3.1Participants and data

The clicks in this study are part of a larger corpus of clicks, produced by 20 native speakers of Spanish from Spain, discussed in Pinto and Vigil (2018)Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. . For the present analysis, we excluded two of the original speakers due to concerns about their gestures not being discernable in the video, giving us a total of 18 speakers. Although we do not take into consideration the variables of gender, age and profession, there are nine females and nine males ranging from 30 to 62 years of age whose professions include politics, journalism, and the arts. These participants appeared in interviews that were publicly available online. In order to ensure a robust corpus of clicks, we selected these interviews by applying the minimum baseline that a speaker must produce at least five clicks in the first five minutes of the interview. This selection process is not meant to yield findings that are necessarily representative of the general Spanish population, especially with regard to frequency of click usage. The total amount of actual speaking time of the interviewees for the 18 videos was 411 minutes (191 for females, 220 for males). While we used the video footage for our analysis, we also converted the videos to mp3 audio files, allowing us to analyze different aspects of the interviewee’s speech using Praat (Boersma and Weenink 2007Boersma, Paul, and David Weenink 2007 Praat: Doing Phonetics by Computer (Version 4.5.14). http://​www​.praat​.org/).

In Pinto and Vigil (2018)Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. and Wright (2011b)Wright, Melissa 2011b “The Phonetics-interaction Interface in the Initiation of Closings in Everyday English Telephone Calls.” Journal of Pragmatics 43: 1080–1099. , clicks in Spanish and English respectively can be produced with both alveolar/dental and bilabial points of articulation. To represent the clicks in the examples taken from our data corpus, we opt for the symbol © which does not specify place of articulation.

3.2Classification of clicks

The 281 clicks in this study comprise the two categories that occurred with the highest frequency in our previous study, ‘Search’ (n = 173) and ‘Search-Stance’ (n = 108). For that study, the coding process that we implemented entailed two evaluators assigning functions individually. At the completion of the coding process, we carried out a Cohen’s Kappa inter-rater reliability test which yielded [ p. 88 ]a rating of Kappa = .84, considered to be very good agreement (Altman 1991Altman, Douglas G. 1991 Practical Statistics for Medical Research. London: Chapman and Hall.). The subset of clicks selected for the present study occur at the turn-medial position, meaning some other linguistic element (words or fillers) occurred immediately before the click. While the larger corpus used in Pinto and Vigil (2018)Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. did contain turn-initial clicks, this particular category of ‘opening clicks’ was not included in this present study. ‘Search’, frequently referred to as a word search, has been reported as a common function for English clicks (González Temer 2014González Temer, Verónica 2014 “Clicks in Chilean Spanish Conversation.” In Proceedings of the First Postgraduate and Academic Researchers in Linguistics at York, ed. by Theodora Lee, 74–99. www​.york​.ac​.uk​/language​/ypl​/PARLAY2013​/PARLAY%20Proceedings​.pdf; Ogden 2013Ogden, Richard 2013 “Clicks and Percussives in English Conversation.” Journal of the International Phonetic Association 43: 299–320. ; Wright 2005Wright, Melissa 2005 “Studies of the Phonetics-interaction Interface: Clicks and Interactional Structures in English Conversation.” Doctoral Dissertation, University of York, UK., 2011aWright, Melissa 2011a “On Clicks in English Talk-in-interaction.” Journal of the International Phonetic Association 41: 207–229. ). Given that searches can involve momentary delays as the speaker looks for the best way to articulate his or her thoughts, we prefer to call this category ‘Search’ rather than ‘Word Search’. To illustrate an example of the search function, (1) shows an excerpt taken from an interview with a Spanish journalist who specializes in international soccer. In this segment he talks about a time earlier in his career when he wanted to exchange videos of soccer games with people in different countries. With this in mind, he placed an advertisement in a sports magazine to see if others were interested in trading videos. The speaker’s utterance here contains two searches, the first for recibí ‘I received’ without a click, occurring after a pause of 0.73 seconds, and the second for contestación ‘answer’, appearing after a pause of 0.97 seconds, a click and the slightly elongated filler eh ‘uh’. For the second search the speaker also exhibits a thinking face.

