Variable Acceptability of Differential Object Marking in Bilingual Galician–Spanish Speakers: An Exploratory Study

Abstract

This paper presents an initial study of the acceptability of differential object marking (DOM) by Galician–Spanish bilinguals in Galicia. The research explores judgments provided by these bilinguals (n = 69) on DOM in both Galician and Spanish and it also explores data from a monolingual Spanish control group (n = 12). The surveys target contexts covering key syntactic and semantic–pragmatic contexts for DOM in Galician and Spanish, based on the existing literature. The Galician data reveal a tendency towards reduced acceptability of DOM compared to Spanish, but without a generalized rejection of DOM in any of the contexts. The Spanish data show variability in both groups. The study contributes insights from an under-studied language pair and aims to open avenues for further work. More generally, it enhances our understanding of DOM in bilingual grammars, particularly in microcontact situations.

1. Introduction

This study investigates differential object marking (DOM) (Bossong 1991; a.m.) in bilingual speakers of Galician and Spanish, in order to explore the degree of acceptability and variability in their responses to equivalent written inputs in both languages using an online survey.

DOM is mandatory in Spanish with certain direct objects. Consider the contrasts between the direct objects in the examples in (1):

(1)	a.	Juan visitó el museo
		‘Juan visited the museum’
	b.	Juan visitó a su hijo DOM
		‘Juan visited his son’

Both examples involve the same subject and the same verb. However, only the direct object in (1b) requires the DOM marker/P a (lit. ‘to’). Typically, this contrast is explained in terms of the properties of the direct object itself. In (1a), the direct object is an object, thus it is not human/not animate. In (1b), the direct object is a human/animate, definite/specific referent. We provide more information on factors favoring DOM in Section 2.

Contrary to Spanish, the limited literature on DOM in Galician (e.g., López Martínez 1999; Cidrás Escáneo 2006) suggests that the equivalent to (1b) in Galician does not display the DOM marker (2):

(2)	Onte detiveron o fillo de María (López Martínez 1999, p. 558)
	‘Yesterday they arrested María’s son’

However, the same literature, which we discuss in Section 2 below, reports a variable use of DOM.

Within Romance languages, DOM is attested with different degrees of grammaticality (see Gerards 2023 for a recent overview). In Spanish, its use is extensive (Pensado 1995; Fábregas 2013; a.o.), but it is more limited in Italian, e.g., in some topicalized DPs (Belletti 2018) and in other Romance languages (Iemmolo 2010 for Sicilian and other languages; Ledgeway et al. 2019, on southern Italo–Romance languages in Italy; Cyrino and Irimia 2019, on Brazilian Portuguese).

Since DOM is not present in all languages, bilingual speakers pose interesting scenarios, since both of their languages or only one of them may display DOM (Mardale and Karatsareas 2019, p. 3). Studies on DOM in bilingual speakers have received growing attention, including DOM among heritage bilingual speakers (e.g., Sorgini 2021; Friulian in Argentina, Montrul 2022) or DOM among regional bilinguals (e.g., Rodríguez-Ordóñez 2020, on Spanish–Basque). Perpiñán (2018) found that Catalan–Spanish bilinguals in Spain exhibited generalized variability in the use of DOM in all contexts tested and also the expansion of DOM in the grammars of all types of Catalan–Spanish bilinguals, leading to a new bilingual grammar.

Unlike other bilinguals, Galician–Spanish bilinguals are under-studied (see Sheehan et al. 2019 for insights into inflected infinitives within this bilingual group). This paper aims to address this gap by specifically documenting the acceptability of DOM among Galician–Spanish bilinguals, in order to test the degree of variability of such acceptability, if any, in both their languages. The contrasts we have briefly seen in (1) and (2) above suggest that some differences should arise. In addition, to provide a background for any possible variable use of DOM, a monolingual group of Spanish speakers was also included in our study. In Section 2, we provide an overview of DOM triggers in Spanish and Galician based on the existing literature, however a detailed account of DOM in Spanish and Galician is outside of the scope of this paper.

2. DOM in Spanish and Galician

2.1. DOM in Spanish

Differential object marking is a well-known and well-studied phenomenon in Spanish, both synchronically and diachronically (cf. the different chapters in Pensado 1995; Aissen 2003; Leonetti 2004; López 2012, among many others). For the purposes of this paper, we will follow the detailed critical descriptions of DOM triggers provided in RAE-ASALE (2009, pp. 2630–49) and in Fábregas (2013).

In Spanish, DOM is mandatory in some configurations (see example 1b above) and is open to variation in others, with or without semantic differences. The different explanations about DOM in Spanish have put forward a number of syntactic and semantic triggers. Some of them focus on specific features of the objects themselves. Key triggers include animacy and human-likeness of the object (e.g., humans, but also personified animals or objects), definiteness and/or specificity of the object, usually triggering DOM use along a scale (Aissen 2003, a.o.). This is what we have seen in (1b). Other factors adduced in the literature include the semantics of certain verbs and some syntactic configurations, such as left dislocation/topicalization of the direct object and secondary predication (see Fábregas 2013 for a critical overview of the limitations of those factors to account for the use of DOM as a whole).

Let us briefly illustrate some typical contexts of DOM in Spanish, which will serve as the basis for the comparison with Galician. DOM is mandatory with tonic personal pronouns (3a); with interrogative, exclamative, relative and some indefinite pronouns, if they are [+human] (3b), and with definite [+human] direct objects (3c):

(3)	a.	La vi a ella (RAE-ASALE 2009, p. 2632)
	‘I saw her’
	b.	No obligaremos a nadie (RAE-ASALE 2009, p. 2632)
	‘We won’t force anyone’
	c.	He visto a tu hermano (RAE-ASALE 2009, p. 2633)
	‘I have seen your brother’
	d.	He visto a Juan
	‘I have seen John’

Example (3c) is similar to (1b), namely a direct object that is [+human], [+animate] and [+definite]. Human proper nouns are also marked with DOM (3d).

