Schizophrenia is a mental disorder that impairs accurate perception and manifests through negative and positive symptoms. "Word salad" is Schizophrenia disorganized speech, where communication becomes jumbled and incomprehensible. Advances in Natural Language Processing (NLP) and artificial intelligence can analyze individuals' schizophrenia speech, offering insights into their thought disorder. NLP research shows promise in quantifying the magnitude of disordered thought by measuring language complexity, coherence, and the ability to maintain topics in conversations.
Whenever we decide we want to say something our brains engage in a complex process of formulating the message. The brain must select the right words to convey that message, and assemble the words in strings that represent our message. This process involves higher order parts of the brain, such as the language areas in the frontal and temporal lobes; as well as structures that mediate memory and emotion, such as the hippocampus and amygdala. Even before a word is uttered, the brain is busy formulating the message to be spoken.
The messages we convey, and the way we convey them, reveal much about how we perceive ourselves and the world around us. Nowhere is this more evident than for people who suffer debilitating disorders of thought, such as schizophrenia.
Schizophrenia robs people of their ability to accurately perceive themselves, others, and the world around them. It is a serious mental disorder that can manifest with negative symptoms, like withdrawing from social interactions, not maintaining personal hygiene, and an inability to experience pleasure. And it also can manifest as positive - in the additive sense - symptoms, including delusions, hallucinations, abnormal voluntary movements, and disorganized thought. This last symptom, disorganized thought, is evident in the speech of some individuals with schizophrenia, which is colloquially referred to as “word salad.”
The term “word salad” conjures up images of discrete words tossed randomly together that carry no larger meaning or clear message. In addition to reflecting a disorder of thought, the jumbled speech is incomprehensible to others, preventing one from effectively communicating and further reducing quality of life.
Analyzing the speech of individuals with schizophrenia offers a window into the debilitating disorder of thought they experience. Recent advances in Natural Language Processing and artificial intelligence are beginning to enable speech analytics to quantify the magnitude of disordered thought. Natural language processing, or NLP, is a branch of computer science that leverages computational linguistics– or the rule-based modeling and statistical probabilities of language– and machine-learning technology. For example, imagine a clinical tool that could help those at risk predict an oncoming worsening of their disorder–such as a psychotic break from reality–or improvements in thought related to treatment.
To date, NLP research has shown promise in measuring the complexity and coherence of the language produced, especially in a social context such as conversation. Engaging in conversation requires the ability to maintain a topic of discussion without randomly jumping to other topics, and respond in appropriate ways to the comments and questions of the conversational partner. NLP analytics can characterize the relevance of responses to questions, how well topics are maintained, and the overall coherence of the conversation.