Spanish phraseology in formal and informal spontaneous oral language production

2.1 Gran Hermano and CREA oral 3.2 (sub-)corpora

For this study a corpus that contains exactly 30,041 words obtained from randomly recorded fragments of the reality contest Gran Hermano broadcast in Spain by the private television station Tele 5 was analysed. These fragments belong to the 65th, 67th, and 76th days of the 16th season, which was broadcast live 24/7 via online streaming between September and December 2015. Besides, in order to avoid, as far as possible, the influence of the intervention of the TV production, only fragments from the 24-hour online streaming of Gran Hermano have been analysed, including exclusively conversations among the participants. In the moment when the fragments were recorded, from the original eighteen contestants only ten remained in the house, specifically four women and six men. Although the number of words of the speakers in the corpus is very unequal, every one of them has at least 1500 words (see Tab. 1). Moreover, their average age was 22.9 years and they could be classified as middle- to upper-middle-class, as illustrated by the fact that three of the contestants (Han, Sofía, and Marta) were pursuing higher education at the time they participated in the show. The fragments were not fully transcribed; only the PUs were excerpted after several viewings of the fragments that were analysed. Finally, the number of words uttered by every speaker was counted, so that the frequency of PUs per word could be calculated.

On the other hand, the subcorpus of the a priori formal spontaneous spoken language includes 10,000 words obtained from eleven interviews with two female and nine male speakers broadcast on Spanish TV and radio between 1990 and 1996. There is a wide range of profiles, from politicians or artists to businessmen or “normal people”, i.e. persons who are not used to speaking in public. Finally, every speaker in the CREA subcorpus has between 585 and 1089 words (see Tab. 2).

2.2 Taxonomy of Spanish phraseology

One of the main issues faced by specialists in phraseology is that it is rather exceptional to find two scholars who, even within the same linguistic approach or national tradition, make use of the same phraseological taxonomy and terminology. After the evaluation of many existing taxonomies in the Spanish tradition, three basic categories of PUs are proposed for this research.

1. Idioms (“locuciones” in Spanish). Although the boundaries of this group are not absolutely defined, in the sense proposed in this research an idiom would be a unit “[…] that is fixed and semantically opaque or metaphorical, or, traditionally, ‘not the sum of its parts’, for example, kick the bucket or spill the beans” (Moon 1998: 4). Idioms such as partirse de risa^[9] (‘to crack up’) are generally considered the most salient type of PUs and the only one which most researchers agree on. Moreover, only expressions denoting additional naming are accepted (see Dobrovol’skij and Piirainen 2005: 18). Idioms are further divided into seven subtypes according to their function in the sentence: adjectival phrases, adverbial phrases, conjunctive phrases, discourse marker phrases, nominal phrases, prepositional phrases, and verbal phrases.

2. Routine formulae (also known as pragmatic phrasemes), i.e. a heterogeneous group of multi-word expressions frequently used for pragmatic purposes such as ¡madre mía! (‘oh my!’) or ¿sabes lo que te quiero decir? (‘you know what I mean?’). Unlike idioms, PUs of this sort are independent speech acts, expressing a complete thought. In the manner of the idioms, routine formulae can also be divided into subtypes. Since they are independent speech acts, the classification does not respond to their function in the sentence, but to the kind of interaction established between the speakers, i.e. to pragmatic-discursive criteria (see López Simó 2016: 178–226): interpersonal-relation formulae, personal formulae, impersonal formulae, and metainteractive formulae.

3. Proverbs (including sayings and quotes), i.e. an independent maxim with a deontic function such as Spanish “piensa el ladrón que todos son de su condición” (‘knaves imagine nothing can be done without knavery’).

3.1 Frequency distribution

One of the major aims of this research on phraseology as it is used in spontaneous communicative situations is to shed light on the frequency of its use, and this is for two main reasons: on the one hand, because this is a question often commented on in studies on phraseology and colloquial register, but that has rarely been the object of systematic investigation, and on the other hand, to attempt to give a response to the question of the extent to which phraseology is a typical feature of colloquial registers, as has been suggested by some specialists (see Čermák et al. 2009: 9).

