RESEARCH on politeness has grown steadily over the past several decades and represents a truly multidisciplinary (if not interdisciplinary) field. Linguists, social psychologists, cognitive psychologists, communication scholars, sociolinguists, and others have all made important contributions to this endeavour. Experimental research on politeness, the focus of this chapter, represents a relatively small subset of this work. It has, however, made important contributions, particularly in terms of evaluating theoretical proposals. The purpose of this chapter is to provide a relatively broad overview of experimental research on linguistic politeness, describing both methodological techniques as well as some of the major findings and their theoretical implications. I provide first a brief overview of the major theoretical orientations to politeness. This is followed by a review of early empirical research on Brown & Levinson’s (B&L) politeness theory as well as the issues raised by this research. I then consider more recent research examining (1) the role of politeness in reasoning and the communication of uncertainty, and (2) the cognitive and neural processes involved in the processing of politeness.
Although there is a long history of scholarly interest in politeness (e.g., see Watts, 2003), experimental work on politeness followed the appearance of maxim-based models in the late 1970s and early 1980s. Two of these models, those of R. T. Lakoff (1973, 1979) and Leech (1983), adopted the Gricean (1975) view of communication, but expanded Grice’s set of conversational maxims to include a set of politeness maxims designed to account for why some linguistic forms were preferred over others. Both theorists proposed a Politeness Principle (PP) that worked in conjunction with Grice’s Cooperative Principle (CP). For Leech, the PP included the maxims of generosity, tact, approbation, modesty, agreement, and sympathy. Lakoff’s (1973) politeness maxims included: give options, don’t impose, and make the other feel good. Politeness occurs, according to these models, because speakers are constrained by these politeness maxims.
The most popular maxim-based theory, of course, was B&L’s politeness theory. Their theory, published first as a chapter in a volume on Questioning in 1978, and then later as a stand-alone book in 1987, continues to be the standard approach, against which all new theoretical developments are compared. The unique contribution of their theory was the inclusion of face-work as a universal motive for regulating politeness. Borrowing the concept of face from Goffman (1967), and merging it with Durkheim’s (1915) positive and negative rites, B&L postulated the existence of two universal human desires: negative face (desire for autonomy) and positive face (desire for approach by others). Humans are assumed to be oriented to both positive and negative face, and politeness is the linguistic (and non-linguistic) means for symbolically attending to these two dimensions. Equally important in their theory was the inclusion of broad social variables (power, distance, and act imposition) designed to influence the perceived weightiness of an act, and hence the politeness with which that act is performed.
Although popular and generative, numerous criticisms of the B&L model resulted in the development of discursive or post-modern approaches that began to gain currency in the 1990s (Watts, 1992, 2003; Eelen, 2001; Locher, 2004) and today remain a vibrant alternative to the B&L approach. Fundamental to these approaches is the distinction between politeness as a lay or folk concept, termed first-order politeness (politeness1, and politeness as a technical, sociolinguistic variable, or second-order politeness (politeness2). Researchers taking a discursive approach generally argue that there is a fundamental divergence between first- and second-order politeness, and that it is the former rather than the latter that should be the central concern for sociolinguists. On this view, there are no universal motives or mechanisms for politeness; rather, politeness is completely situated. More recently, there have been attempts to articulate a middle ground between an overarching politeness model (à la B&L) and situated discursive views of politeness. These middle-ground or interactional models (Terkourafi, 2005; Arundale, 2006; Haugh, 2007) view politeness less as a strategy and more as expected behaviour, but behaviour that nonetheless is partially a function of more general interactional constraints.
Early experimental work on the B&L model focused primarily on two issues: the proposed ordering of politeness strategies and the role played by social variables in determining politeness levels. In terms of the former, B&L proposed the existence of four linguistic super-strategies that constituted a universal continuum of politeness. The ordering was based on the degree of face-threat associated with each of the following strategies; bald-on-record (no face support) was the least polite (and hence most face-threatening), followed in ascending order by positive politeness (emphasize solidarity with hearer), negative politeness (indicate respect for hearer’s autonomy), and off-record (maxim violation with deniability) politeness (the least face-threatening). Experiments designed to test this ordering focused primarily on requests and the general research strategy was to have participants rate tokens of the four super-strategies in terms of politeness and related interpersonal constructs (e.g. liking of the speaker). Partial support for the theory’s predicted ordering of the super-strategies was reported by several researchers (Fraser & Nolan, 1981; Hill et al., 1986; Blum-Kulka, 1987; Bauman, 1988; Holtgraves & Yang, 1990). More fine-grained politeness orderings were examined as well. Clark & Schunk (1980; see also Holtgraves & Yang, 1990) examined the perceived politeness of a set of negatively polite requests and found perceived politeness to vary as a function of the implied cost (i.e. threat) to the hearer. For example, ‘Could you x?’ was perceived as less costly, and hence more polite, than ‘Would you x?’
