FACE-ENHANCING COMPLIMENTS IN INFORMAL CONVERSATIONS IN VALENCIAN SPANISH

This article presents the results of an analysis conducted on politeness in informal conversations in Peninsular Spanish, specifically on the use of compliments in casual Valencian Spanish with face-enhancing effects. Face-enhancing compliments typically have the primary function of constituting and reinforcing the face of the addressee. The compliments that were analyzed consisted of positive comments about the complimentees’ belongings, people related to them, their actions, and their qualities. Most of these face-enhancing compliments were exchanged in situations defined by a high degree of communicative immediacy (Koch & Oesterreicher 1985) and conventionality. The settings where these compliments were produced, plus their discursive, structural and formal features, illustrate their formulaic nature. Among the most important of these features is the fact that there is a predominance of faceenhancing compliments formulated as declarative, copulative and exclamatory sentences. Most of these compliments contain a short repertoire of adjectives with a positive meaning, and are modified by intensifiers that help to reinforce and guarantee the faceenhancing effect of the speech act. They usually consist of isolated utterances occupying a second turn position in adjacency pairs. Lastly, their addressees rarely acknowledged these compliments explicitly.

Based on the literature mentioned above regarding Peninsular Spanish culture, the starting point for this research is that Spaniards make greater use of face-enhancing politeness than of face-saving politeness in their interactions.A study was hence conducted to achieve a better understanding of FEAs in Spanish informal conversations.The study analyzes a corpus of naturally-occurring data collected in Valencia, Spain in an attempt to answer the following two research questions: (RQ1) what are the most frequent FEAs used by native speakers of Valencian Spanish when interacting in informal conversations?And (RQ2) what are the most salient formal, structural and discursive aspects of these FEAs?The results from the analysis showed that the top three most frequent FEAs from the data were compliments (39%), offers (19%), and expressions of empathy (10%).Due to the higher frequency of faceenhancing compliments, this paper will focus on their characterization.Therefore, the main aim of this paper is to present the results on the use and features of faceenhancing compliments in Valencian Spanish casual speech.
The article is organized as follows.After this introduction, there is a description of the characteristics of face-enhancing compliments in Section 2, while Section 3 explains the materials and methodological principles in which the study is based.In Section 4 the results from the analysis are discussed with illustrative examples.The main findings from the analysis are summarized in Section 5, and Section 6 finalizes with the conclusions and points out future steps in the research on face-enhancing politeness in Spanish.

Face-enhancing compliments
Compliments are speech acts that have been widely studied by scholars.It has been concluded that different cultures make different uses of compliments, not only in the devices used to formulate them but also on their frequency of appearance, the contexts where they are considered appropriate, the responses of the complimentees, the recognition of an utterance as a compliment, and the functions performed by the speech act.Regarding their functions, this paper studies face-enhancing compliments (FE compliments), whose primary function is to constitute and reinforce the face of the addressee.Following Archer's (2015) facework scale, which outlines faceenhancing and face-aggravating levels, FE compliments represent the positive end of this scale since they genuinely intend to benefit the hearer.These compliments may have other perlocutionary effects, such as achieving the communicative goal, having a pleasant discourse, and strengthening the interpersonal relationship between the speakers.Sifianou (2001) explains that the face of the complimenter is also positively impacted after a successful FE compliment "by presenting him/herself as a cooperative interlocutor, who knows and observes social conventions of appropriate behaviour, can share the addressee's pleasure in attaining a goal and/or cares about other people's wants" (Sifianou 2001: 399).