The second category, ‘Search-Stance’, is a hybrid of two functions. Similar to previous research on discourse markers (Brinton 1996Brinton, Laurel J. 1996 Pragmatic Markers in English. Berlin/New York: Mouton de Gruyter. ; Schiffrin 1987Schiffrin, Deborah 1987 Discourse Markers. Cambridge: Cambridge University Press. ), more than one function can be attributed to a particular linguistic device. In conjunction with the search, the stance aspect of this category refers to a variety of stance-taking or affective displays (González Temer 2014González Temer, Verónica 2014 “Clicks in Chilean Spanish Conversation.” In Proceedings of the First Postgraduate and Academic Researchers in Linguistics at York, ed. by Theodora Lee, 74–99. www​.york​.ac​.uk​/language​/ypl​/PARLAY2013​/PARLAY%20Proceedings​.pdf; Ogden 2013Ogden, Richard 2013 “Clicks and Percussives in English Conversation.” Journal of the International Phonetic Association 43: 299–320. ; Wright 2011aWright, Melissa 2011a “On Clicks in English Talk-in-interaction.” Journal of the International Phonetic Association 41: 207–229. ; Reber 2012Reber, Elisabeth 2012 Affectivity in Interaction. Amsterdam/Philadelphia: John Benjamins. ). This description is used for cases in which “the speaker appears to be searching for the right way to say something while also displaying an emotional reaction to the topic that may be potentially offensive, controversial, disappointing, exciting, etc.” (Pinto and Vigil, 2018Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. , 450) In some cases, the stance displayed seems to be more metadiscursive in nature, an attitudinal commentary on the speaker’s word choice or lack of ability to find the right way to say something, rather than a reaction to the actual topic itself.

In Example (2), the interviewer asks a professor about how her background from a working-class family shaped her as a person. She discusses a time from her [ p. 89 ]childhood when her family did their best to help out others who were in need. She also remembers being exposed to other children with lice and getting lice herself, one of many memories that contributed to an awareness of her working-class background. These experiences are framed in a positive light in the sense that, thanks to her humble upbringing, she learned to empathize with the needs of others.

In addition to co-occurring with a search for por suerte ‘luckily’, the click is a manifestation of the speaker’s desire to effectively express the ironic juxtaposition of unpleasant experiences and their beneficial impact, especially in hindsight from an adult perspective. The stance here also seems partially metadiscursive, a passing moment of dissatisfaction as she grapples with the best way to articulate this contradictory notion of feeling fortunate to have faced discomfort in order to learn an important lesson.

3.3Object of search, gestures and click environment

Other studies on searches have used object of search as one of the important aspects to investigate (Wright 2005Wright, Melissa 2005 “Studies of the Phonetics-interaction Interface: Clicks and Interactional Structures in English Conversation.” Doctoral Dissertation, University of York, UK.; Hayashi 2003Hayashi, Makoto 2003 “Language and the Body as Resources for Collaborative Action: A Study of Word Searches in Japanese Conversation.” Research on Language and Social Interaction 36.2: 109–141. ; Brouwer 2003Brouwer, Catherine E. 2003 “Word Searches in NNS-NS Interaction: Opportunities for Language Learning?” Modern Language Journal 87: 534–545. ). The categories we code for are Noun, Adjective, Adverb, Verb and Means of Expression. Noun can be a single noun (ilustraciones ‘illustrations’), or a noun phrase (el presidente Jefferson ‘President Jefferson’). Adjective can be a single adjective (digitales ‘digital’, espectacular ‘spectacular’), or an adjectival phrase (menos frío ‘less cold’). In our data, Adverb refers to adverbial phrases (de esta forma ‘this way’, por suerte ‘luckily’). Verb may include a verb (invadir ‘to invade’) or verb phrase (me ha cautivado ‘has captivated me’). With this first series of categories, the speaker interrupts the flow of speech to engage in the cognitive process of accessing the mental lexicon. The idea here is akin to somebody searching and saying ‘what is that word I’m trying to think of?’ With the last category, Means of Expression, rather than a search for a particular lexical unit (noun, noun phrase, etc.), the speaker is thinking [ p. 90 ]about how to best articulate his/her thought or idea. The search here could be conveyed with the question ‘how can I say this?’ or ‘what do I want to say here?’ The segments that occur with Means of Expression would largely be considered, in grammatical terms, a clause or a sentence. For example, an actor in our corpus paused and clicked while searching for, or thinking about how to express, the following statement: si miro hacia atrás he hecho más drama que comedia ‘if I look back I have done more drama than comedy’. In some instances, the searched-for item is not necessarily the first to appear immediately after the click (e.g. when a filler word such as eh ‘uh’ occurs before a searched-for noun).

