The Functions of the Spanish Approximators Como and Como Que in Institutional and Non-Institutional Discursive Contexts

1 Introduction

The approximators como and como que (“sort of,” “as if,” “kind of,” “seems,” “like”) are lexical items that convey vagueness. They are employed to modify the certainty of an expression when speakers may not have sufficient information (e. g. Cuervo, 1954; Hengeveld & Keizer, 2011; Mihatsch, 2009; Said-Mohand, 2006, 2008; Trujillo Carreño, 1990). Furthermore, como and como que have been found to attenuate the content of an utterance when a speaker does not want to be perceived as too direct or straightforward (e. g. Albelda, Briz, Cestero, Kotwica, & Villalba, 2014; Briz, 2003, 2007; Haverkate, 1994; Montecino, 2004). Camacho (2011) has further suggested that these forms are used by speakers of different Spanish varieties and in several contexts to indicate that the speaker is not willing to commit to the content of the constituent the approximators modify. However, these lexical forms are context dependent and may differ in use among speakers of Spanish varieties and among the discursive events in which they are produced.

Based on the concept of linguistic variation (Labov, 1972, 2004, 2011), Camacho (2011, p. 8) posits that como que has become lexicalized as a free variant of como and that “como que seems to have become a compound, and que does not contribute anything either syntactically or semantically to the whole sequence”. Consider examples (1a) and (2a) produced by two speakers of our study, in which como and como que can alternate without changing the meaning or the grammaticality of the sentence. Examples (1b) and (2b) illustrate how the alternation between the two forms is plausible.

Since both of these forms compete in similar linguistic environments, the aim of this study then is to uncover the sociolinguistic and pragmatic factors that condition the use of one form over the other and to uncover whether the speech event may condition their expression. To our knowledge, there is an absence of research that investigates the use of the approximators como and como que in various Spanish varieties and spontaneous speech using both sociolinguistic and pragmatic variationist approaches. Also, the expression of como and como que has not been examined in one study using different discursive environments. This study, therefore, advances our understanding of these approximators and their functions among several Spanish varieties in two categories of spontaneous speech: narratives of personal experience (i. e. non-institutional discourse) and therapeutic motivational interviews (i. e. institutional discourse).

2 Literature review

Como and como que have been known to serve multiple functions. One of the functions these forms serve is that of approximators. As approximators, como and como que can be employed to convey, depending on context, uncertainty, proximity, or resemblance (e. g. Channell, 1994; Jiménez Juliá, 2003). Thus, approximators are used to convey vagueness. For instance, Channell (1994) has suggested that there are various means by which speakers convey vagueness: By implicature, by choice of vague words, and by vague additives. Channel maintains that vagueness by implicature arises from the notion that even apparently precise phrases can be interpreted as having vague meanings. Vagueness by choice of words consists of lexical items that are intrinsically vague (e. g. thingy, thingummy, whatnot). On the other hand, vague additives (e. g. like, around, kind of, sort of) represent lexical items that alter the degree of certainty or straightforwardness of the constituents they modify. The Spanish approximators como and como que belong to this latter category. Fant (2007) has suggested that these forms tend to modulate and adapt the reference so that it fits the conceptual content that the speaker wishes to express.

Crystal (1995) has suggested that imprecision is part of everyday language as definitions of lexical items are not always clear and determinate. According to Mihatsch (2007), even though accuracy in discourse is important, in many contexts speakers cannot express what they wish by only relying on conventional lexical meanings. For this reason, approximators such as como and como que are commonly employed. That is, approximators as these convey what Lakoff (1973, p. 458) coined “fuzzily defined boundaries” rather than “sharply defined” ones.

Several researchers who have investigated como and como que have noted that these forms represent linguistic resources that allow speakers to reduce the certainty of an expression (e. g. Cuervo, 1954; Hengeveld & Keizer, 2011; Mihatsch, 2009; Said-Mohand, 2006, 2008; Trujillo Carreño, 1990) or to soften or attenuate the content of an utterance (e. g. Albelda et al., 2014; Briz, 2003, 2007; Haverkate, 1994; Montecino, 2004). Thus, these forms, depending on the context, Spanish variety, and discursive setting, can be employed as strategies to avoid fully committing to the true meaning of a proposition when there is lack of information or lack of direct evidence (e. g. Camacho, 2011; Said-Mohand, 2006, 2008).

