“How to become a woman without turninginto a Barbie”: Change-of-state verbconstructions and their role in Spanish as a Foreign Language

2.1 The study of PCOS verbs in Spanish L1

Although the study of PCOS verbs has attracted the attention of many researchers with different theoretical perspectives and research agendas, most of this work revolves around three main problematic areas: the categorization of PCOS verbs, the internal classification of PCOS verbs, and the explanation of PCOS verb-construction alternations.

With respect to the first area, the categorization of PCOS verbs has been the focus of attention in descriptive analyses, grammar books, and phraseological studies. Unfortunately, these studies do not agree on basic issues such as the name for this category or the number and type of its members. For instance, they are known as intransitive reflexive verbs (Crespo 1949), attributive reflexive verbs (Navas Ruiz 1963), auxiliaries and functional verbs (Alba de Diego and Lunell 1988), se constructions (Fernández Leborans 1999) or canonical change constructions (Rodríguez Arrizabalaga 2001) in descriptive studies. However, they are considered idiomatic expressions, pluriverbal lexical units, formulaic or prefabricated sequences or lexical chunks in phraseological studies (Casares 1969 [1950]; Firth 1957; Corpas Pastor 1996; Koike 2001). A look at grammars of Spanish does not offer a different picture. Works such as the Nueva gramática de la lengua española (RAE/ASALE 2009) consider these verbs as a category of their own. The Gramática descriptiva de la lengua española (Fernández Leborans 1999) does not treat PCOS separately, but rather as part of a wider phenomenon: the se constructions. This disparity is also reflected in grammar textbooks. Nilsson et al.’s (2014) article, for example, includes PCOS in the group of intransitive change-of-state verbs, whereas Matte Bon’s (1992) communicative grammar groups them together with other verb types that have an effect on the characteristics of the subject (what this author calls “subject transformations”). As far as category members are concerned, the same variation is found. The list of members and structures provided in Crespo (1949) and Coste and Redondo (1965) has served as the starting point for many subsequent studies. These have reduced the number of members but, once again, in different ways (cf. Navas Ruiz 1963; Fente 1970; Lorenzo 1970; Pountain 1984; Eberenz 1985; Porroche Ballesteros 1988; Eddington 1999; Bybee and Eddington 2006; Morimoto and Pavón Lucero 2007; Conde Noguerol 2013; Van Gorp 2017).

In relation to the second area, the internal classification of PCOS verbs also lacks uniformity. Navas Ruiz (1963) and Coste and Redondo (1965), for example, classify PCOS on so-called extralinguistic criteria such as the essential-accidental value of the verb. These criteria, sometimes criticized (Fente 1970), have nevertheless been followed by many other studies (Crespo 1949; Alba de Diego and Lunell 1988; Morimoto and Pavón Lucero 2007; RAE/ASALE 2009; Nilsson et al. 2014). Other semantic variables have also been used to organize this category. For instance, animacy, degree of intentionality, naturalness, radicalness, predictability of the change or degree of subject involvement. All these criteria are quite subjective and as a result, multiple variants, rules and, consequently, easily detectable exceptions occur.

Finally, the third area concerns PCOS verb-construction alternations. The combination of these verbs with lexical items is somehow unrestricted, in the sense that the same lexical item can be freely combined with any of these PCOS verbs. Recall, for instance, example (2) in Section 1; roja ‘red’ combines with hacerse ‘make.cl’, ponerse ‘put.cl’, volverse ‘turn.cl’, and quedarse ‘remain.cl’. Fente (1970) is one of the first studies to point out that the semantic limits between these verbs are diffuse and that their usages frequently overlap. Different reasons have been adduced for this overlapping: emphasis and expressivity (Crespo 1949), stylistic differences (Navas Ruiz 1963), and even dialectal variation (Eddington 1999). Other authors have offered alternative explanations for these alternations. Eberenz (1985) argues that the change has to be understood as a linear progression. When the change crosses a boundary, the transformation has been completed, and therefore, the rest of that line is to be taken as the result. This way of characterizing PCOS verbs on the basis of their role in the change process has been adopted by authors such as Alba de Diego and Lunell (1988), Conde Noguerol (2013), and Van Gorp (2017). The latter offers a cognitive semantic account where this progression is explained by means of the image schema metaphor changes are movements (Lakoff and Johnson 1999: 147), where the source-path-goal schema corresponds to the beginning, the development, and the achievement of the change of state.