This research, however, raised several issues regarding this ordering. First, a consistent exception to the predicted ordering was that negatively polite forms were often ranked higher in politeness than off-record forms (Blum-Kulka, 1987; Holtgraves & Yang, 1990). Some researchers suggested that off-record forms carry a cost because the recipient must make an effort in order to infer the speaker’s meaning (Leech, 1983; Blum-Kulka, 1987). Others have argued that off-record forms may give the impression of manipulativeness on the part of the speaker (Lakoff, 1973). This issue is part of a larger question regarding the extent to which politeness should be equated with indirectness. If politeness is viewed as any deviation from maximally efficient communication (i.e. not in accord with all Gricean maxims) then there is a rough correspondence between politeness and indirectness. However, indirectness can occur for reasons other than politeness. Pinker and colleagues (Pinker et al., 2008; Lee & Pinker, 2010), for example, have argued that off-record forms are motivated by attempts to negotiate the nature of the relationship between interactants, and that the use of off-record forms accomplish this by providing speakers with a means of achieving plausible deniability. In this view, then, politeness and indirectness do not reside on the same scale.
A related issue concerns the proposed ordering of negative and positive politeness strategies. B&L, following Durkheim (1915) and Goffman (1967), argue that positive politeness (an approach-based strategy) is inherently less polite than negative politeness (an avoidance-based strategy), due to the presumption of closeness inherent in the former. However, some researchers argued that these forms are qualitatively different and hence cannot be ordered on a unidimensional continuum (Scollon & Scollon, 1981; Baxter, 1984; Tracy, 1990; Lim & Bowers, 1991). Still, for directives (threats to the hearer’s negative face) the proposed ordering makes sense theoretically (negative politeness grants the hearer greater autonomy than positive politeness) and is supported by empirical research (Holtgraves & Yang, 1990). On the other hand, for acts that threaten primarily the hearer’s positive face, research suggests that positive politeness may be perceived as more polite than negative politeness (Lim & Bowers, 1991; Holtgraves, 1997). It is possible that there may be a specificity principle at work here such that strategies orienting to the specific type of face threatened will be regarded as the most polite strategy. Thus, negatively polite strategies would be more polite for acts threatening the hearer’s negative face, and positively polite strategies would be more polite for acts threatening the hearer’s positive face.
Finally, several researchers have argued that B&L’s set of super-strategies is incomplete due to its overemphasis on politeness at the expense of impoliteness (Craig et al., 1986; Culpeper, 2011). That is, the least polite strategy in the B&L model is bald-on-record, a strategy which is simply the absence of any face support. No doubt aggressive attacks on others’ face occur and are not handled well within the B&L model.
One of the most important features of the B&L model is the specification of links between politeness and the major social dimensions of power and distance. Their theory assumes that increasing hearer power, relationship distance, and act imposition will increase the overall weightiness of the act (i.e. degree of face-threat), and increased weightiness is expected to result in increased politeness. Researchers have used relatively straightforward role-playing scenario techniques to manipulate power, distance, and imposition in order to examine their predicted impact on the perceptions and production of politeness. Fairly consistent support has been reported for act imposition, with increasing imposition associated with increasing levels of politeness, an effect that has been reported for requests (Brown & Gilman, 1989; Leichty & Applegate, 1991; Holtgraves & Yang, 1992), expressions of gratitude (Okamoto & Robinson, 1997), recommendations vs. reports (Lambert, 1996), and accounts (i.e. explanations for untoward behaviour; McLaughun et al., 1983; Gonzales et al., 1990), as well as other speech acts (Brown & Gilman, 1989; Leichty & Applegate, 1991). Null findings have been reported (Baxter, 1984) but are rare. The power variable has also received experimental support. In general, greater politeness occurs or is expected as a function of increasing power of the recipient relative to the speaker (and hence decreasing power of the speaker). This effect has been reported with requests (Holtgraves & Yang, 1990, 1992; Leichty & Applegate, 1991; Lim & Bowers, 1991), including observational studies of actual requests (Blum-Kulka et al., 1985), messages conveying bad news (Ambady et al., 1996), teasing (Keltner et al., 1998), reminders and complaints (Leichty & Applegate, 1991), criticisms (Lim & Bowers, 1991), accounts (Gonzales, Pederson et al., 1990), and questions (Holtgraves, 1986). Some of these effects have been replicated cross-culturally (Holtgraves & Yang, 1992; Ambady et al., 1996).
