Ni as the introductory particle for expressions of negation in three dialectal variants of Spanish

1.1 Phraseology and discourse markers

Penadés (2012) refers to the field of phraseology by stating that:

En las lenguas existen combinaciones de palabras en cuya utilización el hablante carece de libertad para alterar o modificar la sucesión de elementos y para variar los propios elementos que constituyen la combinación.

Phraseology embraces a wide variety of elements, including collocations, locutions and sayings. Among them are also some discourse markers, sufficiently grammaticalized expressions that operate on the pragmatic level. Described by Schiffrin (1987: 32) as “sequentially dependent elements which bracket units of talk”, their form varies greatly: some discourse markers consist of only one word. For instance, in Spanish, the words bueno or pues, are the focus of several in-depth pragmatic studies.^[1] However, a very broad group of discourse markers formed by two or more words, intersects with the field of phraseology. Ruiz Gurillo (2010: 176) explains this intersection, referring specifically to Spanish, but his observation is equally applicable to other languages:

Es previsible que, tras una fase intermedia, diversos complementos circunstanciales se gramaticalicen como marcadores del discurso de diversos tipos (…). Algunos de ellos son de carácter complejo, es decir, se definen como unidades fraseológicas.^[2]

Corpas (1996) includes a large number of discourse markers in her book Manual de Fraseología Española, mostly under the “locutions” and “routine formulas”^[3] sections. Ruiz Gurillo (2001: 45–46) includes a “marker locution” in her organization. However, García-Page (2008: 89) highlights that there are several groups of locutions, not only one, that have a purpose in discourse. There is, therefore, a meeting point where the study of discourse markers and the study of phraseology converge.

Phrases of rejection of a proposition or a statement, such as the ones that start with ni, the focus of this paper, are certainly a part of this group. The word ni has been studied in grammar and discourse studies as a polyvalent particle with scalar value. As Martí (1998: 79) points out, it is not enough to say that it functions as a conjunction that links negative members. The author differentiates the conjunction ni from a modal particle ni. Albelda Marcos and Gras (2011) focus on the discursive value and the scalarity^[4] of the word, including a function to introduce a negative answer as a response to a previous intervention while providing an example with the adverbial phrases ni hablar, ni pensarlo, ni soñarlo and ni borracho. Although we do not focus on the scalarity structure in this paper, it constitutes the basis for the continuous creation of new expressions, especially non-canonical ones in negation (Amaral and Schwenter 2009).

1.2 Phrases of negation and rejection

This work focuses on the functions of rejection/negation/refusal/disagreement^[5] and the phrases speakers use to express them. There is a wide variety of forms for saying “no” to a statement, a suggestion or a question. In Spanish, some common ones are formed with the conjunction ni followed by an infinitive, like in ni hablar; by a noun phrase, like in ni en sueños; by an adjective, like in ni loco; and sometimes by a verbal phrase, like in ni lo pienses. As Pérez-Salazar (2017: 270) points out:

La renovación de secuencias se produce, en algunos casos, sobre la base sintáctico- semántica de ciertas estructuras; es decir que el español aprovecha modelos fraseológicos para crear fórmulas nuevas (…). Así, puede hablarse, en diacronía, de patrones productivos.

Most of these ni sequences are prototypical of informal speech.^[6] They are considered slightly impolite, and are less studied than other discourse markers, for instance, those used in accepting a proposal or making a statement.^[7] The context of informal ni expressions is usually a spontaneous conversational dialog, as Holgado Lage and Rojas point out in reference to ni hablar (2016). Fuentes Rodríguez mentions that this particular marker absolutely rejects what another interlocutor said (2009: 226), nevertheless, Holgado Lage and Rojas found that it can also work in a monolog (2016: 114). These contexts also naturally host other ni expressions that have become phraseological units (Pérez-Salazar 2009: 38):

Una vez alcanzada esta condición es mi intención mostrar que han adquirido la capacidad de negar expresivamente constituyendo por sí solas un enunciado (esto es, como enunciados fraseológicos) o bien como elementos oracionales, rasgo éste que se tiene como propio de las locuciones.

When looking at the current literature on expressions of negation with ni, the phrase that appears most regularly is ni hablar and some of its variants. This very common marker is the focus of study in Holgado Lage and Rojas (2016), who, in addition to the habitual function, analyze the marker as a positive response (showing acceptance) in Río de la Plata. Frequently, ni hablar is the only ni phrase recorded in general linguistic works such as Casamiglia and Tuson Valls (1999: 239), Perona (2000: 450) or Landone (2009: 287). In Corpas (1996: 196), the same expression is mentioned along with the variant ni hablar del peluquín, which is falling into disuse.^[8]

Some works are a little broader and incorporate other rejection phrases besides the ubiquitous ni hablar. One of the seminal articles in rejection locutions in Spanish, by Pérez-Salazar (2009), analyzes the historical process through which some ni combinations became phraseological units. The author focuses on three particular locutions: Ni hablar, ni pensar, and ni soñar. The same author (Pérez-Salazar 2017) studies negation expressions diachronically, focusing on sequences that are productive, including ni en sueños, ni pensar, ni por lumbres, ni por asomo, and ni por el forro. Ni hablar is also the most detailed ni expression in Santos Río (2003). Olza (2011) studies some phrases of rejection, including ni qué narices, which also starts with ni but requires a different syntactic structure, since it needs to be preceded by the actual object that is being rejected.

