ArticlePDF Available

The Two Spanish Subjunctives: The Required and Default Subjunctives

Authors:

Abstract and Figures

Article link: (https://septentrio.uit.no/index.php/borealis/article/view/6334). Standard Spanish grammar states that desideratives (querer que), directives (aconsejar que), purpose clauses (para que), causatives (hacer que), emotive-factives (alegrarse de que), dubitatives (dudar que), modals (ser posible que), and negated indicative-normative verbs (e.g., no creer que), embed subjunctive complement clauses. However, in spite of this classification, some predicates will exhibit a certain degree of mood variation. For instance, emotive-factives can take indicative complements (Crespo del Río 2014; Faulkner 2021a, 2021b). Similar variability between the moods may also come about in negated epistemic (Bolinger 1991), dubitative (Blake 1981), and modal clauses (Deshors and Waltermire 2019). In the present paper, I propose that such variation stems from the Spanish mood system involving a split between two types of subjunctives: one that is required in non-realistic, preference-based contexts, and another that is the default of propositions that are at least, somewhat realistic. I argue that, whereas non-realistic, preference-based clauses (e.g., desiderative, directive, and purpose clauses) are inflexibly subjunctive, emotive-factive and uncertainty clauses (e.g., negated indicative-normative, dubitative, modal complements) may accept indicative if the speaker intends to add the affirmative or negated proposition to the common ground; i.e., if the speaker intends to assert the complement in question. I close this argument by stating that assertion with the indicative is most likely to occur if the proposition (affirmative or negated) is informative (i.e., new, or unknown to the addressee, important, contrastive, and/or highly likely). KEY WORDS: Non-prescriptive Spanish grammar, semantics, pragmatics, syntax-semantics interface, mood variation
Content may be subject to copyright.
ã Tris Faulkner. Borealis: An International Journal of Hispanic Linguistics, 2022, 11 / 1. pp. 70-100.
https://doi.org/10.7557/1.11.1.6334
This is an Open Access Article distributed under the terms of the Creative Commons Attribution License
(https://creativecommons.org/licenses/by/4.0/legalcode) which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT
SUBJUNCTIVES*
Tris Faulkner
Kalamazoo College
ABSTRACT. Standard Spanish grammar states that desideratives (querer que), directives
(aconsejar que), purpose clauses (para que), causatives (hacer que), emotive-factives
(alegrarse de que), dubitatives (dudar que), modals (ser posible que), and negated indicative-
normative verbs (e.g., no creer que), embed subjunctive complement clauses. However, in
spite of this classification, some predicates will exhibit a certain degree of mood variation.
For instance, emotive-factives can take indicative complements (Crespo del Río 2014;
Faulkner 2021a, 2021b). Similar variability between the moods may also come about in
negated epistemic (Bolinger 1991), dubitative (Blake 1981), and modal clauses (Deshors and
Waltermire 2019). In the present paper, I propose that such variation stems from the Spanish
mood system involving a split between two types of subjunctives: one that is required in non-
realistic, preference-based contexts, and another that is the default of propositions that are at
least, somewhat realistic. I argue that, whereas non-realistic, preference-based clauses (e.g.,
desiderative, directive, and purpose clauses) are inflexibly subjunctive, emotive-factive and
uncertainty clauses (e.g., negated indicative-normative, dubitative, modal complements) may
accept indicative if the speaker intends to add the affirmative or negated proposition to the
common ground; i.e., if the speaker intends to assert the complement in question. I close this
argument by stating that assertion with the indicative is most likely to occur if the proposition
(affirmative or negated) is informative (i.e., new, or unknown to the addressee, important,
contrastive, and/or highly likely).
Keywords: non-prescriptive Spanish grammar; semantics; pragmatics; syntax-semantics
interface; mood variation
RESUMEN. La gramática española estándar establece que las construcciones desiderativas
(querer que), directivas (aconsejar que), de propósito (para que), causativas (hacer que),
factivo-emotivas (alegrarse de que), dubitativas (dudar que; no creer que), y modales (ser
posible que) subordinan cláusulas de complemento de subjuntivo. Sin embargo, a pesar de
esta clasificación, algunos predicados exhiben un cierto grado de variación modal. Por
ejemplo, los factivo-emotivos pueden llevar complementos de indicativo (Crespo del Río
2014; Faulkner 2021a, 2021b). Una variabilidad similar en modo también puede ocurrir en
las cláusulas epistémicas negadas (Bolinger 1991), dubitativas (Blake 1981) y modales
(Deshors y Waltermire 2019). En el presente artículo, propongo que tal variación surge del
sistema de modo español, que implica una división entre dos tipos de subjuntivos: uno que
se requiere en contextos no realistas, basados en preferencias, y otro que es la forma por
defecto de proposiciones que son, al menos, moderadamente realistas. Argumento que,
mientras que las cláusulas no realistas basadas en preferencias (p. ej., cláusulas desiderativas,
* This paper has benefitted from the help and comments of my advisors Drs. Elena Herburger and Paul
Portner. I sincerely appreciate all of their help over the years. I would also like to thank my family and friends
for their unceasing support. Finally, I also am grateful to the National Science Foundation (NSF), as without
their Doctoral Dissertation Research Improvement Award (DDRI), carrying out this research would not have
been possible.
TRIS FAULKNER
71
directivas y de propósito) son inflexiblemente subjuntivas, las cláusulas emotivas y de
incertidumbre (p. ej., indicativo-normativo negado, dubitativo, complementos modales)
pueden aceptar complementos indicativos si el hablante tiene la intención de agregar la
proposición afirmada o negada al terreno común; es decir, si el hablante pretende darle
aserción al complemento en cuestión. Refuerzo este argumento afirmando que es más
probable que ocurra una afirmación con el indicativo si la proposición (afirmada o negada)
es informativa (es decir, nueva o desconocida para el destinatario, importante, contrastiva o
muy probable).