Finally, the results for the effects of relationship distance on politeness have been mixed. Consistent with the theory, some researchers have reported greater politeness as a function of increasing distance between interlocutors (Wood & Kroger, 1991; Holtgraves & Yang, 1992); others have reported the exact opposite (Baxter, 1984; Brown & Gilman, 1989). And some (e.g. Lambert, 1996) have reported no relationship between distance and politeness. Distance, of course, is a multifaceted variable and it has been measured and manipulated in a variety of ways. Slugoski & Turnbull (1988) (see also Brown & Gilman, 1989) argued that researchers sometimes confounded distance (i.e. familiarity) and affect (i.e. liking). Higher levels of politeness have been found to be associated with greater interpersonal distance (i.e. interactants are more polite with people with whom they are less familiar) but also with greater liking (people are more polite with those whom they like).
More recently, the relationship between distance and politeness has been investigated in terms of a popular social psychology theory termed Construal Level Theory (CLT; Trope & Liberman, 2003; Liberman & Trope, 2008). For CLT, something is psychologically distant (temporally, spatially, or socially) when it is not part of one’s direct experience. Psychologically distant stimuli are typically represented at a higher, more abstract level of construal while stimuli that are close are represented at a lower, more concrete level of construal. Stephan et al. (2010) argued that higher level construal is generally more polite than lower level construal. In eight experiments they found increasing politeness to occur as a function of increasing spatial and temporal distance, and increasing politeness to result in inferences of greater temporal and spatial distance. For example, in terms of temporal distance, participants produced more polite messages directed at someone in the distant future relative to someone in the near future. And conversely, the use of more polite forms led to judgements of greater temporal distance than the use of less polite forms. This represents an important extension of politeness theory because it both demonstrates the bidirectional relationship between distance and politeness, and expands the notion of distance to include spatial and temporal distance.
One issue that has been raised regarding the B&L model is the manner in which power, distance, and imposition interact. The model (implicitly) assumes that their effects are additive. Empirical research suggests otherwise. Many researchers who have examined the simultaneous impact of these variables on politeness have reported interactions between them, including Power by Distance interactions (Blum-Kulka et al., 1985; Holtgraves & Yang, 1990; Lim & Bowers, 1991), Imposition by Distance interactions (Leichty & Applegate, 1991; Holtgraves & Yang, 1992), and Imposition by Power interactions (Gonzales et al., 1990; Holtgraves & Yang, 1992). The existence of these interactions simply means that as the effects of one of the interpersonal variables become very large, the effects of the other two variables become much smaller. For example, a person who makes an extremely large request (i.e. high imposition) will tend to be polite regardless of power and distance.
A second issue is whether politeness is influenced by variables other than power, distance, and imposition. Obviously it is. It is important to note, however, that power, distance, and imposition are high-level, abstract variables that should subsume more specific variables. For example, ethnicity, gender, occupational differences, and so on are variables that feed into power and distance, and ultimately, politeness. Even mood states may be incorporated in the model in this way. For example, Forgas (1999a,b) demonstrated that people in sad moods prefer to use greater levels of politeness than people in happy moods, possibly because a person’s mood influences their perceptions of the interpersonal context (power, distance, and imposition). Hence, people in a sad mood may perceive themselves as being relatively low in power, or perceive an act as being relatively more imposing, and it is these perceptions that affect their level of politeness. It may also be possible to use the model to explore individual differences in politeness, an area that has seen relatively little research. Introverts, for example, may perceive relatively greater distance between themselves and others, and hence produce higher levels of politeness. Extraverts, on the other hand, may perceive relatively less distance and hence favour the use of relatively less polite but more approach-based strategies (i.e. positive politeness).
On the other hand, Terkourafi (2001, 2005; Vergis & Terkourafi, 2015) has argued for a frame-based approach to politeness in which different situational contexts, over time, come to be associated with expected politeness forms (i.e. they become conventionalized). Although these expectancies can be overridden by the context, the default meanings of these terms become part of their lexical meanings and the terms do not intentionally convey (im)politeness. This alternative offers a more granular approach, one in which power, distance, and imposition can play a role in politeness, but not the overarching role theorized by B&L.
A third issue concerns the direction of the relationship between power, distance, and politeness. For B&L, relative power and distance are often viewed as determinants of politeness; that is, speakers’ estimation of these variables determines act weightiness and hence the politeness strategy to be used. However, as many have noted, this is a static view and it is likely that the relationship is bidirectional. Specifically, if the use of a particular linguistic form is affected by power and distance, it follows that the use of a particular linguistic form will be informative for observers (including the hearer) regarding the speaker’s perceived power and distance. For example, because high-status speakers use less polite forms than lower-status interactants, the use of less polite forms should result in perceptions of higher speaker status, other things being equal. In a cross-cultural study using participants from the United States and South Korea, Holtgraves & Yang (1990) found that less polite request forms were associated with perceptions of greater speaker power. And there is a fairly substantial literature on what is termed powerful (vs. powerless) language. In general, this research suggests that the use of powerful language (essentially less polite language) results in perceptions of relatively greater power, credibility and persuasiveness (Erickson et al., 1978; Bradac & Mulac, 1984; Gibbons et al., 1991; Hosman & Wright, 1987; Burrell & Koper, 1998; Holtgraves & Lasky, 1999; Blankenship & Holtgraves, 2005), although there is some evidence that these effects may be moderated by speaker gender (Carli, 1990).
The relationship between politeness levels and perceptions of the speaker can be quite complex. For example, when a high-status speaker is extremely polite to a subordinate it will often result in perceived sarcasm (Slugoski & Turnbull, 1988). Moreover, this bidirectional relationship allows people to strategically vary their politeness as a means of negotiating and/or altering the interpersonal context; it is, in effect, an important component of impression management (Goffman, 1959). So, a higher-power person (e.g. a boss) can move from negative politeness to positive politeness in an attempt to negotiate a closer relationship. Similarly, a person in an established relationship may begin to use less politeness as a means of negotiating higher power in the relationship, and so on. This possibility can obviously result in interpersonal misperceptions or misunderstandings. A speaker may assume his politeness level reflects one dimension (e.g. closeness), but his interlocutor may assume it reflects a different dimension (e.g. power). This negotiated nature of politeness-based person perception awaits further empirical investigation.
A new line of research has developed over the past decade that examines the role of politeness in reasoning and the communication of uncertainty (Bonnefon et al., 2011; Bonnefon, 2014; Juanchich et al., Chapter 21 in this volume). This research has demonstrated how politeness can influence the interpretation of probability terms (e.g. possible), connectives (e.g. or, and), and quantifiers (e.g. some, all). Consider first probability terms. Their use can convey varying degrees of uncertainty (e.g. It’s possible you’ll flunk the course). There are, however, other motivations for their use. Specifically, probability expressions can function as a politeness strategy (e.g. as a hedge on an assertion) and used as a means of managing face. Rather than saying ‘You’ll never finish it in time’—an expression that might threaten the recipient’s face—a speaker can hedge his opinion with ‘It’s possible you won’t finish it in time’.
The existence of multiple motives for the use of probability terms can influence the manner in which they are interpreted. This was demonstrated by Bonnefon & Villejoubert (2006) who asked participants to imagine that their family doctor told them they would ‘possibly’ develop either deafness or insomnia in the upcoming year. Participants judged the probability of the more severe disease (deafness) to be significantly higher than the probability of the less severe disease (insomnia). This occurred because participants judged the doctor to be using ‘possibly’ as a face-management device significantly more frequently when the condition was deafness than when it was insomnia, and they adjusted their estimates accordingly. In this experiment it was the hearer’s face that was being threatened, although it is possible for the speaker’s face to be threatened as well, with a similar impact on the interpretation of probability terms (Juanchich et al., 2012).
This effect has been demonstrated also with certain scalar expressions. Scalar expressions can be ordered on a scale with respect to their strength (e.g. <some, all>) (Levinson, 1983) and it has been argued that the use of the weaker, more inclusive, term (e.g. some) implies that the stronger term (e.g. all) does not hold (Levinson, 1983; Horn, 1984). Hence, the scalar implicature for ‘some’ is ‘some but not all’. Bonnefon et al. (2009) demonstrated that face management can influence the likelihood that a scalar implicature is generated, in much the same manner that it influences the interpretation of probability terms. When a situation is face-threatening, ‘some’ can be used as a hedge to politely indicate ‘all’. Consistent with this reasoning, Bonnefon et al. (2009) found that estimates of ‘some’ in ‘Some people hated your party’ were higher than estimates of ‘some’ in ‘Some people enjoyed your party’; people were more likely to assume that ‘some’ was being used in the service of face management in the former situation, and they adjusted their interpretations accordingly. More recently, Bonnefon et al. (2015) demonstrated that this effect varies as a function of the discourse context such that the effect is more pronounced when the scalar term is preceded by a brief silence, a dispreferred marker signalling that bad news is forthcoming (Holtgraves, 2000).
Finally, just as scalar expressions can be ordered on a scale according to their strength, so too can connectives (e.g. <or, and>). And just as with scalar expressions, the interpretation of these terms can be influenced by politeness considerations. Feeney & Bonnefon (2013) manipulated the content connected by ‘or’ such that it was either positive or negative (i.e. face-threatening). They found that exclusive interpretations (i.e. ‘one or the other but not both’) were significantly more likely for positive content than for negative content. There is a potential face management motive for the term when the content is negative, and because of this, people are less likely to generate the exclusive interpretation and instead assume that both interpretations are possible.
This research demonstrates that politeness can influence the interpretation of connectives, probability expressions and scalar terms. Is the production of these terms similarly affected? It appears so. Juanchich & Sirota (2013) asked participants to choose the probability expression that they would use in order to convey a particular outcome, and they manipulated the speaker’s communicative goal such that they were instructed either to be informative, to avoid blame (i.e. speaker face management), or to avoid upsetting the hearer (i.e. hearer face management). As expected, when the outcome was negative, participants chose expressions that conveyed less certainty under face management instructions (both hearer-face and speaker-face) than under instructions to be informative, an effect that was reversed when the outcome was positive. In a follow-up study, Sirota & Juanchich (2015) asked participants to choose probability expressions that they would use to convey negative information, and then assessed their communicative motives (rather than manipulating them) for choosing those expressions. Participants frequently (41.6%) indicated face management as their motive, and when they did so, they were more likely to choose an expression that conveyed a lower probability, relative to when they indicated that their motive was to be informative.
More recently, Holtgraves & Perdew (2016) examined both production and interpretation in the same study. Participants in these experiments were presented with situations that varied in terms of face-threat and were asked how they would communicate potentially threatening information. Both hearer threat (Experiment 1) and speaker threat (Experiment 2) were examined, and participants either chose from a pre-existing set of utterances or responded in an open-ended manner. In both experiments a second set of participants read these utterances and provided judgements as to the degree of uncertainty conveyed by the utterance. In both experiments, messages in the face-threatening condition conveyed greater uncertainty than messages in the non-face-threatening condition, and the probability estimates made by the second set of participants varied as a function of the conveyed uncertainty. This research demonstrates that when examining speakers and hearers together, severe events may be judged less likely (rather than more likely), because speakers tend to hedge the certainty with which they communicate the information.
Variations in politeness can have a range of social and cognitive effects. But how, exactly, do these effects occur? Several different lines of research examining the processing of politeness have been pursued. One line of research has examined the extent to which people encode and retain the wording that conveys politeness. This is an important issue because it is relevant for the claim that at least some forms of politeness are expected, normative behaviour and hence non-salient (Watts, 2003). A frequent finding in the memory literature is that people typically forget the exact wording of an utterance but retain the gist of what was said (Sachs, 1967). One important exception to this is that wording will be remembered well when it has interpersonal implications (Keenan et al., 1977; MacWhinney et al., 1982). Politeness, of course, has clear interpersonal implications and research suggests that politeness wording is spontaneously encoded and retained. For example, Holtgraves (1997; see also Slugoski, 1995) examined incidental memory for wording that varied in politeness and found that people remembered politeness wording at levels exceeding chance. Interestingly, even when the specific wording was not remembered, people appear to have encoded the overall level of politeness and recalled wordings consistent with that level of politeness. In other words, if participants heard an impolite request, when asked to recall that request they tended to recall an impolite (rather than polite) form, even if they could not remember the exact wording. It was also the case, however, that memory for politeness wording was greater for cases of politeness wording that were inconsistent with the social context. For example, participants in a psychology experiment were more likely to remember impolite forms if the speaker was low (rather than high) in status (a graduate student) and polite forms if the speaker was high (rather than low) in status (a faculty member). Such forms violate expectations and hence are remembered well.
Memory studies have a potential limitation in that their results can be ambiguous in terms of actual on-line processing. More recently, researchers have begun to use on-line techniques in order to examine the processing of politeness in real time. One such approach is to use an eye-tracking methodology. For example, Raizen et al. (2015) used an eye-tracking procedure to examine the processing of taboo words (i.e. potential violations of positive face). They found that early fixations on taboo words varied as a function of an interaction between the speaker’s identity and the situational appropriateness of a taboo word. Specifically, taboo words in appropriate situations resulted in longer fixations when uttered by someone unlikely to use taboo words (‘saints’) than when uttered by someone expected to use taboo words (‘sinners’).
Other researchers have used electrophysiological techniques to examine politeness processing. For example, Jiang and colleagues (Jiang et al., 2013, 2015) recorded Event-Related Potentials (ERPs) as participants read conversations in which speaker status and pronoun type (respectful vs. disrespectful) were manipulated. Status-inconsistent pronouns (e.g. disrespectful pronouns from a lower-status speaker) resulted in an enhanced N400—interpreted as an indicator of semantic integration effort—compared to status-consistent pronouns, thereby suggesting that brain activity varies as a function of the pragmatic implications. The status-inconsistent pronoun also resulted in a later negativity (500–800 milliseconds (ms)) but not late positivity (i.e. P600). The authors interpret the failure to observe the latter as indicating support for a contextualized view of politeness (Terkourafi 2001, 2005) rather than an intentionalist view (B&L). These researchers also examined individual differences in reactions to the politeness violation. Male participants, as well as those scoring higher on a measure of fantasy (ability to transpose oneself into the feelings and actions of fictitious others), demonstrated a significantly larger N400 in response to the status-inconsistent pronoun than females and those scoring lower on fantasy.
Hoeks et al. (2013) examined neurophysiological reactions to impolite utterances that were request refusals. Participants listened to dialogues in which one person asked another for a favour (either a request for an action to be performed or a question) and the other responded with a blunt ‘No’ or a ‘No’ that included an apology and reason for the refusal. These researchers found a significant P600 for the former relative to the latter. However, for the request for action, reactions to the refusal condition started earlier, were more broadly distributed, and of a larger magnitude relative to the question condition. The authors interpret this reaction as reflecting participants’ reorganization of their conversation model as a result of linguistic impoliteness. The possibility that this reorganization may reflect also an updating of representations of the speaker (i.e. that he’s rude) was noted.
Recently, Holtgraves & Kraus (2018) used electrophysiological techniques to examine the role of politeness in the comprehension of several different scalar expressions. Prior research using electrophysiological techniques to examine scalar expressions has explored their use in logical contexts (e.g. Some people have lungs) (Noveck & Posada, 2003; Nieuwland et al., 2010; Hartshorne, Snedeker, Azar, & Kim, 2015; Skordos & Barner and Breheny, Chapters 2 and 4, respectively, in this volume). In contrast, Holtgraves & Kraus examined their use in conversational contexts. They had participants read scenarios followed by a target utterance containing a scalar expression (e.g. some) in the first half of the utterance, with a second half continuation of the utterance containing either the pragmatic meaning (e.g. not all) or the semantic meaning (e.g. all). The semantic meaning resulted in a larger P300 than did the pragmatic meaning, and this difference was greater when the situation was face-threatening relative to when it was not face-threatening. This suggests that in conversational contexts, scalar implicatures are generated when the scalar term is encountered, especially when the situation is face-threatening.
Deficit-based research is also relevant for understanding politeness processing. Consider, for example, recent research on language deficits in Parkinson’s disease (PD). Although PD is primarily associated with debilitating extrapyramidal motor dysfunction, it also affects thinking, reasoning, planning, and language functions, and in terms of the latter there is some evidence of pragmatic impairment in PD (e.g. Lewis et al., 1998; McNamara & Durso, 2003), including politeness. To investigate the latter, Holtgraves & McNamara (2010) used a role-playing task and asked participants (those with PD and matched controls) to imagine being in situations in which they were to make a request of another person and to write out exactly what they would say in order to make each request. Overall, the PD participants were less polite than the control participants. More importantly, although increasing imposition was associated with increasing politeness for control participants, this did not occur for PD participants, suggesting a reduced sensitivity to the social context for those with PD. This reduced sensitivity also occurred for speaker power, but only for PD participants who were on higher doses of dopaminergic medication.
What are the potential sources of the politeness impairment in PD? One possibility is that it reflects an overall decline in cognitive capabilities, especially executive cognitive functions (Owen et al., 1992; Lange et al., 1992; Taylor & Saint-Cyr, 1995; Troster & Woods, 2003; McNamara et al., 2008). Reduced cognitive resources in PD could result in an attentional deficit such that variations in request size are not noticed; because they are not noticed there is no corresponding change in politeness. Consistent with this possibility, researchers have demonstrated a clear connection between executive function deficits in PD and the ability to contribute meaningfully (e.g. being appropriately informative) to conversations (Holtgraves et al., 2013). Another possibility is that even when variations in the context are noticed, the cognitive capacities required to produce more polite (and cognitively complex) strategies are hindered in people with PD. In the Holtgraves & McNamara (2010) research, participants on higher doses of dopaminergic medication did notice variations in recipient status (based on manipulation check items) but they failed to produce more polite strategies for a higher-power recipient.
Researchers also have examined politeness processing deficits associated with damage to the right hemisphere (RHD). Pell (2007; Experiment 2) examined politeness judgements of RHD and heathy controls in which both prosody and utterance type (direct, indirect, very indirect) were manipulated. Past research has suggested RHD is associated with a prosody processing deficit. However, in this experiment it was the interaction of prosody and language, more specifically the occurrence of linguistic and prosodic discrepancies as a means of conveying sarcasm, that posed difficulties for RHD participants (but not for healthy controls). In addition, the distinctions made by RHD participants based on linguistic cues alone were not as fine as those made by healthy controls, a finding similar to that reported for PD individuals.
Experimental studies of politeness continue to be a small but relatively critical component of politeness research. After a flurry of experimental activity in the 1980s and 1990s designed to test predictions derived from the B&L model, experimental studies of politeness have expanded into new areas as described in this chapter. Looking forward, I expect there will be a continuing focus on the processing of politeness. The use of electrophysiological techniques in this regard is only beginning and has the potential to provide relevant information regarding the processes involved in comprehending politeness. Experimental studies of the role of politeness in reasoning and the communication of uncertainty should continue as well. Hopefully, the success of this endeavour will prompt researchers to consider other areas for which the analysis of politeness processes may be relevant. There is, for example, no shortage of topics in social psychology (e.g. persuasion, person perception, relationship negotiation, etc.) for which politeness plays an important but as yet unexamined role. Other important areas for future experimental research include the relationship between politeness and indirectness, the manner in which non-verbal politeness interacts with and modifies verbal politeness, individual differences in politeness, and of course cross-cultural similarities and differences in politeness, all areas that could be examined experimentally.
Experimental studies of politeness should continue to remain an important component of politeness research because they allow for the evaluation of theoretical predictions and tests of competing models. For example, the results of early empirical research on the B&L model suggested several ways in which that model needed to be modified, including the ordering of politeness strategies and the interactive effects of social variables. Experimental studies of the processing of politeness have much to offer in this regard. Or consider the claim that politeness is normative and expected and hence typically not noticed (Watts, 2003). Experimental memory research suggests that this may not be the case, and that people do encode some representation of a speaker’s level of politeness, even when it is expected (Holtgraves, 1997). Or consider the manner in which ERPs to status pronoun violations provide support for a contextualist rather than intentionalist view of politeness (Jiang et al., 2015).
The downside of experimental studies of politeness is their potential artificiality and decontextualized nature. Trade-offs are involved, of course, and gains in experimental control are often paid for with a drop in realism. Still, it is possible to make experimental stimuli fairly realistic, by, for example, collecting actual discourse samples to then be used in experimental research. And the development of new methodologies may eventually allow for the examination of real-time situated politeness via the use of electrophysiological techniques as individuals engage in (constrained) natural language use (see, e.g., Hoeks & Bouwer, 2014). The gain in precise experimental control, coupled with the back and forth between theory and data, can facilitate advances in our understanding of certain facets of politeness, an understanding that can contribute to and compliment advances made with non-experimental techniques. In short, experimental research on politeness should not be ending anytime soon.
The writing of this chapter was supported by a grant from the National Science Foundation (BCS-1224553).