In addition to the previously mentioned function related to vagueness, como and como que also serve other grammatical functions which this study does not address. For instance, Jiménez Juliá (2003, p. 117) posits that como is “one of the most versatile particles in Spanish”, which suggests that expressing approximation is simply one of its several uses. Como may also be used as an interrogative or exclamatory pronoun, as a conditional conjunction, as a comparative adverb, or as a quotative (Camacho, 2011; Hernando Cuadrado, 2002; Jiménez Juliá, 2003; Kern, 2014; Moreno Ayora, 1991; Said-Mohand, 2006; Schwenter, 2001; Trujillo Carreño, 1990 among others).

That said, one way of distinguishing the approximator function of como and como que from their other possible grammatical and discursive functions is by examining the speakers’ communicative intention in using these approximators.

Concerning the speech events in which we examine the alternation between como and como que (i. e. oral narratives of personal experience and an institutional discourse of therapeutic interviews), we maintain that both interactional discourses represent natural, spontaneous speech. Therefore, the study examines the uses of the approximators in similar naturalistic interactions but different interactional settings.

Following Hymes (2005), we define speech events as activities that follow the rules for the use of speech. In other words, speech events are established based on the use of language. For instance, Duranti (1985) noted that there are events in which speech is crucial for them to take place, such as class lectures or phone conversations. This aspect is also the case for narratives and therapeutic interviews since both events are manifested through speech and are characterized by specific uses of language. Concerning the oral narratives of personal experience and, following a sociolinguistic framework, our study centered on identifying several of the linguistic and social variables that mediate the use of these forms. The reasoning behind the use of oral narratives was that these were found to reduce the effects of observation to a minimum (Labov, 2001). Since narratives are delivered with a similar organization in a wide variety of societies and cultures, we found the discourse to be fit for an examination of como and como que.

The corpus in which we also exemplify the use of these approximators relies on what is known as Institutional Talk. In particular, the use of como and como que was extracted from what is called “Motivational Interviews” (MIs) in which the therapist reviews adherence to treatment and the status of the clients’ well-being. These interviews are known as a change-inducing strategy, a modality that aims at enhancing client motivation towards change (Von Wormer, 2007). Bardovi-Harlig and Hartford (2005, p. 13) defined institutional discourse as “spontaneous authentic language use by speakers who are speaking as themselves, in genuine situations, with socio-affective consequences”. That is, this discourse is not motivated by experimental prompts (e. g. discourse completion tasks) that lead the speakers to respond in a specific manner. Agar (1985) maintains that these discourses are characterized as instances in which one person who represents an institution encounters another to provide a service. Agar (1985, p. 149) also notes that institutional discourse has to accomplish three things: Diagnose the client, provide a limited number of ways to describe people, their problems, and the solutions (i. e. institutional frames), and shape the discourse of the institutional representative so that it adheres to the client’s ways of talking.

3 Theoretical frameworks

We have noted here that the approximators como and como que are known to convey vagueness in oral discourse because the content of the constituents they modify is not made exact and precise (Channell, 1994). Fuentes Rodríguez (2008) has further pointed out that forms such as como and como que present information that is inexact but close to the truth and allow speakers to voluntarily introduce both imprecision and proximity at the same time. Such imprecision can be motivated by the need to modify the certainty of an expression due to insufficient information, or by the desire to attenuate the content of an utterance in order to avoid being too direct.

Given the aforementioned observations, which have posited that the approximators como and como que compete in similar linguistic environments, serve similar discursive functions, convey similar meanings, and are pervasive in speech, the problem we investigate is the following: Do speakers exhibit preferences in the use of one of these forms?

To approach this problem, the present study was informed by two frameworks: Variationist sociolinguistics (Labov, 1972) and variational pragmatics (Schneider & Barron, 2008). As noted by Labov (1972, 2004, 2011), variationist research examines the different forms that represent ways of saying the same thing (i. e. variants) and aims to uncover the conditions that influence the use of one variant over the other. Thus, for the sociolinguistic framework, we focused on determining the linguistic and external variables (i. e. educational attainment, Spanish variety, and years of residence in the U.S.) that may condition the use of the variants under examination. For the pragmatic variational framework, we attended to the situational and linguistic factors (e. g. discursive interaction, power relations, distance) that characterize each speech event under examination, to address the different pragmatic meanings conveyed by como and como que (Schneider & Barron, 2008). It should be noted that while variation has been attested in sociolinguistic research (e. g. syntactic, phonological), variation at the discourse-pragmatic level has not received similar attention. In particular, matters in pragmatic variation have been examined regarding regional differences in one or more varieties of a language and, Schneider and Barron (2008) have called for empirical work such as studies conducted by Márquez Reiter and Placencia (2005). Márquez Reiter and Placencia (2005, pp. 192–193) define socio-pragmatic variation as the way in which speakers “vary their use of language in similar situational contexts with similar communicative purposes and thus exhibit different interactional features/patterns”, which remains one of the goals of this study.