Comparatively, DOM is not always grammatical with indefinite [+human] DPs (4). This contrast has traditionally been explained as specificity vs. non-specificity, although such semantic differentiation has been called into question (RAE-ASALE 2009, pp. 2638–41; Fábregas 2013, p. 19). A well-known case is illustrated in (4); in (4a), the unspecific reference of the doctor (any doctor) is compatible with no DOM, whereas in (4b), DOM marks the specific reference of the doctor, it is a known one (a specific one).

(4)	a.	Buscaba un médico especialista (RAE-ASALE 2009, p. 2640)
	‘I was looking for a specialist doctor’ (any)
	b.	Buscaba a un médico especialista (RAE-ASALE 2009, p. 2640)
	‘I was looking for a (specific) specialist doctor’

Other typical contexts involve relative clauses with indicative vs. subjunctive (5). The indicative in (5a) is linked to the specific reading of the modified noun, as the continuation makes explicit, since the proper name of the doctor is known in this context. The opposite takes place in (5b), where the subjunctive is used instead – that mood in relative clauses is linked to non-specificity (Fábregas 2013, p. 19):

(5)	a.	Busco a un médico que trabaja aquí. Su nombre es Juan.
	‘I’m looking for a doctor that works here. His name is Juan’
	b.	Busco un médico que trabaje aquí. No me importa quién sea.
	‘I’m looking for a doctor that works here. I don’t care who s/he is’
	c.	Busco a la persona que tenga más puntos (Fábregas 2013, p. 19)
	‘I’m looking for the person who has more points’

However, while many of the previous contrasts are clear for speakers, the indicative–subjunctive contrast is not full-proof. In (5c), the speaker may not identify the winner/person by name yet, but knows that one exists, yet the subjunctive tenga is used. This illustrates the fact that in some contexts, DOM may be open to variation for Spanish speakers.

Finally, certain verbs such as preceder a ‘to preceed’ and seguir a ‘to follow’) (RAE-ASALE 2009, pp. 2647–48; Fábregas 2013, pp. 15–16, 27–29) (6a), topicalization (6b) (Fábregas 2013, pp. 32–33), and secondary predication (Fábregas 2013, p. 34) (6c) may trigger DOM regardless of animacy, human-likeness, definiteness, etc.:

(6)	a.	El sujeto precede al verbo (Fábregas 2013, p. 15)
	‘The subject goes before the verb’
	b.	A la cautiva Zenobia (la) presentaron al vencedor (Fábregas 2013, p. 33)
	‘The prisoner Zenobia, they offered her to the winner’
	c.	Quise imaginarme a un grano de trigo aislado de los demás (Fábregas 2013, p. 34; Roegiest 1980, p. 145)
	‘I wanted to imagine a grain of wheat isolated from the rest’

2.2. DOM in Galician

Unlike in Spanish, DOM in Galician is under-studied. To our knowledge, in addition to some brief notes in grammars of contemporary Galician (e.g., Álvarez and Xove 2002, pp. 97–98) and some recent formal studies on some contexts (Gravely 2021; Gravely and Irimia 2022), only a few linguists have studied it in some detail, including López Martínez (1999) and Cidrás Escáneo (2006). We will follow these works in this section and will introduce the key contexts that guided our study.

López Martínez (1999, p. 553) states that DOM is mandatory with tonic personal pronouns (7), like in other Romance languages; in all other cases, DOM is either absent or open to variation (8), with semantic notions such as animacy and definiteness (determinación), or the semantics of the verbs, involved as key factors:

(7)	a.	Collíame a min (López Martínez 1999, p. 554)
	‘He grabbed me’
	b.	Deixaba ir a ela co ganado (López Martínez 1999, p. 554)
	‘He allowed her to go away with the cattle’

(8)	a.	Ca, xa non ven a ninguén, non ven máis que un ó outro (López Martínez 1999, p. 555)
	‘Because they no longer see anyone, they only see one another’
	b.	E ti non vías ninguén e despois diante vías un carneiro (López Martínez 1999, p. 555)
	‘And you would not see anyone and then in front of you you would see a sheep’

Cidrás Escáneo (2006, pp. 154–56) adds that DOM is completely grammaticalized with quantifiers such as todos (9a) or calquera (9b), and tends to be present with demonstratives (9c) when their referent is human:

(9)	a.	Isto último foi o que mais acontentou a todos (Cidrás Escáneo 2006, p. 154)
	‘This latter thing was what satisfied them all the most’
	b.	Para o posto de encargado non van contratar a calquera (Cidrás Escáneo 2006, p. 155)
	‘For the manager position they are not going to hire whoever’
	c.	Nunca vin diante miña a eses (Cidrás Escáneo 2006, p. 156)
	‘I never saw them in front of me’

López Martínez (1999, p. 554) indicates that DOM is common with [+human] proper nouns (10) and is more limited with common nouns. Nevertheless, even human proper nouns are attested with no DOM in some parts of Galicia and in some subtypes (11); for instance, when the proper noun includes an article (12a) or a title (Sir, Madam, etc.) (13) (López Martínez 1999, pp. 556–57; Cidrás Escáneo 2006, pp. 156–59):

(10)	a.	¿Conoces (coñeces) a Marcelina? (López Martínez 1999, p. 556)
	‘Do you know Marcelina?’
	b.	A boa muller ollaba a Pedriño e xa lle parecía un santo (Cidrás Escáneo 2006, p. 156)
	‘The good woman looked at Pedriño and he already seemed to her to be a saint’

(11)	Onte vin Pepe no bar (López Martínez 1999, p. 556)
	‘Yesterday I saw Pepe in the pub’

(12)	a.	Onte fun ver o Pepe da Sara que está na cama (López Martínez 1999, p. 556)
	‘Yesterday I went to see Sara’s Pepe, who is in bed’
	b.	Onte vin ó Pepe pola rúa (López Martínez 1999, p. 556)
	‘Yesterday I saw Pepe on the street’

