Character in the Corpus: Empathy

Identifying empathy

Empathy is a staple of both moral education and literary studies. It both fits an Aristotelian prescription that links virtue and emotion (Battaly, 2008) and appeals to cognitive appreciations that treat literary texts as peculiarly experiential artefacts. As a result, the literary study of empathy has often drawn on thinking in moral education and vice versa.

Their common ground is most evident in their empirical scrutiny. Studies have ranged from looking at the specific linguistic devices that cue empathy (Van Lissa et al., 2016) to the types of literary text that generate empathy (Kuzmičová et al. 2017). One study found that readers positioned themselves according to characters’ mental states and their moral evaluation of them, meaning “empathy is morally sensitive” (Fernandez-Quintanilla, 2020: 140). This suggests lucrative potential for the A-level classroom and accordingly the associated touchstone discussions focus on the ways in which authors represent their characters’ mental states and moral standpoints.

Questions of whether readers and writers can appropriately empathise however have informed both empirical and theoretical studies. One of the limits to empathy noted in the empirical research is its biases, including that of proximity; whereby we more readily empathise with people who are like us. American writer Eudora Welty makes the claim for writer empathy:

What I do in writing of any character is to try to enter into the mind, heart, and skin of a human being who is not myself. Whether this happens to be a man or a woman, old or young, with skin black or white, the primary challenge lies in making the jump itself. It is the act of a writer’s imagination that I set most high. (1983: xi)

Although Zadie Smith argues “fiction had empathy in its DNA, that it was the product of compassion” (2019), the shortcomings of empathy in fiction seem to stem from its potential to be vicarious without being virtuous. A commonplace of literary criticism is that villainy has a particular attraction for readers and writers. Examples, like Blake’s (1790) assessment that Milton’s Satan in Paradise Lost had the best lines, show the hazardous seductiveness of rhetoric and how language is at the heart of the debate as to whether reading can enrich or corrupt. Just because language is compelling, does not necessarily mean that the ideas it conveys are too.

For a complex character like Milton’s Satan, such linguistic acrobatics force readers to negotiate carefully between clashing virtues (e.g., in Paradise Lost, humility and courage, service and wisdom) to carefully develop their understanding of those virtues. As is illustrated in the touchstone passages, literature is populated with villains and unreliable narrators who can be rumbled through attending to modes of discourse presentation, which thereby develop key literacy skills involved in comprehension.

The question, then, is empathy virtuous or simply vicarious? Nussbaum (2001) claims that literary fiction’s ability to transport readers is uniquely placed to eradicate biases associated with empathy as well as posits that perspective taking in itself enables readers to adopt appropriate moral action, and this brings us to a further empirical finding of relevance to virtue literacy. Significantly, the prosocial effects of empathy strongly correlate with supported learning, whether that be from parenting (Konrath et al., 2015) or from teachers (Keen, 2013).

Defining empathy

Philosophical accounts of empathy tend to shift semantic furniture, defining empathy according to previous research, with an extensive consideration of its definition, historical provenance and use across various academic fields. But what if a definition were to be based on its actual use amongst speakers at large; a descriptive, rather than prescriptive, definition?

Corpora enable a word to be analysed by its occurrence in context: whether that context be genres, speakers, geographies, historical periods, or even co-text (the linguistic company a particular word keeps). Such an examination of language also tells us something of individuals’ mental associations between concepts. The corpus thereby becomes a tool by which one can both statistically interrogate language usage, and, by implication, the way in which concepts (such as empathy) are understood.

When looked at over time, empathy’s cognates show a shifting usage (and therefore understanding) of the term. From the noun descends its adjectival form empathetic (and the later empathic), suggesting its growing use in moral evaluation. Next comes its verb form, to empathise, suggesting an evolving understanding of empathy as a process in which individual agents participate. This finally evolves into the notion of the empath, also a noun, now not abstracted but fully personalised. This evolution tracks how empathy has been understood, from abstraction through evaluation to practice and, finally, character.

A further definition may be discerned by examining British and US English. In British English empathy collocates with with, in American English with for. But what’s in a preposition? Prepositions are after all functional rather than semantic. Yet this distinction belies grammatical words’ semantic role because they are the way that language expresses the (often logical) relationship between words; they glue bits of the language together. Thus, there is a difference in connotation between the two: the British usage suggests that empathy is aligned to another individual, whereas the American usage suggests something more akin to sympathy.

A final key feature of empathy is its prosody: the negative or positive contexts in which a word is used. Empathy is frequently negative, collocating with ‘lack of’ – it is perhaps not surprising why critics spill so much ink trying to define it. This lack is further evident in its exclusive appearance in academic discourse up until the 1980s (British National Corpus). Indeed, it remains most present in academic discourse (3.91 words per million) with its lowest usage in speech (1 in a million). From a virtue perspective, this lack suggests a deficit that virtue literacy can remedy, and in which fiction can play a leading role.

Generating empathy

Much of the research into narrative fiction investigates what textual features evoke empathy within the reading experience. Various stylistic features have been proposed as empathy-inducing. Keen (2017) notes that these features principally attach to characterisation processes (they include naming, description, traits, types, narrative role, quality of speech, and mode of consciousness presentation), narrative situation (point-of-view, narrator relationship, and internal/external presentation) as well as broader, discourse-level features (length of novel, genre expectations, vivid settings, metanarrative, serialisation and slowed pace). These, along with other tropes, provides an inventory of tools, many of which overlap with the techniques studied in the A-level classroom.

This is a crucial point, owing to the fact that empathy itself does not appear that frequently in the A-level corpus: another approach is required. Considering empathy as ‘fellow feeling’ a corpus trawl of the word feel shows extensive usage in the corpus, far outstripping any of the virtue terms considered in this study. It shows authorial usage, being a preferred word of choice for Jane Austen, Theodore Dreiser, and Richard Wright. That their novels are particularly character centred within the context of worldly social commentary, suggests that feeling is being used to effect empathy. Because of the projecting properties of words like feel, special attention is given in the associated touchstones to verba sentiendi, those words that mark character perception.

Indeed, fiction uses feel and its cognates much more frequently than general language users do, therefore lucrative space in terms of talking about the emotions and thereby emotional approaches to virtue. The twenty-first century surge further shows promise in that general usage is catching up, that is, with respect to feelings at least, there is a growing common ground between common language and that of fiction.

Thus, while empathy is also not a term widely attested in fiction this does not necessarily present a problem with using literature to consider empathy. In line with the empirical research, empathy’s power lies in its function as a tool by which to frame classroom discussions of text and ultimately considers empathy, whether a virtue in itself or not, a skill that can enable virtuous behaviours.

Comments are closed.

Blog at

Up ↑

%d bloggers like this: