Generic uses of the English pronoun one and the Spanish pronoun uno in parliamentary debates

Full Text

1. Introduction

The human impersonal pronouns one in English and uno in Spanish are widely considered functionally similar and can refer to the speaker’s personal experience and/or establish generalizations (Gelabert-Desnoyer 2008, Moltmann 2006, 2010, Rasson 2016, Pearson 2022, Gutiérrez-Rodríguez & Pérez-Ocón 2024a, b). However, while most scholars have extensively analyzed their first-person oriented uses, particularly in Spanish data (so-called ‘concealing’ uno), generic uses of both pronouns received less attention, leaving their interpretive properties and a cross-linguistic comparison underexplored. This work seeks to address this gap, thereby advancing our understanding of these pronouns across semantic, pragmatic, and comparative domains.

Pronouns one and uno belong to the class of One-impersonals (Siewierska 2011: 58, see also Givón 1982) found mainly in Germanic and Romance languages, in which the subject¹ is rendered by a pronominalized form of the numeral ‘one’. One-impersonals in turn enter a wider class of R-impersonals, i.e. impersonals triggered by a reduction in referentiality. R-impersonals have “the appearance of regular, personal constructions but [feature] a subject which is human and non-referential” (Siewierska 2011: 57), see (1–2).

(1) These days, one is required to wear a mask on trains. (Pearson 2022: 293, ex.1)

(2) Uno aprende cuando se equivoca. (Gutiérrez-Rodríguez, Pérez-Ocón 2024a: 115, ex.4)

‘One learns when one makes mistakes’

Previous research suggests significant variation in the use of pronouns across functional discourse styles (Gelabert-Desnoyer 2008, Rasson 2016: 244–245, Serrano 2022: 10–13). In line with Gelabert-Desnoyer’s (2008) findings, we focus on parliamentary debates as this register demonstrates higher frequency of generic uses of One-impersonals. Our data comes from the Europarl parallel Corpus, comprising official proceedings of the European Parliament from 1996 till 2011, and the methodology is partly taken from Gast’s (2015) study on the German pronoun man.

The aim of this paper is to examine the distribution of generic uses of English one and Spanish uno, comparing their similarities and distinctions, and to investigate the role the generic component plays in shaping cross-linguistic correspondences. The research questions we seek to address are as follows:

1. To what extent can theories of genericity be applied to the usage patterns of English one and Spanish uno?
2. How are generic uses of English one and Spanish uno distributed in these typologically distinct languages based on political discourse data?
3. How do varieties of generalizations (rules vs. inductive generalizations) interact with the first-person orientation?
4. What is the correlation, if any, between the generic meaning component in the use of One-impersonal in one language and the strategy employed to render the same meaning in another language?

The paper is organized as follows. In section 2 we review considerations about English one and Spanish uno in previous studies and examine observations from research on genericity relevant to our analysis. Section 3 addresses methodological issues. We present the results in Section 4. A general discussion of the results is provided in Section 5. Section 6 summarizes the main conclusions of the paper. 

2. Theoretical background

In this section we review the findings from previous works that are relevant for our analysis. In particular, we look at the observations that have been made about the uses of English one and Spanish uno in discourse in connection to the first-person orientation and generalization and the interaction of these meaning components and discuss properties of generic statements.

2.1. Previous approaches to the analysis of ONE

English pronoun one has been explored in the literature from various viewpoints (Moltmann 2006, 2010, van der Auwera et al. 2012, Malamud 2012, Mignot 2015, Pearson 2022, among others). Let us start with the most prevalent Moltmann’s (2006, 2010) approach who describes two main strategies of the interpretation of one: (i) inference from the first person, as in (3), and (ii) inference to the first person, as in (4).

(3) One can see the picture from the entrance. (Moltmann 2010: 440, ex. 1)

(4) One should not lie. (Ibid.: 441, ex. 2)

For Moltmann, both strategies include first-person orientation and genericity, but they diverge in their direction of fit. Inference-from-the-first-person strategy involves “generalization based on a first-person application of the predicate or first-person” (Ibid.: 447), i.e. the speaker’s experience is generalized to other individuals. On the contrary, inference-to-the-first-person strategy involves “an (already established) generalization that is to allow for an immediate application to the first person in the reasoning relevant in the context” (Ibid.). Moltmann notes that within the latter strategy generalization is internalized, but independent, and “potentially applied in a first-person way by whoever accepts the sentence, in particular the addressee” (Ibid.). It is used in deontic sentences: laws, general requirements, or general recommendations, which are meant to play a role in speaker’s or addressee’s reasoning for their actions.

Pearson (2022) examines Moltmann’s claim that one encodes first-person orientation and generalization and finds its usage similar to impersonally used second-person pronoun you. She discusses examples with experiential predicates like (5) and notes that in (5a) it is infelicitous to use pronouns one and you unless the speaker participated in the marathon herself, i.e. the first-hand experience is required. However, with other predicates or modal verbs this requirement does not hold. She concludes that first-person orientation is necessary only for a subset of utterances with one, namely for those with experiential predicates.

(5)   a. One feels/You feel exhausted after running a marathon. #But I’ve never run a marathon before.
        b. People feel exhausted after running a marathon. But I’ve never run a marathon before. (Pearson 2022: 298, ex. 15)

Van der Auwera et al. (2012) analyze one from a typological perspective along with other human impersonal pronouns in English, Dutch and German. In their model they set a number of parameters for sentences, in which the pronoun appears, and for the interpretation of pronouns. For instance, a sentence can be (i) generic or episodic, and (ii) modal or veridical, while a human impersonal pronoun can be (i) generic or existential; (ii) definite or indefinite; (iii) singular or plural; (iv) exclusive or inclusive; (v) collective or individual. For English one, they describe five uses, illustrated in (6–10).

(6)     When one travels, the umbrella has to come along. (S: Epi, Mod, HP: Gen, Incl) (Van der Auwera et al. 2012: 21, ex. 57)

(7)     One saw that again later when he tried to comfort a little boy who has AIDS. (S: Epi, Ver, HP: Gen, Excl/Incl) (Ibid.: 21, ex. 60)

(8)     One married young in the Middle Ages. (S: Gen, Mod/Ver, HP:Gen, Excl) (Ibid.: 21, ex. 58)

(9)     One only lives once. (S: Gen, Mod/Ver, HP: Gen, Incl) (Ibid.: 22)

(10)   One doesn’t want to set quotas. One doesn’t want to set diktats, but one does want to maintain a dialogue and one does want to maintain pressure. (S: Gen, Mod/Ver, HP: Exst, Def, Sg 1) (Ibid.: 21, ex. 61)

In four out of five uses one is generic in the sense that it can be paraphrased as everybody or anybody, and the first-person orientation does not play a key role in its interpretation (as it does in Moltmann’s account). However, the authors also distinguish the fifth use where one refers to the speaker only. They argue that this use is relevant “especially if the speaker wants to represent his/her behaviour as a result of general rules, as politicians often do” (Ibid.: 21), and the example (10) they provide is uttered precisely by a British politician.

In Mignot’s (2015) corpus study the author distinguishes three uses of one: (i) one1 means ‘everybody’ and refers to people in general; (ii) one2 means ‘everybody including myself’ and implies that “the generalization expressed by one stems from a specific situation” (Ibid.: 281); (iii) one3 means ‘I’ and refers to the speaker. She collected and annotated a corpus of examples with one from the British National Corpus and other sources. Mignot notes that some examples can be ambiguous, and the difference between the first two uses is gradual. Also, for the second use “the particular person who is included in one is not ‘I’ but ‘you’ or even ‘he/she’, i.e. a third person’, therefore, the more precise meaning is ‘everybody including a particular person” (Ibid. 283). The results show that one1 appears in 41% of cases, one2 — in 56% of cases, and one3 — only in 3% of cases, hence, most of her examples are generic, which empirically supports the claim about generalization as a crucial meaning component for one made in theoretical works.

2.2.   Previous approaches to the analysis of UNO

In the literature dedicated to the analysis of the Spanish pronoun uno, we distinguish two approaches. First, there is a first-person oriented approach suggesting that in the utterances with uno the covert reference to the first person is obligatory, and generic interpretation is built over it in specific contexts (Flores-Ferrán 2009, Serrano 2022, Fábregas 2024, Gutiérrez-Rodríguez & Pérez-Ocón 2024a, b). Under this account, uno is always used as a desubjectivizing strategy referring indirectly to the speaker. The covert referent can be the speaker alone (concealing use), as in (11), where uno encodes the speaker, or the speaker and some other people (generic use), as in (12) where the speaker’s experience is extended to the relevant set of individuals and the statement becomes generalizing, which reminds of Moltmann’s inference-from-the-first-person strategy.

(11)    A — Qué bien te has librado, ¿no?

B — Uno, que es muy listo. (Fábregas 2024: 2, ex. 2)

‘You managed to escape quite well, didn’t you?’

Lit: ‘One, that is very smart’ (‘I am very smart’)

(12)    Uno puede aprender a cantar en unos pocos meses. (Gutiérrez-Rodríguez & Pérez-Ocón 2024b: 118, ex. 3)

‘One can learn to sing in few months’

We might further distinguish a contextual approach claiming that both first-person orientation and generalization are optional elements and the interpretation of uno is determined by extralinguistic context and the surrounding linguistic material (Holænder Jensen 2002, Gelabert-Desnoyer 2008, Rasson 2016, de Cock 2020). This approach is also adopted in our study. Its proponents demonstrate that first-person orientation is not always present in the uses of uno. De Cock (2020) provides naturally occurring examples that can’t be understood as first-person oriented (De Cock 2020: 101).

Gelabert-Desnoyer (2008) describes the following four uses: (i) self-referential (=concealing) use when the referent is exclusively the speaker; (ii) self-referential experiential use when the first-person experience is generalized to other referents; (iii) omnipersonal use where generalization is established independently of the speaker; (iv) other-referential use when uno refers to a particular referent who is not the speaker. Gelabert-Desnoyer (2008) demonstrates that these uses are distributed differently in two different genres. On the one hand, in his parliamentary discourse corpus composed of 44 examples the most common uses are other-referential (52.3%) and omnipersonal (40.9%), while self-referential and experiential are found only in 2.3% and 4.5% of cases accordingly. On the other hand, in his control corpus composed of 128 examples from oral interviews, obtained from the online version of traditional Spanish newspapers, the most common is self-referential use (68%), and other uses are far less common: experiential use is found in 18.8% of cases, omnipersonal use — in 10.9%, and other-referential — in 2.3% of cases.

Gelabert-Desnoyer’s (2008) classification was further refined by Rasson (2016). She describes nine main uses of uno and intermediate uses, making distinction between uno referring to the speaker, addressee or the third person, generic use referring to all humans and reduced generic use (sp. generico reducido) referring to the limited group of people (i.e. professional or social class, age category, etc.), and generalizations based on the speaker’s, addressee’s or the third person’s experience (Rasson 2016: 247)². In her study, she examines the distribution of uses in three genres: spontaneous conversations (28 examples), academic texts (11 examples), and forum Yahoo (100 examples), and shows that for conversations, the most common use is generalization based on the third person’s experience, for academic texts, it is reduced generic use, and for Yahoo forum, it is generic use. While the sizes of her samples are not very large, it is meaningful that the most common uses in all three genres still involve generalization.

To summarize, most scholars agree that the pronouns one and uno show a connection to the first person and generalization, but their exact contribution is a matter of discussion. It seems reasonable to accept that first-person orientation and generalization are crucial meaning components for one and uno, but neither of them is obligatory. There are uses of one and uno where only one component is present, such as reference to the speaker alone or independent generalization targeted at the addressee or the third person, and uses where both components are combined, such as first-person based genericity and first-person targeted genericity. Since we want to look more closely at uses of one and uno involving generalization, now let us discuss some properties of generic sentences that we later apply to our data.

2.3. Generic statements

Generic statements express common conceptual knowledge about entities and their characteristics and are pervasive in our everyday speech. They can be exemplified by (13–16).

(13) Dogs bark.

(14) Gold is a precious metal.

(15) Bishops move diagonally.

(16) A gentleman pays his debts.

Generic statements have been discussed extensively in the literature both from a formal semantic perspective and a cognitive perspective (Carlson 1977, 1995, Krifka et al. 1995, Cohen 2001, 2022, Leslie 2008, Krifka 2012, Prasada et al. 2013, Leslie & Lerner 2016, Filip 2024, among many others). It is still a question under discussion whether a unified analysis for all generics is achievable. There are, however, some points most scholars agree upon. First, genericity includes two subdomains: kind reference and generic sentences. Kind reference sentences express generalizations over properties of kinds and may contain kind predicates, e.g., be(come) common / widespread / extinct. Generic sentences express non-accidental regularities over individuals or situations. While each subdomain has its own properties, they can coexist in one utterance (Krifka et al. 1995). Next, Carlson (1977) distinguishes the following properties of generics: (i) they are aspectually stative, i.e. they lack reference to particular situations³; (ii) they are intensional in the sense that they express regularities, and they describe not only observed facts, but also have a predictive force and a law-like; (iii) generic statements tolerate exceptions.

Generic statements by definition express generalizations, but the type of generalization is still a matter of debate. To account for their truth-conditions, two approaches have been put forward in the literature: the inductivist view and the rules-and-regulations view, see Carlson (1995) and Cohen (2016). The inductivist view suggests that generics express inductive generalizations based on observed facts in the world: “after ‘enough’ instances have accumulated, the generic form can be truly asserted”, as in (13–14). The rules-and-regulations view states that generics express rules and depend on causal relations between entities (15). Cohen (2001: 193) also notes that “the rule may be physical, biological, social, moral, etc.”, i.e. different varieties are possible, see also Krifka (2012). While inductive generalizations can be made only if the relevant situations actually took place, rules are valid even if the described situation never occurred. For instance, if a generic statement like "Boys don’t cry" is analyzed as an inductive generalization, it implies that there was a case when a particular boy didn’t cry. However, if analyzed as a rule, it could be that all real boys have cried, even though they should not have. While these two approaches were originally opposed to each other, Cohen (2001) notes that they can be combined: some generics are better analyzed as inductive generalizations, others fit into rules-and-regulations set. This is a stance we take here.

We argue that these observations are relevant for generic statements with one and uno. Below we examine the distribution of inductive generalizations and rules for generic uses of pronouns in both languages, and analyse how the properties of generic statements with one and uno in one language is related with the way it is transmitted to another language.

3. Data and methodology

This study employs a parallel corpus approach to examine generic uses of impersonal pronouns one in English and uno in Spanish. Using parallel corpus data proves to be an effective approach for analyzing one and uno, as it allows for the examination of their use in substantially equivalent contexts. This enables the comparison of units that are described in the literature as functionally analogous, helping to identify both their similarities and differences. Additionally, exploring strategies employed to convey the same meanings as impersonal pronouns in parallel texts can offer valuable insights into how these meanings are realized through different linguistic forms.

We adopt Gast’s (2015) methodological framework, which demonstrates how parallel corpora can enhance contrastive studies of linguistic expressions’ interpretation and distribution patterns. In his analysis of German impersonal man and its English equivalents, Gast introduces the concept of heterophrases defined as “pairs or sets of sentences [in a pair of different languages — EV & OCh] that are intended to render (approximately) the same meaning, in the same context, irrespective of the source and direction of translation” (ibid.: 9).

As shown above, English one and Spanish uno both belong to the One-impersonals class and can either refer to first-person experience or express generalizations. These findings suggest that the pronouns serve analogous functions, generating two testable implications: (i) statistically similar patterns of first-person oriented/generic uses across corpora; (ii) mutual preference as functional equivalents in parallel contexts (the meaning rendered with one in English is transmitted as uno in Spanish and vice versa).

Our analysis draws on the Europarl corpus (Koehn 2005), available through Sketch Engine, which contains approximately 60 million words per language (1996–2011) of European Parliament proceedings in 21 languages, representing formal political discourse.

We compiled two randomized samples (250 concordances each): (i) English instances of one and their Spanish equivalents, e.g. (17), and (ii) Spanish instances of uno and their English equivalents, e.g. (18).

(17)    EN: Following the recent elections there are one-third fewer women in Silesian politics. So how can one speak of women’s equality?

SP: Tras las recientes elecciones, la proporción de mujeres en la política de la región ha descendido en un tercio. Por tanto, ¿cómo podemos hablar de igualdad de las mujeres? (#2610077)

(18)    SP: Por ejemplo, si uno introduce el código postal en su página web puede saber cuál es el proveedor más barato de la zona.