With regard to the positive effect that compliments can have in rapport management, Wolfson (1983: 86) describes compliments as "social lubricants serving to create or maintain rapport," and Holmes (1986: 486) defines them as "affective speech acts directed to the addressee which serve to increase or consolidate the solidarity between the speaker and addressee."Furthermore, compliments have been said to be a "positive politeness strategy that express solidarity and esteem and which contributes to soften and harmonize human relationships" (Ferrer & Sánchez Lanza 2002: 43), 3 and they have also been described as an "expressive act that aims to transmit solidarity and esteem to the other interlocutor" (Haverkate 2003: 67). 4 Lastly, Sifianou (2001: 403) holds that "the major function of compliments is to consolidate, increase or negotiate solidarity between interlocutors." As mentioned above, it is important to note that compliments are not always FEAs and can also be used for different goals.Brown and Levinson ([1978] 1987) mention that compliments are potentially face-threatening acts.For instance, they can threaten the addressee's negative face (i.e.their freedom from imposition and freedom of action), since they can incur debts (Brown & Levinson 1987: 66).Drawing on this view, Holmes (1988: 448) remarks that "[c]ompliments can be regarded as facethreatening to the extent that they imply the complimenter envies the addressee in some way or would like something belonging to the addressee."Such interpretation would mean that the recipient of a compliment has to offer the praised object to the complimenter, as has been attested in Samoan culture (Herbert 1990: 219).Archer's (2015) study on strategic facework moves shows that compliments can also function as insults in context.In her data on English and German compliments, Golato (2005) found face-threatening compliments used to criticize, reproach and interrupt.Besides, compliments can also be used to apologize or to reinforce an apology (Wolfson 1981: 123), and to mitigate a face-threatening act, like some requests (Albelda 2005: 104).
The explanation for different interpretations of the same speech act is simple: speech acts are multifunctional and can serve several goals for the speaker.The interpretation of the functions of a particular speech act greatly depends on the interlocutors' culture and on the setting where it is performed.Hence only a complete analysis and a reference to the whole speech situation can reveal the function performed by a particular compliment.This study will interpret a compliment as a FEA when its production enhances the other's face in the first place, so the compliment is not used to mitigate a FTA and there are not indicators to consider it as a FTA, because the complimentee, or any other interlocutor, does not react negatively to it.

Materials and Methods
The data used for this study is a corpus published in 2002 by Briz and Val.Es.Co.Group.It contains the transcripts of nineteen informal conversations in Spanish gathered in the city of Valencia, Spain and its urban area.The authors of the corpus granted access to the audio-recordings of these conversations.The length of each conversation fluctuates between seven and fifty minutes, with a total of four hundred and fifteen minutes of audio-recordings.The number of interlocutors actively participating in each conversation of the corpus varies from two to four, making a sample of seventy-two informants (N = 72), all of them native speakers of Valencian Spanish.There are 30 males and 42 females and three groups of ages: informants who are aged between 18-25 years old (N = 22), between 26-55 years old (N = 40), and older than 55 years old (N = 10).Regarding their socioeconomic status, informants represent three different levels: high (N = 22), middle (N = 24), and low (N = 26). 5ith regard to the relationship among the interlocutors in each conversation, 34 of them are friends, 27 are family members, and 11 maintain other kinds of relationships, such as coworkers or shop-assistant/client.This corpus was chosen because it consists of naturally-occurring data, not of elicited data especially collected for the purposes of this research.The selection of the data included in the corpus followed the parameters established in Briz (1995: 15-19) for providing an adequate and representative sample of the linguistic practices found in this speech community (see Briz and Val.Es.Co. Group [2002] for the selection criteria).Therefore, it is reasonable to think that the conclusions of the research could better reflect the actual use of FEAs in Valencian Spanish interaction.Valencian Spanish is a variety of Catalan Spanish, which is the dialect used in the bilingual regions of Catalonia, the Valencian Community and the Balearic Islands.According to Moreno Fernández (2009), there are around 12,500,000 speakers of Catalan Spanish, making it the largest among all of the dialects of Peninsular Spanish.Another reason why this corpus was chosen for the study of FEAs in Spanish is that it is a representative sample of Spanish casual speech, and informal conversations are the most frequent type of verbal interaction speakers engage in every day, so it will be interesting to know more about Spaniards' use of face-enhancing politeness in this type of discursive genre and register.Finally, the transcription system6 devised by Val.Es.Co.recovers prosodic information (e.g., utterance hesitations, laughter, intonation, and pronunciation), which is extremely relevant to examining politeness (Culpeper 2011).
To obtain a complete and real image of FEAs in informal conversations in Valencian Spanish, the data was qualitatively analyzed in two steps: first, a general qualitative analysis searching for the FEAs in the entire corpus.In order to identify these FEAs, the transcriptions of the conversations were studied together with their audio-recordings.To assess the perlocutionary effects of these speech acts, they were analyzed as utterances produced within a specific speech situation, so their production was studied together with the responses given by the addressees, taking into consideration both context and co-text. 7Once the FEAs in the data were identified, a second step in the qualitative analysis was to study these FEAs in detail, analyzing their formal, structural and discourse features.In order to finalize the qualitative study, a basic quantitative analysis was conducted, counting features and assigning frequencies and percent distributions to some of the patterns found in the data.This quantitative information showed that compliments were the most frequent speech acts used by the informants to enhance the interlocutor's face.

Results
The first step of the qualitative analysis led to the conclusion that many different ways to politely enhance the interlocutor's face existed, such as using signs of agreement and empathy, offering help to the other interlocutor, inviting him or her to attend a party, among others.This Section will focus on part of the results from the second step of the qualitative analysis; more specifically, on the description of the formal (see Section 4.1), structural (see Section 4.2) and discursive (see Section 4.3) characteristics of the most frequent FEA in the data: compliments.From a total of 64 documented compliments, 58 of them were FE, while the remaining 6 were used to repair the other's face after a FTA.Analyzed FE compliments consist of positive comments on objects belonging to the addressee, people related to him or her (who may be present or absent in the setting), his or her actions, and his or her qualities.The following two excerpts (see 1 and 2) will be used to illustrate the most common features of face-enhancing compliments in the data.Both excerpts are part of longer conversations from the corpus. 8In ( 1), interlocutors M, A and E make a series of compliments about the beauty of the dress that R is sewing and the fabric chosen by her.They also highlight how pretty R was going to look in it and her good decision complementing this dress with a jacket.This kind of linguistic behavior enhances R's face, because the interlocutors are stressing R's possessions, natural beauty, sewing skills and good choice.In this framework, compliments do not constitute a threat, but rather boost the addressee's face.(1) Setting: four female friends and neighbors (tagged as M, A, R and E), are gathered in R's house.They are all housewives of low socioeconomic status, aged between 26-55.In this sequence, R has just shown her guests the dress she is sewing to attend a wedding.This gesture brings out several compliments from the guests to the host: 1. R: pues→ pues→ pues bien/ yo me-yo-yo he idoo a comprar hilo y agujas pequeñas/ por lo deel-lo del vestido 9 § 2. A: § ¡ay! 3. M: [es muy majo] 4. A: [¡aay!este queda] muy bonito 5. M: es precioso § 6. A: § esto queda muy bonito/ Roge 10 § 7. R: § por si acaso 8. A: sí 9. R: así que→ 10.E: esto queda muy bonito puesto interpretation of them toward reliance on sequences of talk and co-text for the interpretation of utterances." 8 Original conversation passages are followed by their approximate translation to English. 9R shows the dress she is sewing. 10R's name.The excerpt in ( 2) starts with V's request to see her sister's new lamps.The request already shows an interest in A, which enhances A's face.12As V sees the lamps, she makes positive comments on their beauty, originality, functionality and modernity, so the compliments also generate A's face-enhancement, because they praise her good taste and wise choice. (2) Setting: this conversation takes place in the apartment of A, who has just gotten married.Her sister (tagged as V) is visiting the house, and A is showing her the new lamps.Both are women between 26 and 55 years old of a high socioeconomic status. 1. V: ¿no tienes la lámpara que parece una escultura?§ 2. A: § no↓ ¡uy!/ esa va a tardar un mes § 3. V: § ¿solamente tienes esta? 4. A: no § 5. V: § ¿no? cuenta// la verdad es que es muy bonita ¿se parece a la mía?/// (3") ¿eh? ¿tú has visto la mía?// ¡OSTRAS!¡QUÉ MOONA!// ¡QUÉ GRACIOOSA!§ 6. A: § mira cómo se ((queda)) 7. V: ¿cómo la has encendido↑?