This brief review clearly reveals the degree of difficulty that the analysis of PCOS verbs entails. A difficulty enhanced by the absence of a unified account of PCOS verbs. As a consequence, the implementation of these verbs in the language classroom is also challenging.

2.2 The explanation of PCOS verbs in SFL

The multiplicity of explanations for the behavior of PCOS verbs is mirrored in the wide variety of explanations found in SFL pedagogical materials Cheikh-Khamis (Submitted). Most of these materials follow the guidelines established in the Curriculum Plan of the Cervantes Institute (Instituto Cervantes 2007), which specifies CEFRL descriptors for the teaching of Spanish. The CPCI recommends the introduction of these verbs at the B2 level. It is at this moment when the student has to deal with at least five different verbs to express the notion of change, namely hacerse ‘make.cl’, volverse ‘turn.cl’, ponerse ‘put.cl’, quedarse ‘remain.cl’, and convertirse en ‘convert.cl in’.

Together with the intrinsic difficulty of having different verbs to describe a change of state, the lack of appropriate SFL materials make the learning of these verbs quite challenging for the learner. In a recent study, Cheikh-Khamis (Submitted) examines 35 SFL materials (27 textbooks and 8 grammar books) on the basis of 10 pedagogical criteria and concludes that the explanations and rules offered in most of these materials for PCOS are heterogeneous, imprecise, and insufficient to clarify their meaning and structure.

Some of these SFL materials for PCOS verbs are problematic instead of helpful. For example, a recurrent problem is the inclusion of different kinds of verbs of change under the same group. Most SFL materials include the teaching of four PCOS verbs —hacerse ‘make.cl’, volverse ‘turn.cl’, ponerse ‘put.cl’, and quedarse ‘remain.cl’— together with other verbs such as convertirse en ‘convert.cl in’, transformarse en ‘transform.cl in’ or llegar a (ser) ‘arrive to (be)’. These are all change verbs but the latter are both formally and semantically different from PCOS verbs. As far as their structure is concerned, they require a prepositional phrase headed by the preposition en ‘in, on, at’ or a ‘to’, whereas PCOS take a noun or an adjectival phrase. With respect to their meaning, these verbs do share the general meaning of change, but there are crucial differences, as illustrated in (3).

Whereas in (3.a) the change is prototypically understood as physical (the place was turned into an establishment which maintains a collection of wild animals), in (3.b) the change can only be interpreted as metaphorical (the situation was quiet but it suddenly became a madhouse).

Another problem in SFL materials has to do with the explanations offered to guide students in their learning process of the PCOS. These materials usually offer a short description of how PCOS verbs work together with an illustrative example. The procedure itself is adequate, as it is a good combination of theory and practice. The problem, however, lies in the type of explanation. It is typically brief and above all, vague. Hedges such as a veces ‘sometimes’, normalmente ‘normally’ or suele ‘usually’, together with subjective evaluative information based on the length, willingness, and social appraisal of the change, find their way into these explanations. Such elements leave doors open to exceptions, contradictions, and alternative explanations. Let us examine a couple of examples.

Verdía et al. (2005: 183) characterize the PCOS verb hacerse ‘make.cl’ as “Suele presentar cambios que decide el sujeto” (‘It usually covers changes decided by the subject’). Unfortunately, this explanation does not hold for an instance such as (4), because becoming old is a natural process, impossible to be controlled by the subject.