Previous research on searches and clicks has shown that certain types of body language play a role in revealing information about a speaker’s mental state or attitudes (e.g., Goodwin 1983Goodwin, Marjorie Harness 1983 “Searching for a Word as an Interactive Activity.” In Semiotics, ed. by John N. Deely, and Margot D. Lenhart, 129–138. New York: Plenum.; Goodwin and Goodwin 1986Goodwin, Marjorie Harness, and Charles Goodwin 1986 “Gesture and Coparticipation in the Activity of Searching for a Word.” Semiotica 62: 51–75.; Eckman 1979Eckman, Paul 1979 “About Brows: Emotional and Conversational Signals.” In Human Ethology: Claims and Limits of a New Discipline: Contributions to the Colloquium, ed. by Mario von Cranach, Klaus Foppa, Wolfgang Lepenies, and Detlev Ploog, 3–55. Cambridge: Cambridge University Press.; Ogden 2013Ogden, Richard 2013 “Clicks and Percussives in English Conversation.” Journal of the International Phonetic Association 43: 299–320. ; Argyle and Cook 1976Argyle, Michael, and Mark Cook 1976 Gaze and Mutual Gaze. Cambridge: Cambridge University Press.; González Temer 2014González Temer, Verónica 2014 “Clicks in Chilean Spanish Conversation.” In Proceedings of the First Postgraduate and Academic Researchers in Linguistics at York, ed. by Theodora Lee, 74–99. www​.york​.ac​.uk​/language​/ypl​/PARLAY2013​/PARLAY%20Proceedings​.pdf). Based on some of these findings, we selected six categories of gestures to analyze. While these categories do not exhaust all the possible options, we determined that they are among the most observable in our corpus. The six categories are: gaze/thinking face, head movement, brow movement, lip movement, blinking, and shoulder movement.

The literature on word searches has highlighted the importance of gaze aversion and the commonly referred to “thinking face” (Argyle and Cook 1976Argyle, Michael, and Mark Cook 1976 Gaze and Mutual Gaze. Cambridge: Cambridge University Press.; Goodwin 1983Goodwin, Marjorie Harness 1983 “Searching for a Word as an Interactive Activity.” In Semiotics, ed. by John N. Deely, and Margot D. Lenhart, 129–138. New York: Plenum.; Goodwin and Goodwin 1986Goodwin, Marjorie Harness, and Charles Goodwin 1986 “Gesture and Coparticipation in the Activity of Searching for a Word.” Semiotica 62: 51–75.; Hayashi 2003Hayashi, Makoto 2003 “Language and the Body as Resources for Collaborative Action: A Study of Word Searches in Japanese Conversation.” Research on Language and Social Interaction 36.2: 109–141. ). When speakers engage in a word search, they often withdraw their gaze from the listener. According to Goodwin (1983Goodwin, Marjorie Harness 1983 “Searching for a Word as an Interactive Activity.” In Semiotics, ed. by John N. Deely, and Margot D. Lenhart, 129–138. New York: Plenum., 130), when looking away, the “speaker adopts a gesture which is recognizable in our culture as a thinking face, a gesture which embodies the activity of trying to remember a word.” In our study, since they occur hand in hand, we combine gaze aversion and thinking face into one category.