Palabras clave: gramática española no prescriptiva; semántica; pragmática; interfaz sintaxis-
semántica; variación de modo
1 Introduction
In the present paper, I argue that Spanish has not one, but instead two subjunctives. The
idea that the Spanish mood system is characterized by two subjunctives is not a novel
concept. As far back as 1847, Andrés Bello put forth that Spanish had both a COMMON
and a HYPOTHETICAL subjunctive. Similarly, Gili Gaya (1980) suggested that there
exist both a POTENTIAL and an OPTATIVE subjunctive in Spanish. That being said, the
proposal that is most similar to that of the current study is Lozano (1972)
1
, which states
that Spanish has both an optative and a DUBITATIVE subjunctive. In the present article,
I argue that, whereas one of Spanish’s two subjunctives is required, the other is default and
can be overridden by the indicative. I will refer to the required subjunctive as
SUBJUNCTIVE 1, and the default, “overridable” subjunctive as SUBJUNCTIVE 2. The
differences between the two relate to both the predicates with which they occur, as well as
the (im)possibility of being substituted by the indicative. While subjunctive 1 is described
as being triggered by the core subjunctive, comparative or preference-based predicates
(e.g., desideratives, directives, and purpose clauses), subjunctive 2 is the default of
emotive-factives and verbs of uncertainty (i.e., verbs that express the speaker’s
apprehension, doubting, or refutation of the proposition in question, such as, dubitatives,
negated indicative-normative predicates, and modals). Unlike the complements in which
subjunctive 1 is used, subjunctive 2 appears in clauses that may alternate with indicative if
the speaker’s goal is to ASSERT the embedded proposition or its negation. In this way, the
affirmative or negated proposition gets added to the common ground (and, consequently,
to the addressee’s mental model). I argue that this is most likely to occur if the complement
or its negation is thought to be INFORMATIVE (i.e., new or unknown to the hearer/reader
in question, important, contrastive, and/or highly probable). As will be discussed in section
4, this delineation between the two proposed subjunctives is summarized as relating to one
main distinguishing feature: whereas the predicates that select subjunctive 1 compare a
non-realistic p or ¬p (i.e., the proposition p or its negation), the predicates that take
subjunctive 2 have complements that are at least somewhat realistic. As such, the assertion
of p or ¬p is acceptable only in the case of the latter. Emotive-factives are argued to be a
unique hybrid between the two, in that they are able to do both; assert p or ¬p and compare
p or ¬p. However, they are described as taking subjunctive 2 since the propositions they
compare are at least somewhat realistic (which, like verbs of uncertainty, makes them
1
Section 5 discusses Lozano (1972) and its connection to the current paper.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
72
accepting of assertion with the indicative). This distinction between the subjunctives is
represented in table 1 below:
Table 1: The Two Spanish Subjunctives
2
Subjunctive 1
Subjunctive 2
Contexts used
Non-realistic and preference-
based
Somewhat realistic and uninformative
Predicate types
Volitional
Verbs of uncertainty
Emotive-
factives
Desideratives
Directives
Purpose clauses
Dubitatives
Negated indicative
predicates
Modals
Emotive-
factives
Use
Required
Default
Default
Compare p or ¬p
Can assert p or ¬p
The sections to follow provide some background on the subjunctive-indicative contrast,
with the objective of arriving at the main argument that Spanish has two subjunctives. They
are organized as follows:
2) The Two Verbal Moods: Indicative and Subjunctive Section 2 introduces sub-
sections 2.1 and 2.2, which discuss the contexts in which indicative and subjunctive
are normatively used.
3) Mood Variation in Subjunctive-normative Environments Section 3 discusses
several subjunctive-normative environments in which the use of the indicative is
occasionally acceptable; i.e., emotive-factive, negated epistemic, dubitative, and
modal clauses.
4) The Required and Default Subjunctives Section 4 elaborates on the main
differences between the environments in which the two proposed subjunctives
(required and default) tend to be found: whereas subjunctive 1 is required in
volitional, preference-based complements, subjunctive 2 is the default of
uninformative, emotive-factive and uncertainty clauses.
5) Revisiting Lozano’s (1972) Two Subjunctives In section 5, I discuss how
Lozano’s (1972) +optative and ±dubitative subjunctives are both similar to and
different from the two subjunctives on which the present paper is focused.
6) Conclusion – Section 6 summarizes the points made throughout sections 1 to 5 by
reiterating that the Spanish subjunctive is not a homogenous form. In other words,
the subjunctive used in volitional contexts is not the same as that used with verbs
2
First published in Faulkner (2021b).
TRIS FAULKNER
73
of uncertainty or emotive-factives (i.e., in the case of the latter, predicates which
reveal a truth or an experience, as assessed by the speaker or matrix subject).
2 The Two Verbal Moods: Indicative and Subjunctive
Spanish, being part of the Romance language family, distinguishes between two main
categories of verbal mood: indicative and subjunctive. In accordance with standard Spanish
grammar, predicates that embed indicative complement clauses include commissives
(prometer ‘to promise’), verbs of communication (decir ‘to say’), verbs of certainty (estar
seguro/a de que ‘to be certain/sure that’), verbs that indicate a happening or occurrence
(ocurrir ‘to occur’), fiction verbs (soñar ‘to dream’), predicates of knowledge and belief
(pensar ‘to think’, creer ‘to believe’), perception predicates (sentir ‘to sense’), neutral
factive predicates (recordar ‘to remember’), and predicates of mental judgment (entender
‘to understand’). On the other hand, the nominal clauses that house subjunctive verbs
include desiderative (querer ‘to want’), directive (aconsejar ‘to advise’), causative (hacer
que ‘to make that’), emotive-factive (estar contento/a de que ‘to be happy that’), dubitative
(dudar ‘to doubt’), and modal complements (ser posible que ‘to be possible that’). The
subjunctive is also said to be triggered by negation (e.g., no creer que ‘to not believe that’;
no decir que ‘to not say that’; no estar seguro/a de que ‘to not be sure that’), as well as
required in purpose clauses (para que ‘so that’).
2.1 Normative Variation Between Indicative and Subjunctive
Although the two moods tend to be in complementary distribution, there are certain
environments in which their use will overlap. For instance, it is widely known that choice
of mood may vary in conditional clauses (Si tengo/tuviera tiempo‘If I have/had time…’),
relative clauses (Compraré el vestido que me gusta/guste ‘I will buy the dress that I
(already) like/(will) like [perhaps after looking at a few options]’), as well as with several
other non-verbal subordinators (such as, aunque ‘even though’ or ‘even if’; porque
‘because’; el hecho de que ‘the fact that’; cuando ‘when’, etc.: e.g., - Cuando me levanto,
te llamo ‘When I wake up, I (always) call you’ vs. Cuando me levante, te llamo
‘When(ever) I wake up, I (will) call you’). In addition to these environments, the Real
Academia Española ‘Royal Spanish Academy’ (2011), henceforth RAE, identifies various
selecting verbs whose complements may exhibit mood variation. They describe the
following contrasts as being the most noteworthy.
i. Communication vs. Influence: with verbs like decir ‘to say’, the indicative is used
to communicate a fact, while the subjunctive is used to present orders, suggestions,
and requests (RAE 2011: 479).
