Microaggressions in Context: Linguistic and Pragmatic Perspectives

Microaggressions are conversation turns directed toward another person that cause harm. Although initially identified several decades ago (Pierce, 1974), there has been a tremendous upsurge in research on microaggressions over the past 15 years. Much of this work has been devoted to classifying types of microaggressions and exploring their short- and long-term effects on recipients. Somewhat surprisingly, research on microaggressions from a linguistic and, in particular, pragmatic perspective has been rare. This is surprising because (most) microaggressions are meaningful communicative acts occurring in a particular context, in effect the domain of pragmatics. In this commentary I argue that there is much to be gained by considering microaggressions from a linguistic-pragmatic perspective. More specifically, I conceptualize microaggressions as nonconventional indirect speech acts and consider the role of context and intentionality in their use.

Pragmatics is a subarea of linguistics concerned with the examination of natural language as it is used in context. That is, rather than considering sentences in isolation, pragmatics scholars focus on utterances in the interactional contexts within which they occur and how that context plays a role in the meaning that interactants convey. Hence, pragmatics researchers examine how features of the context (broadly construed) contribute to how people convey and comprehend meaning when they use language. Recent advances in the field include the development of experimental methods, including electrophysiological and imaging techniques, for exploring how this occurs (e.g., Noveck, 2018). As such, the field has much to contribute to our understanding of microaggressions.

A particularly thorny issue for research on microaggressions has been definitional. This was noted by Lilienfeld (2017). What exactly constitutes a microaggression? Sue et al. (2007) defined microaggressions as “brief and commonplace daily verbal, behavioral, or environmental indignities, whether intentional or unintentional, that communicate hostile, derogatory, or negative racial slights and insults toward people of color” (p. 271). More recently Sue and Spanierman (2020) described them as “brief, everyday exchanges that send denigrating messages to certain individuals because of their group membership (e.g., people of color, women, or LGBTQ persons)” (p. 36). In Sue’s formulation, microaggressions are divided into three types: microassaults, microinsults, and microinvalidations. My focus here is primarily on microinsults and microinvalidations; microassaults are less ambiguous, and their role has been deemphasized in recent formulations (e.g., Sue & Spanierman, 2020; Williams et al., 2021). Moreover, Freeman and Stewart (2021) argued that the inclusion of microassaults as a type of microaggression undermines and weakens the very concept of microaggressions.

Critics of the microaggression concept, Lilienfeld (2017, 2020) in particular, have argued that the concept is defined too broadly to be scientifically useful. Specifically, Lilienfeld argued that the definition is so loose as to allow any behavior that someone finds offensive to be considered (by someone) a microaggression. What is needed, according to Lilienfeld, is a tighter, operational definition that specifies in advance when and how an utterance will be identified as a microaggression by a substantial number of people. Such specification is currently lacking. The debate regarding the operational definition of microaggressions involves multiple issues (see, in particular, Mekawi & Todd, 2021), some of which I argue can be addressed by adopting an experimental pragmatics approach.

The first thing to note about microaggressions, specifically instances of microinsults and invalidations, is that they are inherently ambiguous. In pragmatic terms, they are indirect utterances, and as such, the aggressive meaning must be derived in some manner. Many classic pragmatic theories (e.g., Grice, 1989; Searle, 1979), as well as related psychological approaches (e.g., Clark, 1985, 1996), distinguish between different levels of utterance meaning. Direct (or literal) meaning refers to the meaning of an utterance apart from the context in which it is used, or what Clark (1985) referred to as sentence meaning. To use a classic example, the direct meaning of “Can you pass the salt?” is a request for information regarding the hearer’s ability to pass the salt. When uttered by a speaker at the dinner table who is some distance from the saltshaker, the indirect (or speaker) meaning is roughly “Please pass the salt.” Applying this concept to a sample microaggression such as “You are so articulate,” the direct meaning of this utterance is an assertion that the recipient is able to communicate their ideas fluently and coherently. It is the indirect (or speaker) meaning, then, that constitutes the microaggression, and this meaning must be derived. Note that indirect meaning is not rare but rather quite common and in fact pervades everyday language use. And it occurs for multiple reasons, including efficiency (Levinson, 2000), manipulation (Pinker et al., 2008), and politeness (Brown & Levinson, 1987). Metaphors, bribes, polite requests, and so on are all instances of indirect meaning. Microaggressions can be added to that list.