In two studies on clicks, Ogden (2013)Ogden, Richard 2013 “Clicks and Percussives in English Conversation.” Journal of the International Phonetic Association 43: 299–320. in English and González Temer (2014)González Temer, Verónica 2014 “Clicks in Chilean Spanish Conversation.” In Proceedings of the First Postgraduate and Academic Researchers in Linguistics at York, ed. by Theodora Lee, 74–99. www​.york​.ac​.uk​/language​/ypl​/PARLAY2013​/PARLAY%20Proceedings​.pdf in Spanish, head movement is mentioned as a possible gesture that may accompany the click. In our analysis, we consider head movement to include moving the head forward, up and down or side to side. In cases in which head movement is a natural result of gaze aversion, we do not count it as a separate action. In addition to head movement, González Temer (2014)González Temer, Verónica 2014 “Clicks in Chilean Spanish Conversation.” In Proceedings of the First Postgraduate and Academic Researchers in Linguistics at York, ed. by Theodora Lee, 74–99. www​.york​.ac​.uk​/language​/ypl​/PARLAY2013​/PARLAY%20Proceedings​.pdf also observes lip protrusion as a gesture that can occur with stance clicks. Different types of lip movement are also prevalent in our data, not only as lip protrusion but also as lip pursing, lip licking and raising the corner of the lip. Regarding lip pursing, this gesture often coincides with the prolonged use of mmm.

The last three gestures entail brow movement, shoulder movement, and blinking. According to Eckman (1979)Eckman, Paul 1979 “About Brows: Emotional and Conversational Signals.” In Human Ethology: Claims and Limits of a New Discipline: Contributions to the Colloquium, ed. by Mario von Cranach, Klaus Foppa, Wolfgang Lepenies, and Detlev Ploog, 3–55. Cambridge: Cambridge University Press., brow movement is common with word searches. In these cases, “the brow actions may serve to signal the listener not to interrupt or take over the speaker turn.” (Eckman 1979Eckman, Paul 1979 “About Brows: Emotional and Conversational Signals.” In Human Ethology: Claims and Limits of a New Discipline: Contributions to the Colloquium, ed. by Mario von Cranach, Klaus Foppa, Wolfgang Lepenies, and Detlev Ploog, 3–55. Cambridge: Cambridge University Press., 186). While we did not see shoulder movement and blinking reported in other studies on searches or [ p. 91 ]clicks, we observed in our initial analysis of the data that these could be important. For shoulders, we consider noticeable upward movement of the shoulders, which could be referred to as shrugging. Regarding the last category, blinking, we count either heavy and exaggerated blinking, or multiple blinking, as in a fluttering of the eyelids. Lastly, although hand movement is widespread in our corpus, given the fact that the hands are not visible in all of the videos for the duration of the interview, we opted not to code for this.

One question that surfaced during our analysis of gestures was how far from the click a given gesture would be considered relevant to the function of the click. Since all of the clicks in this study involved searches, we concluded that the window for coding gestures would correspond to the duration of the speaker’s search process. In order to determine the duration of the search, we relied on linguistic and non-linguistic evidence, such as gaze aversion and the use of fillers or pauses that indicate that the flow of speech was being interrupted.