In total, 1501 occurrences and 550 different PUs (types) were documented in the 30,041 lexical units of the Gran Hermano corpus.^[10] Logically, many PUs appear more than once in the conversations (2.7 times on average). As might be expected, the two most frequent PUs are two discourse marker phrases frequently used as pet phrases – o sea (approximately ‘I mean’) 69 times and es que (a neutral expression that could be (with difficulty) translated into English as ‘the thing is…’, ‘well…’, or even ‘it’s just that…’) 113 times.^[11] Moreover, since 1501 occurrences were documented in a corpus of 30,041 words, on average one PU is used every 20.0 words. In addition, 550 types (i.e. different PUs) were identified, which means that a different type is used every 54.6 words.

The data extracted from the CREA oral subcorpus representing formal oral Spanish shows that one occurrence is used every 28.8 words (347 occurrences in a total of 10,000 lexical units);^[12] in other words, the frequency of use (in terms of occurrences) in Gran Hermano is approximately 30% higher than in the formal subcorpus.^[13] This would speak for a tendency, according to which the more informal the discourse context, the higher the frequency of use of PUs. This hypothesis could be partially confirmed by the only study (at least in Spanish) to have previously investigated the general frequency of phraseological use on the basis of an oral corpus. Marcin Sosiński (2010, 2011) documented 2816 occurrences in his analysis of the interviews and dialogues that form the PRESEEA GRANADA corpus – one of the PRESEEA corpora, composed of 82,141 lexical units – i.e. one occurrence every 29.2 words. Since the corpora of the PRESEEA project vary in a range between “formal” and “semi-formal” (see Penadés Martínez 2015b: 252), a frequency higher than 20 words per occurrence (Gran Hermano) and lower than 29 words per occurrence (CREA oral) would be expected.^[14] According to the data obtained, phraseology is almost equally frequent in both (semi-)formal corpora, which is not exactly what was expected, since the PRESEEA corpora are supposed to be slightly less formal than the CREA oral subcorpus, but neither is it very surprising. In this sense, it has to be admitted that the comparison between my data and that proposed by Sosiński should be taken with precaution. Although the methodology used by Sosiński for his study is apparently very similar to the one used in my analysis (see Sosiński 2010: 138–140), when the time comes to decide whether a combination of words is or is not a PU, significant differences can be observed. According to the examples presented in his papers, this would probably not affect the most salient subtypes of PUs, such as verbal phrases (2011: 124), but it could be a key factor regarding other less salient subtypes, such as conjunctive and discourse marker phrases, and especially all kinds of routine formulae (routine phrases in his taxonomy). Moreover, Sosiński (2010: 143–144), in contrast with the methodological principles of my study, seems not to respect the additional naming criterium proposed by Dobrovol’skij and Piirainen (2005: 18) by treating terms such as fin de semana (‘weekend’) or primera comunión (‘first communion’) as nominal phrases.

3.2 Typological distribution

Another relevant statistic when it comes to shedding light on the characteristics of Spanish phraseology as it is used in formal and informal spontaneous language production is the typological distribution of the PUs documented in both oral contexts.

The results of the typological analysis of the PUs registered in the (sub-)corpora, in terms of occurrences, are presented in Tab. 3.

As can be observed in Tab. 3, the presence of the type idioms is clearly dominant among the PUs that were documented in both oral contexts. However, there is a very significant discrepancy regarding the distribution of idioms and formulae, with the latter being 30% higher in the Gran Hermano corpus. As a result of this, it could be concluded that these pragmatically-oriented expressions are not representative of oral conversations in general, but that they are a distinctive feature of the phraseology used exclusively in spontaneous informal language production. In this sense, the data provided by Sosiński (2010, 2011) based on the PRESEEA GRANADA corpus fulfils the expectations of a (semi-)formal corpus – 18.9% of the PUs are categorised as routine phrases, i.e. the number is situated between the data obtained for the informal (36.1%) and the formal (5.8%) (sub-)corpora, but is closer to the latter (however, again, the data proposed by Sosiński should be treated with caution). Finally, the peripheral role of proverbs and sayings in both kinds of oral contexts has been confirmed.

3.3 Stylistic distribution

It is obvious that in informal discourse contexts speakers will tend to use more informal expressions than in more formal situations. However, in this section it is intended to reveal in detail the similarities and differences between the configurations of the PUs uttered in these two kinds of oral spontaneous language production.