(1) Le dijo que actuaba/actuara de buena fe.
Him tell.PAST.INDIC.3SG that act.PAST.INDIC/SUBJ.3SG of good faith.
Indicative: ‘S/he/you told him/her that s/he/you acted in good faith.’
Subjunctive: ‘S/he/you told him/her/you to act with good faith.’
(RAE 2011: 479)
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
74
ii. Thought vs. Intention: whereas with verbs like pensar ‘to think’, the indicative is
used to manifest what the speaker or subject thinks or believes, the subjunctive
relays the speaker or subject’s stance towards a particular event (RAE 2011: 479).
(2) Inicialmente pensaba que
Initially think.PAST.INDIC.1SG that
participaron los dos equipos sevillanos con un tercer
participate.PAST.INDIC.3PL the two teams Sevillian with a third
rival.
rival.
‘I was initially thinking that the two Sevillian teams and a third rival participated.’
(Adapted from Razón, as cited in RAE 2011: 479)
(3) Inicialmente estaba pensando que
Initially be.PAST.INDIC.1SG think.GER that
participaran los dos equipos sevillanos con un tercer
participate.PAST.SUBJ.3PL the two teams Sevillian with a third
rival.
rival.
‘I was initially thinking that the two Sevillian teams and a third rival should
participate.’
(Razón as cited in RAE 2011: 479)
iii. Understanding vs. Evaluation or Empathy: whereas with verbs like comprender
‘to understand’, the indicative presents information that is either correct, perceived to be
correct, or new to the discourse environment, the subjunctive is used to agree with or accept
information that the other conversational participants already know (RAE 2011: 479).
(4) Comprendí que estaba equivocada.
Understand.PAST.INDIC.1SG that be.PAST.INDIC.1SG wrong.
‘I realized that I was wrong.’
(5) Comprendo que estés molesto conmigo.
Understand.PRES.INDIC.1SG that be.PRES.SUBJ.2SG annoyed with-me.
‘I admit/accept that you are annoyed with me.’
(RAE 2011: 479)
iv. Perception vs. Intention: with verbs like ver ‘to see’, the indicative reveals that
the speaker or subject has perceived or seen a particular event, while the subjunctive is
used to relay their intent to make said event occur (RAE 2011: 479).
(6) Siempre veía que cada cosa estaba
Always see.PAST.INDIC.1SG that each thing be.PAST.INDIC.3SG
en su lugar.
in its place.
‘I always saw that everything was always in its place.’
TRIS FAULKNER
75
(RAE 2011: 479)
(7) Siempre veía que cada cosa estuviera
Always see.PAST.INDIC.1SG that each thing be.PAST.SUBJ.3SG
en su lugar.
in its place.
‘I always saw to it that everything was always in its place.’
(RAE 2011: 479)
v. Assertion vs. Justification: with the indicative, the verb explicar ‘to explain’ acts
as a verb of saying (verba dicendi) and is, thus, used to communicate a particular
happening or event. With the subjunctive, on the other hand, it relays the cause of or
reason for something (RAE 2011: 479).
(8) El maestro te ha explicado que las cosas
The teacher you have.PRES.INDIC.3SG explain.PP that the things
no son como tú pensabas.
not be.PRES.INDIC.3PL as you think.PAST.INDIC.2SG.
‘The teacher explained to you that things are not as you think.’
(RAE 2011: 479)
(9) Eso explica que las cosas no sean
That explain.PRES.INDIC.3SG that the things not be.PRES.SUBJ.3PL
como tú piensas.
as you think.PRES.INDIC.2SG.
‘That explains why things are not as you think.’
(RAE 2011: 479)
vi. Affect vs. Assertion: According to the RAE (2011), although emotive-factives
generally take subjunctive clauses, variation with the indicative does also occur. They
state that this happens more frequently in Latin American, as compared to European
Spanish (p. 480). With these expressions (e.g., alegrarse de que ‘to be happy that’;
preocuparse de que ‘to be worried that’), the speaker’s use of the indicative is to
highlight the informativeness of the subordinate proposition. This idea ties in with the
main argument of the present paper, which is that, in certain normatively, subjunctive
environments, alternations with the indicative may occur if the complement or its
negation is informative (i.e., new/unknown to the addressee, important, contrastive,
and/or highly probable) (see: section 4.2).
Conversely, when the complement appears in its default, subjunctive mood, the
RAE states that the proposition’s affectivity is instead the focus (p. 480).
(10) Se quejó de que el citado
REFL complain.PAST.INDIC.3SG of that the cite.PP
individuo había propalado infundios
individual have.PAST.INDIC.3SG divulge.PP lies
por la comarca
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
76
through the region.
‘S/he complained that the cited individual had spread lies through the
region.’
(RAE 2011: 480)
(11) Se queja de que la impiedad
REFL complain.PRES.INDIC.3SG of that the impiety
pretenda corromper el estudio de las ciencias naturales.
try.PRES.SUBJ.3SG corrupt.INF the study of the sciences natural.
‘S/he is complaining that impiety may corrupt the study of the natural
sciences.’
(RAE 2011: 480)
vii. Fear vs. Suspicion: With the predicate temer ‘to fear’, the indicative puts forth
that the subject suspects that an event either will happen or has happened. However,
with the subjunctive, it is understood that the subject is afraid that the particular event
will/or has already transpire(d) (RAE 2011: 480).
(12) Me temo que va a haber dificultades.
Me fear.PRES.INDIC.1SG that go.PRES.INDIC.3SG to have.INF difficulties.
‘I fear (suspect) that there are going to be some difficulties.’
(Caballero Bonald as cited in RAE 2011: 480)
(13) Temo que algún deslenguado lo sepa
Fear.PRES.INDIC.1SG that some foul-mouthed it know.PRES.SUBJ.3SG.