§ 8. A: § tocando (3") 13 1) and ( 2): -Declarative sentences: este queda muy bonito 'this one looks very beautiful' (line 4 of Example 1), esto queda muy bonito 'this one looks very beautiful' (line 6 of Example 1), esto queda muy bonito puesto 'this one looks very beautiful on you' (line 10 of Example 1), esto queda precioso puesto 'this one looks lovely on you' (line 12 of Example 1), quedará bonito con la-con el contraste 'it will look beautiful with the-with the contrast' (line 25 of Example 1), pues sí/ sí que ilumina 'so yeah/ it does light up' (line 9 of Example 2); ilumina un MONTÓN/ para lo pequeñita que es 'but it lights up a LOT/ for its small size' (line 20 of Example 2).In all of the examples from excerpt (1) the speakers made use of the verb quedar 'to look' plus an adjective.In ( 2), the speaker tagged as V uses declarative sentences to stress a good quality for a lamp, its good lighting.
With regard to the adjectives chosen by participants to formulate their compliments, bonito/a 'beautiful' is found in 40% of the FE compliments in the data.This can be seen in (1), where this adjective is constantly used by the speakers; its reiteration contributes to the success of the compliment, since all participants agree with the same positive quality.In the data, other adjectives used with similar meaning are precioso 'gorgeous,' guapa 'good looking,' and the more informal majo 'nice,' mono 'pretty' and monada 'lovely.' Lastly, the analysis reveals that 70.7% of the FE compliments were reinforced via different intensifying modifiers.These devices strengthen the positive force of these utterances and, hence, maximize their face-enhancing effect.Table 1 summarizes the intensifiers found in the data, with examples from ( 1) and ( 2).In sum, the results from the analysis of the formal features of FE compliments in Valencian Spanish showed that (1) the most frequent syntactic schemes are declaratives, copular and exclamatory sentences; (2) informants used a short repertoire of adjectives and nouns with a positive semantic load; and (3) the main communicative strategy used by the speakers who produced the compliments was intensification, as a means to magnify the polite effect of the FEA.

Structural features of face-enhancing compliments
The results from the analysis show that 53.7% of the FE compliments are isolated utterances, while 35.2% are combined with different speech acts, and 11.1% are reduplicated (two or more compliments in the same line).Examples of isolated utterances can be seen in lines 3,5,10,16,18,20 and 25 of the excerpt in (1), and there are examples of duplicated or triplicated compliments in line 21 of (1) and in line 16 of (2).With regard to combined compliments, there is one example in line 4 of (1), where the compliment formulated by speaker A is strengthened by a previous sub-act.That sub-act is composed of one exclamation that attracts the attention of the participants in the conversation and shows A's interest and amazement with R's affairs.Vowel lengthening intensifies the exclamation.With these strategies, A tries to guarantee the face-enhancing effect of her compliment.Other possibilities consist of strengthening the compliment by addressing the recipient with forms that encode the positive attitude of the speaker towards him or her, as in line 6 of (1), where A uses R's abbreviated name (Roge, from Rogelia) after the compliment.When using a first name as a vocative, the speaker is pointing out one particular person, which guarantees the right reception of the FEA by its addressee.Also, the use of the first name denotes affection and shows that A knows R very well, as it shows closeness between them.Several authors (Miranda Poza 1998;Placencia 2005) have analyzed the use of first names in different contexts and found that calling someone by his or her first name can involve a sense of intimacy, which contributes to the intensification of solidarity between the interlocutors and to the conversational harmony.Thus, together with the strengthening of the illocutionary force of the compliment, the first name in line 6 of (1) helps to tie the bonds of friendship between the interlocutors.Finally, other devices used in the data for supporting the face-enhancing effect of a compliment are imperatives that fulfill a vocative function (see line 23 of example 1), signs of amazement or surprise (see line 5 of example 2), question tags that attract the others' attention, looking for the other's agreement and reinforcing the assertions (see line 5 of example 1 and line 11 of example 2), and informal structures that reaffirm the statement made by the compliment (see line 24 of example 1).
