ArticlePDF Available

Abstract and Figures

The western variety of Peninsular Spanish possesses a type of causative construction in which an intransitive lexeme is used transitively. This phenomenon, called lability, is attested in three specific verbs: caer (‘to fall’), quedar (‘to stay’) and entrar (‘to enter’). As a consequence, they can induce a direct object at the expense of the standard forms tirar (‘to throw’), dejar (‘to leave’) and meter (‘to put in’). Lability has not been studied in depth for Spanish and, with this paper, I attempt to pinpoint its current extension as well as the possible semantic factors that prompt the transitivisation of these verbs.
Content may be subject to copyright.
ıctor Lara Bermejo
Abstract. The western variety of Peninsular Spanish possesses a type of causative
construction in which an intransitive lexeme is used transitively. This phe-
nomenon, called lability, is attested in three specific verbs: caer (‘to fall’), quedar
(‘to stay’) and entrar (‘to enter’). As a consequence, they can induce a direct
object at the expense of the standard forms tirar (‘to throw’), dejar (‘to leave’) and
meter (‘to put in’). Lability has not been studied in depth for Spanish and, with
this paper, I attempt to pinpoint its current extension as well as the possible
semantic factors that prompt the transitivisation of these verbs.
1. Introduction
This article aims to provide an initial overview of a grammatical
variation attested in Peninsular Spanish within a series of causative
constructions, in which certain unnacusative verbs are employed tran-
sitively. Standard Spanish can express causation through different
strategies, being the main ones a periphrastic construction formed by
the verb hacer (‘to make’) plus an infinitive (1); and through lexical pairs,
one of which expresses causation and the other effect (2). Likewise,
Spanish relies on a construction formed by the reflexive pronoun (se) plus
a tensed verb and an experiencer dative. This strategy is mainly employed
to express a nuance of lack of intention on the part of the author of the
action (3).
(1) Le hice ver la pel
ıcula, aunque no
3SG.DAT make.PST.1SG see.INF the film although NEG
‘I made him watch the film, although he did not want to.’
(2) a) Tir
e el vaso El vaso (se) cay
throw.PST.1SG the glass the glass (REFL) fall.PST.3SG
‘I threw the glass.’ ‘The glass fell.’
b) Dej
e los libros en la mesa Los libros (se)
leave.PST.1SG the books on the table the books (REFL)
quedaron en la mesa.
stay.PST.3PL on the table
‘I left the books on the table.’ ‘The books stayed on the table.’
Studia Linguistica 73(2) 2019, pp. 203–247. ©2018 The Editorial Board of Studia Linguistica
Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK, and 350 Main Street, Malden,
MA 02148, USA
c) Met
ı el coche en el garaje El coche
put in.PST.1SG the car in the garage the car
o en el garaje.
enter.PST.3SG in the garage
‘I put the car into the garage.’ ‘The car entered the garage.’
(3) Se me cay
o el vaso mientras caminaba.
REFL DAT.1SG fall.PST.3SG the glass while walk.PST.1SG
lit. ‘To me the glass fell while I was walking.’
However, some vernacular varieties of western Peninsular Spanish can
modify the lexical strategy by using the intransitive lexeme to also express
causation (4):
(4) a) Ca
ı el vaso.
fall.PST.1SG the glass
lit. ‘I fell the glass.’
b) (Me) qued
e los libros en la mesa.
(REFL.1SG) stay.PST.1SG the books on the table
lit. ‘I stayed the books on the table.’
c) Entr
e el coche en el garaje.
enter.PST.1SG the car in the garage
lit. ‘I entered the car into the garage.’
A verb’s ability to act both transitively and intransitively is called
lability and, as a consequence, this type of verb is called labile (I will
focus on this phenomenon in 5.1). In this article, I will try to show the
current distribution of this phenomenon as well as its semantic causes.
But before going any further and for the sake of argument, I will make
clear what I understand by transitivity, transitivisation, causation, causativi-
sation and valency, since these terms will appearall throughout. First of all, it
must be clear that causation and transitivity are not synonymous, so the use
of these terms is not interchangeable. Transitivity, on the one hand, is the
verb’s ability to have a subject and a direct object. Hence, transitivisation is
the conversion of a non-transitive verb into a transitive one. On the other
hand, a causative verb is one which expresses the ability of a subject to
provoke a change of state on a patient: causation. While all causative verbs
are by definition transitive, not all transitive verbs are causative. Therefore,
causativisation is the conversion of a non-causative verbinto a causative one.
Lastly, valency is used as a synonym of argument, that is, the obligatory
complement(s) of a verb. Therefore, a transitive verb has two valencies, or
arguments, whereas an intransitive verb possesses only one.
2. State of the art
The transitive use of these intransitive verbs in Spanish has not been
researched in depth, since the literature has not gone beyond the simple
204 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
mapping of the geographical extension of this phenomenon. Specifically,
Zamora Vicente (1970), Alvar (1996), Montero (2006), Ariza (2008) and
Real Academia Espa~
nola (RAE) (2009) state that this situation is
attested in the provinces of Burgos, Le
Avila, Zamora, Salamanca,
Valladolid, C
aceres, Badajoz, Toledo and Ciudad Real (map 1).
In addition, Jim
enez Fern
andez and Tubino Blanco (2014) assure that
this phenomenon is attested in Andalusia (the most southern region in
Spain). However, due to a number of reasons their statements should and
will not be taken into account: firstly, their analysis is not applied to
recorded data or any fieldwork results, and the origin of the sentences
they use as examples is unknown; secondly, they believe that the
Andalusian dialect is uniform throughout Andalusia. Alvar (1996),
Frago (1993), Narbona (2003), Mond
ejar (1991), Penny (2004) and
nez (2011) have demonstrated that, linguistically speak-
ing, Andalusia can be perfectly divided into at least two varieties: eastern
and western. Both areas are characterised by different phonetic,
morphologic and syntactic particularities. Only central areas may share
several patterns of both sides. This raises again the question about the
origin and validity of the instances Jim
enez Fern
andez and Tubino
Blanco (2014) provide in their article. However, I will return to some
interesting points of their study later in 5.4.
Despite the efforts made to delimit the geographical extension of this
phenomenon, the bibliography mentioned above has failed to provide a
linguistic explanation or an in-depth investigation of the actual use of
these verbs as transitive. The only reference to a possible linguistic reason
can be found in Montero (2006). Nevertheless, this author hardly devotes
Map 1. Coverage of the phenomenon in the literature
Lability in Western Peninsular Spanish 205
©2018 The Editorial Board of Studia Linguistica
two lines to suggesting that lability emerges in unintentional actions
although this may be only valid for the case of caer (‘to fall’). According
to her, and always as a conjecture rather than based on exhaustive
analysis, the semantic difference between (5) and (6) resides in the fact
that the latter, unlike the former, emerges to express a voluntary action
caused by the agent.
(5) Ca
ı el vaso.
fall.PST.1SG the glass
‘I dropped the glass.’
(6) Tir
e el vaso.
throw.PST.1SG the glass
‘I threw the glass.’
However, Spanish already possesses quite a common strategy to express
lack of volition, as (3) indicates. Montero’s hypothesis does not make
clear the differences between (3) and (5) as well as the possible factors
that play a role in the transitivisation of the verbs entrar (‘to enter’) and
quedar (‘to stay’).
Moreover, if a more precise search is carried out throughout the
diverse linguistic atlases and dialect corpora (ALCyL, ALEANR, Alec-
Man and COSER) which can be counted on nowadays, it is possible to
verify that the lability phenomenon is witnessed in the same places
described by the abovementioned literature. As a matter of fact, some
authors even suggest that the use of the labile is more widespread for
certain verbs than for others. Specifically, Alvar (1996), Ariza (2008) and
ıa Mouton (1994) affirm that entrar (‘to enter’) is the most extended
causativised verb, whereas caer (‘to fall’) and quedar (‘to stay’) are found
in fewer areas. Also, it seems that caer (‘to fall’) is slightly more reduced
in its extension than quedar (‘to stay’).
The labile phenomenon is not restricted to Spanish in the Iberian
Peninsula. In fact, the Catalan and Portuguese linguistic domains
exhibit the same phenomenon, but in principle with the lexical pair
morir (‘to die’) matar (‘to kill’) (Said 1931, Nunes 1945, Wheeler
et al. 1999). Although many languages show the pattern of lability (I
will refer to specific studies in 5.1), most of them detransitivise certain
verbs to also employ them as intransitive. In other words, most
languages, when applying lability, usually turn transitive verbs into
intransitive. The Spanish case exhibits the opposite effect: an intran-
sitive verb is used transitively even though there exists a corresponding
transitive lexeme to express the same cause. In order to understand the
semantic factors that are related to this phenomenon and its current
geographical validity, I have carried out fieldwork, whose methodology
I will describe below.
206 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
3. Methodology
With the aim to collect current data about the causative phenomenon in
order to pinpoint its semantic factors as well as its current geographical
diffusion, I have carried out fieldwork (map 2 and table 1) throughout
the areas referred to by the literature (map 1).
Map 2. Fieldwork
Table 1. Localities of the survey
Province Localities
on Le
on, Bembibre
Zamora Toro
Valladolid Olmedo
Segovia Riaza
Salamanca Lumbrales, Pe~
naranda de Bracamonte
Palencia Villmuriel de Cerrato
Avila Piedrah
Burgos Aranda de Duero, Briviesca
Toledo Consuegra, Talavera de la Reina
Ciudad Real Pozuelo de Calatrava
aceres Guadalupe, Valencia de Alc
antara, Plasencia
Badajoz Badajoz, Valle de Santa Ana, Bienvenida, Villanueva de la Serena
Lability in Western Peninsular Spanish 207
©2018 The Editorial Board of Studia Linguistica
As it deals with a non-standard phenomenon, I have surveyed speakers
with a lower level of formal education, since they represent a social
profile more inclined to maintain vernacular or non-standard phenom-
ena. In table 2, I will detail the number of informants and the
occurrences I have obtained.
The method I employed for eliciting has always tried not to prime the
informants. Due to the difficulty in recording the emergence of this
phenomenon during a semi-conducted interview, through the sociolin-
guistic interview or by means of indirect questions that lack spontaneity
(Gilquin 2010), I filmed a series of scenes in which a person carried out
certain activities that implied the use of the verbs under study. The
speakers were asked to describe spontaneously the scenes they had to
watch. Several scenes were devised for each lexical pair, taking into
account different types of patients and agents. In table 3, I detail the
number of scenes for each verb and the semantic nuances foreseen.
As can be observed, not all the verbs induce the same number of scenes
for each semantic nuance. The reason is the lack of linguistic information
I have found in the existing literature. However, because Montero (2006)
suggests a possible divergence between lability and the usage of the
standard transitive verb, based on the degree of control on the part of the
subject, I divided the scenes in three different semantic nuances anchoring
them to the subject: wilful human subject, unwilful human subject and
non-human subject. The likelihood of non-human subjects to emerge as
authors of the verbs under study is considerably low, so I prioritised the
scenes in which a human subject performed a given action. In the case of
entrar (‘to enter’), the creation of scenes in which a human subject put
something into a container non-intentionally was difficult to apply and
also to elicit. Consequently, this verb was provided with more scenes in
Table 2. Data of the survey
Informants Occurrences
200 1.938
Table 3. Number of scenes of the fieldwork
Entrar (‘to enter’)
(‘to stay’) Caer (‘to fall’)
Wilful human subject 6 out of 10 4 out of 10 4 out of 10
Unwilful human subject 2 out of 10 4 out of 10 4 out of 10
Non-human subject 2 out of 10 2 out of 10 2 out of 10
208 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
which a wilful human subject carried out a given activity that implied the
use of this verb.
As a way of clarifying the dynamics of the survey, I will transcript a
sample of the elicitation of two of the scenes.
INTERVIEWER: Please, tell me what happens in this scene, what is the girl
you are watching doing?
[We can see a girl in an office. She walks to an armchair which has a
bottle on one of its arms; she sits down and unintentionally knocks the
bottle off. The girl realises this and bends down to pick up the bottle in
order to place it on the table she has in front of her.]