EN: For example, if you type in your postcode on their website you can find out who is your cheapest supplier. (#20581984)

Corpus Query Language (CQL) queries (19–20) excluded most of irrelevant sequences (e.g., Spanish cada uno, English the one):

(19)    [word! = “Cada|cada”]{1}[word = “Uno|uno”]{1,2}[tag = “V.*”]

(20)    [word! = “The|the|No|no|First|first|This|this|Each|each”]{1}[word = “one”]{1,2} [tag = “V.*”]

Through manual review, we excluded non-target uses, resulting in final datasets of 187 contexts with one and 198 contexts with uno.

The annotation parameters considered were as follows:

1. primary interpretation as either first-person orientation or generalization as the main meaning component;
2. when generic components were present, the generalization type was coded as either inductive generalizations or rules;
3. for rules, specific types were identified (biological, legal, metalinguistic, moral, or social);
4. for primarily generic uses, we noted whether first-person experience supported the generalization;
5. cross-linguistic realization through heterophrases (Spanish equivalents for English one and English equivalents for Spanish uno).

In our annotation, we relied on observations from Rasson (2016), Serrano (2022), and Fábregas (2024), which indicate that the generic reading of uno arises in the presence of specific genericity inductors. These include impersonal constructions, the use of hay and hay que, temporal constructions with cuando ‘when’, conditional constructions with si ‘if’, the second-person pronoun tu in a generic sense, clitic se, quantifiers like siempre ‘always’ and a menudo ‘often’, expressions with modal semantics, and lexical items with generalizing semantics, such as gente ‘people’ and población ‘population’. Rasson also identifies expressions that hinder generic interpretation, such as personal pronouns and noun phrases coreferential to uno. For English examples we used the observations from van der Auwera et al. (2012) and Moltmann (2006, 2010). Contextual information was also taken into account during the analysis.

This annotation framework enables identification of four distinct combinations of generalization and first-person orientation in impersonal pronouns: (i) generalizations without first-person reference, (ii) generalizations supported by first-person experience, (iii) generalized first-person experience extended to others, and (iv) non-generalized first-person reference. The first three types constitute the generic uses that form our primary analytical focus. Through examining their interaction with various generalization patterns (inductive generalizations and rules) and analyzing the strategies for conveying equivalent meanings in parallel texts (heterophrases), we uncover both the shared characteristics and divergent behaviors of English one and Spanish uno, particularly in their generic uses.

4. Results

This section presents our analysis of occurrences of English one and Spanish uno in the sampled Europarl corpus data. We specifically investigate the distribution between uses containing a generic component and those lacking it, while exploring how genericity interacts with first-person orientation. Furthermore, we analyze how these meaning components relate both to the varieties of generalization and to the selection of corresponding heterophrases in parallel texts.

4.1. The analysis of ONE

Our sample of of occurrences of English one and their Spanish equivalents comprises 187 instances. Among these, 125 cases (66.84%) exhibit a generic component, while the remaining examples demonstrate reference to speaker’s experience without generalization. Figure 1 illustrates the distribution between generic and non-generic uses in the one-sample.

Figure 1. Distribution of generic (GEN) and non-generic (NON-GEN) uses in the one-sample

We now examine in greater detail the generic and non-generic uses of the English one, with particular attention to how genericity interacts with first-person orientation. As shown in Figure 2, the distribution reveals four distinct uses: speaker’s experience without generalization (first-person primary, non-generic) accounts for 62 cases (33.15%), generalized speaker’s experience (first-person primary, generic) comprises 46 cases (24.59%), independent generalizations unsupported by speaker’s experience (generalization primary, generic) total 43 cases (22.99%), generalizations supported by speaker’s experience (generalization primary, generic) represent 36 cases (19.25%). Of particular interest is the substantial proportion of non-generic first-person references in English one (33.15%) — a usage pattern that has not been sufficiently documented in existing linguistic descriptions. This finding challenges accounts that primarily emphasize one’s generic functions while overlooking its significant role in expressing speaker-anchored, non-generic meaning. 

Figure 2. Distribution of genericity and first-person experience in the one-sample

 Examples (21–24) below illustrate different uses of one where genericity and first-person experience contribute variably to the semantics of the utterance. The interpretation of these uses is supported by functionally equivalent Spanish text segments and contextual background knowledge.

(21)    first-person orientation (non-generic)

EN: Of course we are all against discrimination, but one dare not question the route we are taking here for fear of being pushed into a corner.

SP: Claro que todos estamos en contra de la discriminación, pero no me atrevo a cuestionar la ruta que estamos tomando por temor a ser arrinconado. (#7695300)

(22)    generalized first-person experience

EN: Mr President, when I first entered this House 20 years ago, this debate would not have taken place, because this House was a consultative parliament at the time, and many of us have battled over the last 20 years to make it what it is today, a co-legislator parliament. However, when one co-legislates, one must be responsible. It is an exercise in responsibility that is now in need.

SP: Señor Presidente, cuando yo entré en este Parlamento, hace veinte años, este debate no habría tenido lugar, porque este Parlamento era entonces un parlamento consultivo y la batalla de muchos de nosotros en estos veinte años ha sido convertirlo en lo que es hoy, un parlamento colegislador. Pero cuando uno colegisla, uno tiene que ser responsable. Y aquí tenemos que hacer un ejercicio de responsabilidad. (#16780136)

(23)    generalization supported by first-person experience

EN: We will have to have a short course for Members to learn that the blue card has a specific use, and that in order to ask to speak during ‘catch the eye’, one has to raise one’s hand, raise one’s white card, or use some other mechanism rather than the blue card.

SP: Vamos a tener que hacer un cursillo para que los diputados comprendan que la tarjeta azul tiene determinada utilidad y que, para pedir la palabra en el “catch the eye”, se levanta la mano, se levanta la tarjeta blanca o se utiliza cualquier otro mecanismo, menos la tarjeta azul. (#11824425)

(24)    independent generalizations unsupported by speaker’s experience

EN: Unfortunately, many citizens still believe that education is something one acquires in the first part of one’s life.