With this in mind, our study was guided by the following questions:

1. Do speakers exhibit a preference in the use of the approximators como and como que? And, if a preference exists, is it conditioned by the speech event in which they are employed (i. e. oral narratives as opposed motivational interviews (MIs)?

2. Can we detect whether the use of como and como que is conditioned by external variables, such as Spanish variety, years of residence in the U.S. (i. e. exposure to English), and the educational level of the speakers?

3. Are the pragmatic functions of the approximators (e. g. conveying uncertainty) conditioned by the discursive setting in which they are produced?

4 Method

5 Findings

In this section, we first address our research questions by way of a quantitative analysis. In the subsequent section, we provide a qualitative analysis. Research question one focused on whether the use of como and como que would differ according to the discourse (i. e. speech event). The second question inquired whether the use of como and como que was conditioned by social/external variables.

5.1 Findings: The quantitative analysis

This section addresses the intersection between several linguistic and social variables concerning the use of como and como que. We report on the statistical significance in the results of the cross-tabulations conducted in the study in order to address whether there is a correlation between the variables being examined.

5.1.1 The use of como and como que according to the type of discourse

To respond to the first research question that inquires about the use of como and como que according to the discourse (i. e. narrative vs. motivational interviews), we present Table 1 which illustrates the patterns in the use of the approximators in each type of speech event.

Table 1:

The distribution of como/como que according to speech event (n=473).

Discourse	Approximator		Total
	Como	Como que
Narratives	75.0 % (72)	25.0 % (24)	100.0 % (96)
MIs	54.6 % (206)	45.4 % (171)	100.0 % (377)
Total	58.8 % (278)	41.2 % (195)	100.0 % (473)

1. (p=0.000)

A Chi-Square test of independence was performed to determine whether there was a statistically significant relationship between the use of como and como que and the variable related to the discourse. A significant correlation was found (p=0.000). Thus, we make several observations from Table 1. First, and in general, we find that como was the preferred form in narratives and interviews, representing 58.8 % of the tokens. More specifically, 75.0 % of the tokens were issued using como in the narratives and, 54.6 % in the interviews. In other words, we uncovered a preference to use como among the speakers in both discursive contexts. Second, como que was issued at a higher frequency (45.5 %) in the MIs when compared to its use in the narratives (25.0 %). That is, the discourse setting (i. e. speech event) conditioned the use of these forms.

To further illustrate these findings, Figure 1 presents the overall preference in the use of como over como que in both discourse categories and further shows the frequent use of como que in the MIs rather than in narratives.

Figure 1:

The use of como/como que in the narratives and MIs.

As seen in Figure 1, como was issued more frequently in narratives and interviews. However, como que was employed more frequently (almost double) in the MIs than in narratives, We address this latter finding in more detail in the qualitative analysis.

5.1.2 The use of como and como que according to social/external variables

Our second research question focused on the relationship between the use of the approximators and several external variables: Speakers’ Spanish variety, educational attainment, and years of residence in the U.S. We first examine how como and como que were employed by speakers of each variety represented in the study. We explain these findings with a note of caution since a random sampling method of speakers within the corpus produced only a few speakers of several of the varieties. Nonetheless, we found a statistically significant relationship between the use of the approximators and the speakers’ Spanish variety at a p value of 0.000. In other words, the variety of speaker conditions the use of these forms.

Table 2 reveals a pattern in the use of the approximators. For instance, of the 118 tokens issued by the Mexican speakers, 72 % were produced using como. A similar pattern can also be found among the Puerto Rican and Colombian speakers. That is, speakers of the Puerto Rican variety produced this form in 71.3 % of the instances in which como que could have been expressed and, the speakers of the Colombian variety, 84.0 %. Therefore, we can suggest at this juncture that speakers of these varieties tend to favor the use como more than como que.

Table 2:

The distribution of como/como que according to speakers’ variety (n=473).