INTERVIEWED: This girl is walking..., she sits down... and she FALLS
the bottle. Now she is taking it. She puts it on the table and, she lies back
in the armchair.
INTERVIEWER: Please, tell me what happens in this scene, what is the girl
you are watching doing?
[We can see the same girl who is working. She takes a look at her watch
and decides to leave. She takes all her belongings but forgets to take her
sunglasses. When she is heading down the stairs, she touches her head, as
though she were looking for something and she realises she has left her
sunglasses in the office.]
INTERVIEWED: Here is the same girl..., it seems she is working...., now
she is looking at her watch, she stands up, she is leaving, she’s fed up with
working... she takes her things... ah, she STAYS her glasses on the
table..., you see? Now she noticed she HAS STAYED the glasses on the
To summarise, this modus operandi has enabled the qualitative and
quantitative production of instances of lability, and in the following
epigraphs I will present and analyse the results.
4. Results
4.1. Geographical distribution
In the first place, it is necessary to highlight that the results obtained in
the fieldwork have shown that the phenomenon under study is attested in
all the ages and genders surveyed, but it would be quite useful to extend
the investigation to more educated informants, in order to verify the
social extension of this non-standard phenomenon. Below, I show the
spatial distribution that the three lexical pairs occupy nowadays
(Nerbonne et al. 2010).
When maps 3, 4 and 5 are compared to map 1, a dramatic geograph-
ical decrease in the use of lability is easily perceived. On the one hand, I
have not found any occurrences in Burgos, southern Valladolid, Segovia
and Le
on. Likewise, the Toledo area that borders
Avila has not provided
Lability in Western Peninsular Spanish 209
©2018 The Editorial Board of Studia Linguistica
examples of this phenomenon either and it is in eastern Extremadura
where the non-standard linguistic behaviour starts to fade. Furthermore,
its incidence in and around the zone of Zamora has been less than
throughout Salamanca, C
aceres or Badajoz: provinces that have given
the highest number of instances. Therefore, the findings recorded in the
bibliography have to be updated, by taking into account the drastic
geographical limitation to more western areas; specifically, from southern
Map 3. Current extension of caer (‘to fall’) as labile
Map 4. Current extension of quedar (‘to stay’) as labile
210 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
Zamora up to western Andalusia. Below, I will detail in table 4 the
number of informants who have produced the labile phenomenon and
those who have not.
Table 4 shows that 80 speakers (out of 200) produced at least once the
verb entrar (‘to enter’) as labile, 62 resorted at least once to quedar (‘to
stay’) as labile and 38 did so with caer (‘to fall’). Clearly, table 4 suggests
that the most widespread labile usage corresponds to entrar (‘to enter’),
followed by quedar (‘to stay’) and caer (‘to fall’) in the last place.
Despite the fact that the extension of the verbs entrar (‘to enter’) and
quedar (‘to stay’) begins to be displaced by the standard pattern, it is this
first verb (entrar, ‘to enter’) that spreads the most eastwards. The choice
of quedar (‘stay’) instead of dejar (‘to leave’) is the most widespread in
northern areas, while caer (‘to fall’) instead of tirar (‘to throw’) is the
most uninterruptedly widespread along the Roman Silver Route, from
southern Zamora up to western Andalusia. The diffusion pattern
responds to the wave model (Wolfram & SchillingEstes 2003), since
Map 5. Current extension of entrar (‘to enter’) as labile
Table 4. Number of informants and their linguistic behaviour
Entrar (‘to enter’) Quedar (‘to stay’) Caer (‘to fall’)
Labile 80 (40%) 62 (31%) 38 (19%)
Non labile 120 (60%) 138 (69%) 162 (81%)
Total 200 (100%) 200 (100%) 200 (100%)
Lability in Western Peninsular Spanish 211
©2018 The Editorial Board of Studia Linguistica
the urban points have not been more standard than the rural ones; in
other words, the demography of a municipality does not intervene in the
spatial diffusion of the lability. This can also be argued because of the
relic areas produced by the spread of the standard model, as its
distribution is purely geographical. Hence, this phenomenon has
decreased and the standard pattern has started to become more common.
The overall geographical extension of the lability strategy occurs along
the Silver Route founded by the Romans and which has always been a
main road link in Spain until the present (map 6).
As a matter of fact, the region where this phenomenon is attested was
repopulated by Leonese settlers and western Castilian colonials who were
given lands and power in exchange for their settlement across the recently
conquered territories (Cano 2004, Men
endez Pidal 2005). Their linguistic
features were exported and nowadays this whole area shares certain
dialect particularities that can only be explained due to that repopulation
and the communication and contact they have always maintained.
Specifically, western Peninsular Spanish is mainly characterised by the
confusion of clitic personal pronouns. While standard Spanish relies on
case to choose the clitic, western Peninsular Spanish usually selects the
clitic on the basis of animacy, definiteness or the opposition between
countable and uncountable (Fern
nez 1999) (I will refer to
this phenomenon later in 5.4). Also, the morphology of certain clitic
pronouns, like vos instead of the Peninsular standard os, is shared (Lara
2012), as well as the pronunciation of velars, the velarisation of /n/ (Cano
2004, Men
endez Pidal 2005) or the use of present tenses to express future
Map 6. Silver Route
212 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(Lara 2016). Fern
nez (2011) has also proved that the
development of Peninsular Spanish depends on the generalisation of
different patterns that spread along two different paths: from east to west
and vice versa (mainly in morpho-syntax), and from north to south and,
to a lesser extent, vice versa (mainly in phonetics). So, western Peninsular
Spanish, according to Fern
nez, represents a dialect block
that shares specific dialect particularities which every now and then have
also generalised throughout the standard.
In the next subsection, I will describe the semantic behaviour of the
results I have obtained in the fieldwork.
4.2. Semantic distribution
In table 4 I provided the number of speakers who produced lability in the
different verbs under study. However, if a more exhaustive analysis is to
be applied, the prompt of the labile verb is restricted by various semantic
nuances. These nuances have caused the informants to express lability in
some cases, to resort to the standard transitive lexeme in other cases or to
be consistent with the labile behaviour. Let us now focus on this by first
observing the cartographic distribution of lability depending on diverse
semantic factors (maps 712).
Map 7 shows that the entire extension of the verb quedar (‘to stay’)
coincides at least with the trace of an un-wilful agent (78).
(7) Se ha quedado las gafas en la mesa.
REFL have.PRES.3SG stay.PCP the glasses on the table
‘She has left the glasses on the table.’
(8) Ha quedado la luz encendida.
have.PRES.3SG stay.PCP the light switch.PCP
‘She has left the light on.’
In addition, map 8 shows that the entire zone where the verb caer (‘to
fall’) as transitive is attested reflects a non-volitional reading (910).
(9) La chica ha ca
ıdo el vaso sin querer.
the girl have.PRES.3SG fall.PCP the glass without want.INF
‘The girl has dropped the glass not on purpose.’
(10) Ha ca
ıdo la botella al sentarse.
have.PRES.3SG fall.PCP the bottle to+the sit.INF+REFL
‘She has dropped the bottle when she sat’
Lability in Western Peninsular Spanish 213
©2018 The Editorial Board of Studia Linguistica
Moreover, map 9 demonstrates that the use of caer (‘to fall’) over tirar
(‘to throw’) presents a slightly smaller diffusion, specifically in sentences
with an inanimate subject (11).
(11) El viento ha ca
ıdo la calabaza.
the wind have.PRES.3SG fall.PCP the pumpkin
‘The wind has thrown the pumpkin.’
Map 7. Quedar (‘to stay’) unintentionally
Map 8. Caer (‘to fall’) unintentionally
214 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
Map 10 draws attention to the fact that only two areas within the zone
of quedar (‘to stay’) instead of dejar (‘to leave’) are characterised by also
having a volitional reading (12).
(12) La chica ha decidido quedar los libros en la
the girl have.PRES.3SG decide.PCP stay.INF the books on the
‘The girl has decided to leave the books on the table.’
Map 9. Caer (‘to fall’) non-human subject
Map 10. Quedar (‘to stay’) intentionally
Lability in Western Peninsular Spanish 215
©2018 The Editorial Board of Studia Linguistica
Map 11 shows that, in the case of caer (‘to fall’), the intentional action is
only attested throughout an area which is located inside the area where
its transitive use is unintentional. However, the volitional reading has
caused the phenomenon to decrease dramatically (1314).
Map 11. Caer (‘to fall’) willingly
Map 12. Entrar (‘to enter’) telic readings
216 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(13) La chica ha cogido el vaso y lo
the girl have.PRES.3SG take.PCP the glass and ACC.3SG
ha ca
have.PRES.3SG fall.PCP
‘The girl has taken the glass and she has thrown it.’
(14) ¿C
omo que qu
e ha hecho? Caerla
how that what have.PRES.3Gdo.PCP fall.INF+ACC.3SG
[la botella].
[the bottle]
‘What has she done? Throw it.’
In the case of entrar (‘to enter’), the nuances referred to animacy or
intentionality have not revealed anything of significance. Indeed, it is the
fact of having a telic sentence that makes the phenomenon decrease, as it
is only located in the most south-western area (15).
(15) Ha entrado el folio en el sobre.
have.PRES.3SG enter.PCP the paper in the envelope
‘She has put the sheet of paper into the envelope.’
If maps 3 to 5 are compared to maps 7 to 12, it is possible to notice that
the lability decreases or increases depending on the semantic reading. As
a consequence, unwilling actions are for instance more inclined to
prompt lability than volitional ones. This means that there are
informants who can produce both lability and the standard pattern at
the same time, and that this production is closely related to the nuances I
have been referring to. In fact, table 5 features this behaviour even more
plainly. This table only includes the behaviour of the speakers who, at
least once, produced lability in either of the verbs under study; in other
words, speakers who never produced lability are excluded from the
following table.
Table 5 demonstrates that the informants that use lability do not do
so in every case. The totality of them uses it with non-volitional
subjects (in the cases of quedar ‘to stay’ and caer ‘to fall’) and atelic
sentences (in the case of entrar ‘to enter’). But these same informants
have chosen either lability or the normative transitive lexeme (meter ‘to
put in’, tirar ‘to throw’ and dejar ‘to leave’) in the rest of semantic
nuances (1620).
(16) La chica coge la botella y la tira
the girl take.PRES.3SG the bottle and ACC.3SG throw.PRES.3SG
a la papelera.
to the bin
‘The girl takes the bottle and throws it into the bin.’
Lability in Western Peninsular Spanish 217
©2018 The Editorial Board of Studia Linguistica
Table 5. Number of informants and lability, based on the relevant semantic nuances
Caer (‘to fall’) Tirar (‘to throw’) Quedar (‘to stay’) Dejar (‘to leave’) Entrar (‘to enter’)
(‘to put in’)
Unintentional 38 (100%) 0 (0%) 62 (100%) 0 (0%) Atelic 80 (100%) 0 (0%)
29 (76,3%) 11 (13,7%) 51 (82,25%) 11 (17,75%) Telic 17 (21,25%) 63 (78,75%)
Wilful 8 (21%) 30 (79%) 13 (20,9%) 49 (79,1%)
218 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(17) La chica coge el vaso y lo tira
the girl take.PRES.3SG the glass and ACC.3SG fall.PRES.3SG
al suelo.
to+the floor
‘The girl takes the glass and throws it onto the floor.’
(18) Al final deja los libros en la mesa y
to+the end leave.PRES.3SG the books on the table and
se va.
‘She eventually leaves the books on the table and goes away.’
(19) Mete el papel en el sobre.
put in.PRES.3SG the paper in the envelope
‘She puts the sheet of paper into the envelope.’
(20) Mete los folios en el archivador.
put in.PRES.3SG the papers in the folder
‘She puts the papers into the folder.’
If these uses are analysed, it seems clear that, in the case of tirar (‘to
throw’), the subject has carried out the action willingly (1617). It is the
same behaviour that informants have when they choose dejar (‘to leave’)
(18), whereas meter (‘to put in’) arises in telic contexts (1920).
