Empathy, or understanding another’s thoughts and feelings, is a quintessential construct in society. How accurately one empathizes (i.e., empathic accuracy; EA) relates to well-being. EA is especially important in romantic relationships to maintain understanding and closeness. One construct crucial to both romantic relationships and EA is language because it fosters communication. However, specific word use categories have never been empirically related to EA. First-person plural pronoun use (we-talk) and language style matching (LSM) should reflect the cognitive and emotional components of EA and also covary with well-being. Therefore, this dissertation associated we-talk and LSM with EA, and all of these constructs to well-being. Seventy-seven committed romantic couples were assessed through retrospective and momentary self-report as well as the Electronically Activated Recorder (EAR) to record everyday word use. Chapter I explains the conceptual parallels between word use and EA. Chapter II focuses on we-talk, EA, and well-being at the individual level. Own we-talk negatively related to own EA, while partner’s we-talk positively related to EA. Further, negative EA (i.e., correctly empathizing with partner’s negative emotions) was the strongest predictor of well-being, though this was also found for positive EA. Chapter III focuses on LSM, EA, and well-being at the couple level. The most likely association was between cognitive LSM (matching cognition words such as pronouns) and negative EA. Negative emotion LSM (matching negative emotion words) positively related with well-being, along with both positive and negative EA. Again, negative EA most strongly related to well-being. Relations between word use and EA were partially supported directing research toward better understanding under what conditions demonstrate this association. Importantly, the current study was naturalistic, meaning participants engaged in any conversations they desired. However, the topic of conversation can change word use and its subsequent relation with EA. This was the first study to attempt relating word use and EA and doing so in a naturalistic setting. As such, more work is needed to further guide under what contexts and conditions it occurs and its relation to downstream well-being. Empathy permeates most any social context and further research on it can have widespread, and positive, consequences.

This dissertation investigated the language individuals use to describe their pain. First, a novel tool was developed and tested that captured and categorized pain words. There were three theoretical pain word categories: sensory pain words (e.g., pain, hurt), affective pain words (e.g., ouch, excruciating), and medical words (e.g., arthritis, ibuprofen). Next, the tool was used to examine naturally occurring pain disclosure in a group of individuals living with rheumatoid arthritis (RA) and a group of individuals living with breast cancer (BC). This study found that pain words accounted for 1% of words spoken in the RA sample and 0.7% in the BC sample. In addition, the study found that RA and BC participants disclosed pain to their significant others, as well as to many other people in their social networks such as family and friends. Next, the tool showed that in online forums designated for discussion of specific illnesses characterized by pain, pain disclosure seemed to elicit greater support. This was observed in a greater number of replies and better engagement for initial posts that mentioned pain compared to initial posts that did not mention pain. Last, a study examined expressive writing about the most painful experience of individuals who were currently suffering from pain and those who reported no current pain. For the current pain group, a small stable relationship emerged, where greater use of affective pain words related to worse psychosocial outcomes. In addition, those in pain mentioned a greater loss of social support, described fewer instances of support, and more instances of having their pain dismissed than the non-current pain group. This dissertation found across three diverse studies that sensory pain words were used more than twice as often as affective pain words, yet affective pain words were tied to more outcomes. Pain is a continuous subjective experience and the findings in this dissertation indicate the specific words used to describe pain are associated with important psychosocial outcomes.

Rheumatoid arthritis is a chronic autoimmune disease that does not have a cure. Therefore, it is important for patients to receive support, which would allow them to ask questions and express their feelings. This study examined online social networks for patients with rheumatoid arthritis to better understand the emotional valence of their initial posts and whether there was an association between posts with negative emotional valence and requesting support/information. We hypothesized that the majority (more than 50%) of the emotional valence of initial posts would be negative, and that there would be an association between negative emotional valence and support/information. Nine hundred eighty-six initial posts from a rheumatoid arthritis online social network via Reddit were coded as either positive, negative, neutral, or mixed. In addition, the initial posts were coded as either requesting support/information, offering support/information, neither requesting nor offering support/information, or both requesting and offering support/information. Negative was the most common emotional valence in the initial posts followed by mixed, neutral, and positive. There was also an association between initial posts that had a negative emotional valence and requested support/information, and initial posts that had a negative emotional valence but did not request support/information. As a result, the implications of this study indicate the need for additional information and support to be provided to patients with rheumatoid arthritis, so they can have a better experience and an easier way to cope with their illness.