(13)	Dixo el “quen me dera ver a siña (señora) María (López Martínez 1999, p. 557)
	‘He said: I wish I could see Mrs María’

With [+human] common nouns, DOM is mostly absent, regardless of the animacy or definiteness of the object (14). DOM is also absent with family nouns, even in cases of high definiteness (15) (López Martínez 1999, p. 557). A couple of counter-examples are listed in (16):

(14)	a.	Había que ir buscar un curandeiro (López Martínez 1999, p. 557)
	‘We had to go look for a healer’
	b.	Conoceu (coñeceu) unha rapaza de Cariño (López Martínez 1999, p. 557)
	‘He met a girl from Cariño’
	c.	Eu criei catro fillos en seis anos (López Martínez 1999, p. 557)
	‘I raised four kids in six years’
	d.	E cando mataron un fillo de Terito, o fillo máis vello (López Martínez 1999, p. 557)
	‘And when they killed one of Terito’s sons, the oldest one…’

(15)	a.	Prefería cambiar ela a nena (López Martínez 1999, p. 558)
	‘She preferred to change the little girl herself’
	b.	Onte detiveron o fillo de María (López Martínez 1999, p. 558)
	‘Yesterday they arrested María’s son’

(16)	a.	Visitei a papá (Cidrás Escáneo 2006, p. 161)
	‘I visited dad’
	b.	Ir en bicicleta ver a unha moza (López Martínez 1999, p. 560)
	‘To cycle to go see a young woman’

While the authors do not include examples with [+human] common nouns modified by relative clauses with the indicative and subjunctive, we can conclude that no DOM is expected, regardless of the mood.

Several other factors may play a role in DOM in Galician, as it does in Spanish, for example the presence of secondary predication (17), or certain verbs such as chamar, preceder, and subsituír (18) (López Martínez 1999, p. 559; Cidrás Escáneo 2006, pp. 164–65):

(17)	a.	Mandaron ó fillo maior busca-la súa irmá (López Martínez 1999, p. 560)
	‘They sent the oldest son to search for his sister’
	b.	Eu xa tiña á miña nai moi enferma (López Martínez 1999, p. 560)
	‘I already had my mom very sick’

(18)	a.	Que tome a decisión de chamar precisamente a este home, a Macín Picallo, o
	escultor (López Martínez 1999, p. 560)
	‘May he make the decision to call precisely this man, Macín Picallo, the sculptor’
	b.	O outono precede ao inverno, e a primavera ao verán (Cidrás Escáneo 2006, p. 165)
	‘Autumn comes before winter, and spring comes before summer’

However, topicalization/dislocation is not sufficient to force DOM (López Martínez 1999, p. 558; Cidrás Escáneo 2006, pp. 156, 162):

(19)	a.	Os fillos tenos todos casados (López Martínez 1999, p. 558)
	‘The kid, they have them all married’
	b.	E outra filla téñoa nas Vascongadas (López Martínez 1999, p. 558)
	‘And another daughter, I have her in the Basque Country’
	c.	Eses nunca os vin diante miña (Cidrás Escáneo 2006, p. 156)
	‘Those, I never saw them in front of me’

Based on the discussion of the contexts indicated above, López Martínez (1999, p. 561) concludes that DOM in Galician is not grammaticalized, other than with personal pronouns “a diferencia do castelán” (‘unlike Spanish’; our translation). The contact between Galician and Spanish is undeniable (see Gugenberger et al. 2013 for an overview). This has led to some prescriptive push against DOM in Galician (Cidrás Escáneo 2006). However, López Martínez (1999) shows that DOM is attested historically in Galician and, therefore, regardless of any contact with Spanish, it is not an exogenous feature. The evolution of DOM in Galician is thus similar to that attested for Catalan by Pineda (2021) (see also Gerards 2023, p. 26).

Table 1. Summary of factors that favor DOM in Galician and Spanish.

DOM Expected with ___	Personal Pronouns	+Human +Definite	+Human -Definite	-Human	Certain Verbs	Topicalization	Secondary Predication
SPAN	Yes	Yes	Variable	No	Yes	Yes	Yes
GAL	Yes	Common with proper nouns Mostly absent with common nouns	Mostly absent	No	Yes	No	Yes

below presents a summary of the factors that favor the use of DOM in Galician and Spanish.

2.3. Motivation of the Current Study

The sociohistorical and linguistic backdrop in the previous sections sets the stage for our investigation into the acceptability of DOM in Galician–Spanish bilinguals, with data from monolingual Spanish speakers included as well, aiming to uncover potential divergences and nuances in their linguistic judgments. Motivated by the discussions on Galician DOM contexts, particularly highlighted by López Martínez (1999), who concludes that DOM in Galician, unlike in Spanish, is not grammaticalized except with personal pronouns, our study aimed to shed light on the grammar of Galician speakers, all of whom are bilingual with Spanish.1 Thus, based on the differences noted above, we embarked on data collection with some specific contexts in mind.

We carefully selected contexts that would highlight the key differences noted between the two languages. These contexts were chosen to reveal how Galician–Spanish bilinguals accept DOM, with a monolingual control group of Spanish speakers from a different region in Spain, thereby shedding light on the interplay between the two linguistic system.

Our study is exploratory in nature, with the primary goal of understanding and describing patterns in the data, rather than making inferential claims. Our emphasis lies on a qualitative exploration and descriptive insights. As Tizón-Couto and Lorenz (2021)2 highlight, it is important to remember that a statistical model is not the reality itself, but an abstract representation of it. This representation should highlight relevant aspects to help us understand the phenomenon of interest. As such, detailed statistical analyses would require a large and sociolinguistically representative corpus of present-day Galician with all that is not (yet) available (cf. Fernández Rei 2010), and thus, are forcefully reserved for subsequent, more extensive investigations. The ensuing section explains the methodology employed in our study.