Existing research on clicks highlights the importance of co-occurring elements in the environment contiguous to the click such as fillers, prolonged words, pauses and repetitions (Wright 2005Wright, Melissa 2005 “Studies of the Phonetics-interaction Interface: Clicks and Interactional Structures in English Conversation.” Doctoral Dissertation, University of York, UK.; Pinto and Vigil, 2018Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. ). These features, in conjunction with the click, may provide further evidence that the speakers are in the process of searching or monitoring their speech. For each click in our data, we determined whether the preceding element was a non-filler lexical item, a filler or a pause (a minimum length of 0.5 seconds). The same coding system was employed in the post-click environment. A separate category was utilized for repetitions across the click when the same word was repeated immediately before and after the click. Lastly, if the word or filler before and after the click was elongated over 0.5 seconds, we coded this separately.

3.4Research questions

The quantitative research question that we set out to answer in our analysis is as follows:

1. What are the characteristics of the click functions of Search and Search-Stance and, applying the criteria described below, is there statistically significant evidence that they are distinct groups?

   1. Object of search (Noun, Adjective, Adverb, Verb and Means of Expression)

   2. Gestures (gaze/thinking face, head movement, brow movement, lip movement, blinking, and shoulder movement)

   3. Contiguous environment (fillers, prolongation and pauses in both pre- and post-click position; repetitions across the click)

The qualitative research questions that guide our analysis are as follows:

1. What do clicks contribute to a search episode?

2. How do Search and Search-Stance differ and how does multimodality play a role in conveying this difference?

4.1Quantitative results

The first research question concerns the characteristics of the click functions of Search and Search-Stance. Starting with the element being searched for, our goal is to determine whether or not there is statistically significant evidence that Search and Search-Stance are distinct groups. Although she does not provide quantitative findings, Wright (2005)Wright, Melissa 2005 “Studies of the Phonetics-interaction Interface: Clicks and Interactional Structures in English Conversation.” Doctoral Dissertation, University of York, UK. observed that nouns were typically the most searched-for item in her English data set. Table 1 displays the frequencies and percentages of the object of search for both Search and Search-Stance. We ran two-proportion z-tests for each of the five categories, with the null hypothesis being that there would be no difference (α < 0.05). The results of the tests yielded statistically significant differences for Noun (p = 0.000) and Means of Expression (p = 0.001). From a total of 173 Search clicks, 44 were used for Noun, accounting for 25% of the searched for objects, whereas only 6 of the 108 Search-Stance clicks (or 5%) were used for Noun. On the contrary, from the 108 Search-Stance clicks, 83 (or 77%) were used for Means of Expression, versus only 102 out of 173 (or 59%) of the Search clicks.

With regard to gestures, we set out to discover if results from the z-tests provide evidence indicating that Search and Search-Stance are distinct groups (α < 0.05). Table 2 contains total occurrences and percentages for the six categories that we analyzed. Head movement shows a statistically significant difference [ p. 93 ]between Search and Search-Stance (p = 0.013). While head movement was more common with Search-Stance, accounting for 54% of these clicks, it occurred with only 39% of Search clicks. For the rest of the gestures, the findings are not statistically significant.

When taking into consideration both Search and Search-Stance, gaze and lips are the most frequently used gesture. As mentioned above, averting gaze has been widely reported in other studies on searches, and the phenomenon commonly involves two aspects, both gaze aversion and thinking face. Although a statistically significant difference for gaze is not observable between the two groups, usage is high for both Search (92%) and Search-Stance (95%). Lip movement was the second-most frequent gesture in our corpus, 77% for Search and 71% for Search-Stance, a result that may be partially attributed to the high occurrence of lip protrusion coinciding with the common filler mmm.

For the last group, the elements found in the environment contiguous to the click, our objective is once again to determine whether or not the results from z-tests show evidence indicating that Search and Search-Stance are distinct groups (α < 0.05). In Table 3 we include the nine categories that we identified as potentially relevant to click usage. The three categories that demonstrate statistical significance are the following: pre-lexical, repetitions, post-filler and post-prolongation (small sample size).