Lastly, when all the maps are compared, different linguistic continua
can be inferred. In accordance with the results, the informants that
transitivise caer (‘to fall’) should also do it with quedar (‘to stay’) and
entrar (‘to enter’) (i).
i. Entrar > quedar > caer
The hierarchy in (i) can be explained as follows: if a speaker causativises
the verb quedar (‘to stay’), he or she may causativise entrar (‘to enter’),
but not yet caer (‘to fall’). The transitivisation of a verb in the hierarchy
implies the transitivisation of those on its left, but not on its right. This
can be observed in map 13.
Likewise, as the lability strategy is subjected to semantic factors, these
can be ordered as follows (iiiii):
ii. un-wilful human subject > non-human subject > wilful human subject
iii. atelic > telic
The hierarchies reproduced in (ii) and (iii) indicate that the informants that
use any of the three verbs as labile in sentences with a non-human subject,
also do so in sentences in which the subject is human and un-wilful. In
Lability in Western Peninsular Spanish 219
©2018 The Editorial Board of Studia Linguistica
addition, if the informants resort to lability in atelic readings, they may
also do so in telic sentences (15). The hierarchies are implicational and the
labile strategy always moves rightwards along the hierarchy.
In the next section, I will focus on the linguistic analysis of the results I
have presented in this section.
5. Analysis
Below I will analyse the results of my fieldwork from a typological
perspective. Firstly, I will address the emergence and development of lability
and what factors bring causativisation and transitivisation together; then, I
will discuss the tendency of Spanish to split intransitivity and, finally, I will
argue that western Peninsular Spanish is characterised by split ergativity.
5.1. Causative strategies: lability
The phenomenon of lability in western Peninsular Spanish is not an
isolated case. In principle, cross-linguistically attested causative con-
structions can be expressed through three different means: analytical
causatives (such as the use of the verb ‘to make’ plus an infinitive);
morphological causatives, in which an intransitive verb adopts certain
affixes that denote causation (in Turkish,
ol, ‘to die’ can acquire an affix
and become
ur, ‘to kill’); and lexical pairs, with a transitive lexeme
to express causation and an intransitive lexeme to express effect
(Comrie 1981). However, the lexical strategy is sometimes expressed by
Map 13. Continuum in vernacular causativisation
220 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
the use of the same lexeme that can be resorted to in intransitive and
transitive sentences. This is called lability (Haspelmath 1993, Letuchiy
2004 and 2009) and it is relatively uncommon. Concretely, the term
began to appear in grammars of Caucasian languages (Klimov &
Alekseev 1980,
Sejxov 1987) and a bit more frequently in some
academic contributions published mainly in the 1990’s (see Nichols
1984, Dixon 1994, Song 1996, Kulikov 1999a). As a consequence, the
labile resource has drawn less attention in the literature than the rest of
causative constructions (the majority of the literature has focused on
the alternation between inchoative / resultative, as in English I broke
the vase / the vase broke (Aikhenvald & Dixon 2000, Shibatani 1976)),
but there are some studies that can help us understand the situation in
western Peninsular Spanish.
First of all, it is necessary to know in which type of lability Spanish fits.
Letuchiy (2009) establishes a typology of the different kinds of labile
verbs and he distinguishes the following ones: anticausative, reflexive,
reciprocal, passive and converse. Based on this typology, the verbs in
(4ac) seem to fit anticausative labile verbs, whose lability is patient-
preserving. The different behaviour regarding the semantics of the subject
in lability is addressed by Letuchiy and Creissels (2014), who classify
labile verbs in agent-preserving (A-alignment) (21) and patient-preser-
ving (P-alignment) (22).
(21) John drinks (tea).
(22) I broke the vase. / The vase broke.
While (21) possesses the same subject (always the agent), one of the
alternatives of (22) exhibits a patient subject. The Spanish cases are
clearly correlated to the P-alignment model, since the intransitive verb
arises to connote the lack of intention or lack of agentivity of the subject.
Actually, the semantics of the verb as well as of its arguments is crucial
for the emergence, development or later decay of lability. It is not
surprising that Spanish presents this phenomenon, as many other
languages (related or not between each other) have a higher or lower
degree of lability. Virtually all the articles that deal with it point out that
verbs of movement or change of state are more inclined to be labilised
(see Karantzola & Lavidas 2014, Jim
enez Fern
andez & Tubino Blanco
2014, Letuchiy 2004, Letuchiy 2015).
Furthermore, as for the likelihood of labilising a verb, the semantic
features of volition and agentivity are usually crucial (Van Valin 1990,
Larjavaara 2000). If the examples of lability in western Spanish are
analysed, it is possible to discover that all the informants that produce it
start doing so with non-intentional subjects or in subjects which lack
agentivity (2326). Once the lability is established in these semantic
readings, it spreads to agentive and wilful subjects (2728).
Lability in Western Peninsular Spanish 221
©2018 The Editorial Board of Studia Linguistica
(23) Se ha quedado los libros en la mesa.
TEFL have.PRES.3SG stay.PCP the books on the table
‘She has left the books on the table.’
(24) Ha quedado la l
ampara encendida.
have.PRES.3SG stay.PCP the lamp switch.PCP
‘She has left the lamp on.’
(25) La chica se ha tropezado y ha
the girl REFL have.PRES.3SG slip.PCP and have.PRES.3SG
ıdo el vaso.
fall.PCP the glass
‘The girl has slipped and dropped the glass.’
(26) El viento ha ca
ıdo el papel.
the wind have.PRES.3SG fall.PCP the papers
‘The wind has thrown the sheet of paper.’
(27) La chica quiere quedar los libros en la mesa.
the girl want.PRES.3SG stay.INF the books on the table
‘The girl wants to leave the books on the table.’
(28) La chica ha cogido el vaso y lo
the girl have.PRES.3SG take.PCP the glass and ACC.3SG
ha ca
ıdo a la papelera.
have.PRES.3SG fall.PCP to the bin
‘The girl has taken the glass and she has thrown it into the bin.’
It is perhaps because of these characteristics that verbs of achievement
are the likeliest to develop lability, as Van Valin (1990) puts forwards.
As a matter of fact, the three verbs affected in Spanish are verbs of
achievement and the intransitive verbs that have been transitivised
both in standard (sortir ‘to go out’, apprendre ‘to learn’) (2932)
and dialect French (tomber ‘to fall’) belong to this classification too
(29) Je sors de l’
a 17h.
1SG go out.1SG.PRS.IND of the school at 17h
‘I go out of school at 5pm.’
(30) Je sors la poubelle tous les jours.
1SG go out.1SG.PRS.IND the rubbish all the days
‘I throw the rubbish away everyday.’
(31) J’ apprends
a parler l’ anglais.
1SG learn.1SG.PRS.IND to speak.INF the English
‘I learn to speak English.’
222 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(32) Le professeur m’ apprend
a parler l’
the teacher ACC.1SG learn.3SG.PRS.IND to speak.INF the
‘The teacher teaches me to speak English.’
(33) Je tombe sur le sol.
1SG fall.1SG.PRS.IND over the floor
‘I fall onto the floor.’
(34) Je tombe la chemise sur le sol.
1SG fall.1SG.PRS.IND the shirt over the floor
‘I throw the shirt onto the floor.’
These instances show that the three verbs may develop a direct object
although their original meaning was intransitive. While (2933) are
perfectly standard, (34) is not accepted by the norm and belongs to
vernacular French. It must be noted that all of them are unaccusative and
verbs of achievement; in addition, tomber (‘to fall’) coincides with the
Spanish phenomenon. As for tomber (‘to fall’) in French, Larjavaara
(2000) and Bilous (2011) draw attention to a fixed collocation in which
the acceptability of tomber as transitive is full among speakers and is
completely included in the normative variety: tomber un adversaire (‘to
beat an opponent’). Also, other unacussative verbs such as bouillir (‘to
boil’) are gaining ground in non-standard French at the expense of the
analytical strategy faire (‘to make’) plus infinitive (faire bouillir).
Consequently, according to Bilous (2011), there must be some intra-
linguistic and cross-linguistic tendency toward transitivisation that is
based on certain parameters (I will refer to this in 5.2).
In spite of the importance of the semantics for lability, syntax has been
also found to be relevant in this respect. Kulikov (1999a, 1999b and 2003)
and Letuchiy (2004) have remarked that lability is favoured by tense,
mood or even phonetic proximity of closely-related verbs (as in many
labile verbs formed in Estonian: Kehayov & Vihman 2014). In fact, the
behaviour of entrar (‘to enter’) as labile in current western Peninsular
Spanish follows the opposition perfectiveimperfective rather than the
semantics of the subject or the object. Karantzola and Lavidas (2014)
have even found that certain constructions with a post-verbal neuter
noun phrase have triggered lability in Greek.
Studies carried out have certified that labilisation is often produced
with the conversion of a transitive verb into an intransitive too. The
better examples can be found in English (Visser 1970, McMillion 2006) or
even in Latin (Gianollo 2014). However, the Spanish lexemes affected by
this process are intransitive verbs that have been converted into
transitive. This pathway from intransitive to transitive is hardly
Lability in Western Peninsular Spanish 223
©2018 The Editorial Board of Studia Linguistica
documented and it is rarely attested cross-linguistically. When Letuchiy
(2015) researched the development of labilisation in modern Russian, he
acknowledges that all verbs that have undergone this process are
transitive verbs that ended up being also used as intransitive. He only
finds one verb (kapat / kapnut: ‘drop / fall / throw’) which has evolved
from intransitive to transitive, and its usage as labile is restricted to
certain social strata and is not widespread. The scarcity of instances of
the conversion into transitive raises the question of which are the
requirements for transitivisation. A look at the literature indicates that
causativisation and transitivisation are closely related.
5.2. Causativisation and transitivisation
As has been repeatedly argued, semantics is crucial for the syntactic
construction of causation and this is supported by many typologists. For
example, Comrie (1981), Comrie and Polinsky (1993), Shibatani (1976)
and Aikhenvald and Dixon (2000) propose a causativisation continuum
which suggests that the likelihood of a verb becoming causativised
depends on the following features: 1) animacy of the subject, 2) volition
of the agent, 3) control of the subject, 4) prominence and degree of causer
over the causee. Giv
on (1976 and 2001) agrees on these parameters and
he adds the distinction between direct and indirect causation. According
to him, the more direct causation is, the likelier a lexical strategy will be
used; on the contrary, the less direct the causation is, the likelier an
analytical strategy will be used. These semantic differences favour a type
of case-marking in many languages. For instance, Hungarian can have a
patient marked in accusative or in instrumental, depending on the
control of the causer on the causee. So, Hungarian is said to follow a
continuum [instrumental >accusative] to express the degree of control of
the causer over the causee. Accusatives express more control while
instrumentals are usually interpreted as an indication of lack of volition
and control (see 3536, taken from Hetzron 1976).
(35) Az
o minden nap egy
at s
altata }
the nurse every day one hour.ACC made:walk he.ACC
‘The nurse walked him for an hour every day.’
(36) Az orvos minden nap egy
at s
altatott vele.
the doctor every day one hour.ACC made:walk he.INSTR
‘The doctor had him walk for an hour every day.’ [as a prescription]
Moreover, this author makes a clear distinction between the agent and
the author of causation: while authorship does not imply intention,
agentivity does. Ramchand (2011) has found the same pattern in Hindi
and Urdu; and, Daniel et al. (2012) state that Agul, a language from the
Caucasus, selects a periphrastic causative construction to express less
224 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
control on the causee but it chooses a lexical strategy if there is more
control. Kim (2012) affirms that in Korean the causer is not the same as
the agent and the case-marking of a causative construction depends on
the volition of the subject and the affectedness of the object.
The most remarkable feature of the Spanish phenomenon is that the
process converts intransitive verbs into transitives. Hopper and
Thompson (1980) share a well-known article on the factors that favour
the emergence of a transitive verb. According to them, some of its
conditions are agentivity, aspect, volition, kinesis or affectedness. The
stronger these features are, the likelier a transitive verb will be chosen.