‘I fear (am afraid) that some foul-mouthed person knows about it.’
(RAE 2011: 480)
Examples (1) through (13) show that, similar to the case of conditional, relative, or
adverbial clauses (etc.), variation between subjunctive and indicative occurs in the
complements of several selecting verbs and expressions. With these examples, the RAE
(2011) provides very detailed intuitions regarding how the meaning of each mood differs
depending on the predicate at hand. For instance, whereas after the expression alegrarse
de que ‘to be happy/pleased that’, the indicative’s role is to highlight the subordinate
proposition’s informativeness, following the verb temer ‘to fear’, its role is to relay
suspicion. Similarly, whereas the subjunctive after the verb explicar ‘to explain’ discloses
the reason for or cause of the proposition, after the predicate ver ‘to see’, its role is to reveal
a particular intention. However, the fact that the meaning of each mood is defined
differently based on the particular verb in question, suggests that many of these expressions
are polysemous. Thus, choice of mood with the matrix verbs in (1) to (13) appears to
distinguish the specific meaning that is intended to be portrayed. In some cases, mood
choice changes the lexical meaning of the verb in question (e.g., comprender ‘to
understand’, ver ‘to see’), and in others, the compositional meaning of the entire sentence
(e.g., pensar ‘to think’, explicar ‘to explain’).
TRIS FAULKNER
77
3 Mood Variation in Subjunctive-normative Environments
In addition to standard variation between subjunctive and indicative, are the non-
standard fluctuations between moods that occur in certain “subjunctive-requiring”,
nominal clauses. For instance, although generally inclined to taking the subjunctive,
emotive-factives (e.g., ser bueno/malo que ‘it is good/bad that’) may at times accept
indicative complements (Bolinger 1991; Blake 1981; Crespo del Río 2014; Farkas 1992b;
Faulkner 2021a, 2021b; García and Terrell 1977; Gregory and Lunn 2012; Kowal 2007;
Lipski 1978; Lope Blanch 1958; Quer 1998, 2001; RAE 2011; Silva-Corvalán 1994;
Terrell and Hooper 1974). This is said to take place in all varieties of Spanish (Crespo del
Río 2014). Similar variability with indicative also comes about with negation (e.g., no creer
que ‘to not believe that’; no estar seguro/a de que ‘to not be sure that’) (e.g., Bolinger
1991; Borrego et. al 1989; Kowal 2007; Portner, 2018; Quer 1998, 2001, 2009; Waltermire
2017), in dubitative clauses (e.g., dudar que ‘to doubt that’) (Blake 1981; Kowal 2007;
Waltermire 2017), and with modal expressions (e.g., ser posible que ‘to be possible that’)
(Borrego et. al 1989; Deshors and Waltermire 2019; Kowal 2007). The following sections
corroborate these points with examples.
3.1 Emotive-factives
3
Although regularly described as requiring the subjunctive (Alonso-Cortés 1981;
Giannakidou 2017; Gili Gaya 1960; Ruiz Campillo 2001, 2008; Villalta 2008), it is
relatively widely known that Spanish emotive-factives can also allow for indicative (Blake
1981; Crespo del Río 2014; Farkas 1992b; Faulkner 2021a, 2021b; García and Terrell
1977; Gregory and Lunn 2012; Lipski 1978; Quer 1998, 2001; Silva-Corvalán 1994; RAE
2011; Studerus 1995; Terrell and Hooper 1974). Thus, while (14) embodies their default
choice of mood, indicative complements, like that of (15), are not uncommon (Blake 1982;
Crespo del Río 2014; Terrell and Hooper 1974; RAE 2011).
(14) Es triste que se vaya tan pronto.
Be.PRES.INDIC.3SG sad that REFL leave.SUBJ.3SG so soon.
‘It is sad that s/he is leaving so soon.’
(Terrell and García 1977: 221)
(15) Me sorprendió que ustedes no pudieron
Me surprise.PAST.INDIC.3SG that you-all not can.PAST.INDIC.3PL
ayudarnos con el plan.
help.INF-us with the plan.
‘It surprised me that you couldn’t help us with the plan.’
(Terrell and García 1977: 221)
3.2 Negated Indicative-normative Predicates
A factor often claimed to trigger the subjunctive is negation (e.g., Borrego et. al 1989;
Farkas 1992b; Giorgi and Pianesi 1997; Portner 2018). It is, thus, a commonly held view
that certain indicative-normative predicates take subjunctive when negated (Portner 2018;
Quer 2001). However, although subjunctive may be the default after negation (Portner
2018), the use of the indicative is not ruled out (Bolinger 1991; Farkas 1992b; García and
3
Emotive-factives are also referred to as ‘evaluative’ predicates.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
78
Terrell 1977; Giorgi and Pianesi 1997; Quer 2001, 2009; Studerus 1995). Thus, either
subjunctive or indicative may be admissible in examples (16) and (17) to follow, which
involve negated epistemic verbs.
(16) Marta no cree que Pedro sea/es
Marta not believe.PRES.INDIC.3SG that Pedro be.PRES.SUBJ/INDIC.3SG
culpable.
guilty.
‘Marta does not believe that Pedro is guilty.’
(Borgonovo and Prévost 2003: 151)
(17) La gente muchas veces no cree que
The people many times not believe.PRES.INDIC.3SG that
puede/pueda llegar a ese nuevo nivel que Dios
can.PRES.INDIC/SUBJ.3SG arrive.INF to that new level that God
tiene para ellos.
have.PRES.INDIC.3SG for them.
‘People many times don’t believe that they can get to that new level that
God has for them.’
(Adapted from Davies’ Corpus del Español 2016)
3.3 Dubitatives
Like emotive-factives and negated indicative-normative predicates, dubitatives are also
regularly described as being governed by the subjunctive (Borrego et al. 2013; Ramsey
1956; Villalta 2008).
(18) Dudo que sea culpable.
Doubt.PRES.INDIC.1SG that be.PRES.SUBJ.3SG guilt
‘I doubt that s/he is guilty.’
However, although this may be the normative pattern, they do at times accept indicative
subordinate clauses (Kowal 2007; Quer 1998, 2001, 2009; Waltermire 2017).