SP: Desafortunadamente, muchos ciudadanos todavía creen que la educación es algo que se adquiere en la primera etapa de la vida. (#1736991)

Example (21) illustrates a non-generic use of the one-construction with reference to the speaker’s personal experience. The reference is to a specific situation, as emphasized by the adverb here. Furthermore, the surrounding context features a first-person pronoun without generic meaning, and the parallel Spanish text employs a heterophrase with a first-person singular pronoun, explicitly pointing to the speaker. In examples (22) and (23), the one-construction conveys both first-person experience and genericity, but the hierarchy of these semantic components differs. In (22), the speaker describes his personal experience in the Parliament, on the basis of which they draw a generalization using the genericity inductor when. In example (23), by contrast, the utterance describes an established rule (how one should behave during the ‘catch the eye’ procedure) with which the speaker is familiar but whose enforcement does not depend on them. Example (24) lacks any reference to first-person experience: the politician, the speaker, refers to an opinion widespread among citizens, which he himself does not share.

When the impersonal pronoun is identified as carrying generic meaning, the variety of generalization can be classified as either an inductive generalization or a rule. For English one, we observe a statistically significant predominance of rule-type generalizations (binomial test, p = 0.0006). As illustrated in Figure 3, our sample of English one contains 125 generic uses, with rule-type interpretations accounting for 81 cases (64.8%) and inductive generalizations for 42 cases (33.6%). 

Figure 3. Distribution of rules and inductive generalizations in the one-sample

 We now turn to examining the interaction between first-person orientation in generic utterances with English one and varieties of generalization. Figure 4⁴ presents an association plot (Cohen-Friendly plot), showing that inductive generalizations are overrepresented in uses of one when expressing generalized first-person experience.

Our analysis identifies four distinct categories of rules in the one-sample: social, moral, legal, and metalinguistic. These types of rules demonstrate significant variation in their distribution frequency. Social rules constitute the most frequent category (44 instances, 54.32%), followed by moral rules (21 instances, 25.93%), legal rules (13 instances, 16.05%), and metalinguistic rules (3 instances, 3.7%). Figure 5 illustrates this distribution pattern, clearly showing the predominance of social rules among all rule types attested for English one.

Figure 4. Rules and inductive generalizations across GEN/FP configurations in the English one 

Figure 5. Distribution of rule types in English one 

The examples below demonstrate the use of one for expressing different rule types: social (25), moral (26), legal (27), and metalinguistic (28).

(25)    EN: Mr President, “one is not born a woman, one becomes one”.

SP: Señor Presidente, “no se nace mujer, se llega a serlo”. (#14709351)

(26)    EN: As for the rest, to come back to a few specific issues, Prime Minister, notably the issue of the budget, one has greater responsibility on leaving the Presidency than on entering it.

SP: Por lo demás, volviendo a algunos temas concretos, Primer Ministro, en especial el tema del presupuesto, uno tiene mayor responsabilidad al dejar la Presidencia que al ocuparla. (#13653231)

(27)    EN: As a train driver, one is responsible for passenger safety.

SP: El maquinista de un tren es responsable de la seguridad de los pasajeros (#10503585)

(28)    EN: I have just two remarks on an issue raised by very many of you — namely, what is irregularity, what is fraud, and how should one approach recovery.

SP: Tengo dos observaciones acerca de dos cuestiones que han destacado muchos de ustedes: la definición de irregularidad y fraude y cómo se deberían abordar las recuperaciones. (#2063333)

The distribution of rule types across the three categories of generic uses of one (defined by degree of speaker experience involvement) reveals marked qualitative differences. As shown in Figure 6, social rules — the most frequent type overall — occur disproportionately in contexts of generalized first-person experience, whereas legal rules are entirely absent from such uses. Notably, these distributional patterns, while theoretically suggestive, do not reach statistical significance in our data.

Figure 6. Distribution of rule types across generic uses in English one

 Finally, we examine potential correlations between generic/non-generic uses of the English one and their Spanish equivalents in parallel texts. Our data reveals that the English impersonal one is rendered in Spanish through the following strategies: the impersonal pronoun uno, impersonal constructions, passive voice, first-person singular/plural verb forms, quantifiers, non-finite verbs, descriptive expressions, or complete rephrasing (see Vilinbakhova & Chuikova 2024 for corpus examples).

Figure 7 demonstrates the difference between strategies for generic and non-generic uses of one. In generic contexts, first-person forms show significant underrepresentation. Conversely, in non-generic contexts, first-person strategies are overrepresented.

Figure 7. Spanish strategies for English one in generic vs. non-generic uses

The analysis reveals no statistically significant difference in Spanish heterophrases for English one when conveying inductive generalizations versus rules. While impersonal constructions show notable underrepresentation in legal rule contexts, this pattern does not reach statistical reliability (p > 0.05). The observed distribution suggests a potential tendency toward avoidance of se-constructions for legal formulations, though the effect remains statistically unsubstantiated in our corpus (see Figure 8).

Figure 8. Spanish strategies for English one referring to various rule types

4.2. The analysis of UNO

We now turn to examining Spanish uno with respect to its behavior in generic constructions. Our data contains 198 instances of Spanish uno with their English equivalents, revealing that 133 cases (67.17%) demonstrate generic use, while the remaining 65 (32.83%) show first-person oriented non-generic reference. This distribution closely parallels the pattern observed for English one, suggesting similar functional distributions across both languages. Figure 9 presents the distribution of generic versus non-generic uses in the Spanish uno-sample.