In conclusion, the structural analysis reveals some of the syntactic schemes that are frequently used in the data for face-enhancing compliments, with more than half of them being isolated utterances, whereas 46.3% of these compliments were reinforced by other sub-acts or reduplicated.The use of sub-acts and reduplication intensifies the face-enhancement effect of the compliments.

Discursive features of face-enhancing compliments
This Section will focus on the discursive properties of face-enhancing compliments in the data, with the aim of gaining a better understanding of their interactive features and their connection to politeness.First, the analysis studied the specific position that FE compliments take up in talk-in interaction.Research shows that compliments can occupy either a first or a second turn position in adjacency pairs, so the identified FE compliments were classified as starting/initial turns or as reactive turns.A starting or initial turn is aimed at eliciting a reaction by the other interlocutor, whereas a reactive turn is used to show agreement or disagreement with the other's statement, to confirm his or her turns, and so on.Also, in oral interaction it is possible to find reactivestarting turns, that is, the reaction to a previous turn and the starting of a new one (Briz 2007).The analysis leads to the conclusion that compliments can be situated at the beginning, at the end, or even in the middle of interactions, with most of them being reactive turns and reactive-starting turns (78%) that are made following others' comments, questions for opinion, or previous compliment refusals.This feature can be seen in the two examples above (see excerpts in 1 and 2).For instance, in (2) V's compliments are reactions to A's explanations about the lamps.This also happens in (1), where the compliments appear right after R explains that she went to buy thread and needles for the dress she is sewing and shows them the dress.The complimenters praise the other's possessions and taste while overlapping their turns or with minimal pauses between them.Rather than a full sentence on a different topic, these overlaps normally include just a few words of encouragement or elaboration on the same subject, so they come to constitute the interlocutors' faces in accordance with their roles.These are examples of what Tannen (1984) calls cooperative overlaps, defined as simultaneous speech that shows support.Starting-turn compliments were also recorded in the data (22%), mainly at the beginning of a conversation when participants first met.In this setting, speakers used compliments as a strategy to start the interaction, praising the other's qualities, new look, and belongings.Additionally, some starting-turn compliments were used to fill silences during conversation.
With regard to the interactive patterns, the analysis showed four different ways to respond to a compliment: a) compliment-no answer; b) compliment-downplayed acceptance; c) compliment-acceptance; and d) compliment-refusal-insistence.Omitting a verbal acknowledgment after the face-enhancing compliment is the most frequent pattern documented in the data (48.4%).Some examples of this can be seen in the excerpts from ( 1) and ( 2).For instance, in (1) R does not answer thirteen of the compliments that are enhancing her face; she just seems to ignore them and continues showing her work.However, she does answer the rest of the non-polite comments or questions made by the interlocutors (see lines 1-11 and 20-32 of example 1).Similarly, in (2) interlocutor A does not produce any answer to the compliments, at least not a verbal answer, but continues showing the lamps and describing their usefulness (see lines 5-6, 9-10, 14-17, and 22-23 of example 2).