For investigating the stylistic distribution of PUs, first it must be clearly established which styles we are referring to here and where their boundaries are. Nevertheless, to establish what is colloquial, vulgar, popular, neutral, literary, etc. is a very complicated task that goes far beyond the scope of this study. For this reason, the stylistic characterisations provided by the dictionaries will be employed, in this specific case, the labels proposed by the most complete and up-to-date work among the Spanish phraseological dictionaries published so far, i.e. Diccionario fraseológico documentado del español actual (DFDEA) by Seco et al. (2017).^[15] Unfortunately, this is a major drawback because only those PUs included in the dictionary can be considered. From the 1501 occurrences documented in the approximately 30,000 words of the Gran Hermano corpus only 1033 (68.1%) are registered in DFDEA. On the other hand, from the 347 occurrences documented in the 10,000 words of the CREA oral subcorpus 284 (81.8%) are included in DFDEA. These percentages confirm that this issue is especially problematic when analysing informal oral language production owing to the fact that DFDEA collects exclusively PUs documented in written corpora (see Seco et al. 2017: XI–XII).

Following the categorisation used in DFDEA, eight stylistic labels will be proposed: literary, semicultured, neutral, colloquial, juvenile, uncultured, jargonistic, and vulgar. The stylistic distribution of the PUs documented in the (sub-)corpora is presented in Tab. 4.

According to the data presented in Tab. 4, there are two basic features shared by both kinds of oral context: a) literary and semicultured PUs are highly infrequent in both kinds of spontaneous language production (just two occurrences in the CREA oral subcorpus – padre de la patria and a nivel de); b) most of the Spanish PUs used in oral conversations (at least among those listed in DFDEA) are stylistically neutral.

On the other hand, in the Gran Hermano corpus 41.6% of all the excerpted PUs can be considered informal, i.e. situated in the table “under” the neutral register (colloquial, juvenile, uncultured, jargonistic, or vulgar). On the contrary, a mere 15.5% of the PUs registered in the CREA oral are labelled as informal. In other words, the presence of informal PUs is almost three times more frequent in conversations uttered in informal discourse contexts than in formal situations.^[16]

In this sense, the great majority of the uncultured PUs are due to just one pet phrase, o sea – 69 occurrences in Gran Hermano and all the occurrences (14) in the CREA oral subcorpus. The discourse marker phrase a ver (18 occurrences)^[17] and the adverbial phrases a lo mejor (17) and en plan (13) are the most commonly used colloquial PUs in Gran Hermano.^[18] In the case of the CREA oral subcorpus the most-repeated PUs are the discourse marker phrase no sé (eight occurrences), used to pick up the thread of the conversation, and the adverbial phrase a lo mejor (seven occurrences). According to the data, these two idioms are the only colloquial PUs often used in formal discourse contexts – the personal formula qué va, with just two occurrences, occupies the third position in the list. Besides, it is clear that adverbial phrases and those discourse marker phrases used as pet phrases are the two most frequent types of informal PUs in both oral corpora. However, they are comparatively much more commonly employed in informal oral language production. Besides, while o sea and a lo mejor seem to be characteristic of both oral discourse contexts, a ver and en plan are distinctive to Gran Hermano’s conversations.

A last but very significant divergence between the data obtained for both kinds of spontaneous language production is that more than 6% of all the PUs documented in the Gran Hermano corpus are labelled as vulgar.^[19] On the contrary, there is not even one vulgar PU documented among the 10,000 words that form the CREA oral subcorpus.

According to all this, it could be concluded that one of the main disagreements between formal and informal oral contexts is the frequency of use of colloquial or uncultured PUs, but that an even more salient feature is the complete absence of vulgar PUs in the formal discourse contexts.

3.4 Individual distribution

Following Briz and Albelda (2013), formal and informal are not two absolute parameters, but, on the contrary, a contextual, gradual matter. Besides, even under the same conditions different people can behave in a linguistically dissimilar way. In this section the personal distribution in both (sub-)corpora will be shown in order to determine the tendencies when relating (in-)formal discourse contexts and individual linguistic behaviour. Specifically, three aspects will be considered. First, the individual general frequency will be presented and discussed, then the data regarding exclusively the informal PUs; finally, the frequency of vulgar PUs will be the object of attention.