Figure 9. Distribution of generic and non-generic uses in the uno-sample

 Let us now analyze how Spanish uno’s genericity relates to first-person orientation. Figure 10 displays the quantitative distribution across four usage categories: speaker’s experience without generalization (65 cases, 32.82%), independent generalizations unsupported by speaker’s experience (60 cases, 30.30%), generalizations supported by speaker’s experience (38 cases, 19.19%), and generalized speaker’s experience (29 cases, 14.64%). Of particular significance is the observation that the two most prevalent categories constitute either pure reference to first-person experience or general knowledge without combination of these components. This distributional pattern suggests that Spanish uno tends to maintain clear functional separation between first-person oriented and generic meaning components rather than combining them. This distinguishes Spanish uno from English one, where the predominant usage types are those in which first-person experience holds primary status. Examples (29–32) illustrate these usage patterns of uno.

Figure 10. Distribution of genericity and first-person experience in the uno-sample

 (29)    first-person orientation (non-generic)

SP: Uno puede preguntarse si el rechazo de la corresponsabilidad no puede compensarse con uno o más fondos de la Unión Europea.

EN: I wonder whether or not it could be possible to compensate for the refusal to share responsibility out of some European Union fund or other. (#57051884)

(30)    generalized first-person experience

SP: Señor Presidente, señoras Comisarias, señores Comisarios, estimadas y estimados colegas, hablo como experto en temas de presupuesto y en esta calidad a veces uno tiene la sensación de que hay que decir muy claramente, inclusive nosotros, los que nos ocupamos de las finanzas en relación con la ampliación, que estamos totalmente a favor de esta ampliación.

EN: Mr President, Commissioners, ladies and gentlemen, I speak as someone who is involved with setting the Budget, and people in that position sometimes have the feeling that they have to start by saying very clearly that we, too, who are involved with the financial side of enlargement, are in complete support of that enlargement. (#10254942)

(31)    generalization supported by first-person experience

SP: Cuando concluimos la primera lectura todos estábamos llenos de optimismo ya que la nueva forma de actuar decidida de la Comisión nos entusiasmó y creímos que nuestras enmiendas podrían encontrar — tendrían que encontrar — apoyo también en el Consejo pues debía tener lugar una rápida aplicación. Pero casi siempre las cosas salen de modo diferente a como uno piensa.

EN: When we concluded the first reading, we were full of optimism, inspired by the Commission’s new-found élan, and we believed that our amendments would be — must be — approved by the Council too, because speedy implementation was supposed to follow. But then things do not usually happen quite as one expects them to. (#5893496)

(32)    independent generalizations unsupported by speaker’s experience

SP: Creo que también se puede configurar el entorno político en tales regiones de un modo razonable para que se proporcione a tales Gobiernos la fuerza política para salir de este círculo que significa que uno quiere por lo visto protegerse pero, a la postre, se perjudica a la propia población.

EN: I also believe that there must be a proper political framework in these regions so as to give such governments the political power to escape from this vicious circle, which means that, ostensibly, in trying to provide protection for themselves they ultimately harm their own population in the long term. (#3395733)

Note that both examples (30) and (31) contain two meaning components: generalization and first-person orientation. However, in (30), the speaker constructs a generalization about the feelings of a person in a certain position based on his own experience in that very position, with the genericity inductor a veces ‘sometimes’ contributing to the generic interpretation. On the other hand, in (31), the speaker recounts events in which he himself participated, yet statements such as things do not usually happen quite as one expects them to belong rather to the background of common knowledge, approximate proverbial wisdom, and remain valid irrespective of the speaker’s individual experience.

In examining the distribution of generic uses of Spanish uno, we observe a key contrast with English one: while rules dominate in English, Spanish shows no statistically significant frequency difference between rules (62 cases, 48.82%) and inductive generalizations (65 cases, 51.18%) (see Figure 11).

Figure 11. Distribution of rules and inductive generalizations in the one-sample

Further analysis examines how varying roles of first-person experience (primary, secondary, or absent) correlate with different generalization types (rules vs. inductive generalizations) in Spanish uno-constructions. The association plot in Figure 12 reveals a statistically significant distributional asymmetry: inductive generalizations demonstrate marked overrepresentation, while rule-type generalizations show corresponding underrepresentation in contexts of generalized first-person experience.

Figure 12. Rules and inductive generalizations across GEN/FP configurations in the Spanish uno

Our analysis reveals four rule types in the uno-sample: social, legal, moral, and biological. Mirroring English one, social rules constitute the most frequent category (31 instances, 50%). However, the remaining types show divergent distribution patterns: legal rules emerge as the second-most frequent (16 instances, 25.8%), followed by moral rules (13 instances, 20.97%), with biological rules being the least attested (2 instances, 3.23%). Notably, metalinguistic rules are entirely absent from the sample. The observed distribution of rule types is graphically represented in Figure 13.

 

Figure 13. Distribution of rule types in Spanish uno

The following examples demonstrate the use of uno to represent four types of rules: social (33), legal (34), moral (35), and biological (36).

(33)    SP: Esto se debe a que la política es el arte de lo posible, Señorías. Uno no puede obtener todo lo que se desea.

EN: That is because politics is the art of the possible, ladies and gentlemen. You cannot obtain everything you want. (#26594603)

(34)    SP: Para mí, la norma más difícil es que uno ha de tener una fotografía de pasaporte en la que no sonría.

EN: For me, the most difficult regulation is that you have to have a passport photo in which you do not smile. (#24759348)

(35)    SP: Ruego excusen mi vehemencia, pero a veces, para alcanzar un objetivo político, uno tiene que acostumbrarse a la verdad, y ésta es que no hay acuerdo en cómo organizar las relaciones entre las normas máximas sociales y medioambientales por un lado y la competitividad por otro, pero vamos a trabajar en ello.