Some sources do not incorporate any locutions that start with ni. For instance, Martín Zorraquino and Portolés (1999), contribute to the field with one of the original lists of discourse markers in Spanish, but do not provide expressions of rejection. Their work includes, however, some of the opposite function, for instance desde luego, por supuesto or claro. The Diccionario de Partículas Discursivas del Español^[9] (Briz, Pons and Portolés 2004) also does not focus on the function of rejection.

However, there are some other dictionaries of Spanish discourse markers that include a higher number of expressions of rejection.^[10] Fuentes Rodríguez (2009) records a list similar to the one mentioned by Pérez-Salazar (2009): ni hablar (del peluquín), ni pensarlo, ni soñarlo. In Holgado Lage (2017) there is a longer list of ni phrases: ni de broma/coña, ni en sueños, ni pensarlo, ni por todo el oro del mundo, ni loco/a, ni soñarlo, ni hablar. Seco et al. (2004) include an ample repertoire of ni expressions in their Diccionario Fraseológico Documentado. Not all of them can be included in the group that is pertinent to this study.^[11] The list of expressions that follow the productive pattern used to strongly reject what was said before includes: ni amarrado, ni atado, ni de coña, ni en broma, ni en sueños, ni hablar, ni hablar de eso, ni hablar del asunto, ni hablar del peluquín, ni lo pienses, ni lo sueñes, ni loco, ni pensarlo, ni por broma, ni por pienso, ni por sueños, ni por todo el oro del mundo and ni soñarlo. This is the most exhaustive list we have found in the literature.

Two of the most relevant general dictionaries of Spanish were also consulted, since they are the tool that most speakers would use if they wanted to find a locution. A search in the Diccionario de la Real Academia Española (Real Academia Española 2014) shows that there is not one expression of rejection under the word ni.^[12] However, one can find some of these locutions when looking by the most relevant word. For instance, ni hablar under “hablar”, ni pensarlo under “pensar”, ni soñarlo under “soñar”, ni en sueños/ni por sueños under “sueño”, ni de broma/ni en broma under “broma”, ni loco under “loco”, ni de coña under “coña”, ni borracho under “borracho” or ni de fundas under “fundas”. It is difficult to know if there are others included (for instance, some outdated ones that are not widely used, or some regional ones) since they are not grouped together anywhere within the dictionary. The second-most renowned general dictionary of Spanish, Diccionario de Uso (Moliner 1966: vol. 2, 507), includes several expressions under the definition of ni: ni por mientes, ni pensarlo, ni por pienso, ni por sombra, ni soñarlo, ni lo sueñes, ni en sueños. And it adds that there might be other ones not included in the list.

While none of the sources mentioned incorporates every one of these expressions^[13] – an arduous task, anyhow, since they are constantly appearing (García-Page 2008: 120) – this is also not the main objective of most of them.

4.1 General results

First, we analyzed the ni expressions that appeared in the general survey, contrasting the three varieties. We were expecting to see a preference for some discourse markers above others in the dialectal variants. Fig. 3 shows the average value for each expression (1–5) on the vertical axis. As can be seen, there was a lot of variation between dialectal variants.

Fig. 3:

Frequency of ni expressions from the survey in the three different varieties

For example, the Peninsular Spanish speakers demonstrate a high level of comfort with ni de coña (value 4.42), and ni hablar (value 3.96). These are the highest values given to any marker in the three dialectal variants, as can be seen in Fig. 3. Ni de coña is rejected by the other two dialectal variants (Mexico 1.07 and Colombia 1.21). The answers from Spain attribute the lowest possible value to ni madres (1) and to ni modo (1.03). The most neutral phrase, according to the chart, would be ni pensarlo, as the three groups of participants all rated it high (3.27 Spain, 3.27 México and a little higher, 2.92 in Colombia). The one phrase that is not known or used on a regular basis by any of the dialectal variants is ni en pedo (rated 1.08 in Spain, 1.58 by the Mexican participants^[21] and 1.25 in Colombia).