Fig. 1 undoubtedly confirms that, in terms of general frequency of use, personal linguistic preferences can be a key factor. Let us focus first on the Gran Hermano corpus; six out of the eleven contestants present a frequency very close to the average, being one PU used every 19 up to 21 words. Han (16) and Sofia (17) employ PUs slightly more than the average, while the contrary can be said about Suso (23). The only contestant who, according to the data, behaves as if he were speaking in a formal discourse context is Daniel Vera, since his ratio in terms of words per occurrence (32) is even higher than the average in the CREA oral subcorpus. On the other hand, the speakers in the formal discourse contexts seem to be much more heterogeneous in terms of general frequency. In most cases, their ratio is more distant from the average (one PU every 29 words) than among the contestants of Gran Hermano. However, these divergences between these two kinds of oral language production are undoubtedly affected by the different number of words of which the (sub-)corpora are compounded, as will be discussed below.

Fig. 1:

Individual general frequency of PUs in Gran Hermano and CREA oral corpora

That said, the most interesting data for the purposes of this study is probably not the general frequency, but the numbers related to the individual use of informal PUs, i.e. colloquial, juvenile, uncultured, jargonistic, and vulgar PUs. Fig. 2 shows two clear tendencies.

Fig. 2:

Individual frequency of informal PUs in Gran Hermano and CREA oral corpora

First, although some significant divergences among the contestants of Gran Hermano can be observed, their frequencies are relatively alike (Han uses informal PUs 2.3 times more often than Marina), at least when compared with the numbers obtained for the speakers in the formal subcorpus. So, Agustín Jorge Giner, a businessman and the president of a modest football club in the Spanish third division, utters nine colloquial PUs and one uncultured one in just 585 words, i.e. he makes use of informal PUs 18.5 times more often than the former Spanish prime minister José María Aznar does (just one informal PU in 1089 words, a lo mejor) and his ratio is as low as the second lowest ratio in Gran Hermano. However, again this data should be treated with caution, since the number of words uttered by all the speakers in the CREA oral subcorpus is smaller than in the case of the contestants of Gran Hermano (see Tabs. 1 and 2).^[20] Nevertheless, examples like Giner and Aznar demonstrate that speakers in formal contexts would tend to behave linguistically more heterogeneously than in more informal language production.

Second, it cannot be stated that informal language is used exclusively in informal discourse contexts. Some speakers tend to use informal expressions even in highly formal discursive situations, such as TV or radio interviews. Following the data, three possible profiles can be identified. On the one hand, there are speakers like Giner, who are not used to speaking in public and probably lack the proper linguistic tools or experience to behave “adequately” in these formal oral contexts. On the other hand, there are artists, such as the well-known musician Peret or the actress and model Amparo Muñoz, who seem not to be interested in “respecting” the linguistic prescription appropriate for formal language production, maybe because they wish to look natural and close to the audience, and consequently they behave on television as they would do in other less formal spaces (although probably not as informally as they would do if they were not being recorded). Finally, there are some interviewees who are used to talking in public and are perfectly aware of the importance, in their role as the public faces of companies or political parties, of using highly formal speech in these discourse contexts (some of them have even probably been trained for this purpose). Logically, people like Aznar, Abad, or Zarraluqui use almost no informal PUs in their discourse. Moreover, some speakers who are probably not so used to speaking in public often, but are able to produce formal Spanish when it is required, could be added to this third group, such as Cardona or Solé.

Finally, in Fig. 3 the frequency of use of vulgar PUs can be observed. The first obvious remark is that in formal discourse contexts the use of these rude expressions is drastically avoided. So, even those speakers who do behave linguistically as if they were talking in fewer formal oral contexts, such as Giner, Peret, or Muñoz, manage to avoid using vulgar PUs. On the contrary, all the contestants of Gran Hermano make use of profanity to a greater or lesser extent. Moreover, here the personal preferences are more visible than in terms of general and informal frequency. While some speakers, especially Suso, Marta, and Sofía, utter a vulgar PU approximately every 300 words, Niedziela does so every 1600 words and Han almost every 2700 words, i.e. Suso makes use of vulgar PUs 5.6 times more often than Niedziela and 9.4 times more often than Han does. The reasons for such individual divergences go beyond the scope of this study. However, a few suggestions can be proposed. First, it is not a gender question, since two out of the three contestants who most often make use of vulgar PUs are women (Marta and Sofía). Besides, personal circumstances can play an important role. This can be observed in the fact that the two least rude speakers in the Gran Hermano corpus have foreign backgrounds. Thus, although Niedziela was born in Spain, her family is Polish. Han was born in China and moved to Spain with his mother at the age of twelve. In addition to this, while Niedziela could be categorised as a native speaker of Spanish, Han presents some characteristics typical of a proficient^[21] foreign speaker, including inaccurate versions of PUs such as *soltar de melena instead of soltarse la melena and *me importa un pedo instead of me importa un bledo.