Similarly, volverse ‘turn.cl’ is described as an unavoidable change. In Verdía et al.’s (2005: 183) words, “A veces usamos volverse para cambios más definitivos y negativos” (‘Sometimes volverse is used for more permanent and negative changes’) (Chamorro-Guerrero et al. 2006). Once more, this explanation is incompatible with example (5), where the change is positive and its duration depends on how this person manages his fortune.

The two problems we have just highlighted seem to be quite widespread in SFL textbooks and reference materials, and they reflect the lack of uniformity in those theoretical descriptions discussed in Section 2.1. Now, the question is: why are PCOS verbs so problematic? And, secondly, is it possible to offer a unified account of PCOS verbs? The rest of this paper is devoted to tackling these questions.

3.1 Two problems, one solution: The family of PCOS constructions

From a syntactic perspective, PCOS verbs are neither fully copulative nor predicative. They are mainly inaccusative but they also allow accusative structures. These multiple possibilities are partly responsible for the lack of a unified account. All PCOS are different and therefore, it is not possible to syntactically classify them in a single category without finding exceptions. A possible solution is to focus on those characteristics shared by these PCOS: their middle-voice structure and their intransitive, pronominal character. According to Castañeda Castro (2006), these characteristics are responsible for a shift in the focus of the PCOS verbs from the cause of the change onto the subject itself. On the one hand, the subject loses some of the control features typical in an active agent, but it nevertheless participates in the COS event. On the other hand, the subject itself is the locus of the COS described in the verb, since it is the one that experiments the change. In other words, the subject is simultaneously both agent and experimenter (see Maldonado 1999).

From a semantic perspective, the main challenge lies in the subtle meaning differences among PCOS verbs. Characteristics such as animacy, intentionality or gradualness have been used before, but they are not enough. We argue that it is necessary to consider the specific meaning of the PCOS verbs. This is crucial because the lexical verb is responsible for the subtle differences in meaning for the interpretation of the change-of-state event (that is, the speaker’s own view of how the change is like, see examples in (2)), and for the physical and metaphorical interpretation of the PCOS structure.

A possible way to capture these common and distinctive characteristics is to propose a “family of PCOS constructions” in the sense of Goldberg’s Construction Grammar (1995, 2006). Constructions are defined as form and meaning pairings which are conventionalized and usually non-compositional. In this framework, form and meaning are taken broadly, that is, “form” refers to any kind of structure (phonemic, morphosyntactic, prosodic, etc.) and “meaning” to any kind of semantic, pragmatic, and discursive information. The main idea is that constructions themselves are carriers of meaning independently from the specific linguistic items that saturate (i. e. integrate) them. On some occasions, the choice of linguistic items is restricted (e. g. tightly-bound idiomatic expressions), but on others, it is less constrained, which results in fully productive patterns. The latter may create a family of constructions that interact with each other by means of different mechanisms, called inheritance relations (see Golberg 1995: 72–81), such as polysemy links (same construction with different associated meanings), subpart links (a subpart is also a separate construction), metaphorical extension links (constructions related by metaphorical mappings), and instance links (a construction is a fully specified case of a construction).

In the case of the PCOS construction, its general meaning is “X undergoes a change of state Y”. Notice that the interpretation of the change itself is intrinsically metaphorical since it is understood as a progression (path) from an initial state (source), up to the resulting final state (goal) (Lakoff and Johnson 1999; Van Gorp 2017). The PCOS construction form is represented as [SUBJ Vclitic OBJ]. The configuration of the PCOS construction is represented in Figure 1.

Figure 1:

PCOS construction in Spanish.

This general PCOS construction will then be saturated by different linguistic items that will shape its specific characteristics. All subconstructions will then inherit all the characteristics of the matrix construction by means of instance links, while adding the information of the specific lexical items.

We propose that these subconstructions can be organized in several levels of specificity. A first-level family of constructions is triggered by the choice of a PCOS verb and thus, by its meaning, as shown in (6).