Ni hablar, although it appeared in almost every source that included ni expressions, and many times was the only one, is not the most preferred expression in any of the dialectal variants, with Mexico rating it at 2.63 and Colombia at 2.35. Ni madres is the most preferred phrase in Mexico (3.62), with a low value in both Spain and Colombia (1.33). Ni modo is given a high value in both Mexico (3.00) and Colombia (2.85), but not as high as others are. This might be because although sometimes used for rejection, the main function of the marker is to express acceptance with resignation, so it is possible that for many speakers, answering ni modo to the example given did not fully mean a strong rejection, as the other expressions did.^[22]

In Tab. 2 there is a summary of the values given to every item in the three countries.

One of the most innovative features of the survey was the opportunity for participants to provide new expressions that were not part of the original list and that they know or use on a regular basis. The differences between dialectal variants were conspicuous as shown in the next three sections.

4.2 Results by country: Spain

As seen in Fig. 3, the prototypical ni expression for the Peninsular dialectal variant is ni de coña (value 4.42), with ni hablar and ni pensarlo ranked really high as well. This is interesting as ni de coña was not listed in most of the sources consulted^[23]. The reason could be that it is slightly vulgar. Participants from Spain provided seven different new ways to reject a proposition like the one given as an example. In Tab. 3 is a list of how many times they appear in the results:

As seen in the chart, the phrases ni lo sueñes / ni en sueños (very similar variations) are the most popular, with three instances in total, followed by ni loco/a. Both of these locutions appear in some of the sources consulted.

4.3 Results by country: Mexico

Mexican speakers follow a very similar pattern to that described in 4.2. There is a prototypical expression, ni madres (value 3.62) ranked much higher than in the other two countries. Ni modo also had a high rate^[25] along with ni pensarlo. The locution ni en pedo was rated much lower than expected, only 1.58, but another variation was provided three times, as shown in Tab. 4, along with the other new phrases recorded by speakers of the Mexican variety.

As shown above, some of the answers are very similar to the ones in Peninsular Spanish, especially ni loco/a or ni de loco/a (4 instances in total) and the variations ni sueñes and ni soñando (2 instances in total). Some new discourse markers that do not appear in the other two dialectal variants are ni de pedo (3 instances), which also did not appear in any of the sources, and ni en broma/ni de broma (2 instances).

4.4 Results by country: Colombia

The data collected from participants in Colombia are slightly different from those collected in the other two countries. It appears that there is no prototypical expression generally used in the Colombian area that is valued significantly lower in the other two varieties. Ni pensarlo and ni modo have relatively high values, but at 2.92 and 2.85 they are far from the highest-rated locutions in Spain and Mexico, plus they are also highly accepted in the other two countries. This lack of a high-rated marker translates into a higher rate of responses for expressions that were not a part of the main list: ten additional phrases were provided, some of them similar to the ones from Mexico and Spain, some of them completely new. In Tab. 5 the complete list is included:

The most common locution provided was ni loco/a (4 instances) which was also used in the other two countries. Ni de fundas appears twice, as does ni en sueños. The latter phrase also appeared in the results from Spain and some variations that included the word verb soñar^[27] were recorded in Mexico. However, seven of these expressions only appear in the Colombian surveys (eight if we include ni a palo, for which there is a very similar variety, ni de palo, in Spanish from Spain), which might be an indication of the wide variety of options that the speakers can choose from. This also provides an explanation for the lack of a prototypical phrase that is highly rated in the survey by most speakers.^[28]

4.5 Age and gender of participants

4.5.1 Age

One of our research questions was whether the age of the participants was a relevant factor when providing new expressions, since there is a link between young age and innovation in language.^[29] On analyzing the data, we found ten new locutions^[30] that did not appear in the bibliography or the first part of the survey. Most of them were recorded in Colombia and Mexico.^[31] The list includes: ni de palo (Spain), ni de chiste (Mexico), ni de desmadre (Mexico), ni a bate (Colombia), ni a palo (Colombia), ni po’el putas (Colombia), ni puel (Colombia), ni de riesgo (Colombia), ni en joda (Colombia) and ni se te ocurra (Colombia). While it was of great interest to analyze the speakers’ age and gender, the results did not match our hypothesis.

Fig. 4 shows participants per age group who provided new expressions. Thirty-nine participants included some ni expressions in the optional second part of the survey:

Fig. 4:

Age of participants providing new ni expressions

Only two of the participants were younger than 25 years old, with most of the speakers in the 25–34 and 34–44 sections. While this did not match the proposed theory, the numbers match the age distribution of the entire pool of participants, as seen in Fig. 5:

Fig. 5:

Age of participants, total

It seems that the participants that decided to provide new expressions with ni were a similar small percentage from every age group.^[32] The only exception would be the 55–64 age group. The percentage is slightly higher among the participants that provided new locutions (15%) than among the general pool of participants (7%). Therefore, there is not a reliable age group that provides a higher number of new markers, but every range of age included some participants that took the time to fill out the second part of the survey and were familiar with some ni expressions that were not included in the original list.