Fig. 3:

Individual frequency of vulgar PUs in Gran Hermano and CREA oral corpora

A last remark regarding the individual distribution of PUs has to do with what could be called phraseological idiolects; in other words, with PUs that can be treated as personal discourse markers. According to the results of the analysis of Gran Hermano, the following phraseological idiolects can be pointed out: as already commented, the second most common pet phrase is o sea. In this sense, no other contestant uses a PU as often as Han utters o sea – 23 times in 2677 words, i.e. once every 116 words. Something similar can be stated for the habitual use of the discourse marker phrase es que by Niedziela (once every 219 words) and Carlos (once every 244 words).^[22] Carlos also tends to use the adverbial phrase a lo mejor (once every 515 words) and the profanity me cago en la puta/leche/virgen (every 1123 words). This is the second most frequent vulgar PU used by one speaker just after the verbal phrase no tener ni puta idea, uttered by Sofía every 851 words. Marta’s favourite expression seems to be the interpersonal relation formula ¿sabes lo que te digo/quiero decir? (once every 554 words). Aritz’s distinguishing feature is the use of the impersonal formula ¡qué va! (once every 701 words). With nine occurrences, Marina’s most common pet phrase is again es que (once every 284 words), but she also often uses the following PUs – a veces (seven occurrences), o sea (six), dar la razón (five), and ya está and no sé (four each). Ricky utters the interpersonal relation formula no pasa nada once every 348 words and Daniel Vera diverse versions (11 occurrences, i.e. once every 138 words) of personal formulae with qué – ¡qué guay/bueno/asco/bien/divertido/guapo/peste/rabia! It is interesting that Suso does not use the most common pet phrases extremely often (es que four times, o sea twice, and a ver once in 2278 words), but he seems to like the verbal phrase saber mal (lit. ‘taste bad’) for expressing that something is unpleasant to him, since he uses it in three different conversations – once every 759 words (this PU is also uttered once by Niedziela).

In the CREA oral subcorpus it is more difficult to observe such a high frequency of certain PUs. In part this is due to the lower number of words uttered by every speaker that is registered, but in part also to the fact that pet phrases are much less frequent in formal contexts. According to the results of the analysis, the speakers can be divided into two main groups. The only three artists among the speakers stand out for the frequent use of o sea (Ceferino Moreno Sandoval once every 240 words, Amparo Muñoz once every 259 words, and Peret once every 339 words, i.e. more often than most of Gran Hermano’s contestants, representing 79% of all the occurrences of this uncultured discourse marker phrase in the CREA subcorpus). On the contrary, Agustín Jorge Giner, although he is the speaker with the most informal speech in this subcorpus (see Fig. 2), does not use pet phrases often (no sé twice and o sea once). However, he utters the colloquial adverbial phrase a lo mejor five times, i.e. once every 117 words. The second group would include those speakers who employ rather formal Spanish. In this case, the frequency of pet phrases is extremely low. However, two tendencies can be observed. Some of the interviewees often employ non-expressive PUs as tools to carry out and organise their speech. For instance, 32% of the PUs uttered by the hospitality manager Marta Solé are prepositional phrases (especially dentro de, four times and a través de, three times). Josep Miquel Abad constantly employs formal discourse marker phrases (41% of all his PUs), such as es decir (six times), por (lo) tanto (five times), or dicho esto and por ejemplo (twice each). A similar situation, but less extreme, can be observed in the speech of the coffin manufacturer Cecilio Cardona, with 28% of formal discourse marker phrases (he uses por ejemplo once every 148 words). These three speakers utter 70% of all the formal discourse marker phrases and 57% of all the prepositional phrases in the CREA subcorpus. The remaining speakers in this second group also employ rather formal discourse, but they rarely use prepositional phrases and formal discourse marker phrases. This group includes Luis Zarraluqui, who utters a veces every 206 words, and the three remaining interviewees, who do not use any PU more than twice – the former priest Ramón, Miguel Durán, and José María Aznar.