Attention to empathy (E) and compassion (C) in the health-related literature is burgeoning, a trend that can be expected to continue, given the determinant role relationships have been shown to play in the clinical healthcare picture as evidenced by Del Canale et al.’s landmark study in Parma, Italy linking physician empathy and positive clinical outcome for patients with diabetes mellitus [1]; and, Kelly, Kraft-Todd, Schapira, Kossovsky, and Riess’ meta-analysis of randomized control trials associating the quality of patient-clinician relationships to a statistically significant role in healthcare outcomes [2].

These studies suggest a need to look more closely at how E and C are represented in the health-related literature since empathic and compassionate relationships and communities of care hold such great promise for reducing burden on public health.

I performed 353 search queries in the PubMed database between February 1–5, 2018. I focused solely on PubMed because I was interested in identifying trends and gaps in the literature rather than reviewing individual publications, and regarded PubMed as a reliable representation of the health-related literature, a service of the National Library containing more than 28 million citations for biomedical literature.

Tables 1 and 2 display the distribution of results related to E and C in the literature.

Tables 3 and 4 show the distributions of noteworthy results specific to E and C within nursing and medicine.

Results suggest three main findings.

Finding 1 (semantic overlap) helps support the hypothesis in the light of the other findings. The numbers indicate that while C appears in the literature without E, there are no instances in the results where E appears without C. This suggests the distinction commonly made between E and C for theoretical purposes may not be as useful in practice as in theory.

Finding 1 is consistent with results presented by Sulzer, Feinstein, and Wendland [3] in their review of 109 studies aimed at assessing empathy development in medical education. The authors found that most of the studies they reviewed were characterized by internal inconsistencies and vagueness in the conceptualization of E, pointing out that the methods most commonly used to measure E relied on self-report and cognition divorced from action, and as a result may not be able to predict the presence or absence of E in clinical settings. Their conclusion suggests the need for humanized perspectives of E.

Finding 2 (bias toward psychology) confirms the hypothesis. It implies that the literature associates E and C with the psychological tradition of studying non-naturally occurring data (NND), data elicited with researcher interventions such as prompts, reports, and surveys as opposed to data gathered without researcher intervention as it naturally emerges in human speech communities, known as naturally occurring data (NOD).

Wolfson [4] and Beebe and Cummings [5] were first to highlight the limitations of NND for purposes of understanding discourse features critical to meaning and intention well enough to serve as a sound basis for second language acquisition education. Later, Félix-Brasdefer [6] showed that using role-play as a data elicitation method for requests in Mexican Spanish was insufficient because it did not generate certain types of requests revealed through an analysis of NOD.

As these studies suggest, NND serves many study purposes and provides easier and faster access to data without the collection and transcription challenges of NOD, but it is not a substitute for NOD when the study goal is to explain complex human behaviors in ways that are accurate and meaningful to real-world pragmatic understanding.

Finding 3 (lack of significant association in the literature between E and C and language/linguistics and other disciplines and topics crucial to a pragmatic perspective of E and C) is the most salient finding in the review and supports the hypothesis.

I was perplexed to find an insignificant association between E/C and language (5%/4% respectively of the total results returned) or E/C and linguistics (1%/1%) since modern linguistics promotes the view that language is social behavior [7], a view that is valid in the light of research based on NOD involving real people and authentically evolving relationships and social contexts and circumstances.

I was equally perplexed to discover an insignificant association in the results between language and emotions and E/C (1%/1% respectively) since I had uncovered evidence of this association in an early ethnographic discourse analysis study of expressions of disapproval [8] I conducted where I demonstrated how intimates commonly convey and respond to negative emotional meaning (criticism) using linguistic forms (syntactic and lexical-semantic surface structures) inconsistent with literal meaning, a finding I was able to quantify with the help of computer analysis at Yale University’s Haskins Laboratory: I showed how the differences in voice pitch and frequency (paralinguistic/prosodic language symbols) that I had found to be determinant to communicative function in interpersonal behavior were measurably distinct [9].

Finding 3 is consistent with Finding 2 (bias toward psychology) and the tradition in research psychology to study emotions apart from social context, which has led many eminent psychologists to see language as detached from emotions, where it is seen to play only a descriptive role [10–12].

The oversight suggested by the lack of significant association in the literature between E and C and topics important to pragmatic understanding such as relationship dynamics, conscious mind, rumination, self-talk, implicit bias, linguistic structures, resistant thought, and discrimination reinforce this objectified and dehumanized view of emotions and becomes especially meaningful in the light of philosopher and cognitive scientist Chalmers’ distinction between the easy and the hard problems of consciousness [13].

Chalmers explains how cognitive science is well suited to explaining the easy problems of consciousness, which he associates with cognitive abilities, functions, and behaviors that can be explained in terms of computational or neural mechanisms. He points out that the hard problem persists even when the performance of all relevant functions has been explained because it relates to the subjective way we process experience through perception.

Perception is critical to making sense of E and C from the perspective of communicative function and integral to the way we use language literally and symbolically to shape emotional values [14]. Yet, the topics mentioned above, which are critical to perception, were not significantly associated with E and C in the findings, further suggesting that E and C are understood from objectified and reductionist perspectives in the health-related literature.

The gap in pragmatic perspective revealed in this review suggests a need in the health-related literature on E and C to bridge theoretical understanding with the practical understanding needed for understanding how and why we shape and inhibit empathic and compassionate relationship experiences and communities of healthcare.

The disciplines and topics associated with E and C in the literature that point to this gap reflect perspectives that are too narrow to account for the communicative function of E and C, the complicated and nuanced ways we shape meaning and intention within our own minds that we discriminately share with different others, constrained by social variables such as relationship (e.g., spouse or supervisor) in the light of our immediate and longer-term motivations and goals.

There exists an opportunity to understand E and C as emotionally symbolic expressions of language, integral to communicative function within the larger pragmatic picture of interpersonal conflict and connection [14]. This larger pragmatic picture provides a potentially useful framework for connecting the dots between the communicative function of E and C and the plethora of topics critical to it that were disassociated from E and C in the literature such as bias, discrimination, frustration, resentment, resistant thought, self-talk, and social dynamics.

It is essential that we understand these topics in relation to E and C from a humanized perspective well enough to serve useful and positive ends in healthcare contexts if we are to reduce their burden on public health. For instance, we can benefit from understanding E and C in ways that enable us to actively mitigate bias and discrimination that is rooted in race and gender identity, which have been shown to adversely impact health outcomes at the population level [21–23].

Significantly, this paper reflects an understanding of the cultural construction of empathic and compassionate interpersonal relationships within rich and complicated contexts specific to American English speech communities, understanding essential to evolving effective public health policy and especially useful to an international audience interested in advancing cross-cultural comparisons to further a fuller understanding of E and C in the human community.

It is my sincere hope it will prompt unusual interdisciplinary collaborations and new lines of scientific inquiry related to the communicative function of E and C within and among divergent human speech communities, and ultimately lead to the advancement of the science informing health policy.

History teaches us that dramatic changes in how we understand the world and our place in it do not make their way easily into our minds and behaviors. Yet retrospectively, we can appreciate the benefits to scientific advancement of embracing complementary paradigms of understanding that offer potential for making sense of familiar topics and solving persistent challenges in novel ways.