Based on the collected occurrences, it is visible that the fall of the
transitive lexeme is favoured in readings of scarce or null agentivity
(11), of lack of volition (710) or of little affectedness of the patient (7
8). The more intentional, animate (1618) or telic (1920), the more
likely it is that informants will tend to select the normative transitive
lexeme. The lexical pair meter entrar (‘to put in’ ‘to enter’) is the one
that best exemplifies the factor of telicity, since informants have
distinguished between (37) and (38).
(37) La chica mete la botella en el bolso.
the girl put in.PRES.3SG the bottle in the purse
‘The girl puts the bottle into the purse.’
(38) La chica abre la puerta y entra la
the girl open.PRES.3SG the door and enter.PRES.3SG the
silla en el despacho.
chair in the office
‘The girl opens the door and puts the chair into the office.’
While (37) shows a telic action (the informants produce it when the action is
completed), in (38) the chair is dragged into the room. This means that
while (37) happens in a single moment, (36) needs a longer period of time to
be completed, and this is why the informants express the development of
this action or its lack of conclusion through the intransitive verb. As the
informants were asked to describe simultaneously what they were
watching, the choice of meter (‘put in’) might have arisen if the sentence
had been construed in the past. This is underpinned by the examples
provided by the informants when they produced the tokens once the action
had been completed. In this sense, regardless of whether it was the chair or
the bottle, if the object was already inside its container, the informants were
more likely to produce the standard transitive meter (‘to put in’) in order to
express a perfective action (3941).
(39) La chica ha metido el papel en el sobre.
the girl have.3SG.PRS put in.PCP the paper in the envelope
‘The girl has put the paper in the envelope.’
Lability in Western Peninsular Spanish 225
©2018 The Editorial Board of Studia Linguistica
(40) Acaba de meter la silla en el despacho.
end.3SG.PRS of put in.INF the chair in the office
‘She has just put the chair into the office.’
(41) Te ha metido en el coche.
ACC.2SG have.3SG.PRS put in.PCP in the car
‘She has put you into the car.’
As stated by Hopper and Thompson (1980 and 1982), atelic sentences
belong to low transitive contexts, since the action carried out by the
subject may have not yet affected the patient. Again, a low transitivity
context has prompted the emergence of lability. In the case of entrar (‘to
enter’), the factor has been the opposition between perfectiveimperfec-
tive rather than other semantic features such as agency or volition.
The traditional concept of transitivity is ever more questioned. As Bilous
(2011) puts it, there are transitive constructions that do not follow the
traditional model at all. He exemplifies Hungarian, as this language may
take a direct object marked in accusative though it may be not very affected,
referential or individualised. As a result, this accusative in the direct object
is not a prototypical direct object of the prototypical transitivity. There are
enough examples, according to this author, that contradict the definition of
transitivity. In other words, transitivity cannot be necessarily the transfer
of an action on the part of an agent over a patient, because there are
transitive constructions where the direct object is hardly affected or the
subject is not truly an agent. Indeed, it is possible to have impersonal
transitivised expressions. Observe the occurrences (4246) in English,
French and Ukrainian provided by Bilous (2011).
(42) Il pleuvait des bombes.
3SG.MASC.NOM rain-3SG.PST PART bombs
‘It was raining bombs.’
(43) It is raining cats and dogs.
(44) There came three men.
(45) Il est venu trois hommes.
3SG.MASC be.3SG.PRS come.PCP.MASC.SG three men
‘Three men came.’
(46) Pryj
sl-o dvi div
come.3SG PST.NEUT two girls
‘Two girls came.’
In these examples, the subject is expletive and singular and turns the verb
into transitive, as it agrees with the expletive and not with the object,
which can be plural or singular without shifting the verb inflection.
226 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
For Bilous (2011) and Descl
es (1998), it is necessary to distinguish
between semantic and syntactic transitivity. The semantic transitivity
implies control on the part of the subject that brings the action about.
This action affects a patient, animate or not, which changes its position
or state. Nevertheless, the syntactic transitivity implies the participation
of two arguments, one of which is the agent, highly individualised and
receiving nominative case, normally conscious and controlling the event;
the other is a patient, which appears in the accusative case, normally non-
animate and affected by the action brought about by the verb.
Nevertheless, there is evidence that the syntactic transitivity not always
reflects a corresponding semantic transitivity. Observe (4748).
(47) He so~
nado que me ca
ıa por un
have.1SG.PRS dream.PCP that REFL.1SG fall.1SG.PST off a
‘I have dreamt that I was falling off a cliff.’
(48) He o
ıdo un ruido.
have.1SG.PRS hear.PCP a noise
‘I heard a noise.’
Instances (4748) show syntactic transitive sentences. However, their
semantic transitivity may be questionable. After all, the subject of so~
(‘to dream’) is not volitional and, rather than an agent, it is the
experiencer of dreaming. Likewise, there is no affected object or any
patient that undergoes any action, so the degree of transitivity of so~
(‘to dream’) is minimal. The same applies to o
ır (‘to hear’). Paying
attention to a noise is escuchar (‘to listen’), but never o
ır (‘to hear’), since
one may have the ability to perceive a sound without having decided to
do so or without any affected patient. Needless to say, the transitivity of
all these verbs is far from clear. In order to better comprehend how
transitivity may work, it is necessary to focus on the type of subject and
object it may develop. Below, I will attempt to demonstrate that both the
kind of subjects and/or the objects prompt or disfavour both transitivity
and lability.
5.2.1. Subjects
For Bilous (2011), instead of dealing with the dichotomy between agent
and patient, it is necessary to understand the verb’s arguments as internal
and external arguments. According to him, every verb has an internal
argument, rather than a patient, and not always an external argument (in
the place of agent). Without doubt, (4246) show the possibility of
transitivisation without the need of an external agent, despite the
insistence of the traditional definition of transitivity, which promotes the
Lability in Western Peninsular Spanish 227
©2018 The Editorial Board of Studia Linguistica
role of an agent that provokes an action. Bilous (2011) puts forward that
every verb is able to be transitive and, as a result, the universal grammar
has traits de transitivit
e(‘features of transitivity’). The emergence of these
features is subjected to the existence of an external argument or a causer.
Compare (4952).
(49) El chico es alto El chico lo es.
the boy be.3SG.PRS tall-MASC the boy CLITIC be.3SG.PRS
‘The boy is tall. The boy is.’
(50) La chica es alta La chica lo es.
the girl be.3SG.PRS tall-FEM the girl CLITIC be.3SG.PRS
‘The girl is tall. The girl is.’
(51) El atleta corre los 100 metros El atleta
the athlete run-3SG.PRS the 100 metres the athlete
los corre.
‘The athlete runs 100 metres. The athlete runs them.
(52) La empleada trabaja la madera La empleada
the employee work-3SG.PRS the wood The employee
la trabaja.
‘The employee works the wood. The employee works it.’
Examples (4952) show four sentences: the first two with a copulative
verb and the last two with an unergative intransitive verb. Notice that the
predicate of the copulative verb can be replaced by a clitic that coincides
with the direct object in Spanish (the same applies to French, Italian or
Portuguese). However, the choice of this clitic is not identical for
copulative and direct objects. On the one hand, the clitic replacement
does not add any new valency, it simply refers to the predicate. On the
other hand, the clitic in the copulative does not shift depending on the
number and gender of the element to which it alludes. As can be
observed, the copulative clitic is always lo regardless of the adjective it
substitutes, whether it is feminine, masculine, singular or plural. This is
not so in direct objects, whose corresponding clitic must (at least in the
standard) be inflected based on the gender and number of the direct
The examples of the unergative verbs behave differently. They may
develop a direct object every now and then without breaking the
standard rules. Once they do, the clitic that may replace the new valency
functions exactly the same way as a direct object clitic does. Notice that
in (5152), the clitic pronoun inflects in number and gender, depending
on the phrase it replaces. Unergative verbs are in principle more likely to
develop a direct object, owing to their agentive subject, but as has been
228 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
shown, unaccusative verbs can also do so, despite their lacking an
agentive subject. According to Bilous (2011), the feature of transitivity or
detransitivity of a certain lexeme is the emergence of a controller. If one
emerges, the transitivisation is more likely, while if the event lacks a
controller, the verb is more likely to be intransitivised. The second
possibility is the existence of a causer or an argument that allows the
development of an event. If the action lacks a controller or a causer, a
certain verb is very unlikely to be transitivised. This is why an unergative
verb, which by definition possesses an external argument that controls a
situation, prompts transitivisation with greater probability. Unacus-
satives, on the contrary, contain an internal argument which may be
promoted to the subject position in syntax, but which has to be recycled
in the position of object if the action foresees an external argument or
some type of controller that causes the event.
Let us return to some unergative examples (5355).
(53) Corro la marat
run.1SG.PRS the marathon
‘I am running the marathon.’
(54) Trabajo la madera.
work.1SG.PRS the wood
‘I work the wood.’
(55) Vuelo la cometa.
fly.1SG.PRS. the kite
‘I am flying the kite.’
In (5355), the intransitives become transitive, but in these examples
there exists an external agent. As for unaccusatives, when an expletive
arises, it obliges the verb to agree with it instead of with the argument
laid after the verb. The reason why the argument of unaccusative verbs is
put after the verb responds to its patient-like behaviour rather than being
an agent of the event. Thus, as it is semantically the patient, its syntax
tends to place it in a post-verbal position. Following this argument,
unaccusative verbs lack an agent or an element that prompts the action
denoted by them. However, logic may lead to suggest that there is always
an external argument, a prompter of the event, though we may not know
what or who it is. In other words, when something or someone falls, there
is something that has prompted this event, regardless of its agency,
volition or animacy.
This suggestion implies that all verbs may have an external argument
that, in the case of unaccusatives, is not explicitly expressed either
because it is irrelevant or because we do not know it. This triggers its
omission in speech, leaving the internal argument as the only valency of
the verb. Take note, this is highly important since it produces certain
Lability in Western Peninsular Spanish 229
©2018 The Editorial Board of Studia Linguistica
rearrangements in syntax. In this respect, Blake (2004) points out that
cross-linguistic phenomena usually follow a specific pattern both
throughout the grammatical cases and the syntactic functions, as
reproduced in (5657).
(56) Nominative >accusative >dative >ablative >genitive
(57) Subject >direct object >indirect object >oblique
The hierarchies represented in (5657) pinpoint that a specific phe-
nomenon usually appears in the nominative case or in the subject and
later spreads rightwards in the hierarchy. For instance, passivisation and
relativisation follow this model. If a language can relativise the indirect
object, then it can do so with the syntactic functions on the left. Likewise,
if a language can promote the indirect object of an active voice to the
subject of a passive voice, it can do the same with the direct object of the
active voice. Causativisation is also an excellent example of this tendency.
According to Comrie (1976 and 1981), Turkish observes both hierarchies
in causative constructions. Hence, if a monovalent sentence possesses a
subject inflected in nominative and has to add a new valency, this will
occupy the new subject and will be inflected in nominative. Therefore, the
old subject becomes the direct object and, as a consequence, is inflected in
accusative. If one more valency is again added, this becomes the subject
and agrees in nominative, while the old subject becomes the direct object
and agrees in accusative and the old direct object turns into the new
indirect object and agrees in dative (58).
(58) a. Hasan
Hasan.NOM die-PST
‘Hasan has died.’
b. Ali Hasan-ı
Ali.NOM Hasan.ACC die.CAUS.PST
‘Ali has killed Hasan.’
In the case of the Spanish lability, the logical subject is the semantic
author of the event, which becomes syntactically the subject. Thus, the
old subject (the internal argument which is the semantic patient of the
event) needs to follow the hierarchies depicted in (5657) and turns into
the direct object.
The same applies to the usage of expletive pronouns in unaccusative
verbs. The patients of these verbs are normally promoted to the subject
position or maintained in the object position, but working as the
grammatical subject. However, Ukrainian and French can exhibit two
alternatives when they deal with unaccusative predicates. On the one
hand, they can promote the patient to a subject position and turn it into
the grammatical subject (5960).
230 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(59) Dvi div
cyny pryj
two girls come.3PL.PST
‘Two girls came.’
(60) Trois hommes viennent
three men come.3PL.PRS
‘Three men come.’