However, not all indirect meanings are equivalent. They vary in terms of how they are performed, the underlying motivation for their use, how they are processed, and their ultimate effects. In this regard, one important distinction has been made between generalized and particularized implicatures (i.e., inferences). Generalized implicatures are highly conventionalized, and the indirect meaning is obvious; often it is the only meaning that is activated. Idioms are a classic example. When someone says “He let the cat out of the bag,” the hearer recognizes automatically and instantaneously that the speaker means that someone revealed a secret and not that someone released a cat from a bag (Gibbs, 1980; Keysar, 1989). Likewise, “Can you pass the salt?” is highly conventionalized, and most recipients recognize immediately that the speaker is requesting the salt rather than asking about one’s ability to pass it.

Particularized implicatures, on the other hand, are more ambiguous and typically do not have a conventional meaning associated with their use. They are nonconventional indirect speech acts. For example, indirect replies (i.e., not directly answering a request for an opinion) can be used to indirectly convey a negative opinion. When John asks Mark what he thought of his presentation and Mark replies “It’s hard to give a good presentation,” John will likely infer that Mark’s opinion is negative. Here the interpretation must be inferred; there is nothing inherently negative in the meaning of the utterance. It is only in the context of the preceding question that the reply can be considered to convey a negative meaning. Research suggests that these meanings are routinely recognized and require effortful processing to do so (Holtgraves, 1998, 1999; Bašnáková et al., 2014).

Most of the sample microinsults and microinvalidations in the literature would be regarded as particularized implicatures and hence nonconventional indirect speech acts. For example, there is nothing inherent in the meaning of “You are so articulate” that is insulting. Context is critical for the interpretation of microaggressions, as all researchers of microaggressions would acknowledge. What is missing is a systematic exploration of how context plays a role in the interpretation of microaggressions.

Specifically, what are the processes through which recipients generate an aggressive interpretation of a (potential) microaggression? Are there any systematic means for identifying nonliteral meanings as a function of context? Relevance theory may be particularly useful in this regard (Sperber & Wilson, 1986, 2002). In this approach, all utterances are assumed to come with a presumption of relevance. That is, hearers assume that a speaker’s remark is in some way relevant for the current verbal exchange, and relevance is defined explicitly in terms of processing considerations. Specifically, an utterance is relevant to an individual “to the extent that the contextual effects achieved when it is optimally processed are large” and “to the extent that the effort required to process it are small” (Sperber & Wilson, 1986, p. 145). In general, then, people are assumed to search for the first interpretation (least effort) consistent with the principle of relevance (maximum contextual effects). Consider, for example, how ironic sarcasm can be conveyed in this manner (Wilson & Sperber, 2012). A speaker who says “Lovely weather we’re having” in the context of a downpour typically intends to convey the opposite of the literal meaning of the utterance. The hearer assumes the speaker’s remark is relevant in some manner, even in a situation that directly contradicts its literal meaning. As a result, the hearer infers, on the basis of the utterance in context, that the speaker actually means the opposite of what they are saying literally.

One of the means for achieving relevance is to attempt to understand the reason for why a speaker is uttering a particular remark. That is, particularized implicatures are triggered by the context in some manner, and their occurrence can often prompt the recipient to search for a reason why the speaker is saying what they are saying (Holtgraves, 1998). For example, a likely—and possibly only—reason for saying “It’s lovely weather” during a downpour is to convey one’s observation that the weather is lousy. It is reasonable to assume that such processes will play a role in the recognition of microaggressions. For example, interpreting the utterance “You are so articulate” as a microaggression requires recognition by the hearer that the literal meaning is not maximally relevant; to make it relevant requires the hearer to infer that the speaker is surprised about their level of articulateness. Importantly, those who have been victims of discrimination or marginalization are more likely to be primed to detect microaggressions in just this way. Thus, the first interpretation with maximum contextual effects for them likely would be that the remark is insulting.

To take another example, consider the meaning of the phrase “Black Lives Matter.” As I have described elsewhere (Holtgraves, 2020), the intended meaning of this phrase depends on a recognition of its meaning in context. On its surface (i.e., its literal meaning) the phrase is a truism and hence conveys little information. To make it maximally relevant, one needs to recognize the context of the phrase, which is to call attention to the police killing of African Americans, in effect a reminder that the lives of Black people matter, too, just as much as the lives of other people.