3. Methodology

3.1. Method

The method used to elicit answers was an acceptability task in the form of a forced choice task with three answer options: only A, only B, maybe both. In every instance, one sentence featured DOM, while the other did not. To be able to reach as many participants as possible, we opted to conduct a written survey on Qualtrics. Figure 1 displays a screenshot depicting how participants see the survey.

In order to facilitate a comparison of results between bilinguals and monolinguals, we created one version of the survey in Spanish and one in Galician. The Galician–Spanish bilinguals were invited to respond to the survey that combined Spanish and Galician examples, which were shuffled in order to represent the bilingual reality of day-to-day life in Galicia (cf. Johns et al. 2018). In total, the survey included 31 pairs of examples, illustrating different key contexts of DOM that we wanted to analyze. Some pairs counted toward more than one context. Due to our oversight, one example was repeated in the survey, so we opted to omit the second occurrence from our study. Two contexts exploring common nouns were inadvertently omitted from the Galician survey and one Galician example was not present in Spanish, and was likewise removed from our study. Consequently, we have a total of 29 valid contexts for Galician and 31 contexts for Spanish.

Following the criteria established by López Martínez (1999) and Cidrás Escáneo (2006), we identified specific contexts of interest, summarized as follows. Please note that the complete lists of examples for both languages, along with the corresponding numerical and alphabetical identifiers for each example, are available in the documents provided in the Appendix A:

A.

Personal pronouns, indefinite pronouns [+human] (three pairs of examples; see Appendix A numbers 1–3 in the Spanish examples, letters A–C in the Galician examples). Example:

(20)	A.	Pepe chamoume min onte.
	B.	Pepe chamoume a min onte.
	‘Pepe called me yesterday’

B.

Proper nouns [+human] (one pair of examples; number 4 in the Spanish examples, letter D in the Galician examples). Example:

(21)	A.	Onte vin Pepe
	B.	Onte vin a Pepe.
	‘Yesterday I saw Pepe’.

C.

Common nouns [+human] (24 pairs of examples in Galician [letters F-cc], 26 pairs in Spanish [numbers 6–31]). Example:

(22)	A.	Preciso atopar aquela enfermeira que pasaba a mañán con miña nai. Onde estará?
	B.	Preciso atopar a aquela enfermeira que pasaba a mañán con miña nai. Onde estará?
	‘I need to find that nurse that used to spend the morning with my grandma. Where might she be?’.

D.

Other syntactic/selectional factors (six pairs of examples; numbers 3, 5, 15, 17, 18, and 28 in the Spanish examples, letters C, E, O, Q, R, Z in the Galician examples). Example:

(23)	A.	Nun desfile militar un soldado sigue outro.
	B.	Nun desfile militar un soldado sigue a outro.
	‘In a military parade one soldier follows another’.

3.2. Design and Its Challenges

We should note that designing a feasible and useful online survey to assess DOM in Galician posed some challenges. The foremost challenge pertained to the poor internet accessibility and connectivity issues within the internet infrastructure in Galicia, particularly in less urbanized areas (see Sheehan et al. 2019, for a detailed discussion). Due to these concerns, we opted for a written survey exclusively, as embedding audio on Qualtrics could potentially pose technical challenges for participants, as previously noted by Sheehan et al. (2019). Another obstacle was ensuring the geographical representativity of the informants. Given the available channels for participant recruitment, equal participation from all areas could not be guaranteed. This, along with the curiosity about potential geographical variation or variation stemming from other extra-linguistic variables, underscores the importance of collecting sociolinguistic background information. As such, towards the conclusion of the survey, participants were requested to provide background linguistic information, including details such as gender, place and year of birth, parents’ place of birth, places where the participant had lived, job, education level, age of initial use of Galician and Spanish, years of schooling in Galician and in Spanish, language used in school, language used at home, language used by the participants’ parents with the participant; use of Galician, Spanish, or other languages with friends, use of Galician, Spanish, or other languages in school or at work, self-reported proficiency in spoken Galician and in Spanish, self-reported level of comprehension of Galician and Spanish, other languages spoken by the participant, and level of proficiency. Similar questions were part of the Qualtrics survey for the monolingual group; in this case, for obvious reasons, no questions about their knowledge and use of Galician were added; instead, the participants were asked whether they knew any other languages and at what level, and whether they considered themselves to be native speakers of other languages apart from Spanish.

A final challenge related to the standard spelling conventions for Galician, which impact DOM. In Galician, contractions between the DOM marker “P a” and the definite article have two accepted spellings; some speakers write “ao” (literally, “to-the masculine”), while others use “ó.” We consistently opted for the latter due to its broader recognition across all age groups. It is worth noting that although the pronunciation of “ao” and “ó” is the same, the subtle difference in vowel openness between the determiner “o” and the contraction “ao” or “ó” complicates the identification of when a speaker is employing DOM. This should be considered in the interpretation of the results.

3.3. Participants

We obtained 69 (58 female, 11 male) complete responses in Galicia (out of 103 participants), using the friend of a friend approach. The responses from the sociolinguistic background questionnaires revealed that all participants identify as bilingual in Galician and Spanish to a certain degree. Among them, 47 individuals self-assessed their Galician proficiency as 8–10, 17 reported a proficiency score of 5–7, and only five indicated a lower score between 0–4 (Figure 2). The majority are early bilinguals, with ten reporting a later acquisition of Galician, defined as being from the age of six or older; two of those, who also reported having lower levels of spoken Galician proficiency, reported having learned Galician in their 20s. Of the 69 participants, 58 rated their understanding of Galician as 10 out 10, five of them reported it as 9 out of 10, three reported 8 out of 10, one reported 7 out of 10, and two reported 6 out 10 (Figure 3). In terms of home language use, 23 participants reported using only Galician, 20 reported using Spanish, 22 indicated using both languages, and four reported using another language. Geographically, responses were obtained from various areas around Galicia. Finally, the age range of participants spanned from 1977 to 2003, with the oldest participants being in their early 40s at the time of testing. Figure 2 and Figure 3 below illustrate their reported use (spoken proficiency) and understanding (comprehension proficiency) of Galician and Spanish, showing a slight tendency to speak and understand Spanish better.

