Sounds, words, and sentences are the building blocks of language, but most communication involves longer stretches of speech. People take turns in a conversation or argument, hold the floor for perhaps an hour during a classroom lecture or sermon, give directions on how to drive to a point of interest or tie a pair of shoes, perform a marriage ceremony or give a eulogy, or tell a story about what happened on the way to St. Ives. All of these speech events are forms of discourse, involving a level of organization above that required to formulate or comprehend a single word or sentence. Some patients with Frontotemporal degeneration (FTD) may be able to speak in a fashion that seems normal at first blush, yet if they suffer from impairments of organizational skills or attention, they may also be impaired in the ability to communicate effectively.
The Penn FTD Center conducts several studies that investigate discourse processing in patients with FTD and healthy adults. In order to engage in discourse, individuals must use language mechanisms that support the semantic knowledge of single words and the grammatical processing of sentences. In addition, speakers use organizational skills to organize sentences into a narrative.
One way in which FTD patients may be impaired is in the ability to organize a discourse into a beginning, middle, and end. For example, in telling a story with a picture book as a guide, FTD patients frequently lose the thread of the story. As the story progresses, they may forget what is the point of the story, so when they arrive at the climax, they do not recognize it as such.
Executive mechanisms that are involved in working memory are important for maintaining the thread of a conversation or story. Thus another characteristic of impaired story-telling, when “reading” a story from a wordless picture book, is that patients may describe the scenes one by one as they occur but fail to relate each picture to the overall purpose of the story. The result is that the story is not told as a coherent narrative, but is only a series of unrelated pictures. We see corresponding changes that limit narrative comprehension in FTD.
Finally, we hypothesize that cognitive mechanisms beyond language processing are involved in discourse. Social mechanisms, such as those that support taking the perspective of another individual, are important for tailoring one’s speech to the audience. For example, someone might tell a story differently to their doctor than they would to a friend. Individuals with FTD do not appear to recognize these differences in social context, and have limited flexibility to express these differences. Similarly, FTD patients may exhibit deficits of attention. They may have difficulty focussing on the features of a picture that are relevant to the story and become distracted by irrelevant elements. In conversation, they may not keep to the subject at hand, but rather disregard the subject and begin speaking about a tangential matter. Such behavior is not reflective of a problem with language per se, but it is seriously disruptive to the act of communicating.
Investigators in the Penn FTD Center test these hypotheses about discourse in an effort to inform cognitive neuroscience models of natural communication. This allows us to develop quantitative and objective measures of the communication deficits observed in FTD. Finally, we relate these deficits to changes in brain structure and function. Patient studies have demonstrated that discourse is supported by a large-scale network involving multiple brain regions. For example, behavioral variant frontotemporal dementia (bvFTD) patients have intact language, but their executive and organizational limitations impair their ability to organize a story and engage in a conversation. We have used structural and functional neuroimaging to relate discourse deficits to a large-scale network involving disease in frontal and parietal cortex. To provide cross-validation of this neuroanatomic model, we have performed fMRI studies of discourse in healthy adults. These show recruitment of similar brain regions when healthy adults understand and tell a story.