After not answering a compliment, the second most frequent uptake in the data is to softly accept it (31.3%),which means to give a response that downplays one's agreement with the praise. 16A list of several strategies was found to downplay the compliment acceptance, which includes: a) transferring the qualitative praise into a functional one; b) awarding a third person/circumstance the merit; c) temporarily limiting the worth of the praise; d) justifying the merit; e) answering in a roundabout way that does not specify agreement or disagreement with the compliment; f) appealing to the speaker's good intentions; and g) showing doubt about the truth of the compliment. 17There is an example of the first strategy in line 19 of ( 2), where the compliment response consists of transferring the qualitative praise into a functional one.In contraposition with the beauty of the lamp highlighted by V, A's response emphasizes the small space the lamp occupies.V takes advantage of this feature to make a new compliment in line 20, showing agreement with A and adding a new good quality of the lamp, its utility.This last compliment is not properly answered by A, who continues the dialogue explaining its mechanism.As with the lack of response, addressees choose to downplay their own face-enhancement to correct the interactional imbalance or to avoid its prolongation.However, it should be specified that some downplayed acceptances are just a preamble to full acceptances, which is something that also happens with some of the refusals to compliments (see lines 12-14 of example 1).
Next, a total of 12.5% of the face-enhancing compliments in the data were fully accepted by their addressees, making it the third most frequent compliment response.It is important to notice the continual search for the equilibrium of faces by the participants, because constantly refusing or ignoring the compliments could also put the conversational harmony at risk.For instance, after receiving tens of compliments, the addressees in examples ( 1) and ( 2) finally show agreement with the complimenters' opinions about the dress (see lines 16-19 of example 1) and lamps (see lines 11-13 of example 2).
Finally, the interactive pattern compliment-refusal-insistence was found in 7.8% of the documented compliments.There is an example of compliment-refusal-insistence in (1), where R repeats the adjective lovely used by her interlocutor in the previous turn to show skepticism with E's statement, thus refusing the compliment.This turn generates a reaction from a different interlocutor, M, who tries to convince R about the truthfulness of their words.Hence, R's modesty produces a polite insistence pattern, that is, the prolongation of the face-enhancing compliments (see lines 12-14 of example 1).Based on this pattern and other examples found in the data, it might be claimed that some of the refusals actually work as a confirmation of the compliment and as a continuity strategy.
In conclusion, the data shows that most FE compliments were reactive or reactivestarting turns that generated four different (but not mutually exclusive) types of responses: acceptance, downplayed acceptance (minimum agreement), refusal, and not answering explicitly to the compliment but providing contributions to divert attention.This last one was the complimentee's most frequent reaction.

Discussion
The majority of the compliments documented in the corpus (90.6%, N = 58) are FEAs: they help the addressee to be seen as someone with valuable goods and qualities, which makes him or her stand out from the rest of the members of the group.Also, these compliments show that the speakers have a close relationship, which helps to confirm the interlocutors' right to treat and be treated with confianza.Therefore, these compliments were considered as FEAs, because they had a good impact on the recipient's face, helping to constitute and reinforce it.Additionally, while speakers enhance the face of others they can also achieve their own faceenhancement, which has an effect on their personal, communicative and social success.The remaining 9.4% of the documented compliments can be categorized as face-saving acts, since they were primarily used by the informants to repair facedamages after a FTA.
With regard to the expression of FE compliments, there is a short repertoire of adjectives, strategies and syntactic forms in the data.This finding is in accordance with the results from other studies that stated the formulaic nature of compliments (Holmes 1986;Herbert 1989;Golato 2005).Besides, more than half of the compliments were simple, brief, and isolated utterances, that is, compliments that enhanced the face of the other not being accompanied by or modified by sub-acts or supporting movements.However, a noteworthy number of reinforced compliments was found (46.3%).Reinforcing devices supported the face-enhancing effect of the compliment, and consisted of doubling or tripling the compliment in the same turn or accompanying it with supporting movements that strengthened their illocutionary content.Some of those supporting movements were signs of empathy, agreements, enlisting a third person to verify the truthfulness of a compliment, presenting one's own opinions as the only valid ones, and making suggestions.