This strategy leads the verb to agree with the internal argument because,
despite its patient-like behaviour, it is syntactically treated as the subject.
On the other hand, they can resort to an expletive subject and position
the internal argument in the place of the object. This last strategy leads
the internal argument to be syntactically treated as the direct object and
the verb to agree with the expletive subject (6162).
(61) Pryj
sl-o dvi div
come.3SG PST.NEUT two girls
‘Two girls came.’
(62) Il vient trois homes.
EXPL come.3SG.PRS three men
‘Three men come.’
Notice that the reassignment of case markers and syntactic functions
works exactly as Blake (2004) indicates. The monovalent predicate, in
which the internal argument is the only argument of the verb has to be
the subject or receive nominative case. However, if a new valency is
added (in this case, the expletive), the new valency becomes the subject
and the nominative, converting the internal argument in the syntactic
direct object and (in the case of rich inflection languages) a marked
In the case of Spanish, as it lacks a rich case inflection (except in
pronouns), the semantic distinctions regarding the type of subject need to
be codified through other means. In other words, since standard
transitive verbs are too transitive (following the transitivity scale of
Hopper & Thompson 1980 and 1982) (they have an agentive subject,
normally human and an affected patient that usually experiences a
change of state), the employment of an intransitive as a transitive implies
the notion of an unintentional subject, unable to control the situation
produced by the change of state (interestingly, Letuchiy 2010 defends
that labile verbs usually emerge in low transitivity). This is why I have
obtained from the same informants, sentences such as (64) if the action
was not done willingly, and sentences like (63) if the action was done on
Lability in Western Peninsular Spanish 231
©2018 The Editorial Board of Studia Linguistica
(63) La chica ha cogido el vaso y lo
the girl have.PRES.3SG take.PCP the glass and ACC.3SG
ha tirado al suelo.
have.PRES.3SG throw.PCP to+the floor
‘The girl has taken the glass and has thrown it onto the floor.’
(64) La chica se ha tropezado y ha
the girl REFL have.PRES.3SG slip.PCP. and have.PRES.3SG
ıdo el vaso.
fall.PCP the glass
‘The girl has slipped and dropped the glass.’
Sentence (63) has a typical transitive construction and therefore, the verb
tirar (‘to throw’) is preferred and (64) expresses lack of willingness or
intention, but the intransitive verb has emerged to turn the participant
that has carried out the action into the subject of the sentence. The glass
is still the patient, but the girl is the author. Hence, the verbs caer (‘to
fall’), like quedar (‘to stay’) and entrar (‘to enter’), are characterised by
being capable of choosing two types of subject (agentive or not),
depending on the degree of control over the given activity.
5.2.2. Objects
One of the characteristics that may affect the usage of lability is also
found in the type of object that the event denotes. Hopper and
Thompson (1980 and 1982) already drew attention to the likelihood of
transitivity depending not only on the features of the subject, but also on
those of the object. As they put it, less individualised or affected objects
are classified in the column of low transitivity. Levin (1999) underpins
that transitive verbs possess two kinds of objects: those belonging to
complex event structures and those which pertain to simple event
structures. The difference between these two event structures resides in
the fact that the former are causative and the latter are not. In complex
structures, one argument behaves in such a manner that it brings about
the change of state of a second argument. In simple structures, one
argument acts a certain way, undergoes a certain state or achieves a
certain change, but the other argument that may emerge does not really
suffer any modification. This distinction is somehow pertinent, as, for
instance, only some English transitive verbs are also transitive in a
number of Caucasian languages but many others do not have a
corresponding transitive counterpart in these languages. The reason for
this limitation seems to rely on the aforementioned semantic distinction.
Levin (1999) remarks that verbs of change of state such as open, break or
melt are classified as causative verbs, but transitive verbs such as sweep,
rub or wipe, which denote motion or surface contact cannot be
232 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
considered as causative and, therefore, present a simple event structure.
As a result, transitive verbs can also be split into two types: core
transitive verbs and non-core transitive verbs (this has also been pointed
out by Bilous 2011).
Following this premise, non-core transitive verbs such as those of
motion or surface contact are more inclined to a low transitivity
treatment, since their objects are less affected. If this theory is applied to
Spanish, entrar (‘to enter’) and quedar (‘to stay’) are predisposed to
lability because their objects are less affected, they simply change their
location, refer to motion or a surface contact. On the contrary, caer (‘to
fall’) does imply a complex event structure and is genuinely causative. As
lability emerges in low transitivity contexts, entrar (‘to enter’) and quedar
(‘to stay’) tend to become labile before caer (‘to fall’) due to their simple
event structure. Once they have arisen as labile, caer (‘to fall’) also
becomes so in the semantic contexts I have already mentioned. After all,
the object of caer (‘to fall’) takes part in a complex event structure and,
thus, it is induced by a more core transitive pattern than the other two.
Therefore, it does not turn out to be strange that entrar (‘to enter’) is
the most widespread labile verb in Spanish and, according to RAE
(2009), can also be attested as labile in America. This is evidenced by the
labilisation of this same verb and similar ones in other languages. Notice
that in English to enter is transitive, and in French, motion verbs such as
monter (‘to go up’), descendre (‘to go down’) or sortir (‘to go out’) can
also be transitive. They are normalised throughout the whole French-
speaking area and do not break the standard. Notice also that in
Sardinia, these same verbs are used both transitively and intransitively as
in French, but are not accepted in standard Italian, which prefers an
analytical construction with fare (‘to make’) plus a preposition or adverb
(Piredda 2017).
The same cannot be applied to quedar (‘to stay’) and caer (‘to fall’), at
least sensu stricto. Although quedar (‘to stay’) also denotes a contact
surface or belongs to a non-core transitive verb type, its emergence as
labile is less frequent. The explanation can be found in the double
meaning of standard dejar (‘to leave’), which can also denote a change of
state (6566).
(65) He dejado los libros en la mesa.
have.1SG.PRS leave.PCP the books on the table
‘I have left the books on the table.’
(66) Me has dejado anonadada.
ACC.1SG have.2SG.PRS leave.PCP stun.SG.FEM
‘You got me stunned.’
So, dejar (‘to leave’) can both express a causative meaning and a non-
causative one. The emergence of its intransitive counterpart quedar (‘to
Lability in Western Peninsular Spanish 233
©2018 The Editorial Board of Studia Linguistica
stay’) as labile is constrained by the non-core transitive meaning. As a
matter of fact, of the few examples of dejar (‘to leave’) as core-transitive I
have collected, only a couple of them correspond to the labile quedar (‘to
stay’) and always in the most south-western area where lability is
witnessed (6768).
(67) La ha quedado sorprendida.
ACC.FEM.SG have.3SG.PRS stay.PCP surprised.FEM.SG
‘He got her surprised.’
(68) La ha quedado triste.
ACC.FEM.SG have.3SG.PRS stay.PCP sad.FEM.SG
‘He got her sad.’
Lastly, tirar (‘to throw’) belongs to the core transitive verb type and,
thus, is less likely to appear in a low transitivity context. This is why caer
(‘to fall’) as labile may appear in the last place and can only start arising
as such in the above argued semantic readings.
5.3. Split intransitivity
I have already referred to the fact that lability in western Peninsular
Spanish is P-alignment type. Although the P-alignment strategy has been
related to ergative languages and many studies have been carried out on
case valency changes from the evolution of Latin to Romance and even
from Old French to Modern French (Heidinger 2014, Gianollo 2014), in
most of these cases it is the transitive verb that can be used intransitively.
The phenomenon dealt with throughout this paper extends the intran-
sitive valency to transitive. Indeed, it turns unaccusative, but not
unergative, into transitive verbs.
The different treatment between unaccusative and unergative verbs has
been called split intransitivity. In fact, when the behaviour of lability in
the examples collected is analysed, we realise that all the verbs affected
are unaccusative. This turns out to be vital in the linguistic behaviour. In
principle, unaccusative verbs lack a subject’s control (the subject behaves
as a patient rather than as an agent) (Dowty 1991, Ackerman & Moore
2001), whereas unergative verbs possess a volitional subject and,
consequently, an agent. For Perlmutter (1978), the unaccusative verbs
have subjects with similar traces to those held by the objects of transitive
verbs. This distinction brings to mind the ergative-absolutive alignment
pattern. The essential difference between ergative-absolutive languages
and nominative-accusative languages resides in the way they mark the
relation between subject and object. Thus, an ergative language possesses
a specific case-marking for a transitive subject (A) and a different case
marker for intransitive subjects (S), patients or objects (O) (Dixon 1994);
on the contrary, accusative languages combine S and A under the same
234 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
label (the subject) and O, under a different one. However, as Dixon
affirms, there are no purely nominative-accusative languages or purely
ergative-absolutive languages, as most of them use a combination of
these two patterns. This linguistic behaviour has been called split
ergativity and even the word order in Spanish applies this distinction: the
Spanish unaccusative verbs usually place the subject after the verb, as
though they were a kind of object or patient instead of a true subject
(Mendikoetxea 1999). Compare the following examples taken from this
(69) Los ni~
nos juegan.
the kids play.PRES.3SG
‘The kids are playing.’
(70) Los j
ovenes trabajan.
the youngsters work.PRES.3SG
‘The young people are working.’
(71) Existen problemas.
exist.PRES.3PL problems
‘There are problems.’
(72) Siempre vienen mujeres.
always come.PRES.3PL women
‘The women always come.’
Sentences (6972) show an intransitive verb with different types of
subject. While (6970) are characterised by an unergative verb because
they possess an agent, (7172) present an unaccusative verb that sends
the subject (in this case, the patient) to a location held prototypically by
the object or patient. Van Valin (1990) and Merlan (1985) remark that
this feature is better defined as split intransitivity, since the languages
characterised by this show a differentiation in syntax regarding the nature
of the intransitive verb, rather than the consistent distinction between
agentive or non-agentive in all their linguistic behaviour. This split is
manifested in Georgian or Italian, specifically in verbs of achievement or
unaccusative lexemes with patient-like subjects (Harris 1982, Hewitt
1987, Van Valin 1990). Likewise, the tense, aspect and mood are also
important, since, for example, Dutch perfects exhibit two auxiliaries,
depending on the telicity or atelicity contrast; Vedic (Kulikov 1999a,
1999b and 2003) and Greek (Karantzola & Leonidas 2014) are subjected
to mood, tense and, in the latter case, even to the gender of the noun
phrase dependant on the verb. Acehnese and Tsova-Tush languages also
take into account agentivity to develop split intransitivity (Durie 1985).
Unergative verbs may even develop a direct object without breaking
the standard variety. This possibility is explained due to the agentful
feature of their subjects. As Bilous (2011) argues, unergatives possess
Lability in Western Peninsular Spanish 235
©2018 The Editorial Board of Studia Linguistica
controller but hardly ever induce a controlled. However, in Spanish or in
English, their transitivisation is possible and not dialectal (7376).
(73) Correr el riesgo
run.INF the risk
‘To run the risk’
(74) Trabajar la madera
work.INF the wood
‘To work the wood’
(75) Volar una cometa
fly.INF a kite
‘To fly a kite’
(76) Run a company
Unlike unaccusative verbs, which possess an internal argument from the
semantic point of view, unergative verbs are characterised by having an
external argument. According to Bilous (2011), this is highly relevant,
since unaccusative subjects are patients which are positioned post-
verbally in the underlying structure but are promoted to a pre-verbal
position after syntactic rearrangements.
Italian is the most investigated language when studying split intran-
sitivity. According to Burzio (1981, 1986), this language applies the split
by the use of the clitic ne and the selection of the auxiliary verb in perfect
tenses. Compare the (7779), taken from Van Valin (1990).
(77) Angela
e arrivata.
Angela be.3SG.PRS.IND arrive.PCP.SG.FEM
‘Angela has arrived.’
(78) Angela
e morta
Angela be.3SG.PRS.IND die.PCP.SG.FEM
‘Angela has died.’
(79) Angela ha ballato con loro
Angela have.3SG.PRS.IND dance.PCP with them
‘Angela has danced with them.’