Starting with the position immediately before the click, pre-lexical items are more common than fillers or pauses with Search-Stance (47%) than with Search (34%). To illustrate the use of a lexical item preceding a click, in Example (3) below, the word cuando (‘when’) occurs before the click.

Repetitions across the click are more frequent with Search (35%) than with Search-Stance (20%). In the segment in (4), one can see that the speaker repeats the verb son (‘are’) before and after the click:

Post-fillers, fillers appearing directly after the click, are more common with Search (23%) than with Search-Stance (7%). In Example (5), the speaker utters pues (‘well’) after clicking and a short pause.

Lastly, the category of post-prolongation occurs more often with Search (14%) than with Search-Stance (2%). In Example (6), the speaker elongates the y (‘and’) after the click.

To summarize, the quantitative results show the following: Search is more likely to entail trying to find the appropriate noun (Table 1), it is associated with a higher frequency of certain collateral signals (post-fillers, post-prolongations and repetitions across the click) (Table 3), and it is less commonly associated with head movement (Table 2). Search-Stance, on the other hand, shows a higher [ p. 95 ]likelihood to involve searching for a means of expression rather than a noun or single word (Table 1), a more frequent use of head movement (Table 2) and pre-lexical items (Table 3), as well as a weaker association with post-fillers, post-prolongations and repetitions across the click (Table 3). We discuss these findings in more detail below in Section 5.

4.2Qualitative results

The first question we want to address concerns what exactly the click contributes to a search episode. Since searches often occur without clicks, they have to be considered an optional resource that speakers can draw on to communicate additional information. Wright (2005)Wright, Melissa 2005 “Studies of the Phonetics-interaction Interface: Clicks and Interactional Structures in English Conversation.” Doctoral Dissertation, University of York, UK., in her analysis of telephone conversations in English, suggested the possibility that the search click was used to indicate that the speaker is still searching for a word and that the recipient should not interrupt. The author emphasizes the fact that, since her data consist solely of telephone conversations, due to the absence of visual cues like thinking face, the click in these instances carries out an important turn-holding function. In Pinto and Vigil (2018)Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. , we point out that search clicks, and clicks in general, tend to come after a notable pause and shortly before the person resumes speaking. While the pre-click pause average was 0.41s for Search, the post-click pause average was half as long, 0.19s (Pinto and Vigil, 2018Pinto, Derrin, and Donny Vigil 2018 “Clicks as discourse markers in Peninsular Spanish.” Spanish in Context 15(3): 441–464. ). Given that search clicks generally appear further toward the end of the search process (see Example 1 above), it does not seem logical that they are primarily being used to signal holding the floor while the speaker is thinking. This evidence suggests that these search clicks actually indicate something else, such as uncertainty about whether or not the speaker has the right word(s) to express his or her thoughts, comparable to a type of modality marker. Similar to what Roggia (2012Roggia, Aaron B. 2012  Eh as a Polyfunctional Discourse Marker in Dominican Spanish. Journal of Pragmatics 44: 1783–1798. , 1789) points out for eh in Spanish, the click might signal that the upcoming segment of discourse is “something that has been given thought”. Yet another explanation can be found in Ward (2006)Ward, Nigel 2006 “Non-lexical Conversational Sounds in American English.” Pragmatics and Cognition 14: 129–182. , in his study of clicks in English. Ward (2006)Ward, Nigel 2006 “Non-lexical Conversational Sounds in American English.” Pragmatics and Cognition 14: 129–182. refers to the idea of clicks expressing dissatisfaction, whether it be about the topic or something related to the actual speaking process itself. In the case of a search, the speaker can be dissatisfied with his or her own performance (e.g. the inability to find the right words and the need to pause), which may explain why the click appears toward the end of a delay when the speaker detects processing trouble. This notion of dissatisfaction brings us to the next qualitative research question regarding how Search-Stance differs from Search and the role that multimodality plays.