The Spanish monolingual group included 12 complete answers from Albacete, Spain (out of 14 participants) (eight male, four female). This group only received the Spanish version of the questionnaire. Their year of birth ranged from 1952–1998 (three were born in the 50s–60s, one was born in the late 70s, five were born in the 80s, and three were born in the 90s).

Overall, the makeup of the bilingual participants is diverse, reflecting the linguistic situation of Galicia. This diversity makes it difficult to pinpoint any extralinguistic factors that may play a role in the study, such as socioeconomic status, educational background, or regional linguistic influences. The range of ages, geographic distribution, and varying degrees of language proficiency and use among the participants introduce numerous variables that could potentially impact the results. Consequently, while this diversity enriches the data, it also poses challenges in isolating specific extralinguistic factors that might influence bilingual language practices and preferences. Thus, in the next section, we will focus on the linguistic contexts as outlined earlier.

4. Results

The following graphs show the numerical data (see Appendix B for the corresponding tables).3 Context A includes personal and indefinite pronouns; Context B is about proper nouns, Context D includes several syntactic and semantic factors, such as certain verbs, topicalization, and secondary predication, and Context C, the largest group, focuses on common nouns. We use numbers to refer to the Spanish examples and numbers for the Galician ones (please check the items in Appendix A).

We report first on the Spanish-only results. In Graph 1 and Graph 2 we include the raw numbers for the three examples for Context A for the monolingual group and for the Galician–Spanish bilingual participants (

Table A1. (a) Context A (personal and indefinite pronouns), Spanish answers by monolingual group. (b) Context A (personal and indefinite pronouns), Spanish answers by bilingual group.

(a)
#	SPAN-mon
	No DOM	DOM	Both
1	0	10	2
2	0	12	0
3	0	12	0
SUM	0	34	2
%	0%	94.44%	5.56%
(b)
#	SPAN-bil
	No DOM	DOM	Both
1	1	60	8
2	0	68	1
3	1	68	0
SUM	2	196	9
%	0.97%	94.69%	4.35%

in Appendix B).

Graph 1. Context A (personal and indefinite pronouns), Spanish answers by monolingual group.

Graph 1. Context A (personal and indefinite pronouns), Spanish answers by monolingual group.

Graph 2. Context A (personal and indefinite pronouns), Spanish answers by bilingual group.

Graph 2. Context A (personal and indefinite pronouns), Spanish answers by bilingual group.

This context seems to be quite clear for both groups, both favor DOM for all three examples, with only a few answering Both.

Graph 3 reports the example we included to test proper nouns. As with the previous context, the results are almost identical, as DOM was the only option for the monolingual group (SPAN-mon) and the clear majority for the bilingual group (SPAN-bil). This is in line with the common descriptions of DOM in Spanish (

Table A2. Context B (proper noun), Spanish answers, both groups.

#	SPAN-mon			SPAN-bil
	No DOM	DOM	Both	No DOM	DOM	Both
4	0	12	0	2	67	0

in Appendix B).

Graph 3. Context B (proper noun), Spanish answers, both groups.

Graph 3. Context B (proper noun), Spanish answers, both groups.

Graph 4 and Graph 5 group together the examples that included some other syntactic or selectional factors, that is, verbs that tend to require DOM regardless of the object, topicalization, etc. as explained above. This group includes examples that are also part of other contexts, as the numbers and letters show (e.g., number 3 in Spanish also belongs to Context A). The graphs allow us to visualize the fact that there is agreement with regards to the highest scoring option for each example in both groups. Notice, however, that the number of No DOM answers in the bilingual group is higher than in the monolingual group (

Table A3. (a) Context D (other syntactic and semantic factors), Spanish answers by monolingual group. (b) Context D (other syntactic and semantic factors), Spanish answers by bilingual group.

(a)
#	SPAN-mon
	No DOM	DOM	Both
3	0	12	0
5	0	12	0
15	0	7	5
17	5	1	6
18	1	4	7
28	1	7	4
SUM	7	43	22
%	9.72%	59.72%	30.56%
(b)
#	SPAN-bil
	No DOM	DOM	Both
3	1	68	0
5	1	65	3
15	0	59	10
17	27	8	34
18	12	14	43
28	6	43	20
SUM	47	257	110
%	11.35%	62.08%	26.57%

in Appendix B).

Graph 4. Context D (other syntactic and semantic factors), Spanish answers by monolingual group.

Graph 4. Context D (other syntactic and semantic factors), Spanish answers by monolingual group.

Graph 5. Context D (other syntactic and semantic factors), Spanish answers by bilingual group.

Graph 5. Context D (other syntactic and semantic factors), Spanish answers by bilingual group.

Finally, Graph 6 and Graph 7 report the results for the largest context in our study, Context C (

Table A4. (a) Context C, Spanish answers by monolingual group. (b) Context C, Spanish answers by bilingual group.

