right word, and substitutes another word for what is lost - either a word that sounds like the lost word (e.g. 'purdy' for 'purse') or by describing the function of the lost word (e.g. 'that thing, you know, that thing that women have for their money and stuff'). Sometimes, in attempting to describe the function of the word that is lost, the person gets sidetracked, in searching again for the words to describe the (now forgotten) lost word that started their search. This is so characteristic that I often point it out to students as a way that they can recognize what Alzheimer's disease sounds like.
This approach to observing patterns of spontaneous speech in Alzheimer's disease and its treatment is described in Asp et al., Self-referential tags in the discourse of people with Alzheimer's disease: discourse correlates of a compensatory network? Brain Lang 2005; Sep 5.