The meaning of the lexical verb that saturates the V position is kept constant in all the possible instances of this construction. These first-level constructions are then further specified by means of different complements, usually adjectival phrases (e. g. volverse loco ‘turn.cl crazy’) or noun phrases (e. g. volverse un loco ‘turn.cl a madman’). This originates a second-level family of subconstructions that depend on the semantics of these complements. If needed, further levels of subconstructions could be identified.

In a nutshell, we propose a hierarchical family of PCOS constructions which at the lower levels inherit the meaning and characteristics of the previous level subconstruction by means of instance links. The idea is to create a bottom-up approach that captures both the specificity of fully-saturated constructions (e. g. María se volvió loca) as well as the more general abstract patterns (e. g. [Subject volverse AdjP] → [Subject PCOS-verb Object]). Figure 2 schematizes how this family of PCOS constructions works.

Figure 2:

A family of PCOS constructions in Spanish.

This type of bottom-up family of constructions is meant to provide the student with the necessary tools to interpret the meaning of PCOS constructions, and not only the general meaning of a change of state, but also the specific nuances of each subconstruction. The goal is to provide a unique and motivated account of PCOS constructions to facilitate the SFL learner’s processing and understanding.

3.2 The family of PCOS constructions into practice: [Subject hacerse/volverse Object_NP]

With the purpose of testing the validity of the family of PCOS constructions, we undertook a corpus study that focuses on two PCOS verbs, hacerse ‘make.cl’ and volverse ‘turn.cl’, with NP complements.

3.2.2 Analysis

Our constructional approach proposes a multi-level family of PCOS constructions. Therefore, in this section, we will examine each of these levels from the most general to the most specific. The goal is to test the role and meaning of the items that saturate each construction in order to establish regularities that might offer the SFL student a motivated account of the PCOS construction in Spanish.

As far as the quantitative data are concerned, the distribution of examples in each construction is summarized in Table 1.

Table 1:

Hacerse/volverse+NP in corpus.

hacerse/volverse+NP		# types (%)	# tokens (%)
hacerse+NP		606 (32.70%)	2546 (56.75%)
	hacerse+∅+noun	581 (95.87%)	2478 (97.32%)
	hacerse+det+noun	25 (4.12%)	68 (2.67%)
volverse+NP		1247 (67.29%)	1940 (43.24%)
	volverse+∅+noun	834 (66.88%)	1347 (69.43%)
	volverse+det+noun	413 (33.12%)	593 (30.56%)
Total		1853	4486

Quantitative data in Table 1 show that volverse+NP is the construction with the highest variety of types (1247, 67.29% vs. 606, 32.70%), but that hacerse+NP is the construction with more tokens per type (2546, 56.75% vs. 1940, 43.24%). In both cases, the construction with the bare NP is more frequent, but this preference is more acute in the case of hacerse+NP. This means that the construction hacerse/volverse with a determiner is not very productive which is an interesting fact from a teaching perspective, as will be discussed in the next section.

In general, these figures seem to indicate that volverse+NP is structurally less restrictive than hacerse+NP, but it also shows that the usage of hacerse+NP is more conventionalized. In other words, hacerse+NP presents a smaller number of types, but they are widely used. For example, a quick look at the examples in hacerse+bare NP reveals that hacerse realidad (make.cl reality) ‘to become real’ and hacerse amigos (make.cl friends) ‘to become friends’ cover over a third of all the tokens in this subconstruction (35,55%); 534 tokens and 347 tokens, respectively. This suggests that the degree of idiomaticity, i. e. how fixed the expression is, is higher in hacerse+NP. Both characteristics, frequency and conventionalization, are key issues in SLA (see, Ellis 2002) and therefore, this kind of instantiations of the subconstruction could play an important role in the acquisition of Spanish.