(19) Dudo que tiene razón.
Doubt.PRES.INDIC.1SG that have.PRES.INDIC.3SG reason.
‘I doubt that s/he is right.’
(Fente as cited in Kowal 2007: 58)
(20) Algunos dudan que él lo escribió.
Some doubt.PRES.INDIC.3PL that he it write.PAST.INDIC.3SG.
‘Some doubt that he wrote it.’
(Davies’ Corpus del Español 2016)
(Ramsey 1956: 419)
TRIS FAULKNER
79
3.4 Modals
It is a common claim that Spanish requires that subjunctive be used after expressions
of possibility and probability (Farkas 1992b; Hooper 1975; Ramsey 1956; Villalta 2008).
Thus, if this rule were to be adhered to, only sentences such as (21) and (22) below should
be considered grammatical.
(21) Es probable que lleguen a tiempo.
Be.PRES.INDIC.3SG probable that arrive.PRES.SUBJ.3PL at time.
‘It is likely that they will arrive on time.’
(Laca 2010: 203)
(22) Es posible que hayan terminado ya.
Be.PRES.INDIC.3SG possible that have.PRES.SUBJ.3PL finish.PP already.
‘It is possible that they have already finished.’
(Terrell and Hooper 1974: 487)
However, in spite of these claims, native speakers will sometimes opt for indicative
modal complements (c.f., Borrego et. al 1989; Deshors and Waltermire 2019; García and
Terrell 1977). Examples (23) and (24) illustrate this point further.
(23) Es posible que quiere destacar
Be.PRES.INDIC.3SG possible that want.PRES.INDIC.3SG highlight.INF
que es algo que debe ser
that be.PRES.INDIC.3SG something that must.PRES.INDIC.3SG be.INF
respetado en vez de depositado en el sótano.
respect.PP instead-of deposit.PP in the basement.
‘It is possible that she wants to point out that it is something that must be
respected instead of being dumped in the basement.'
(Davies’ Corpus del Español 2016)
(24) Es probable que los precios allí serán
Be.PRES.INDIC.3SG probable that the prices there be.FUT.3PL
más bajos
more low.
‘It is probable/likely that the prices there will be lower.’
(García and Terrell 1977: 220)
As introduced in section 1, both verbs of uncertainty (negated indicative-normative
predicates, dubitatives, and modals) and emotive-factives are described as co-occurring
with the default, as opposed to the required subjunctive; i.e., subjunctive 2 as opposed to
subjunctive 1. In other words, the subjunctive that can at times be replaced by the
indicative. In section 4, I will propose that the replacement of subjunctive with indicative,
in these contexts, tends to occur when the complement is informative (new or unknown to
the addressee, important, contrastive, and/or highly probable).
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
80
4 The Required and Default Subjunctives
As discussed in sections 2 and 3, variation between subjunctive and indicative occurs
both in environments in which it is standardly allowed, as well as in contexts in which it is
traditionally unexpected. We saw, for example, instances of standard variation between the
moods in conditional, relative, and adverbial clauses, as well as with several indicative-
normative predicates (e.g., comprender ‘to understand’, pensar ‘to think’, etc.). However,
we also saw examples of mood variation that is not traditionally considered to be
grammatical (see: section 3). For instance, in spite of the subjunctive being described as
required with negation, dubitatives, and modals (e.g., Real Academia 2011; Romero 2012),
these environments do at times exhibit variability with the indicative (e.g., Blake 1981;
Borrego et. al 1989; Crespo del Río 2014; Farkas 1992b; García and Terrell 1977; Gregory
and Lunn 2012; Kowal 2007; Lipski 1978; Quer 1998, 2001, 2009; Silva-Corvalán 1994;
Studerus 1995; Terrell and Hooper 1974; Waltermire 2017). The same is the case for
emotive-factives, although their receptivity to indicative complements is more widely
recognized. In the present section, I will argue that the alternations between moods that
occur in the complements of emotive-factives and verbs of uncertainty (e.g., negated
indicative-normative verbs, dubitatives, and modals), relate to the speaker’s intent to assert
their commitment to the subordinate proposition (or its negation). In this way, the
proposition gets added to the current common ground, and the hearer’s attention is called
to its importance or value.
4
Thus, in the case of examples (15), (16), (19), and (23) (repeated
below as 25, 26, 27, and 28), the use of the indicative will be explained as signaling both
speaker commitment to the embedded (affirmative or negated) complement, as well as its
assertion.
(25) Me sorprendió que ustedes no pudieron
Me surprise.PAST.INDIC.3SG that you-all not can.PAST.INDIC.3PL
ayudarnos con el plan.
help.INF-us with the plan.
‘It surprised me that you couldn’t help us with the plan.
(Terrell and García 1977: 221)
(26) Marta no cree que Pedro sea/es
Marta not believe.PRES.INDIC.3SG that Pedro be.PRES.SUBJ/INDIC.3SG
culpable.
guilty.
‘Marta does not believe that Pedro is guilty.’
(Borgonovo and Prévost 2003: 151)
(27) Dudo que tiene razón.
Doubt.PRES.INDIC.1SG that have.PRES.INDIC.3SG reason.
‘I doubt that s/he is right.'
(Fente as cited in Kowal 2007: 58)
4
All arguments posed in this paper stem from native speaker judgments (see: Faulkner 2021b).
TRIS FAULKNER
81
(28) Es posible que quiere destacar
Be.PRES.INDIC.3SG possible that want.PRES.INDIC.3SG highlight.INF
que es algo que debe ser
that be.PRES.INDIC.3SG something that must.PRES.INDIC.3SG be.INF
respetado en vez de depositado en el sótano.
respect.PP instead-of deposit.PP in the basement.
‘It is possible that she wants to point out that it is it is something that must be
respected instead of being dumped in the basement.’
