Depression Language - Research

A new study found that people with depression are more likely than others to use the words "always," "nothing," and "absolutely." Depression manifests itself in everything: in the way you move and sleep, in the way you communicate with the people around you. It can even be seen in the way you speak or express your thoughts in writing. Sometimes this "language of depression" has a strong impact on those around you. Suffice it to recall the impact of the poems and songs of Sylvia Plath and Kurt Cobain - after all, the suicide of both of them was associated with the mental problems they were experiencing

Scientists have long tried to find exact matches between depression and language, and now they have been able to get a better understanding of this with technology. A new study published in Clinical Psychological Science has found an entire class of words that can be used to predict depression with certainty.

Traditionally, linguistic analysis in this area was carried out by researchers who read the written texts and took notes. Modern methods of computerized text analysis make it possible to process huge databases in a matter of minutes. This helps to draw attention to language characteristics that may go unnoticed by a person. The computer calculates the percentage of prevailing words and word classes, lexical diversity, average sentence length, grammatical patterns used, and other parameters.

So far, researchers have studied the personal sketches and diaries of depression sufferers, as well as the writings of famous musicians such as Cobain and Plath. In terms of spoken language, studying snippets of natural language from people with depression also helped to gain some understanding of it. Based on the combined analysis of these two types of linguistic activities, scientists were able to identify clear and consistently observed differences in language use between those who have depression and those who do not.

Content

Language can be divided into two components: content and style. Content is what we express, that is, the meaning or theme of the statements. It is not surprising that people with depression use a large number of words to describe negative emotions, especially negative adjectives such as "lonely", "sad", "miserable", and related adverbs.

More interesting is their use of pronouns. People with depression use significantly more first person singular pronouns - "I", "me", "me" - and significantly fewer second and third person pronouns such as "you", "she", "they". This pronoun pattern suggests that people with depression are more self-centered and less connected to others. For identifying depression, the researchers believe it is more important to pay attention to the use of pronouns rather than words that express negative emotions.

It is known that fixation on one's own problems and social isolation are common signs of depression. However, we do not know if they reflect the attentional or thought patterns of people with depression. Does depression cause people to be self-centered, or do self-centered people develop symptoms of depression?

Style

Language style is how we express ourselves as opposed to what we talk about. Our lab recently conducted textual analysis of big data collected from 64 online forums on mental health. Samples of written speech from over 6400 participants in these forums were analyzed. "Absolutist words" that contain absolute values or probabilities such as "always", "nothing" and "absolutely", in the analysis of communication on forums, were more accurate markers of depression than the use of pronouns or words expressing negative emotions.

We assumed from the very beginning that people with depression would have a more “black and white” view of the world and this would somehow manifest itself in their language. Compared to 19 control forums (such as Mumsnet and StudentRoom), the prevalence of absolutist words is about 50% higher in forums for people with anxiety and depression, and about 80% higher in forums for people with suicidal tendencies.

The use of pronouns showed a distribution similar to the use of absolutist words in different forums, but with less differences. But words expressing negative emotions, paradoxically, are less prevalent in forums of people with suicidal tendencies than in forums of people with anxiety and depression.

Our study also included recovered forums, where participants feel recovered from an episode of depression. They write positive and encouraging posts about their healing. Here we found that words expressing negative emotions are used in a rate comparable to that of the control forums, and words expressing positive emotions are used about 70% more. However, the predominance of absolutist words remains quite noticeable when compared to control forums, although slightly lower than in forums of people with anxiety and depression.

Of fundamental importance is the fact that if a person has previously had symptoms of depression, there is a greater likelihood that they will recur. Consequently, the fact that such people, even in the absence of symptoms of depression, have a stronger tendency towards absolutist thinking indicates it as one of the possible factors leading to episodes of depression. A similar effect is observed for the use of pronouns, but not for the use of words expressing negative emotions.

Practical conclusions

Understanding the language of depression not only can help us understand the way people with depression think, but it also leads to some practical conclusions. Researchers combine automatic text analysis with machine learning (where computers can learn from experience without special programming) to classify a variety of mental health conditions based on natural language text patterns such as blog posts.

This classification is already proving to be of better quality than that carried out by professional therapists. Importantly, machine learning-based classification will become more sophisticated as more data accumulates and better algorithms are developed. This means that the analysis will go beyond the above considered analysis of the patterns of using absolutist words, words expressing negative emotions, and pronouns. Work has already begun on using computers to accurately identify increasingly specific subcategories of mental health problems, such as perfectionism, self-esteem problems and social anxiety.

This, of course, remains the possibility that the person uses the language associated with depression without being depressed. Ultimately, it's only how you feel over time that tells you whether or not you have depression. But since the World Health Organization currently estimates that more than 300 million people worldwide are living with depression, with an 18 percent increase in the incidence since 2005, the discovery of more tools for making such a diagnosis is certainly very important for improving public health and the prevention of tragic suicides such as those committed by Plath and Cobain.