Speakers’ variety	Approximator		Total
	Como	Como que
Mexican	72.0 % (85)	28.0 % (33)	100.0 % (118)
Dominican	41.4 % (12)	58.6 % (17)	100.0 % (29)
Puerto Rican	71.3 % (77)	28.7 (31)	100.0 % (108)
Argentinean	50.0 % (5)	50.0 % (5)	100.0 % (10)
Cuban	44.0 % (74)	56.0 % (94)	100.0 % (168)
Colombian	84.0 % (21)	16.0 % (4)	100.0 % (25)
Honduran	28.6 % (2)	71.4 % (5)	100.0 % (7)
Total	58.8 % (278)	41.2 % (195)	100.0 % (473)

1. (p=0.000)

Conversely, the speakers of the Dominican, Cuban, and the only Honduran varieties preferred the use of como que. For instance, the Dominicans preferred como que in 58.8 % of the utterances under examination, the Cuban speakers 56.6 %, and the only Honduran speaker 71.4 %, although this speaker produced few instances of this form.

In sum, and with some reservation, we can suggest that we attested to differences in the use of como vs. como que according to Spanish variety. Interestingly, and with regard to the speakers who identified as speakers of the Argentinean variety, we cannot claim any preference since they produced few tokens and these were evenly distributed among the approximators under investigation.

Our second research question also attended to another external variable, educational attainment, and whether education conditioned speakers’ use of the approximators.

Table 3:

The distribution of como/como que according to speakers’ education (n=473).

Education attainment	Approximator		Total
	Como	Como que
High School	60.5 % (135)	39.5 % (88)	100.0 % (223)
University	57.2 % (143)	42.8 % (107)	100.0 % (250)
Total	58.8 % (278)	41.2 % (195)	100.0 % (473)

1. (p=0.462)

As Table 3 shows, the p-value of 0.462 suggests that the results of this analysis are not robust. We can observe that como remains as the preferred form and, that it was used in close frequencies by speakers with high school and university-level education, 60.5 % and 57.2 % respectively. A similar pattern can be detected with the use of como que (39.5 % and 42.8 % respectively). Thus, we do not find a strong indication which suggests that a speaker’s education constrains the use of these approximators.

Based on the literature regarding how English contact may influence the uses of other linguistic forms (Otheguy, 2013, among others), we took into consideration another external variable: Years of living in the U.S. We operated under the assumption that the amount of years living in the U.S. directly or indirectly exposes the speakers to English contact. Therefore, the reasoning behind this variable was to determine whether we may be able to detect if English contact may influence the use of these forms even though we excluded, among other cases, instances in which como conveyed the meaning of the English focus like or as the English quotative like. A chi-square value (p=0.009) suggests that there is a statistically significant relationship between this variable (i. e. years of living in the U.S.) and the use of como and como que. Nonetheless, several inconsistencies emerged in this analysis which we elucidate in Table 4.

Table 4:

The distribution of como/como que according to speakers’ years in the U.S. (n=473).

Years in the U.S.	Approximator		Total
	Como	Como que
Less than 10	68.9 % (51)	31.1 % (23)	100.0 % (74)
More than 10	55.7 % (214)	44.3 % (170)	100.0 % (384)
Born in the U.S.	86.7 % (13)	13.3 % (2)	100.0 % (15)
Total	58.8 % (278)	41.2 % (195)	100.0 % (473)

1. (p=0.009)

A careful examination of Table 4 suggests that, with regard to como, the speakers who were born in the U.S. generally expressed this approximator in higher frequencies (86.7 %) than the speakers who lived in the U.S. more than ten years (55.7 %). This finding would suggest that the longer speakers live in the U.S., the more likely they are to use this form. However, a contradiction emerged in these data in that the speakers who had lived in the U.S. less than ten years exhibit an even higher frequency in the use of como (68.9 %) than speakers who have lived in the U.S. more than ten years (55.7 %). Therefore, we cannot fully confirm that these tendencies may be suggestive of English contact since there is no apparent increase in the use of these forms that correlates with years of English contact. Furthermore, we do not have sufficient evidence to claim that the use of como and como que resembles the discourse marker or quotative like in English since we excluded this function while defining the envelope of variation as it was not representative of a site in which the alternation could occur with como que.

We also find a similar discrepancy with the patterns exhibited for como que: Speakers born in the U.S. exhibit lower frequency (13.3 %) in using this form when compared to the participants who lived in the U.S. for more than ten years (44.3 %). Figure 2 is illustrative of these inconsistent patterns.