While (79) is an agentive subject and must agree with avere (‘to have’),
the two first statements in (7778) can only form the perfect tense with
essere (‘to be’) because they are non-volitional, patient-like. Moreover,
there are verbs that can swing between essere and avere, depending on the
agentivity or willingness, as (8081) show.
(80) Ho corso durante un’ ora.
have.1SG.PRS.IND run.PCP during one hour
‘I have run for one hour.’
236 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(81) Sono corso durante un’ ora
be.1SG.PRS.IND run.PCP.SG.MASC during one hour
‘I have run for one hour.’
The difference between (80) and (81) is the degree of intention or control
over the situation. While (80) is a completely agentive action, (81) may
imply that the subject has felt relatively obliged to do so. Although this
differentiation in unergatives is barely attested in Italian, certain dialect
areas may resort to it.
Similarly, Spanish also presented this pattern in Medieval times. In
accordance with Elvira (2001), Medieval Spanish could form perfect
tenses either with ser (‘to be’) or haber (‘to have’). The choice of one or
the other was based on the same parameters that Italian follows
nowadays (8283).
(82) De un d
ıa es llegado antes el rey
of one day be.3SG.PRS.IND arrive.PCP.SG.MASC before the king
don Alfons.
mister Alfons
‘The King Alfonso has arrived one day earlier.’
(83) El fuego quando es encendido sin
the fire when be.3SG.PRS.IND light.PCP.SG.MASC without
‘The fire when it is lit without wind.’
Examples (8283) show that, until nearly the end of the Middle Ages,
unaccusative verbs needed the verb ser (‘to be’) for perfect tenses. This
pattern disappeared five centuries ago and it has not been attested
dialectally in current Spanish. However, the proofs for the unaccusativ-
ity hypothesis not always fulfill the mentioned requirements. There are
verbs that may fall into the category of unergatives in some languages
and into the category of unaccusatives in others. For example, entrar
(‘to enter’) and caer (‘to fall’) are considered unaccusative in Spanish,
but Mandan (a Siouan language) treats to enter as unergative and to fall
as unaccusative (Kennard 1936). The selection of to be as the auxiliary
is not a consistent proof either. As Bilous (2011) puts it, Slavic
languages resort to this verb even for transitive verbs when the perfect is
Needless to say, it is precipitated to state that Spanish has a tendency
toward ergativity. In fact, as Dixon (1994), Creissels (2008) or Elvira
(2012) affirm, there are no purely ergative languages while just a few
accusative languages exhibit a pure accusative behaviour without
ergative influences. Besides, Elvira (2012) reminds us of the fact that
Lability in Western Peninsular Spanish 237
©2018 The Editorial Board of Studia Linguistica
Spanish descends from a highly accusative language (Latin) that even
had several ergative patterns, such as the deponent verbs (Bassols De
Climent 1956, Ba~
nos Ba~
nos 2009). However, Klimov (1977) has pointed
out that languages usually follow a specific development in their
alignment. If a language is mainly ergative or exhibits multiple ergative
patterns, it will go through a series of processes that will bring it closer to
accusativity. In fact, Indo-European presented a great many ergative
particularities (Laroche 1962, Meillet 1931) that decreased in Latin
(Cennamo 2009) and which are even fewer in current Romance languages
(Lehmann 1985). In the case of Spanish, it has gradually eliminated the
syntactic particularities that highlighted ergative or split intransitivity
patterns (Elvira 2012). The only feature that still remains in the standard
is the placement of unaccusative subjects in object positions.
5.4. Split ergativity
Although standard Spanish shows split intransitivity, western Peninsular
Spanish accumulates a series of linguistic features that bring it closer to
ergativity. The distinction between the type of semantic subject is not
only applied to intransitive sentences, but also to clitic pronouns,
reflexive markers and the strategy of lability.
This dialect is characterised by presenting a series of vernacular
particularities based on animacy, agentivity and willingness that differ
from the standard. The most studied topic is le
ısmo, la
ısmo, lo
ısmo (the
mixture of case-marking in clitic pronouns). Standard Spanish relies on
case-marking to select clitic pronouns; in other words, the choice of the
clitic pronoun is based on the fact that the object is accusative or dative,
masculine or feminine and singular or plural. Observe the standard use of
the object clitics in (8488).
(84) A mis nietas les he explicado bien
to my granddaughters DAT.PL have.1SG.PRS explain.PCP well
todas las cosas de *antes.
all the things of *before
‘I have explained to my granddaughters the things of the past.’
(85) Juan la cansa [a Mar
ıa] con sus tonter
Juan ACC.FEM tire.3SG.PRS [to Mar
ıa] with his stupidities
‘Juan gets Mar
ıa tired with his stupid things.’
(86) En el verano, hab
ıa que recoger la hierba y
in the summer, have.3SG.PST that collect.INF the grass and
traerla a casa para el invierno.
bring.INF+ACC.FEM to house for the winter
‘In summer, it was necessary to collect the grass and bring it home
for the winter.’
238 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(87) Tenemos un frigor
ıfico nuevo, lo tenemos
have.1PL.PRS a fridge new ACC.MASC have.1PL.PRS
puesto aqu
put.PCP here
‘We have a new fridge, we have put it here.’
(88) El trabajo la cansa [a Mar
the work ACC.FEM tire.3SG.PRS [to Mar
ıa gets tired of working.’
Western Peninsular Spanish, on the contrary, focuses on the animacy of
the subject and the object and even the opposition between countable and
uncountable. The use of the clitics in the former examples changes when
they respond to the dialect pattern of western Peninsular Spanish (8993)
nez 1999, 2006 and 2007).
(89) A mis nietas las he explicado
to my granddaughters ACC.FEM.PL have.1SG.PRS explain.PCP
bien todas las cosas de antes.
well all the things of before
‘I have explained to my granddaughters the things of the past.’
(90) Juan le/la cansa [a Mar
ıa] con sus tonter
Juan DAT.SG/ACC.FEM tire.3SG.PRS [to Mar
ıa] with his stupidities
‘Juan gets Mar
ıa tired with his stupid things.’
(91) En el verano, hab
ıa que recoger la hierba y
in the summer have.3SG.PST that collect.INF the grass and
traerlo a casa para el invierno.
bring.INF+ACC.NEUT to house for the winter
‘In summer, it was necessary to collect the grass and bring it home
for the winter.’
(92) Tenemos un frigor
ıfico nuevo, le tenemos puesto
have.1PL.PRS a fridge new DAT.SG have.1PL.PRS put.PCP
‘We have a new fridge, we have put it here.’
(93) El trabajo le cansa [a Mar
the work DAT.SG tire.3SG.PRS [to Mar
ıa gets tired of working.’
While (8488) have chosen a specific clitic on the basis of case, gender
and number, (8993) misuse standard clitics by mixing accusative and
dative on the basis of animacy of both the object and the subject. Notice
that (89) exhibits a human object and gives importance to the gender and
Lability in Western Peninsular Spanish 239
©2018 The Editorial Board of Studia Linguistica
number rather than the case parameter. On the contrary, (92) shows a
non-human object and this is why it selects a dative instead of an
accusative. The cases of (90) and (93) are based on the semantics of the
subject and this is why a different clitic is chosen despite their having the
same verb and the same object. Token (91) is a sample of the distinction
between countable and uncountable: although grass is feminine in
Spanish, most of western Peninsular Spanish produces a neuter.
But clitic pronouns are not the single example of particularity on the
basis of volition, animacy or agentivity. More and more studies are
drawing attention to the increase of reflexive markers in western
Peninsular Spanish in sentences in which the subject lacks volition or
agentivity (De Benito 2015). Compare (9495).
(94) Ayer me so~
e que me
yesterday REFL.1SG dream.1SG.PST.IND that REFL.1SG
ıa de un precipicio.
fall.1SG.IMPF.IND of a cliff
‘Yesterday I dreamt [myself] that I was falling off a cliff.’
(95) Hay que esperar a que la fruta se
have.3SG.PRS.IND that wait.INF to that the fruit REFL.3SG
‘We have to wait until the fruit rippens [itself].’
Instance (94) shows that the reflexive marker is produced because the
subject cannot decide or control whether he / she dreams or not. The subject
just dreams. Likewise, (95) demonstrates that the fruit cannot control
whether it ripens or not. This strategy is impossible in standard Spanish
and, in fact, it is scarcely attested dialectally outside the western Spanish
area. RAE (2009) limits its emergence in western Spain and elsewhere in
America, without specifying the region. However, if the corpus CREA is
consulted to this respect, only a few occurrences (fewer than 10) appear.
The ones whose origin is America can be circumscribed to South America.
Although I have ruled out the Jim
enez Fern
andez and Tubino Blanco
(2014) research regarding the geographical extension of lability because
of the above mentioned reasons, they succeed in pointing out the resort
to reflexive markers in the verbs they analyse. The pattern of these verbs
is the same as that of (94) and (95). It is another proof that certain
regions within Spain exhibit reflexives to mark the lack of agentivity,
intention or control. However, their analysis refers to the whole
Andalusian dialect. It is true that this pattern is witnessed in western
Andalusian (RAE 2009), but the authors fail to divide Andalusia in two
very recognisable blocks. It is the western part which may develop
reflexives and, as western, it often adopts the linguistic features that are
attested in western Peninsular Spanish.
240 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
Moreover, they put forward that the use of the reflexive marker (se)is
compatible with lability in Andalusian and codifies the origin of the
event. However, they obviate the tendency in western Peninsular Spanish
to produce reflexives when the action is not controlled by the agent. Their
examples (reproduced in 9698) would be one more proof of this
tendency, but the instances they provide are not appropriate for their
(96) Juan se ha bajado la bicicleta a
Juan REFL.3SG have.3SG.PRS.IND put down.PCP the bicycle to
la calle.
the street
‘Juan has taken his bicycle down to the street.’
(97) Juan se ha subido los juguetes a casa.
Juan REFL.3SG have.3SG.PRS.IND put up.PCP the toys to home
‘Juan has taken his toys up home.’
(98) Juan se ha entrado las sillas en casa.
Juan REFL.3SG have.3SG.PRS.IND enter.PCP the chairs in home
‘Juan has entered his chairs home.’
The authors, in sentences (9698) (one of which includes the causativi-
sation of entrar ‘to enter’), relate the employment of the reflexive se to
causation and allege that the three examples are witnessed in the
Andalusian dialect. As has been stated, the Andalusian dialect can be
divided in two, with clear differences in morphosyntax between the west
and the east. In addition, (9698) are attested nationwide and are
perfectly standard. To my view, they fail to recognise these reflexives as
possessive datives. Spanish usually produces possessive datives (Davis
1968, Dumitrescu 1990) and (9698) can be perfectly replaced by
(99) Juan ha bajado su bicicleta a la
Juan have.3SG.PRS.IND put down.PCP POSS.3SG bicycle to the
‘Juan has taken his bicycle down to the street.’
(100) Juan ha subido sus juguetes a su
Juan have.3SG.PRS.IND put up.PCP POSS.3PL toys to POSS.3SG
‘Juan has taken his toys up home.’
(101) Juan ha entrado sus sillas en casa.
Juan have.3SG.PRS.IND enter.PCP POSS.3PL chairs in home
‘Juan has entered his chairs home.’
Lability in Western Peninsular Spanish 241
©2018 The Editorial Board of Studia Linguistica
So, the instances exemplified by Jim
enez Fern
andez and Tubino Blanco
(2014) fail to account for the resort to reflexives in unaccusative verbs,
that is to say, in verbs in which the subject is the author and not the
agent. Notice that this vernacular feature is similar to the standard
strategy illustrated in (3), in which the verb admits a reflexive marker
preceded by an experiencer in order to express lack of intention or
control of the author.
Lastly, as for lability in this dialect, I have only been able to provide a
first overview of the semantic nuances that influence this phenomenon,
but the research should be furthered, by going more in depth by
considering syntactic parameters, such as tense or mood, since quite a
few languages, such as Hindi-Urdu, select ergativity only in certain tenses
(Garrett 1990, Kulikov 1999b).