EN: Please excuse my vehemence, but there are times when, in order to achieve a policy objective, one has to become accustomed to the truth, which is that there is no agreement as to how to go about organising the relationship between maximum environmental and social standards on the one hand and competitiveness on the other — but we will work at it. (#18362190)

(36)    SP: Una de las cosas que había que aprender era que el volumen diario que uno bebe no se puede dejar para el fin de semana y entonces bebérselo todo de un golpe, pues en semejante caso la concentración de alcohol en la corriente sanguínea alcanzaría el día siguiente un nivel catastrófico.

EN: One of the things that had to be learnt was that the units per day that you drink cannot be concentrated in binge drinking at the weekend because the level of alcohol in next day’s bloodstream is then catastrophically high. (#5841461)

The distribution pattern of rule types across three categories of generic uses for the Spanish uno differs substantially from that observed for the English one. Instances where uno expresses rules while denoting generalized first-person experience are notably infrequent and exclude both moral and biological rules (see Figure 14).

Our data demonstrate that Spanish uno is rendered in English through the following strategies: the pronoun one, the pronoun you, first-person singular/plural verb forms, complete rephrasing, descriptive expressions, the plural they, non-finite verbs, passive voice, or quantifiers. Notably, two strategies — you and one — show significantly higher frequency than all others. Of particular interest is the underrepresentation of the pronoun you as a functional equivalent of Spanish uno in contexts referring to non-generic situations (Figure 15). 

Figure 14. Distribution of rule types across generic uses in Spanish uno

Figure 15. English strategies for the Spanish uno in generic vs. non-generic uses

 Consistent with the findings for English one, the analysis detects no statistically significant difference in English equivalents for Spanish uno when expressing inductive generalizations versus rules. However, we observe a significant association between the type of rule expressed by Spanish uno⁵ and the selection of English heterophrases. Specifically, pronouns one and they demonstrate elevated frequency, while you shows significantly reduced frequency in contexts expressing moral rules, see Figure 16.

Figure 16. English strategies for Spanish uno referring to various rule types

5. Discussion

In our study we addressed the research questions about the distribution of generic vs. non-generic uses of One-impersonals in English and Spanish and the interaction of generalization and first-person orientation as their meaning components. We further applied insights from the literature on genericity to our data, specifically the distinction between inductive generalizations and rules-and-regulations, and further classification of rules as social, moral, legal, biological and metalinguistic, and examined their applicability to our data in both languages. Finally, we analyzed the correlation between the generic meaning component in One-impersonals and the syntactic strategy transmitting the same meaning in the corresponding heterophrase in Spanish and English.

The findings reported in section 4 are highly significant, since they demonstrate distinctive properties of One-impersonals in English and Spanish and provide empirical grounds for evaluating claims and predictions of most of the current approaches. First, the examination of the corpus data demonstrated that both languages exhibit a considerable proportion of first-person oriented non-generic uses of One-impersonals (33.15% for English and 32.83% for Spanish). On the one hand, this result is consistent with the observation of van der Auwera et al. (2012) about the use of one to refer to the speaker, which is common among politicians. Besides, it supports the claim that the concealing uno in Spanish is one of its central uses as noted by the proponents of the first-person oriented approach (see Fábregas 2024 and section 2.1 above). On the other hand, our finding diverges from the results reported in Mignot (2015); in her corpus the proportion of one3 referring to the speaker was only 3%, and generic uses (‘everybody’ and ‘everybody including myself’) comprised 97%. This discrepancy might be explained by the difference in genres (BNC versus political discourse) and annotation procedures (paraphrase versus contextual markers). The proportion of 32.83% for Spanish uno is also not expected, taking into account the findings of Gelabert-Desnoyer (2008) about the distribution of uno in political discourse (he reported 2.5% of self-referential use), which can be partly explained by the moderate size of his sample (44 examples) or different time period.

Next, regarding generic uses of one and uno, the distinction between inductive generalizations and rules-and-regulations proves effective for interpreting our data. As shown above, their distributional patterns differ between English and Spanish. For English, the corpus analysis demonstrated a strong prevalence of rules over inductive generalizations (see Figure 3). This is consistent with Moltmann’s (2010) observation about the use of independent generalizations with one in deontic sentences such as laws or general requirements. Since independent generalization in the English data (combining independent generalizations unsupported by speaker’s experience and generalizations supported by speaker’s experience) is more frequent than generalized first-person experience, this result is expected. This explanation is further supported by the finding that inductive generalizations are overrepresented in uses of one when expressing generalized first-person experience (see Figure 4). In contrast, for generic uses of Spanish uno, rules and inductive generalizations are distributed almost equally. Still, the analysis of the interaction between first-person orientation and type of generalizations revealed similar pattern to English: inductive generalizations demonstrate marked overrepresentation in contexts of generalized first-person experience, and in addition rule-type generalizations show corresponding underrepresentation in contexts of generalized first-person experience. Therefore, we can conclude that while in both languages One-impersonals exhibit a tendency to convey inductive generalizations in generalized first-person experience contexts, their use for expressing rules shows significant cross-linguistic variation. This finding may be attributed to the availability of distinct syntactic strategies for expressing rules in each language. For instance, Spanish systematically employs the simple future tense for encoding obligations (as seen in biblical commands), and norms (as found in legal texts), see Escandell-Vidal (2024: 228) while for English it is not the case.