To reiterate, Ward’s aforementioned comment underscores the link between searching for what to say and the different possible attitudes that surface in conjunction with the search, such as feeling dissatisfied or frustrated about the interruption [ p. 96 ]in the flow of speech, about the word selection or a combination of both. In our data, while searches displaying detectable attitudes such as dissatisfaction were coded Search-Stance, it might be possible to propose that all search clicks indicate some degree of speaker dissatisfaction with his or her performance, regardless of whether or not such an attitude is always perceivable to the listener (or evaluator, in our case). From this perspective, what distinguishes Search from Search-Stance is likely a matter of degree, depending largely on how many linguistic and non-linguistic signals come together to express the stance in question. Here is where multimodality comes into play in the sense that numerous linguistic and extralinguistic factors combine, including topic, lexical/semantic content, pauses, prolongation, fillers, repetition and gestures. In isolation, each factor might not convey the stance in question, but together they have a compounded effect. A particular attitude surrounding a click may only be evident when a certain threshold level is reached, when the co-occurring signals align in such a way that a stance is conveyed. Examples (7) and (8) described below illustrate how multimodality manifests itself with Search-Stance clicks.

In an interview with a politician from the region of Castilla-La Mancha (7), the interviewer asks about the political and environmental lessons learned from an experience involving a recent fire at a landfill with 15,000 tons of discarded tires. After opening with a popular saying about not procrastinating, he uses the filler eh ‘uh’ which is slightly elongated, averts his gaze away from the interviewer, clicks, and then simultaneously furrows his brow while heavily blinking before pursing his lips and continuing to speak by saying es verdad que ‘it’s true that’. This segment of his utterance starts off with a combination of repetitions (que este que que este ‘that this that that this’). It is the sum of interrelated elements here that express the Search-Stance. The linguistic/discursive factors include the sensitivity of the topic (i.e. a costly and politically controversial fire), the expression of an opinion, the click, the elongated filler eh ‘uh’ and the use of repetition. The contributing extralinguistic aspects consist of gaze aversion, brow movement, exaggerated blinking and lip pursing.

In another interview (Example 8), an actor is speaking about the breakup of an important relationship in his past. He suggests that the end of the relationship was one of the contributing factors that impelled him to go on to become a successful actor. In this segment, the speaker looks down and exhibits thinking face and then produces a false start ending with an elongated yo ‘I’ that trails off (en fin que bueno que yo::: ‘anyway it’s good that I:::’). He then begins another attempt with mira ‘look’, purses his lips as he utters a lengthy mmm that lasts 1.2 seconds while simultaneously shaking his head, and raises his brows as he exaggeratedly blinks his eyes. Lastly, the speaker flutters his eyelids between the two repeated tokens of que ‘that’. Here again, we have a case of multimodal elements working together to express Search-Stance. In addition to the aforementioned characteristics, the topic concerns conflicting emotions about the failed relationship together with the serendipity regarding the end of the relationship and the impetus to move to Madrid and pursue an acting career. The gestures that play a part are gaze, brows, blinks and lips.

While we have discussed multimodality in the context of Search-Stance, the phenomenon is also relevant for Search. In Example (9), a representative of the wine industry for Castilla-La Mancha is speaking about budgetary details regarding the wine industry in the region. In this sample, the speaker clicks while apparently searching for establecido ‘established’. The search process begins with gaze aversion right before el patronato ‘the board’, followed by the elongated use of ha ‘has’, a short pause of 0.38 seconds, the click (co-occurring with head movement), the prolonged filler mmm as well as the repeated and lengthened use of ha ‘has’. There is also a heavy, fluttering blink that accompanies mmm and the ensuing ha ‘has’.

While some of the highlighted features in (9) are similar to those used for Search-Stance, the affective involvement of the speaker and the overall emotional tone are noticeably weaker. The topic itself of the budget is more routine, at least in this particular segment, and the series of collateral signals (filler, click, prolongations, repetitions and gestures) merely appear to be manifestations of a processing lapse as the speaker searches for the right word(s).