6. Conclusions
The lability attested in vernacular western Peninsular Spanish, consisting
of the extension of an intransitive verb to a transitive reading,
(specifically of caer ‘to fall’, quedar ‘to stay’ and entrar ‘to enter’), is
still maintained although its spatial diffusion has dramatically decreased,
in comparison to the descriptions available up to now. Based on the
fieldwork carried out, nowadays there is pressure towards standardisa-
tion that has displaced this phenomenon to the most south-western areas.
Likewise, the distribution of entrar (‘to enter’) is wider than that of
quedar (‘to stay’), which is even more widespread than that of caer (‘to
fall’). In this sense, there seems to be a hierarchical continuum in which
the informants characterised by the use of caer (‘to fall’) as transitive
produce quedar (‘to stay’) and entrar (‘to enter’) with direct object too. In
addition, the speakers that use these verbs in sentences with a volitional
agent also employ them in sentences with a non-human subject and an
un-wilful subject. Also, if they use them in telic readings, they also resort
to them in atelic sentences. A priori, the area around Badajoz is the one
where this phenomenon is mostly attested, since it has extended the
intransitive verb even onto volitional readings, but this vernacular
particularity moves along the Roman Silver Route as the way of diffusion
northwards through western Spain.
Lability depends on a series of semantic factors that include agentivity,
control, volition or telicity. Instead of the morphological changes
witnessed in other languages, Spanish chooses to transitivise the
intransitive lexeme as it lacks a rich inflectional system that may
establish such semantic nuances. As the existing standard transitive
implies control, volition or even agency, the opposite semantic nuances
are materialised through lability. This also has to do with the type of
transitivity to which the standard transitive verbs belong. While meter
242 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
(‘to put on’) and dejar (‘to leave’) are non-core transitive verbs, because
they are not causative and simply refer to a motion or a surface
contact, tirar (‘to throw’) does have a causative meaning that includes it
in the core transitive verbs. As lability emerges in low transitivity
contexts, the non-core transitive verbs are by definition more inclined to
favour it. Therefore, entrar (‘to enter’) is the first most widespread
causativised verb, followed by quedar (‘to stay’) and in the last place
caer (‘to fall’). The reason why quedar (‘to stay’) follows entrar (‘to
enter’) and not the other way around may rest on the double meaning
that the standard transitive (dejar ‘to leave’) possesses. As has been
mentioned, this verb can be both non-core and core transitive, since it
can also induce a change of state and, consequently, a causative
meaning. In my corpus, the non-core transitive reading is the first to
favour lability, after which the core transitive meaning adopts lability.
As tirar (‘to throw’) is genuinely core-transitive, it is the verb least likely
to become labile.
The transitivisation of the aforementioned verbs implies the adding of
a new valency. As the unaccusative verbs are inserted into monovalent
predicates and, as a result, their internal arguments work as the subjects,
when a new argument is added (in the lability cases, their logic subject,
their external argument) this becomes the new subject and sends the
former subject to the direct object position. The conversion of the
internal argument into the new direct object when the external argument
is made explicit responds to the cross-linguistic tendency in case-marking
and in syntactic functions argued throughout the article.
Although Spanish has exhibited split intransitivity patterns inherited
from Latin throughout its historical development, its tendency is to
eliminate them. The single mixed model we find nowadays is the placement
of unaccusative subjects in object positions, but the choice of ser (‘to be’) as
the auxiliary for perfect tenses does not exist any longer and the
particularities that coexist in certain dialect areas have not generalised.
On the contrary, the block that western Peninsular Spanish constitutes
really shows a linguistic pattern that may better be defined as split
ergativity. Rather than employing unaccusative verbs in a different way
(which it does), it has developed several strategies to syntactically
differentiate between volitional / agentive participants from non-volitional
or non-agentive ones. These strategies are well represented in the clitic
pronoun selection, the lability phenomenon described throughout, the
placement of unaccusative subjects in object positions and the resort to
reflexive markers.
7. Bibliography
ACKERMAN,F.&MOORE, J. 2001. Proto-properties and grammatical encoding.
Stanford: CSLI Publications.
Lability in Western Peninsular Spanish 243
©2018 The Editorial Board of Studia Linguistica
AIKHENVALD,A.&DIXON, R.M.W. 2000. Changing valency. Case studies in
transitivity. Cambridge: Cambridge University Press.
ALEANR: see ALVAR, M. 19791983.
ANDEZ, F. 19881994.
ALVAR, M. 19791983. Atlas ling
ıstico y etnogr
afico de Arag
on, Navarra y Rioja
(ALEANR). Zaragoza: CSIC.
ALVAR, M. 1996. Manual de dialectolog
ıa hisp
anica. Barcelona: Ariel.
ALVAR, M. 1999. Atlas ling
ıstico de Castilla y Le
on. Valladolid: Junta de Castilla
ALCyL: see ALVAR, M. 1999.
ARIZA, M. 2008. Estudios sobre el extreme~
aceres: Universidad de
NOS, J. M. 2009. Sintaxis del lat
ın cl
asico. Madrid: Liceus.
BASSOLS DE CLIMENT, M. 1956. Sintaxis latina. Madrid: CSIC.
BILOUS, R. 2011. Transitivit
e et marquage d’objet diff
erentiel. Toronto: University
of Toronto.
BILOUS, R. 2012. Transitivity revisited: an overview of recent research and
possible solutions. Proceedings of the 2012 annual conference of the Canadian
Linguistic Association,114.
BLAKE, B. J. 2004. Case. Cambridge: Cambridge University Press.
BURZIO, L. 1981. Intransitive verbs and Italian auxiliaries. Cambridge: MIT.
BURZIO, L. 1986. Italian syntax. Dordrecht: Reidel.
CANO, R. 2004. Historia de la lengua espa~
nola. Barcelona: Ariel.
CENNAMO, M. 2009. Argument structure and alignment variations and changes in
Late Latin. The role of semantic, pragmatic and discourse factors in the
development of case, eds. J. Barddal & S. L. Chelliah, 307346. Amsterdam:
John Benjamins.
COMRIE, B. 1976. The syntax of causative constructions. The grammar of causative
constructions, ed. M. Shibatani, 261312. New York: Academic Press Inc.
COMRIE, B. 1981. Language universals and linguistic typology, syntax and
morphology. Oxford: Blackwell.
COMRIE,B.&POLINSKY, M. 1993. Causatives and transitivity. Amsterdam/
Philadelphia: John Benjamins Publishing Company.
NEZ, I. 19882017.
CREA: see Real Academia Espa~
CREISSELS, D. 2008. Direct and indirect explanations of typological regularities:
the case of alignment variations. Folia Linguistica XLII (1):138.
CREISSELS, D. 2014. P-liability and radical P-alignment. Linguistics 52 (4): 911944.
DANIEL, M. A. et al. 2012. Causatives in Agul. Argument structure and
grammatical relations, eds. M. Shibatani et al., 55113. Amsterdam/
Philadelphia: John Benjamins.
DAVIS, C. J. 1968. The indirect object of possession in Spanish. The USF
Language Quarterly,16.
DEBENITO, C. 2015. Las construcciones con se desde una perspectiva variacionista
y dialectal. Madrid: Universidad Aut
onoma de Madrid. PhD diss.
ES, J. P. 1998. Transitivit
emantique, transitivit
e syntaxique. La
e, dir. A. Rousseau, 161180. Villeneuve d’Ascq: Presses
universitaires du Septentrion.
DIXON, R.M.W. 1994. Ergativity. Cambridge: Cambridge University Press.
DOWTY, D. 1991. Proto-Roles and Argument Selection. Language 67 (3):547619.
DUMITRESCU, D. 1990. El dativo posesivo en espa~
nol y rumano. Revista Espa~
de Ling
ıstica 20 (2):403429.
DURIE, M. 1985. A grammar of Acehnese. Dordrecht: Foris.
244 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
ELVIRA, J. 2001. Intransitividad escindida en espa~
nol: el uso auxiliar de ser en
nol medieval. Estudios de ling
ıstica 15:201245.
ELVIRA, J. 2012. Construcciones y significado: aspectos diacr
onicos de la
transitividad en espa~
nol. Corrientes de estudio en sem
antica y pragm
oricas, ed. R. Maldonado, 128. Madrid: Instituto Men
endez Pidal.
NEZ, I. 19882017. Corpus Oral y Sonoro del Espa~
nol Rural
(COSER). <>.
NEZ, I. 1999. Le
ısmo, la
ısmo y lo
ısmo. Gram
atica descriptiva de
la lengua espa~
nola, dirs. I. Bosque & V. Demonte, 13171397. Madrid: Espasa.
NEZ, I. 2006. Del Cant
abrico a Toledo: el “neutro de materia”
anico en un contexto rom
anico y tipol
ogico I. Revista de Historia de la
Lengua Espa~
nola 1:67118.
NEZ, I. 2007. Del Cant
abrico a Toledo: el “neutro de materia”
anico en un contexto rom
anico y tipol
ogico II. Revista de Historia de la
Lengua Espa~
nola 2:2981.
NEZ, I. 2011. La lengua de Castilla y la formaci
on del espa~
Madrid: Espasa.
IA, J. A. F. 1993. Historia de las hablas andaluzas. Madrid: Arco Libros.
ANDEZ, F. 19881994. Atlas Ling
ıstico (y
afico) de Castilla-La Mancha. <>.
IA MOUTON, P. 1994. Lenguas y dialectos de Espa~
na. Madrid: Arco Libros.
GARRETT, A. 1990. The origin of NP split ergativity. Language 66:261296.
GIANOLLO, C. 2014. Labile verbs in Late Latin. Linguistics 52 (4):9451002.
GILQUIN, G. 2010. Corpus, cognition and causative constructions. Amsterdam/
Philadelphia: John Benjamins.
ON, T. 1976. The syntax of causative constructions: cross-language similarities
and divergences. The grammar of causative constructions, ed. M. Shibatani, 80
120. New York: Academic Press.
ON, T. 2001. Syntax. Amsterdam/Philadelphia: John Benjamins.
HARRIS, A. 1982. Georgian and the unaccusative hypothesis. Linguistics 58:290
HASPELMATH, M. 1993. More on the typology of inchoative/causative verb
alternations. Causatives and transitivity, eds. B. Comrie & M. Polinsky, 87121.
Amsterdam/Philadelphia: John Benjamins.
HEIDINGER, S. 2014. The persistence of labile verbs in the French causative-
anticausative alternation. Linguistics 52 (4):10031024.
HETZRON, R. 1976. On the Hungarian causative verb and its syntax. The grammar of
causative constructions, ed. M. Shibatani, 371398. New York: Academic Press.
HEWITT, B. G. 1987. Georgian: ergative or active? Lingua 71:319340.
HOPPER,P.&THOMPSON, S. 1980. Transitivity in grammar and discourse.
Language 56 (2):251299.
HOPPER,P.&THOMPSON, S. 1982. Syntax and semantics. Studies in transitivity.
New York: Academic Press.
A. & TUBINO BLANCO, M. 2014. Variaci
on sint
actica en la
on l
exica. Revista espa~
nola de ling
ıstica 44 (1):738.
KARANTZOLA,E.&LAVIDAS, N. 2014. On the relation between labilisations and
neuter gender: evidence from Greek diachrony. Linguistics 52 (4):10251059.
KEHAYOV,P.&VIHMAN, V. 2014. The lure of lability: a synchronic and diachronic
investigation of the labile pattern in Estonian. Linguistics 52 (4):10611105.
KENNARD, E. 1936. Mandan grammar. IJAL 9:143.
KIM, H. S. 2012. Transitives, causatives and passives in Korean and Japanese.
Argument structure and grammatical relations, eds. M. Shibatani et al., 241255.
Amsterdam/Philadelphia: John Benjamins.
Lability in Western Peninsular Spanish 245
©2018 The Editorial Board of Studia Linguistica
KLIMOV, G. A. 1977. Tipologija jazykovaktivnogo stroja. Moscow: Nauka.
KLIMOV,G.A.&ALEKSEEV, M. E. 1980. Tipologija kavkazskix jazykov. Moscow:
KULIKOV, L. 1999a. May he prosper in offspring and wealth: a few jubilee remarks
on the typology of labile verbs and Sanskrit p
usyati ‘prosper; makes prosper’.