Considering that the main goal of a face-enhancing compliment is to bolster the other's face, one might expect some kind of acknowledgement from the complimentee.In the corpus, however, that reaction was not always found.The lack of verbal uptake and agreement do not seem to disrupt the interactional balance, probably due to the following four reasons.First, the high degree of communicative immediacy that characterizes 95% of the conversations: compliments are mostly exchanged in familiar settings and between participants that maintain a close relationship, such as friends and family.In fact, when asked about the importance of being polite with family and friends, the native speakers of Spanish participating in Barros García and Terkourafi's (2014b) study on first-order politeness stated that it is unnecessary and even inappropriate to use politeness when there is a close relationship between interlocutors and when interactions are predominantly informal.It is important to clarify that these informants understood politeness as formal language, such as using usted instead of tú, giving thanks and saying please.However, this finding can explain why the speakers in the Val.Es.Co.corpus never replied to the FE compliments with an overt expression of gratitude.While using thanks is the basic rule that parents in many cultures, including Spanish, teach their children and that can be found in travel guides to customs and etiquette (Herbert 1990: 207-208), in the data complimentees never opted for this response. 18However, it seems likely that saying thanks after a compliment is a right and safe (non-face threatening) answer when the complimentee's speaking proficiency is not completely developed, as is the case with children and non-native speakers.Also, there might be more chances to find thanks in formal registers or in communicative situations involving strangers, because formality and lack of closeness recommend the recipient's agreement with the other person.Nevertheless, further research is necessary to support these suppositions.Linked to this last idea of closeness is Sifianou's (2012) observation about the role of disagreements as the preferred response when there is communicative immediacy, as in the data, because they can "be a sign of intimacy and sociability and may not destroy but rather strengthen interlocutors' relationships" (Sifianou 2012(Sifianou : 1554)).
The analysis of the interactive properties of the FE compliments in the data reveals another important reason for the lack of explicit acknowledgement by the complimentees: a high percentage of the FE compliments (78%) are produced after someone's comments, so they are reactive turns responding to previous ones, not expecting a response. 19Moreover, complimenters already worked their own face while behaving in a socially appropriate way.In this sense, the equilibrium between the face-wants of both interlocutors has already been achieved, so there is no need for the complimentee to respond.20 Indeed, a third reason why complimentees may choose to omit an answer after a FE compliment is to avoid the prolongation of their own face-enhancement, because this could generate interactional imbalance (Goffman 1967;Bayraktaroğlu 1991).
The fourth and last reason that explains the lack of acknowledgement after a FE compliment is that, in the data, this pattern was specially found after culturally expected compliments produced by guests to hosts while paying a visit, with the aim to constitute and strengthen the host's face (praising his or her personality, actions, house and neighborhood), as a way to show their appreciation.To illustrate, a thorough analysis of the compliments in examples ( 1) and ( 2) makes it possible to claim that these compliments were somehow expected.The interlocutors are gathered in the houses of the recipients of the compliments; consequently, in accordance with Spanish cultural premises 21 (Hernández Flores 1999, 2002; Bernal 2007; Bravo  2008), they are fulfilling their guest role, which supposes praising the host's possessions.In addition, in (1) R is showing her friends a handmade dress but does not seem to be very sure about the result, so positive comments about the dress are expected from M, A and E to convince their reluctant friend.As Sifianou (2001: 418) explains for her data on Greek compliments, by paying these compliments speakers are boosting the other's morale and increasing solidarity between interlocutors.Lastly, the abundance of compliments in culturally expected settings in the data is in line with the general association of compliments with social conventions of polite behavior rather than with the expression of genuine feelings (Sifianou 2001: 392).
Other responses after FE compliments were recorded, such as downplayed acceptances and refusals, and these normally came after non-expected compliments or after a series of several compliments addressed to the same recipient.In those situations, addressees preferred to remain extremely modest, because their faceenhancement could generate interactional imbalance.Downplayed acceptances and refusals frequently generate a reiterative series of turns that could be called a polite insistence pattern: by repeating the FEA, the speaker aims to obtain the agreement from the addressee and, in this way, politely enhances the interlocutor's face. 22All in all, the patterns found in the data indicate that the informants preferred to avoid selfpraise rather than disagreement, since 87.5% of compliments were ignored, downgraded or refused by the complimentees.Lorenzo-Dus (2001: 110) claims that the preference for avoiding self-praise is characteristic of rapprochement or solidarity cultures, such as Spanish culture, whereas distancing cultures place a higher value upon avoiding disagreement.