(Davies’ Corpus del Español 2016)
My hypothesis puts forth that the contexts in which assertion is most appropriate are
those wherein the speaker assumes the proposition or its negation to be informative (e.g.,
new or unknown to the addressee, important, contrastive, and/or highly probable). Thus, if
(the default) subjunctive had been used in either of the prior examples (i.e., examples 25-
28), the affirmative (e.g., Me sorprendió que x) or negated (e.g., Dudo que x) proposition
would likely have been uninformative; i.e., old/already known, unimportant, non-
contrastive, and/or highly unlikely. That is, taken for granted and, therefore, not warranting
of assertion or highlighting. It is, however, important to reiterate that this choice to use the
indicative is only possible with certain subjunctive-governing predicates. In accordance
with section 3, the examples above demonstrate that this option may be available to
speakers using emotive-factives and verbs of uncertainty (i.e., predicates that express the
speaker or subject’s apprehension, doubting, or refutation of the proposition in question,
for example, negated indicative-normative verbs, dubitatives, and modals). Contrarily, it is
unavailable to a speaker using a non-realistic, preference-based predicate (e.g.,
desideratives, directives, and purpose clauses).
In order to better understand why this may be, a good first step would be to separate
desideratives, directives, and purpose clauses from emotive-factives and verbs of
uncertainty. Let’s place the former into group 1 and the latter into group 2. If we examine
the predicates in group 1 (e.g., desideratives, directives, and purpose clauses), we see that
they make up the core, subjunctive-governing verbs and expressions, which put forth the
speaker or subject’s (dis)preferences; i.e., propositions that are non-realistic and cannot be
assumed to materialize in the actual world. These predicates compare the complement
proposition or its negation (i.e., p or ¬p) to alternative propositions. For instance, in
examples (29) through (31), the speaker or subject’s dis(preferred) outcome is that Mary
gets the job (as compared to some other end result).
(29) Desideratives (compare p or ¬p)
a) I (don’t) want Mary to get the job.
b) We (don’t) want Mary to get the job.
c) S/he (doesn’t) want(s) Mary to get the job.
d) They (don’t) want Mary to get the job.
(30) Directive Clauses (compare p or ¬p)
a) I (don’t) recommend that Mary get the job.
b) We (don’t) recommend that Mary get the job.
c) S/he (doesn’t) recommend(s) that Mary get the job.
d) They (don’t) recommend that Mary get the job.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
82
(31) Purpose Clauses (compare p or ¬p)
a) I’ve been practicing with Mary so that she (doesn’t) get(s) the job.
b) We’ve been practicing with Mary so that she (doesn’t) get(s) the
job.
c) S/he’s been practicing with Mary so that she (doesn’t) get(s) the
job.
d) They’ve been practicing with Mary so that she (doesn’t) get(s) the
job.
Thus, regardless of whether the complement reveals a desire (29), suggestion (30), or
intention (31), the subject’s (dis)preferred end result (as compared to alternative outcomes)
is what is being relayed. That is, that Mary be (or not be) the individual selected for the job
in question.
Different from the predicates in group 1, the predicates in group 2 (i.e., emotive-
factives, negated indicative-normative predicates, dubitatives, and modals) represent a
particular truth (e.g., emotive-factives) or an experience (e.g., a doubted, disbelieved,
refuted, or (im)possible proposition, as assessed by the speaker or matrix subject based on
their experiences); i.e., propositions that are potentially realistic - meaning that the speaker
assumes that they could or have materialize(d) in the actual world. Thus, whereas in
example (32), the subject’s truth is that Mary got the job, in sentences (33) through (35),
the uncertainty regarding Mary’s prospective offer (or rejection) results from the subject’s
experiences; e.g., their experiences regarding Mary’s credentials, personality, application,
potential as a worker, etc.
(32) Emotive-factives (can assert p or ¬p; compares p or ¬p)
a) I am (not) happy that Mary got the job.
b) We are (not) happy that Mary got the job.
c) S/he is (not) happy that Mary got the job.
d) They are (not) happy that Mary got the job.
(33) Dubitatives (can assert p or ¬p)
a) I doubt that Mary got the job.
b) We doubt that Mary got the job.
c) S/he doubts that Mary got the job.
d) They doubt that Mary got the job.
(34) Negated Indicative-normative Predicates (can assert p or ¬p)
a) I don’t believe that Mary got the job.
b) We don’t believe that Mary got the job.
c) S/he doesn’t believe that Mary got the job.
d) They don’t believe that Mary got the job.
(35) Modals (can assert p or ¬p)
a) It is (not/im)possible that Mary got the job.
b) It is (not/im)probable that Mary got the job.
c) It is (not/un)likely that Mary got the job.
Since we have now organized like predicates together (group 1 vs. group 2), we can
now further explain how their grouping relates to their choice(s) of mood. My hypothesis
TRIS FAULKNER
83
is that the predicates in group 1, which include the core group of subjunctive selectors,
compare a non-realistic p or ¬p. As such, they require the subjunctive and have been
deemed ASSERTION-AVERSE (see: section 4.1). Conversely, the verbs and expressions
in group 2 (i.e., emotive-factives and verbs of uncertainty) take propositions that are closer
to the actual world and are, therefore, at least somewhat realistic. Thus, their complements
(whether affirmative or negated) may occasionally appear in the indicative (i.e., p or ¬p
can be asserted). We can, therefore, confer this group of predicates the label of
ASSERTION-INCLINED (see: section 4.2). However, as stated previously, assertion is
most likely to occur when the particular complement is informative. In this way, the hearer
gets alerted to the proposition’s contextual value (i.e., its informativeness, e.g., its newness,
importance, contrastiveness, and/or high likelihood).
A question that may come up relates to why emotive-factives, which have been argued
by some linguists to have a comparison-based semantics (p or ¬p is compared) (e.g., Giorgi
and Pianesi 1997; Villalta 2008), form part of group 2. Although their semantics involves
comparison (see: example 32 above), unlike the predicates in group 1, they are able to
compare propositions that are realistic. Thus, emotive-factives are unique in the sense that,
although assuming a preference-based semantics, they act similarly to verbs of uncertainty,
in that their complements can take the indicative (see: section 4.2). Therefore, as related to
examples (32) through (35), Spanish speakers may choose to assert any of the listed
complements with indicative, especially when informative. Accordingly, whereas with
assertion-averse predicates (29-31; 36-38) the use of the subjunctive is inflexible (i.e., the
subjunctive is required), with the assertion-inclined predicates indicative may at times be
available (25-28).
(36) Quiero que todo el mundo
Want.PRES.INDIC.1SG that everyone
se calle/*se calla.
be-quiet.PRES.SUBJ.3SG/PRES.INDIC.3SG.
‘I want everyone to be quiet.’