There are multiple features of the context that play a role in comprehension, including the roles and statuses of the interactants. Much research suggests that features such as the speaker’s relative status, gender, and social class have an impact on how utterances are comprehended, and these effects occur early in the comprehension process (e.g., Van Berkum et al., 2008). Consider the speaker’s relative status. Because higher-status people have the right to direct the actions of others, their use of a nonconventional indirect request (e.g., It’s warm in here) is more likely to be interpreted as a directive (i.e., open the window) than the same nonconventional request uttered by a lower-status person (Holtgraves, 1994). As several researchers have noted (e.g., Freeman & Stewart, 2021; Hobson, 2021), power differentials can play a role in the communication of microaggressions by reinforcing and reproducing existing social hierarchies.

But how does this happen? The relationship between power and language is often viewed in terms of powerful people being able to say what they want to say, relative to those with less power. However, a reasonable extension of this idea is that powerful people have more options in terms in how they say what they want to say. Powerful people, for example, can use indirect forms to perform requests and still have their intended meaning recognized (Holtgraves, 1994). This feature provides powerful speakers with deniability. A person who produces a microaggression—just like a speaker who produces any intentionally ambiguous remark (e.g., see Pinker et al., 2008)—can simply deny any meaning that a recipient claims to have heard. Hence, if a recipient takes offense to a microaggression, the speaker can simply deny the aggressive meaning and claim that it was the literal meaning that was intended (e.g., I meant it sincerely; you are very articulate). Of course, this puts the recipient in a “double bind” because they are left with no good options. This would be regarded as a secondary moral harm in Freeman and Stewart’s (2021) framework. If the recipient decides to confront the speaker, they may be then subject to secondary microaggressions (Johnson et al., 2021), such as gaslighting, which may lead to the recipient questioning their own view of the situation and reality. Viewing microaggressions within this framework helps highlight this feature.

Related to the issue of context is the role of the speaker’s intention in producing a microaggression. Specifically, is it necessary for the speaker to intend to be aggressive with their remark, or can microaggressions be unintentional? On this point there is some disagreement. In Sue et al.’s (2007) original formulation, microaggressions could be either intentional or unintentional. Lilienfeld (2017) argued that intentionality is required, but the consensus of several authors is that intentionality is not required (see especially Freeman & Stewart, 2021; Mekawi & Todd, 2021; Williams, 2021).

Speech act theory (Searle, 1979) makes a relevant and important distinction in this regard; it views utterances as performing actions at three levels. The first level refers to the locutionary act, or the uttering of a string of recognizable words combined with an appropriate syntax. The illocutionary act, on the other hand, refers to the specific speech act that the speaker intends for the recipient to recognize with this utterance (e.g., “You are so articulate”). It is possible for the speaker to intend for the recipient to recognize this utterance as performing a compliment, or as a microaggression, or both. The third level of meaning is the perlocutionary act, which is the recipient’s reactions to the utterance, including their interpretations of the speaker’s intended meaning. It is always possible for the recipient’s interpretation to differ from the speaker’s intended meaning. A speaker may intend to compliment, but the recipient interprets it as an insult. In one sense, these can be regarded as instances of miscommunication, occasions on which there is a mismatch between the speaker’s intention and the recipient’s recognition of that intention. On the one hand, this is not rare, and it is particularly likely when people are discussing potentially sensitive issues (Holtgraves, 2021). On the other hand, how is this even possible? Does this mean that communication is entirely subjective? Yes and no. Yes, because recipients are free to interpret utterances any way they choose. No, because over time language users develop generalized expectations for how to produce and recognize meanings in various contexts. So again, it is context that is critical.

It is important to note, however, that in some formulations, intentions are not one-shot affairs. Instead, they are realized over a series of conversational turns as conversationalists negotiate their intended meanings (Gibbs, 2001; Haugh & Jaszczolt, 2012). So, rather than focusing on a single isolated microaggression, it is probably more useful to consider them within the context of the verbal exchange of which they are a part. As previously noted, the prior discourse is critical for inferring a microaggression. Just as important, however, may be the reactions of the recipient (and bystanders) to a perceived microaggression and the way in which the speaker’s intention is brought to light and clarified (or not) over a series of turns. Researchers are beginning to empirically examine reactions to microaggressions (e.g., Zou & Dickter, 2013), but much more could be done.