Apart from the reasons explained above for the low frequency of straight acceptances in the data is also explained by the fact that responses other than acceptances may serve to increase or consolidate the solidarity between interlocutors.In Herbert's words (1990: 209), compliments are not only statements of admiration and praise but rather are expressions of solidarity.Directly accepting a compliment may sound awkward in conversations between participants who maintain a close relationship, as in the data.This is also the case in Syrian Arabic, "because it appears to signal the end of the conversation" (Nelson et al. 1996: 429).
In sum, the results indicate that FE compliments present many formulaic features in the data, and that they are common in situations where interpersonal goals are important, as in informal conversations among family members and friends, with the aim of maintaining or strengthening the emotional bonds. 23

Conclusions
The research reported in this paper started with the assumption that face-enhancing politeness is an important feature of Spanish interaction, given the significance that the notion of confianza has for face in Spain and the characterization of Spanish culture as a rapprochement culture.The purpose of the present study was to learn more about the use of FEAs by Spaniards in Valencian Spanish casual speech (RQ1), and the characteristics of those FEAs (RQ2).This article was dedicated to the description of the most frequent FEA found in the data: compliments.The majority of the compliments in the data were used by the speakers as a vehicle for the (re)constitution and reinforcement of face, so they are categorized as face-enhancing acts.The high percentage of FE compliments in the Val.Es.Co.corpus makes this speech act the main mean to show closeness and solidarity in informal conversations in Valencian Spanish.These face-enhancing compliments present some specific formal, structural and discursive features.Among the most important of these features is the fact that there is a predominance of FE compliments (1) that fall into three syntactic patterns: declarative, copulative or exclamatory sentences; (2) containing an also restricted set of semantically positive adjectives; (3) modified by intensifiers that help to reinforce and guarantee the positive effect of the speech act; (4) consisting of isolated utterances; (5) occupying a second turn position in adjacency pairs; (6) responded to by their addressees with other than direct acceptances; and (7) exchanged in situations marked by a high degree of conventionality and communicative immediacy.Although it is not possible to conclude that these features correlate unequivocally with FE compliments, because the interpretation of any speech act depends entirely on the context, the analysis of the data showed that the aforementioned features have the effect of demonstrating involvement and therefore sincerity in the performance of FE compliments, and these are exemplified in the extracts ( 1) and ( 2).Additionally, these features reflect the formulaic nature of compliments.
Getting to know more about the characteristics of FE compliments is useful for recognizing and producing them in speech.In addition, this knowledge helps to predict the use of FE compliments and to recommend this behavior to properly respond to some Spanish cultural expectations.This information could be applied pedagogically to the design of materials focused on the development of the competences of Spanish speakers.By knowing more about how FEAs work in Spanish casual speech, both native and non-native Spanish speakers will be able to make an efficient use of politeness and, therefore, to achieve their communicative goals and to be socially successful.This conclusion mirrors Aston's (1993: 229-230) finding that non-native speakers' success depends more on the right use of strategies that help to establish friendly relationships than on strategies used to avoid facethreats.
Finally, it is important to note that the conclusions of this study pertain to conversational practices found in a regional language variety of Spanish (Valencian Spanish), in one type of discursive genre and register (informal conversations) and for one type of speech act (FE compliments).Future research should contrast the results from this study with works based on other Spanish varieties as well as varieties from other languages, works based on other genres and registers, the characteristics of other FEAs and anti-FTAs in the data, and the use of other types of politeness.

Table 1 :
Intensifiers used to reinforce the face-enhancing effect of the compliment , where the predicate is placed at the beginning of the sentence, instead of the subject, to highlight the praised quality: muy bonita es la tela 'very beautiful the fabric' (line 21 of Example 1).