(a)
#	SPAN-mon
	No DOM	DOM	Both
6	1	4	7
7	0	9	3
8	2	7	3
9	0	6	6
10	0	9	3
11	0	10	2
12	0	8	4
13	2	5	5
14	0	11	1
15	0	7	5
16	1	2	9
17	5	1	6
18	1	4	7
19	0	2	10
20	1	2	9
21	4	2	6
22	1	5	6
23	0	11	1
24	1	2	9
25	2	1	9
26	0	4	8
27	0	8	4
28	1	7	4
29	1	5	6
30	0	6	6
31	1	2	9
SUM	24	140	148
%	7.69%	44.87%	47.44%
(b)
#	SPAN-bil
	No DOM	DOM	Both
6	5	49	15
7	1	53	15
8	5	45	19
9	2	49	18
10	2	67	0
11	1	64	4
12	4	56	9
13	3	54	12
14	0	65	4
15	0	59	10
16	12	13	44
17	27	8	34
18	12	14	43
19	6	20	43
20	12	12	45
21	17	18	34
22	15	10	44
23	2	62	5
24	5	42	22
25	20	12	37
26	3	33	33
27	5	46	18
28	6	43	20
29	4	36	29
30	9	20	40
31	12	18	39
SUM	190	968	636
%	10.59%	53.96%	35.45%

in Appendix B).

Graph 6. Context C, Spanish answers by monolingual group.

Graph 6. Context C, Spanish answers by monolingual group.

Graph 7. Context C, Spanish answers by bilingual group.

Graph 7. Context C, Spanish answers by bilingual group.

These graphs show that most examples produce similar results, with some diverging ones; for instance, in item 24, we see that the monolingual group tended to select Both, whereas the bilingual group tended to select DOM.

Overall, the data reveal that the two groups are not very different in any of the contexts. Graph 8 provides the sum of the total responses and corresponding percentages for better comparison between the two groups (

Table A5. Total numbers and percentages of Spanish answers for both groups.

#	SPAN-mon			SPAN-bil
	No DOM	DOM	Both	No DOM	DOM	Both
SUM	24	198	150	195	1296	648
%	6.45%	53.23%	40.32%	9.12%	60.59%	30.29%

in Appendix B).

Graph 8. Total numbers and percentages of Spanish answers for both groups.

Graph 8. Total numbers and percentages of Spanish answers for both groups.

We only find some diverging tendencies. In Graph 8, we can see that the percentage of No DOM is higher for the bilingual group than for the monolingual one. Likewise, the bilingual group seems to be more categorical in their answers, with a lower percentage of answers of Both than in the monolingual group. However, in general, we find that there does not seem to be any exceptional differences between the Spanish answers of both groups; in fact, in both cases, there seems to be some internal divergence.

Let us focus now on the answers provided for both languages by the bilingual group, that is, by the same speakers, which is the central point in this study. As mentioned before, please note that, due to an omission error, two examples in Spanish were absent in the Galician questionnaire (those contexts were discarded and appear shaded in black in

Table A9. Context C (common nouns), Galician and Spanish answers by bilingual group.

#	GAL-bil			#	SPAN-bil
	No DOM	DOM	Both		No DOM	DOM	Both
F	21	21	27	6	5	49	15
G	15	31	23	7	1	53	15
H	17	24	28	8	5	45	19
I	22	27	20	9	2	49	18
J	26	38	5	10	2	67	0
K	10	53	6	11	1	64	4
L	13	45	11	12	4	56	9
M	18	28	23	13	3	54	12
N	13	47	9	14	0	65	4
O	12	37	30	15	0	59	10
P	36	6	27	16	12	13	44
Q	36	2	31	17	27	8	34
R	27	6	36	18	12	14	43
S	22	14	33	19	6	20	43
T	36	4	29	20	12	12	45
				21
U	29	5	35	22	15	10	44
V	17	33	19	23	2	62	5
X *	18	16	35	24	5	42	22
				25
W	12	23	34	26	3	33	33
Y	19	38	12	27	5	46	18
Z	16	35	18	28	6	43	20
Aa	17	27	25	29	4	36	29
Bb	26	7	36	30	9	20	40

* Please note that the Galician example X is listed before the Galician example W, simply because X corresponds to example 24 in Spanish whereas example W corresponds to Spanish example 26.

in Appendix B).

The first context is Context A, reported in Graph 9 (

Table A6. Context A (personal and indefinite pronouns), Galician and Spanish answers by bilingual group.

#	GAL-bil				SPAN-bil
	No DOM	DOM	Both		No DOM	DOM	Both
A	12	42	15	1	1	60	8
B	3	54	12	2	0	68	1
C	0	69	0	3	1	68	0
SUM	15	165	27		2	196	9
%	7.25%	79.71%	13.04%		0.97%	94.69%	4.35%

in Appendix B).

Graph 9. Context A (personal and indefinite pronouns), Galician and Spanish answers by bilingual group.

Graph 9. Context A (personal and indefinite pronouns), Galician and Spanish answers by bilingual group.

As expected, the personal pronoun is almost completely DOM in both languages; ‘someone’ and ‘no one’ show more divergences, although the majority continues to be DOM.

Context B (+human proper noun) is reported in Graph 10. The difference here is minimal, in both languages the clear majority of answers are DOM (

Table A7. Context B (proper noun), Galician and Spanish answers by bilingual group.

#	GAL-bil				SPAN-bil
	No DOM	DOM	Both		No DOM	DOM	Both
D	2	63	4	4	2	67	0
%	2.90%	91.30%	5.80%		2.9%	97.10%	0%

in Appendix B).

Graph 10. Context B (proper noun), Galician and Spanish answers by bilingual group.

Graph 10. Context B (proper noun), Galician and Spanish answers by bilingual group.

Graph 11 and Graph 12 illustrate the answers to Context D (

Table A8. Context D (other syntactic and semantic factors), Galician and Spanish answers by bilingual group.

#	GAL-bil				SPAN-bil
	No DOM	DOM	Both		No DOM	DOM	Both
C	0	69	0	3	1	68	0
E	2	64	3	5	1	65	3
O	12	37	20	15	0	59	10
Q	36	2	31	17	27	8	34
R	27	6	36	18	12	14	43
SUM	77	178	90		41	214	90
%	22.32%	51.59%	26.09%		11.88%	62.03%	26.09%

in Appendix B). The first examples are C–3,4 which, as reported above, include a personal pronoun and the verb chamar/llamar. Example E-5 includes the verb seguir, which is listed in both languages as lexically requiring DOM, and that is the case for most informants in both languages. Examples O–15 involve another verb said to lexically require DOM (sustitutir); in this case, the differences are larger than with seguir. Examples Q–17 and R–18 test secondary predication with the verb tener (to have someone somewhere), as reported in the literature. Finally, examples Z–28 showcase left dislocation of the direct object, and the answers are quite similar.