(Borrego et. al 1989: 34)
(37) Nos aconsejaron que no nos
Us advise.PAST.INDIC.3PL that not us
acercáramos/*acercamos.
advise.PAST.SUBJ.3PL/PAST.INDIC.3PL.
‘They advised us not to come close.’
(Borrego et. al 1989: 34)
(38) La vida nos presta los años para que
The life us lend.PRES.INDIC.3SG the years for that
la disfrutemos/*disfrutamos y luego
it enjoy.PRES.SUBJ.1PL/PRES.INDIC.1PL and later
se acaba.
REFL end.PRES.INDIC.3SG.
‘Life lends us years so that we can enjoy it and later it ends.’
(Davies’ Corpus del Español 2016)
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
84
We can now argue that the differing behaviors of the assertion-averse versus the
assertion-inclined verbs and expressions, relate to a split between subjunctives. Whereas
the subjunctive used with the core subjunctive, volitional predicates, is required, the
subjunctive used in the complements of the assertion-inclined verbs and expressions is
default and can be overridden by the indicative (especially when informative).
4.1 Assertion-averse Predicates and Subjunctive 1
As mentioned in the preceding sections, assertion-averse predicates require the use of
the subjunctive and, thus, have complements that take subjunctive 1. Following many of
the ideas discussed by Giorgi and Pianesi (1997) and Villalta (2008), I assume that
predicates which select for subjunctive 1 (e.g., desideratives, directives, and purpose
clauses) have a comparison- or preference-based semantics. This means that the matrix
predicate introduces a complement which denotes an ordering of possibilities or a
comparison of alternatives. Whether or not this comparativeness is one-to-one (e.g., the
subject’s desire for Mary to get the job versus not getting the job, as related to example 39
below) or related to various contextual alternatives (e.g., Mary getting this particular job
vs. getting another position elsewhere vs. getting another position at the same location, see:
example 39), the purpose of the complement is to express the subject’s (dis)preferred end
result; i.e., p or ¬p is compared to some contextual alternative. This is the case regardless
of the tense that is elected. Whether the proposition is located in the present or the past, the
complement expresses the speaker or subject’s (dis)favored outcome. Thus, in the case of
examples (39), (40), and (41) to follow, each volitional complement discloses the subject’s
desire for Mary to have gotten the job.
(39) Desiderative Clauses
e) I want(ed) Mary to get the job.
f) We want(ed) Mary to get the job.
g) S/he wants(ed) Mary to get the job.
h) They want(ed) Mary to get the job.
(40) Directive Clauses
e) I recommend(ed) that Mary get the job.
f) We recommend(ed) that Mary get the job.
g) S/he recommends(ed) that Mary get the job.
h) They recommend(ed) that Mary get the job.
(41) Purpose Clauses
e) I practiced with Mary so that she get/got the job.
f) We practiced with Mary so that she get/got the job.
g) S/he practiced with Mary so that she get/got the job.
h) They practiced with Mary so that she get/got the job.
According to Portner and Rubinstein (2020), ‘wanting’ is not understood to implicate
desires that the experiencer is committed to and prepared to defend (p. 14). They state that:
[b]eing committed to a priority entails that the priority has a chance of materializing both
circumstantially, given relevant facts, and preferentially, given other priorities. In other
words, one cannot be committed to inconsistent priorities, and one cannot be committed to
priorities that are circumstantially unrealistic (p. 15). Related to this point, is what Giorgi
TRIS FAULKNER
85
and Pianesi (1997) say about desires and wanting. They argue that bouletic backgrounds
are non-realistic, since we cannot assume that anyone’s desires are realized in the actual
world (Giorgi and Pianesi 1997: 212).
With the previous points in mind, let us now assume that all assertion-averse predicates
have complements that report a (dis)preference, desire, requirement, command, instruction,
suggestion, dissuasion, disapproval, prohibition, or intention. Thus, regardless of whether
the matrix subject wants/recommends/requires/prohibits/does something so that p or ¬p (or
wanted/recommended/required/prohibited/did something so that p or ¬p), the complement
proposition is non-realistic, since it cannot be assumed that their desire will materialize in
the actual world. This explains why indicative is disallowed in the complements of the
predicates that take subjunctive 1. Similar to what Quer (2009) states about conditional
statements, I argue that the indicative can only appear in complements that represent
propositions that are close enough to the actual world.
5
Conversely, complements which
represent propositions that are distant or disjoint from the current world (non-realistic),
require the subjunctive. Therefore, the complements of the subjunctive-1-taking predicates
can never be asserted with indicative, since the matrix subject is not able to commit
themself to situations that are far away from the common ground; i.e., desires or
(dis)preferences, which cannot be assumed to be realistic.
In accordance with the prior arguments, my hypothesis puts forth that assertion is only
likely to take place when the proposition is at least somewhat realistic; i.e., propositions
that the speaker or subject can commit to since they may materialize in the actual world.
In these cases, the complement will report a particular truth (factivity) or experience
(uncertainty, doubt, possibility), as related to the speaker or matrix subject. Such
propositions are close to the current world and can, thus, be asserted so that they get added
to the common ground. As will be discussed further in section 4.2, this is most likely to
occur when the proposition is informative (new, important, contrastive, and/or highly
likely). This appears to be why examples (42), (43), and (44) to follow have been deemed
unacceptable. Desires are too far away from the common ground and, thus, too different
from the hearer’s mental model to be asserted with the indicative.
(42) Desiderative
[Comment extracted from a blog focused on the differences between men and women.
The blogger in question is sharing information about his life as a newlywed. This is a
new blog, so the readers don’t know anything about his relationship, neither do they
know the woman.]
6
*Mi mujer prefiere que el mensaje es
My lady prefer.PRES.INDIC.3SG that the message be.PRES.INDIC.3SG
comunicado de tantas maneras creativas como sea
communicate.PP of so-many ways creative as be.PRES.SUBJ.3SG
5
Quer (2009) states that, “indicative appears in protases that are realistic in the sense that they quantify over
worlds close enough to the actual one […]. By contrast, subjunctive conditional antecedents quantify over
worlds which are more distant from the actual one (present non/factual/counterfactual) […] or even disjoint
from the actual one (past non-factual/counterfactual) […]” (p. 1780).