Further analysis of the varieties of rules expressed by One-impersonals revealed that the most frequent in both languages are social rules that prescribe the acceptable behavioral norms for individuals within a given society. This finding aligns with the discursive properties of parliamentary debates that address the problems of communities, see Van Dijk (2000), Gelabert-Desnoyer (2008), and references therein. However, as shown above, the remaining types of rules are distributed differently. For English, the second most common are moral rules expressing universal ethical obligations, followed by legal rules conveying codified normative prescriptions and then metalinguistic rules concerning the use and understanding of words (see Figure 5). For Spanish, the second most common are legal rules followed by moral rules and biological rules describing natural phenomena (see Figure 13). This result again suggests the existence of alternative strategies for conveying rules of different types in each language. The absence of metalinguistic rules in Spanish and biological rules in English in our data may be attributed to the limited sample size and the low frequency of these rule types overall, though their potential existence in these languages cannot be ruled out entirely.

Finally, we examined the role of generalization as a meaning component in cross-linguistic correspondences of One-impersonals. For Spanish heterophrases of English sentences with one, we observe that first-person forms demonstrate significant underrepresentation in generic contexts, while in non-generic contexts, first-person strategies are, in contrast, overrepresented (see Figure 7). This is an expected result since first-person forms inherently refer to the speaker and their personal specific experience, whereas generic statements convey universal truths applied to all humans. In English heterophrases of Spanish sentences with uno, our analysis reveals underrepresentation of the pronoun you in first-person oriented non-generic contexts (see Figure 15), suggesting that Pearson’s (2022) claim about the comparability and interchangeability of both pronouns one and you (see section 2.1 above) is only relevant for their generic uses. Another interesting finding is the association between the expression of moral rules by Spanish sentences with uno and elevated frequency of pronouns one and they along with reduced frequency of the pronoun you in English heterophrases (see Figure 16). However, this pattern requires verification through larger-sample studies.

6. Conclusions

This study set out to investigate the generic uses of English pronoun one and Spanish pronoun uno. First, we reviewed previous approaches to one and uno, as documented in the literature, highlighting that first-person orientation and generalization are key components of their meaning. Furthermore, we discussed properties of generic sentences that express non-accidental regularities over individuals or situations and looked more closely at the distinction between inductive generalizations and rules-and-regulations that we later apply to our data — two samples of examples with one and uno from the Europarl parallel corpus.

The comparative analysis indicated that while the frequency of generic versus non-generic uses is comparable in both languages, there are significant differences in the distributional patterns of rules versus inductive generalizations. Specifically, for generic sentences with one, there is a clear predominance of rules over inductive generalizations in English, while Spanish exhibits no statistically significant distinction between these two categories. This finding may be attributed to the distinct sets of alternative syntactic strategies available in English and Spanish for encoding rules, resulting in quantitatively different distributions of constructions with One-impersonals serving this function. Another finding reveals the prevalence of social rules in the rule-type generalizations for both languages which can be explained by the parliamentary debates’ orientation toward societal regulation and public affairs. The analysis of heterophrases of One-impersonals in both languages demonstrated that in Spanish heterophases of English sentences with one first-person forms are underrepresented in generic contexts, while in English heterophrases for sentences with uno the second person pronoun you is underrepresented in first-person oriented non-generic uses.

The study contributes to better understanding of the properties of One-impersonals in English and Spanish. It shows the advantages of applying observations from theories of genericity to generic uses of one and uno, which revealed the divergence in the distributional patterns of inductive generalizations and rules in English and Spanish. Besides, the parallel corpus approach highlighted the role of generic component of One-impersonals in cross-linguistic correspondences.

Possible directions for further research include the analysis of generic uses of One-impersonals in other types of discourse, such as academic discourse, experimental investigation of One-impersonals, and diachronic analysis of their semantic and pragmatic features. Further research could also explore the distribution of Spanish uno and its feminine form una across different uses and genres, addressing potential asymmetries in their functional and contextual application.

 

 

_{¹ Although our analysis, consistent with prior work, is restricted to subject position, one and uno are not limited to this function and can appear in other syntactic roles.}

_{² We are grateful to the reviewer for highlighting a relevant cross-linguistic parallel: the French on can be analyzed as a non-referential syntactic device (Tesnière’s récessif intégral), as exemplified by the equivalence between En Chine on enferme les dissidents politiques and its passive counterpart, for the analysis of on see also Creissels (2011).}

_{³ While Carlson’s (1977) claim is English-based, the reviewer rightly observes that generic statements like French En France, on tue tous les jours may exhibit eventive properties, suggesting aspectual variation across languages.}

_{⁴ The association plot is based on a χ²-test, where a p-value < 0.05 indicates a statistically significant association between variables. Residuals show the difference between observed and expected frequencies. They are visualized as: blue rectangles where the observed frequency is higher than expected, red rectangles where it is lower than expected, and grey where the difference is negligible (absolute value below 2). The size of each rectangle corresponds to the relative proportion of the cell in the contingency table.}

_{⁵ While contextual analysis was our initial approach, empirical evidence demonstrated that the interpretation is primarily determined by the properties of uno itself.}

About the authors

Elena L. Vilinbakhova

Institute for Linguistic Studies, Russian Academy of Sciences; St. Petersburg State University

Author for correspondence.
Email: elenavilinb@yandex.ru
ORCID iD: 0000-0002-0539-6230

Dr Habil., is a leading researcher at the Department of Grammar Theory at the Institute for Linguistic Studies of the Russian Academy of Sciences, and a full professor at Ludmila Verbitskaya Department of General Linguistics at Saint Petersburg State University

St Petersburg, Russia

Oksana Yu. Chuikova

Institute for Linguistic Studies, Russian Academy of Sciences; Herzen State Pedagogical University of Russia

Email: oxana.chuykova@gmail.com
ORCID iD: 0000-0003-0626-3410

PhD in Philology, is a senior researcher at the Institute of Foreign Languages at Herzen Russian Pedagogical University and a senior researcher at the Department of Grammar Theory at the Institute for Linguistic Studies of the Russian Academy of Sciences

St Petersburg, Russia

References

Supplementary files