Tipologija i teorija jazyka: at opisanija k ob”jasneniju K 60-letiju A. E. Kibrika, eds.
E. V. Rakhilina & Y. G. Testelets, 224244. Moscow: Jazyki russkoj kul’tury.
KULIKOV, L. 1999b. Split causativity: remarks on correlations between
transitivity, aspect and tense. Tense, aspect, transitivity and causativity. Essays
in honour of Vladimir Nedjalkov, eds. W. Abraham & L. Kulikov, 2143.
Amsterdam/Philadelphia: John Benjamins.
KULIKOV, L. 2003. The labile syntactic type in a diachronic perspective: the case
of Vedic. Journal of linguistics 26:93112.
KULIKOV,L.&LAVIDAS, N. 2014. Typology of labile verbs: focus on diachrony.
Linguistics 52 (4):871877.
LARA, V. 2012. Ustedes instead of vosotros and voc^
es instead of v
os: an analysis
through the Linguistic Atlas of the Iberian Peninsula (ALPI). Dialectologia
Special Issue 3:5793.
LARA, V. 2016. La expresi
on del futuro en las lenguas romances de la Pen
erica. Bolet
ın de la Real Academia Espa~
nola XCVI (CCCXIV):529558.
esence ou absence de l’objet: limites du possible en
ßais. Helsinki: Academia Scientirum Fennica.
LAROCHE, E. 1962. Un ergative en indo-europeen d’Asie Mineure. Bulletin de la
e de linguistique de Paris 57:2343.
LEHMANN, C. 1985. Ergative and active traits in Latin. Relational grammar, ed. F.
Plank, 243267. Berlin: Mouton de Gruyter.
LETUCHIY, A. B. 2004. Lability of verbs and its relations to verb meaning and
argument structure (based on data from Indo-European, Arabic, Turkish and
other languages). Vortrag beim LENCA-2-Symposium an der Universit
at Kasan.
vol. 11(115):15.
LETUCHIY, A. B. 2009. Towards a typology of labile verbs: lability versus
derivation. New challenges in typology: transcending the borders and refining the
distinctions. Berlin: Mouton de Gruyter, 247268.
LETUCHIY, A. B. 2010. Lability and spontaneity. Transitivity: form, meaning,
acquisition and processing, eds. P. Brandt & M. Garc
ıa, 237255. Amsterdam/
Philadelphia: John Benjamins.
LETUCHIY, A. B. 2015. Historical development of labile verbs in modern Russian.
Linguistics 53 (3):611647.
LEVIN, B. 1999. Objecthood: an event structure perspective. Proceedings of the
CLS 35:223247.
MCMILLION, A. 2006. Labile verbs in English: their meaning, behaviour and
structure. Stockholm: University of Stockholm.
MEILLET, A. 1931. Essai de chronologie des langues indo-europ
eennes, la theorie
du feminine. Bulletin de la Soci
e de linguistique de Paris 32:128.
MENDIKOETXEA, A. 1999. Construcciones inacusativas y pasivas. Gram
descriptiva de la lengua espa~
nola, dirs. I. Bosque & V. Demonte, 15751630.
Madrid: Espasa.
ENDEZ PIDAL, R. 2005. Historia de la lengua espa~
nola. Madrid: Instituto
endez Pidal.
MERLAN, F. 1985. Split intransitivity: functional oppositions in intransitive
inflection. Grammar inside and outside the clause, eds. J. Nichols & A.
Woodbury, 324362. Cambridge: Cambridge University Press.
EJAR, J. 1991. Dialectolog
ıa andaluza. Granada: Don Quijote.
246 V
ıctor Lara Bermejo
©2018 The Editorial Board of Studia Linguistica
MONTERO CURIEL, P. 2006. El extreme~
no. Madrid: Arco Libros.
NARBONA, A. 2003. El espa~
nol hablado en Andaluc
ıa. Sevilla: Ariel.
NERBONNE,J.ET AL. 2010. Gabmap A Web Application for Dialectology.
Groningen: University of Groningen.
NICHOLS, J. 1984. Direct and oblique objects in Chechen-Ingush and Russian.
Objects: towards a theory of grammatical relations, ed. F. Plank, 183209.
London: Academic Press.
NUNES, J. J. 1945. Comp^
endio de gram
atica hist
orica portuguesa. Lisbon: Livraria
assica Editora.
PENNY, R. 2004. Gram
atica hist
orica del espa~
nol. Barcelona: Ariel.
PERLMUTTER, D. 1978. Impersonal passives and the unaccusativity hypothesis.
Proceedings of the 4th Annual meeting of the Berkeley linguistic society, 157189.
Chicago: Berkeley University.
PIREDDA, N. 2017. L’italiano regionale di Sardegna. Manuale di linguistica sarda,
eds. E. Blasco Ferrer et al., 495507. Berlin: Mouton de Gruyter.
RAMCHAND, G. C. 2011. Licensing of instrumental case in Hindi / Urdu
causatives. Nordlyb 38:4985.
RAPPAPORT-HOVAV,M.&LEVIN, B. 2000. Classifying single argument verbs.
Lexical specification and insertion, eds. M. Everaert et al., 269304. Amsterdam:
John Benjamins.
NOLA. CREA: Corpus de referencia del espa~
nol actual
[online]. <> [last consulted 26/11/2017].
Real Academia Espa~
nola & Asociaci
on de Academias de la Lengua Espa~
2009. Nueva gram
atica de la lengua espa~
nola. Madrid: Espasa.
SAID, M. A. 1931. Gram
atica hist
orica da l
ıngua portuguesa. Rio de Janeiro:
Livraria Acad
SEJXOV, E. M. 1987. Labil’nye glagoly i labil’naja konstrukcija predlo
zenija v
lezginskom jazyke. E
zegodnik iberijsko-kavkazskogo jazykoznanija 14:282287.
SHIBATANI, M. 1976. Syntax and semantics. The grammar of causative
constructions. New York: Academic Press.
SHIBATANI, M. 2002. The grammar of causation and interpersonal manipulation.
Amsterdam/Philadelphia: John Benjamins.
SONG, J. J. 1996. Causatives and causation: a universal and typological perspective.
London: Longman.
VAN VALIN, R. D. 1990. Semantic parameters of split intransitivity. Language 66
(2): 221260.
VISSER, F. 1970. A historical syntax of the English language. Leiden: Brill.
WHEELER, M. A. et al. 1999. Catalan: a comprehensive grammar. London:
WOLFRAM,W.&SCHILLING-ESTES, N. 2003. Dialectology and Linguistic
Diffusion. The Handbook of Historical Linguistics, eds. Brian D. Joseph & R.
D. Janda, 713735. Malden: Blackwell.
ZAMORA VICENTE, A. 1970. Dialectolog
ıa espa~
nola. Madrid: Gredos.
Received February 3, 2017
Accepted April 3, 2018
ıctor Lara Bermejo
a Ca’ Foscari Venezia
Dorsoduro, 3246
30123 Venezia
Lability in Western Peninsular Spanish 247
©2018 The Editorial Board of Studia Linguistica
Full-text available
While Latin syntax uses overwhelmingly accusative alignment, there are also constructions which show ergative or active alignment.
Full-text available
En: Geolingüistica. Trabajos europeos / Pilar García Mouton (editora) ,1994 En marzo de 1987 presentamos el «Proyecto de un Atlas Lingüístico y etnográfico de Castilla-La Mancha (ALeCMan)» en el marco del 1 Congreso Internacional de Historia de la Lengua Española y en él aludíamos a la necesidad de un atlas de la zona central de la Península1. El ALeCMan se pensó como un atlas regional con planteamientos teóricos muy cercanos a los de Andalucía, Canarias, Aragón, Navarra y Rioja, y Santander, dirigidos todos ellos por Manuel Alvar. La zona que abarca fue estudiada por los investigadores del Atlas Lingüístico de la Península Ibérica en los años treinta y, en los setenta, por los del Atlas Lingüístico de España y Portugal, cuyas encuestas permanecen aún inéditas, pero son la base de la contribución española al Atlas Linguarum Europae y al futuro Atlas Lingüístico Románico. Sin embargo, tanto unas como otras eran tareas para un atlas de gran dominio, con cuestiones generales, válidas en principio para una geografía muy amplia y con una red de encuesta forzosamente menos tupida que la de un atlas regional. Este trabajo se ha redactado dentro del proyecto PB86-0583 de la CICYT Peer reviewed
This is a reference grammar of Acehnese, a regional language of north Sumatra. The introduction gives a description of the geographical location and genetic affiliation of Acehnese. There is an overview of previous works on Acehnese, and a statement of goals and methodology. Chapter two is a description of the phonology. Some acoustic data is given, including a formant plot of the oral monophthongs. Points of general interest are: • The treatment of the Acehnese fricative [S]. • The treatment of murmur and aspiration. • The treatment of 11 funny 11 nasal stops and nasal vowels, which differs considerably from past analyses. A phonemicisation is presented and an orthography, which is followed in the rest of this grammar. Chapter three begins with a description of the terms phrase, word and clitic, which are important for the rest of the grammar. It gives a treatment of allomorphy, free morphological variation, and reduplication. The final section contains an outline of parts of speech. Chapter four is a description of verbs. Detailed attention is paid to verb derivation, and the semantic characteristics of the main verbal arguments Agent and Undergoer. There is also a treatment of nominal incorporation onto verbs, and a justification for having no class of adjectives - a major point of departure from previous descriptions of Acehnese. Chapter five is a description of nominals and nominal phrases. Important sections are the treatment of possession, nominal predicates, pronouns, demonstratives, expressions for measuring, and nominal derivation. Special attention is paid throughout to semantics. The section on pronominal clitics is of particular importance for the syntactic description in later chapters. Chapter six is a description of epistemological classifiers, a special closed class of ,vords used to categorise types of knowledge. One of their uses is to form a wide variety of verbs with the meaning 'to know'. Chapter seven is a description of prepositions, their use and function. Chapter eight is a description of syntax within the clause. The key concept of Core Role is developed here. Acehnese is a language of considerable theoretical interest; it is a language that is neither ergative nor accusative in type. For this reason a special nondescriptive section describes the place of Acehnese in syntactic typology and the problem that it presents for current theories of syntax. This chapter also develops the concept of Core Top£c, a grammaticalised topic marked by word order. There is a description of word order, ellipsis, intonation, argument cliticisation, and of the focus marker di; all these fit in the general category of topicality. After these sections, certain miscellaneous aspects of clausal syntax are presented: clausal adjuncts, the vocative, reflexives, the comparative, and predicate operators. Chapter nine describes syntax beyond the clause. The first part deals with clause embedding. In this the treatments of relative clauses and complementation are important for their illustration of the concepts of Core Role and Core Topic. The final sections deal with conjunctions, non-declarative sentence types such as questions and orders, and words such as exclamations and interjections.
Il presente contributo ha lo scopo di fornire, in breve, una descrizione dell’italiano regionale sardo, osservando sia la produzione linguistica dei parlanti che la loro percezione. Dopo una concisa definizione dell’italiano regionale di Sardegna, il contributo si articola principalmente in due parti. Inizialmente l’italiano parlato in Sardegna viene descritto tramite una serie di tratti fonetici, morfosintattici e lessicali che lo contraddistinguono. Si offre, quindi, una caratterizzazione dell’italiano sardo a livello regionale, evidenziando soprattutto quei tratti che lo rendono singolare, ma allo stesso tempo fornendo anche quelle caratteristiche che lo accomunano a varietà regionali più ampie. Segue poi una seconda parte riguardante principalmente la percezione. Infatti, si è osservato se, come e in che modo l’italiano regionale di Sardegna venga percepito dai suoi parlanti. A livello percettivo, non solo si conferma la capacità dei parlanti sardi di identificare l’italiano regionale a livello locale ma anche di fornirne dei tratti distintivi. La produzione linguistica unita alla percezione consente, in conclusione, di assegnare all’italiano regionale di Sardegna una sua collocazione ben precisa.