As far as qualitative data are concerned, we are going to focus on two subcases to illustrate the explanatory power of the family of the PCOS constructions. One examines instances where both verbs, hacerse ‘make.cl’ and volverse ‘turn.cl’, select the same lexical noun to fill the bare NP. The other explores instances of the subconstructions with bare NPs and with NP+determiner.

The goal of the first subcase is to show the way the subconstruction works at the verbal level. In our corpus, out of the 1,505 different nouns, 209 (13.71%) occur with both hacerse ‘make.cl’ and volverse ‘turn.cl’. ^[2] This group of nouns is suitable to show the role of each verbal subconstruction, since this is the level where the difference lies. Let us examine some examples in (7–9).

(7)

a.

Entonces	el	agua	se	volvió	piedra [PVNZ-0973] [3]
then	the	water	cl.3sg	turned.3sg	stone

‘At that moment water turned into stone’

b.

El	corazón	se	hace	piedra [PHNZ-1226]
the	heart	cl.3sg	makes	stone

‘The heart turns into stone’

(8)

a.

Como	una	oruga	que	se	vuelve	mariposa [PVNZ-0697]
as	a	caterpillar	that	cl.3sg	turns	butterfly

‘As a caterpillar when she turns into a butterfly’

b.

El	gusano	se	hace	mariposa	o	quizá	al
the	worm	cl.3sg	makes	butterfly	or	maybe	to.the
revés [PHNZ-6722]
opposite

‘The worm turns into a butterfly or maybe the opposite’

(9)

a.

Imagínate,	si	no	fuera	por	eso	¡el	pueblo	americano
imagine	if	not	were	for	that	the	village	american

entero	se	volvía	fidelista! [PVNZ-0503]
whole	cl.3sg	turned.3sg	fidelist

‘Just imagine! If it weren’t for that, the whole American nation would have become a supporter of Fidel Castro!’

b.

Pepito	se	hizo	fidelista	en	el	59	[PHNZ-0551]
pepito	cl.3sg	made.3sg	fidelist	in	the	59

‘Pepito became a supporter of Fidel Castro in 59’

In these examples, the subjects —heart/water; caterpillar/worm; the American nation/Pepito— undergo a change that results into a different state —stone; butterfly; a supporter of Fidel Castro. This meaning is inherited from the PCOS matrix construction, which is in turn expanded with that of the verbal subconstruction. In all the a. versions, the volverse subconstruction adds the connotation of a change that radically differs from the previous state. In all the b. versions, the change in the hacerse subconstruction refers to the new state but without profiling the previous one. We argue that these are the key differences between these two subconstructions, rather than the involuntary and swift, and consequently negative, character of the change in the volverse construction, or the positive character of the change in the hacerse construction. These traits are very subjective and cannot be applied to all examples. For instance, to turn into a butterfly cannot be considered either positive or negative, as it is just a natural process.

Instead, we suggest that each construction establishes, and therefore profiles, a different relation between the previous and the new state. Similarly to Van Gorp’s (2017) proposal, we also argue that a PCOS construction can be described as a source-path-goal image-schema-based metaphor; however, we propose that the element profiled in each subconstruction corresponds to a different part of the image schema, or in general terms, to a different part of the change-of-state event. Figures 3 and 4 capture these differences. The bold line represents the parts of the event profiled in each subconstruction.

Figure 3:

Volverse PCOS subconstruction.

Figure 4:

Hacerse PCOS subconstruction.

The second subcase we are going to briefly discuss focuses on a higher level of specificity: the structure of the NP. As mentioned above, PCOS subconstructions may take two types of NPs as complements, a bare NP and an NP with an (indefinite) determiner. We have already shown that these two possibilities differ sharply as far as their frequency and prototypicality, but now we would like to explore whether the presence or lack of a determiner triggers meaning differences. Examples (10–11) illustrate these contrasts, where both verbs and nouns are kept constant, but what changes is the presence or absence of the indefinite determiner un (-a, -os, -as) (abbreviated as indf.det (-m.sg, -f.sg, -m.pl, -f.pl)).