Figure 2:

The use of como/como que according to the speakers’ years in the U.S.

Based on Figure 2, we can only suggest that the speakers in our study who were born in the U.S. show a stronger preference to use como, not como que. The figure also shows the inconsistencies regarding the speakers who had lived in the U.S. more than ten years as opposed to those born in the U.S. That is, speakers born in the U.S. tend to disfavor the use of como que. Therefore, further investigation is needed before we can suggest that there may be a semantic convergence in the use of como since we do evidence a higher frequency in use of this approximator. However, with regard to como que, we cannot make the same claim.

6 Discussion

This study shed light with regard to how Spanish speakers alternate the use of approximators como and como que to convey vagueness. Our first research question addressed the preferences in the use of como and como que according to the speech event in which they were employed (i. e. narratives of personal experience vs. motivational interviews). In general, the analyses revealed that como was preferred over como que. Interestingly, however, we found a more robust tendency in speakers to express como que in MI sessions than in the narratives. Therefore, we found a discursive effect concerning the use of these two approximators.

Meillet (1921) has posited that new forms emerge through the formal resemblance to already established ones. In this study, we were able to confirm Camacho’s (2011) claim regarding the lexicalization of como que as a free variant of como. However, our findings suggest that como may have increased its frequency of use and functions when compared to como que. Nonetheless, we cannot claim that como has categorically supplanted como que since we found evidence of these forms competing in our corpus. Moreover, we found that como que was produced more in the MIs, a finding which may suggest that, while the two forms convey approximation, they may serve distinct discursive-pragmatic functions. That is, como que seemed to be employed in instances in which the speakers lacked certainty or expressed belief.

In addition to the abovementioned finding, the second research question examined external factors (i. e. variety, years of residence in the U.S., and educational attainment) to uncover whether these variables conditioned the use of the approximators. With some reservation, the analyses yielded subtle differences in the use of the approximators according to speakers’ variety. For instance, we found that speakers from the varieties of the Dominican Republic, Honduras, and Cuba exhibited a preference to use como que rather than como. But it is important to note that we are not referring to national boundaries. Rather, we use the linguistic identifiers provided by the speakers of the study (e. g. dialect) to capture regional zones or Spanish variety following Lipski (1994).

The other social variable, years of residence in the U.S., yielded inconsistencies in that an apparent increase in years of exposure did not correlate to a preference of one form over the other. Thus, we note here that further investigation is required. The data also revealed that educational attainment was not significantly correlated with the use of these forms.

Our third research question addressed the pragmatic functions of the approximators como and como que in each speech event. The results obtained from the qualitative analysis yielded other findings with regard to the alternation of these forms and how vagueness was conveyed using them. First, we found that in many instances como was issued to convey numeric approximation and that both forms were used to convey resemblance and proximity when speakers recalled events in their narratives. These observations were found consistently throughout the corpus of oral narratives of personal experience. Concerning the institutional discourse, we found similar functions. However, we found that the functions of como and como que were related to the expression of general approximation to convey imprecision (not numeric approximation), to hedge with respect to medication intake, adherence to treatment, and feelings or behavior-related outcomes. Thus, our findings suggest that while both of these forms functioned to convey vagueness, they did not always share similar pragmatic purposes.

Janney (2002, p. 462) has suggested that “vagueness is a particular way of using language” and, we were able to attest to pragmatic variation in the use of the approximators according to discursive contexts. For example, we found that speakers in the MIs sessions may have employed the approximators to avoid being too straightforward when talking about treatments, medication, and improvement, and to express thoughts, feelings, and emotions that were not clear in the speakers’ minds. Or, said differently, we add the notion that indefiniteness of feelings was also conveyed by using the approximators. Conversely, in the oral narratives, the functions of the approximators were related to a possible failure to recollect an event, to loosely recount an event, or informally skip details in narrative segments.

Thus, we suggest here that pragmatic variation emerged in the manner in which speakers expressed vagueness: In oral narratives of personal experience to informally recount aspects of their stories which did not demand preciseness (e. g. details of quality and quantity). However, in the MIs the purpose of the speakers’ use of the approximators may have been to hedge or to avoid providing exact details about their feelings and their improvement. In other words, while the speakers in the narratives employed the approximators to discuss details from past experiences that they could not recall with precision, the speakers in the MIs used these forms mostly to soften the content of their utterances (e. g. to avoid being too straightforward when discussing emotions related to a treatment or a condition).