Graph 11. Context D (other syntactic and semantic factors), Galician answers by bilingual group.

Graph 11. Context D (other syntactic and semantic factors), Galician answers by bilingual group.

Graph 12. Context D (other syntactic and semantic factors), Spanish answers by bilingual group.

Graph 12. Context D (other syntactic and semantic factors), Spanish answers by bilingual group.

We can see a great deal of coincidence between both languages. Only examples Q–17 are different, which include secondary predication, but a common noun, which may explain the higher degree of No DOM in Galician. However, notice that the number of No DOM in their Spanish answers is high too. If we pay more attention to the overall results, we can note that the percentage of No DOM answers in Galician is larger than in Spanish (22.32% v 11.88%). But, once more, the tendencies in both languages are pointing in the same direction: clear majority of DOM, followed by Both and then No DOM.

The final context we will present here is the most important one for us, Context C. Based on the existing literature, we expected [+human] common nouns to provide the most differences, which is why we include many more examples of this type in our survey. The results appear in Graph 13 and Graph 14 (Table A9 in Appendix B).

Graph 13. Context C (common nouns), Galician answers by bilingual group.

Graph 13. Context C (common nouns), Galician answers by bilingual group.

Graph 14. Context C (common nouns), Spanish answers by bilingual group.

Graph 14. Context C (common nouns), Spanish answers by bilingual group.

One initial striking aspect made obvious in this graph is the existence of variation among the different examples in both languages.

Let us focus on examples with clear mismatches between the Galician and Spanish answers. Examples F–6 show a preference for DOM for their Spanish answer, whereas the Galician example is almost equally divided between Both, DOM, and No DOM.5 Thus, it seems that their judgments exhibit more consistency across individuals in the Spanish example, for which the DOM option is clearly the majority. This is in keeping with the fact that the indicative in the embedded clause favors DOM. The Galician counterpart shows that each option garnered almost one third of the answers, indicating greater inter-individual variation.

(24)	F.	Preciso atopar aquela enfermeira que pasaba a mañán con miña nai. Onde estará?
	Preciso atopar a aquela enfermeira que pasaba a mañán con miña nai. Onde estará?
	6.	Necesito encontrar aquella enfermera que pasaba la mañana con mi madre. ¿Dónde estará?
		Necesito encontrar a aquella enfermera que pasaba la mañana con mi madre. ¿Dónde estará?
	‘I need to find that nurse that used to spend the morning with my grandma. Where might she be?’

Related to this is examples P–16, which test a non-definite common noun modified by a relative clause in the subjunctive:

(25)	P.	Preciso unha enfermeira que pase a mañá con ela.
		Precisa a una enfermeira que pase a mañán con ela.
	16.	Necesita una enfermera que pase la mañana con ella.
		Necesita a una enfermera que pase la mañana con ella.
		‘She needs a nurse who spends the mornings with her’

In this context, we notice a higher percentage of No DOM in Galician compared to Spanish, where the predominant choice is Both. It is noteworthy to mention that Both is also the most selected answer by the Spanish monolingual group for example 16.

A final case with high mismatch that we want to address is in examples X–24:

(26)	X.	A compañía mandou unha persoa de fóra para o posto.
		A compañía mandou a unha persoa de fóra para o posto.
	24.	La compañía mandó una persona de fuera para el puesto.
		La compañía mandó a una persona de fuera para el puesto.
		‘The company sent an outsider for the position’

In this case, DOM is the predominant choice in Spanish, while Both is the highest in Galician. Notice the too high number of No DOM answers in Galician compared to Spanish.

Conversely, three pairs of examples exhibit similar high scores for DOM in both languages. These three sets are:

(27)	K.	Pepe ataca as personas racistas cando as escoita falar.
		Pepe ataca ás personas racistas cando as escoita falar.
	11.	Pepe ataca las personas racistas cuando las oye hablar.
		Pepe ataca a las personas racistas cuando las oye hablar.
		‘Pepe attacks racist people when he hears them talk’

(28)	L.	Pepe entrevistará os cantantes para a televisión rexional.
		Pepe entrevistará ós cantantes para a televisión rexional.
	12.	Pepe entrevistará los cantantes para la tele regional.
		Pepe entrevistará a los cantantes para la tele regional.
		‘Pepe will interview the singers for the regional TV’

(29)	N.	Pepe trata o neno coma se fora un desconocido.
		Pepe trata ó neno coma se fora un desconocido.
	14.	Pepe trata el niño como si fuera un desconocido.
		Pepe trata al niño como si fuera un desconocido.
		‘Pepe treats the kid as if he didn’t know him’

All of these examples have in common the fact that the nouns are part of a DP headed by the definite article, regardless of the actual definite semantics of the article, since K–12 involves a generic interpretation (racist people in general). At the same time, in all these cases, and in many others in the list, the count of No DOM is higher in their Galician responses than in their Spanish counterparts (Graph 15) (

Table A10. Total numbers and percentages of Spanish and Galician answers to Context C (common nouns) by bilingual group.

#	GAL-bil			#	SPAN-bil
	No DOM	DOM	Both		No DOM	DOM	Both
SUM	107	143	164		39	196	179
%	25.85%	34.54%	39.61%		9.42%	47.34%	43.24%

in Appendix B).

Graph 15. Total numbers and percentages of Spanish and Galician answers to Context C (common nouns) by bilingual group.

Graph 15. Total numbers and percentages of Spanish and Galician answers to Context C (common nouns) by bilingual group.