(10)

a.

[…] como	si	temiera	hacerse	grande,	o
[…] as	if	was.afraid	make.cl.3sg	big	or

hacerse	mujer [PHNZ-0983]
make.cl.3sg	woman

‘She was kind of afraid of becoming big, or becoming a woman’

b.

La	niña	está	bien,	haciéndose	una	mujer
the	girl	is	ok	making.cl.3sg	indf.det.f.sg	woman

para	el	futuro [PHNI-0035]
for	the	future

‘The girl is ok, just becoming a woman for the future’

(11)

a.

Se	vuelven	señores	los	ladrones [PVNZ-1193]
cl.3pl	turn.3pl	gentlemen	the	thieves

‘Thieves become gentlemen’

b.

Se	volvería	un	señor	en	pantuflas	y
cl.3sg	turned.3sg	indf.det.m.sg	gentleman	in	slippers	and

pijama [PVNI-0546]
pajamas

‘He would become a gentleman in slippers and pajamas’

In (10), both examples refer to a change related to womanhood, but they profile different aspects of it. In (10a), hacerse mujer relates to the biological evolutive change from girl to woman. In (10b), hacerse una mujer does not refer to the biological change, but to the acquisition of all those traits that characteristically differentiate a female adult from a younger one. Similarly, in (11a), volverse señores refers to a change in behavior (represented by the social class), whereas in (11b), volverse un señor focuses on those properties typical of a certain social class, the bourgeoisie.

Our proposal is that the use of the PCOS subconstruction with a NP+determiner triggers a different reading from the bare NP. In the case of bare NPs, the referential meaning of the noun corresponds to a generic interpretation where the change does not refer to any specific characteristic of the noun but to the general class. For the NPs with a determiner, on the contrary, the meaning of the noun refers to the salient prototypical characteristics that may help in differing one class from another.

Examples (12–14) add further evidence to this proposal.

(12)

Los	labios	de	Mariana	se	volvieron	un
the	lips	of	mariana	cl.3pl	turned.3pl	indf.det.m.sg

hilo [PVNI-0266]
thread

‘Mariana’s lips became very slender and thin’

(13)

Sacks […]	ahora	con	este	libro	se	vuelve	un	Darwin
sacks […]	now	with	this	book	cl.3sg	turns	indf.det.m.sg	darwin

de	sus	tiempos [PVNI-0155]
of	his	times

‘Now, with this book, Sacks […] becomes the new Darwin’

(14)

Se	volvían	unos	Judas	sin	árbol	donde
cl.3pl	turned.3pl	indf.det.m.pl	judas	without	tree	where

colgarse [PVNI-0310]
hang.cl

‘They became like Judas but with no tree to hang from’

In (12), Mariana’s lips did not turn into a thread but, according to the speaker’s own perspective, they became slender and thin. In (13), Sacks becomes like Darwin, that is, an influential person (scientist) thanks to a revolutionary book (theory). In (14), these people become distrustful like Judas (only worse, since they did not have a chance to redeem themselves).

In short, in the NP with determiner PCOS subconstruction, the noun is metonymically interpreted thanks to the indefinite determiner. ^[4] The noun stands for its most salient property. This subconstruction is quite flexible since it works with virtually any noun. There is only one important condition: the shared encyclopedic knowledge, i. e. the speakers’ conventionalized knowledge (see, Valenzuela 2017). The second part of the title in this paper is an illustrative example. “[W]ithout turning into a Barbie” refers to the salient characteristics of this famous doll. However, in order to correctly interpret which characteristics are salient, it is necessary to know that Barbie represents a thin and voluptuous blond woman as well as a stereotypical superficial person who cares about her physical aspect more than anything else.

In sum, the use of the indefinite determiner triggers the most salient and prototypical meaning. However, which part of the meaning gets taken as prominent depends on the encyclopedic knowledge that speakers have in common.