To explain these differences, we suggest that we look further into what the institutional discursive setting represented. In general, we note here that the motivational interviews represent institutional talk. Bardovi-Harlig and Hartford (2005, p. 13) defined institutional talk as “spontaneous authentic language used by speakers who are speaking as themselves, in genuine situations, with socio-affective consequences”. In other words, the use of the approximators may have been motivated by the consequences of feeling very exposed while discussing feelings related to treatment or medication, or not wanting to be perceived as too direct. We can also add that they were sharing their beliefs and confidence in the treatment, in the case of the use of como que. For instance, Agar (1985) maintained that institutional discourse accomplishes three things: First, the institution must diagnose the client. Second, the institution identifies specific clients, their problems, and the solutions (i. e. institutional frames). Third, there is a diagnosis, the part of the discourse where a representative adheres to the client’s ways of talking. Clients, on the other hand, come to an institution with a variety of ways of thinking about themselves, their problems, etc. Thus, the use of the approximators could have been constrained by the speech event, which involved an interaction that is inherent in a therapy session and the power relations between client and therapist.

To this end, we add that, to fully explain what prompts speakers to use approximators we need to align these results with Grice’s (1975) maxim of quantity and quality. The maxim of quantity suggests that speakers should make their contribution as informative as required by the context. That is, approximators can be employed to provide the right amount of information that the speaker can recall when he or she does not have full access to specific details. With regard to the maxim of quality, which states that a contribution must be as truthful as possible, we add that approximators allow speakers to communicate the best way they can, even when the content they have in mind is not clear and determinate.

In addition to using approximators when there is lack of information or direct evidence, Channell (1994) argued that speakers might employ approximators when they deliberately withhold information in situations where there is no desire to fully commit to the information that is given, as we were able to attest in the therapy sessions. Furthermore, speakers may use approximators to avoid being perceived as too direct or too straightforward (i. e. thus they mitigate an utterance), all accounts of what was evidenced in the interview sessions. To substantiate these observations from a linguistic perspective, we note here that in a study of commitment language, Amrhein et al. (2003) documented different patterns of commitment strengths among client groups and, this finding may be suggestive of how approximators may contribute to conveying the strength of commitment of the constituent that follows como and como que.

Based on the findings of this study, we note that there is no quantitative clear-cut bifurcated pattern in the use of these approximators. Rather, we found patterns and preferences. Therefore, we would like to add a concern related to epistemic modality. Following Halliday (1970), a common thread we found with the use of como and como que is that they reflected speakers’ assessments and attitudes towards their thoughts and propositions. In several instances, we found that speakers used these forms to express their degree of certainty regarding the content of a proposition and, in others we found that they expressed the speakers’ attitude.

7 Conclusion

This study was set out to investigate the use of the approximators como and como que in various Spanish varieties and spontaneous speech (i. e. narratives of personal experience and MIs) using both sociolinguistic and pragmatic variationist approaches. As we mentioned earlier, and to our knowledge, the Spanish approximators discussed here have not been fully examined in a study that investigates sociolinguistic and pragmatic variation and in two categories of discourse.

One limitation we uncovered was the number of speakers from different Spanish varieties represented in the study. At this moment, we can only hypothesize that a purposeful rather than random sampling may have yielded different results. That is since the data yielded low frequencies in the use of these forms according to several regional varieties (e. g. Argentinean, Honduran Spanish), more speakers who are representative of these varieties perhaps needed to be included.

Regardless of this limitation, we were able to reveal that the use of approximators generated by the corpus served to convey vagueness but in different ways. For instance, and concerning epistemic modality, the approximators under examination reflected the speakers' sense of probability, their attitudes, and commitment to the propositions. That is, in several instances, we noted gaps in recalling details in the narratives while in others we detected concerns related to negative attitudes towards treatment.

Also, and following Jucker et al. (2003), we found that approximation was expressed frequently, yet it did not interfere with communication. In fact, and as noted by Jucker et al. (2003, p. 1737), using vague expressions “may be more effective than precise ones in conveying the intended meaning of an utterance”. Together with the concerns above related to vagueness, the use of these approximators provided relevant contextual information regarding a speaker’s ability to access old information, as well as to express beliefs and manage the degree of commitment the speakers made to their propositions.