5. Discussion and Conclusions

In our exploration of DOM in Galician–Spanish bilinguals, we found tendencies that both align with and deviate from the existing literature, providing an initial understanding of the grammar(s) of this bilingual community. Given that the evidence based on DOM in Galician is quite limited and dated, our research serves as a crucial preliminary step, i.e., our findings offer an initial glimpse into the grammatical structures employed by this bilingual community. There are some key findings in our study:

(a)

Initial insights into DOM in bilinguals

Our study reveals trends in DOM usage among Galician–Spanish bilinguals. Notably, these tendencies both conform to and diverge from existing literature, providing valuable new insights.

(b)

Contextual analysis: absence of systematic rejection of DOM in Galician

The data suggest an important conclusion, that there is no systematic rejection of DOM with common nouns in Galician. Contrary to expectations based on the literature, the clear-cut divergence between Galician and Spanish, especially in Context C, does not materialize. However, Galician DOM answers tend to favour No DOM at higher percentages than their Spanish counterparts in the bilingual group, indicating partial alignment with existing literature on Galician DOM.

(c)

Variability in Spanish responses

The Spanish answers from the bilingual group exhibit variation similar to that observed in the monolingual group. This indicates that the Spanish responses from the bilingual group are not substantially different from those of the monolingual group. However, within the bilingual group’s judgments of Spanish, there is a higher prevalence of No DOM instances compared to monolingual Spanish speakers. Conversely, monolingual Spanish speakers accept more DOM, and bilingual participants show more categorical preferences in their responses.

(d)

Comparison with other bilingual communities

Our results partially resemble those reported for Catalan–Spanish bilinguals. Odria and Pineda (2023, p. 320) mention studies supporting a bidirectional influence in the use of DOM, where bilinguals incorporate DOM in Catalan beyond normative grammar and tolerate DOM-less constructions in Spanish. The authors argue that there is a difference between “what is accepted by prescriptive grammars and what speakers actually do … in Catalan” (Odria and Pineda 2023, p. 321). In our Galician–Spanish bilinguals we found no generalized rejection of DOM with common nouns, which is against the expectations of the normative use in Galician. We also attest some more No DOM responses in their Spanish answers, although we have already discussed that variation in Spanish DOM is independently documented in the monolingual group.

(e)

Intra-linguistic variation: role of individual idiosyncrasies

We have documented clear intra-linguistic variation in our study. The absence of a generalized rejection of DOM with common nouns in Galician suggests that individual speakers play a pivotal role in shaping morphosyntactic choices. However, additional data are necessary to identify the specific factors involved. This echoes similar findings in colloquial Catalan varieties reported by Pineda (2021), highlighting the importance of individual idiosyncrasies in language use.

There are also several methodological challenges that should be highlighted:

(a)

Challenges in detecting DOM

One challenge we encountered is the difficulty bilinguals may face in detecting the marking of DOM due to subtle differences. Two significant issues include (i) the variability in vowel opening associated with DOM marking in Galician and (ii) the reliance on accent marks in written language to distinguish between DOM and No DOM, a feature that many speakers may overlook. These intricacies, particularly within the bilingual context of Spanish and Galician, introduce methodological complexities that require careful consideration.

(b)

Participant diversity: geographical and demographic variation

Another challenge is the diversity among our participants; our sample comprises individuals with varied bilingual backgrounds, including early and later bilinguals, as well as two new speakers. Additionally, there is considerable geographical and age diversity within the sample, although a notable proportion consists of younger speakers from the Vigo area, primarily due to accessibility reasons. This range of linguistic backgrounds and demographics reflects the broader sociolinguistic makeup of Galician-speaking communities.

Overall, our study is limited by our research question and by the relatively small size of our corpus. Future research should aim to extend the discussion beyond linguistic factors, to explore extralinguistic influences, such as sociocultural dynamics and individual differences, which may contribute to the observed patterns. Our results point to the fact that Galician–Spanish bilinguals display some variation that deviates from what is observed in the Spanish monolingual group. Future research should focus on understanding why this bilingual population diverges from monolinguals and from each other. The role of education should also be explored. A preliminary consultation with several high school teachers of Galician suggests that DOM is usually not explicitly or consistently taught in the classroom. In the future, it would be beneficial to carefully examine both teaching materials and teaching practices. As such, research should aim to determine which methodologies can best account for the observed variation.

An obvious ecological strategy to tackle these challenges involves exploring a corpus that incorporates both oral and written language samples from the same group of speakers in both Spanish and Galician. Such a corpus should also contain extensive information on the sociolinguistic background of the participants. This comprehensive methodology would facilitate a more fine-grained analysis of linguistic patterns and variations across different modalities and speakers. However, we ought to recognize that such an approach demands a substantial investment of time and resources, given the meticulous tasks of collecting, transcribing, and analyzing language samples.

Our findings underscore the significance of investigating contact and bilingualism phenomena from a language–ecological lens. Understanding the nuances of grammatical constraints and recognizing the significance of individual experience and community practices in both judgment and other linguistic tasks (cf. Balam et al. 2020; Valdés Kroff and Dussias 2023; Parafita Couto et al. 2024) amplifies the importance of open access bilingual speech corpora. Without access to community practices, which encompass real-world language usage, we find ourselves severely limited. Ultimately, our study establishes the need for future research, acknowledging challenges and pointing towards future directions in the linguistic exploration of DOM in Galician–Spanish bilinguals, and in Galician more generally.

In sum, our study contributes new data on DOM from an under-studied language combination (Galician–Spanish) to a growing body of knowledge on bilingualism, encouraging a shift towards more context-specific and individualized interpretations of language, particularly evident in the study of microcontact, i.e., contact between two minimally different syntactic systems (grammars), as defined by D’Alessandro (2021, p. 310). Indeed, the difficulties encountered in understanding the variation observed in our study underscore the intricate dynamics between Galician and Spanish, forming a varied bilingual linguistic tapestry.