6
All contexts that precede Spanish examples are invented and were extracted from Faulkner (2021b); i.e.,
the contexts in parentheses that precede examples (42), (43), (44), (47), (48), (50), (51), (52), (56), and (57).
However, all examples following these contexts are authentic. Additionally, as mentioned in Footnote 4, all
arguments regarding the acceptability of each contextualized item stem from native speaker consultations.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
86
posible.
possible
Por eso existen las florerías y negocios de tarjetas.
for that exist.PRES.INDIC.3PL the flower-shops and businesses of cards.
‘My wife prefers that the message be communicated in as many creative ways as
possible. For this reason flower shops and card businesses exist.’
(43) Directive
[Extract from a medical website. A doctor is informing readers that they shouldn’t use
homemade medicines. The readers don’t know how dangerous it is to use homemade
treatments.]
*Aconsejamos que no hacen nada casero.
Advise.PRES.INDIC.1PL that not make.PRES.INDIC.3PL nothing homemade
Es mejor consultar a su centro de salud en caso de
Be.PRES.INDIC.3SG better consult.INF to your center of health in case of
cualquier enfermedad.
any illness.
‘We advise that you don’t make anything homemade. It is better to consult with your
health center in the case of any illness.’
(44) Purpose
[Comment taken from a blog titled: “Plants that help us cure cancer.” The blogger in
question is presenting her new blog to readers who had never visited it. She’s sharing
for the first time the reason for which she created it.]
*Hola a todos, acabo de crear este blog
Hello to everyone, finish.PRES.INDIC.1SG of create.INF this blog
por casualidad. Mi intención es hacer difusión de todo lo
by chance My intention be.PRES.INDIC.3SG make.INF diffusion of all the
increíble de las plantas para que puede llegar esta
incredible of the plants for that can.PRES.INDIC.3SG arrive.INF this
información a mucha gente.
this information to many people.
‘Hello everyone, I just finished creating this blog by chance. My intention is to let
everyone know how incredible plants are so that this information can get to many
people.’
Since their complements are assumed to be non-realistic, no matter how informative
they may be, variation with the indicative (the assertion of the subordinate proposition) is
never possible; i.e., they are assertion-averse. Subjunctive-1-taking predicates require the
subjunctive so as to signal their propositions’ distance from the common ground.
It may, nonetheless, be argued that there are circumstances in which the assertion of a
volitional statement is possible. For instance, in example (45) below, y lo hará ‘and she
will’ is an assertion. However, different from venga ‘come’, which forms part of the
desiderative complement, y lo hará can stand on its own. In other words, ‘John wants Mary
to come and she will’ is a compound sentence, made up of two independent clauses; i.e.,
‘John wants Mary to come’ and ‘She will’. In the case of the present paper, my claim is
TRIS FAULKNER
87
that it is the complement of the volitional predicate that cannot be asserted, as opposed to
a following conjunctive sentence.
(45) Juan quiere que María venga y
John want.PRES.INDIC.3SG that Mary come.PRES.SUBJ.3SG and
lo hará.
it do.FUT.3SG
‘John wants Mary to come and she will.’
4.2 Assertion-inclined Predicates and Subjunctive 2
As mentioned in sections 1 and 4.1, assertion is most likely with complements that
represent events that are close enough to the actual world. Events that are close to or
coincide with the common ground include those that:
i. are presupposed to be true (factive),
ii. are negated, doubted, or refuted (as assessed by the speaker or subject, based
on their experiences), or
iii. are (im)possible (as assessed by the speaker or subject, based on their
experiences).
This is why Spanish has emotive-factives and verbs of uncertainty (negated indicative-
normative predicates, dubitatives, and modals) sometimes taking the indicative. With these
verbs, the speaker is able to assert their commitment to the complement (or its negation)
since it may denote a proposition that is realistic (i.e., close to the actual world and, thus,
close to or coinciding with the common ground). In asserting the subordinate clause, p or
¬p gets incorporated into the set of propositions that the speaker and hearer have accepted
during the course of the conversation. Thus, the proposition’s content is added to the
communicative context, with the possible effect that it influences the direction of
subsequent conversation (see: Stalnaker (2002)).
Since emotive-factives and expressions of uncertainty have subjunctive as their default
(i.e., not required and, thus, overridable by the indicative), in asserting and, consequently,
adding the proposition to the common ground, the hearer’s attention gets called to it
newness, importance, contrastiveness, and/or high likelihood. Put differently, the use of the
indicative can signal the embedded proposition’s informativeness. Whereas with emotive-
factives, this is the case when the complement is assumed to be new to the hearer or reader,
with negated and dubitative predicates, both the addressee’s presumed unfamiliarity with
the negated proposition, as well as its contrastiveness (and/or importance), may cause the
speaker to opt for assertion with the indicative. In the case of modal predicates, we will see
that informativeness relates to the proposition’s likelihood. The sections to follow detail
this further.
4.2.1 Emotive-factive Clauses
Emotive-factive or evaluative predicates introduce the attitude of the speaker or matrix
subject towards a particular event (Becker 2010; Portner 2018). Evaluative predicates fall
under the factive classification since the truth of their complements is presupposed (e.g.,
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
88
Farkas 1992b; Portner 2018). Thus, in the example (46) below, Mary is assumed to have
gotten the job.
(46) Me alegra que María haya
Me make.happy.PRES.INDIC.3SG that Mary have.PRES.SUBJ.3SG
conseguido el trabajo.
get.PP the job.
‘I am happy that Mary got the job.’
In addition to their factivity, emotive-factives have also been analyzed as having a
comparative (preference-based) semantics (e.g., Faulkner, 2021a; Giorgi and Pianesi 1997;
Villalta 2008). For instance, if we look again at example (46), we can infer that the matrix
subject’s happiness stems from their evaluation of possible alternatives; e.g., The fact that
Mary got the job is better than or preferred to the possibility of her not having been offered
the job.
As a result of their factivity, emotive-factives tend to be used to communicate old or
known information (e.g., Gregory and Lunn 2012; Lunn 1989; Quer 2001; Ruiz Campillo
2001, 2008; Sessarego 2016). This means that in using an emotive-factive, the speaker
assumes that their addressee is already in the know about what the complement affair
entails. As Faulkner (2021a) puts it: It would be unusual […] for John to tell Tim that he’s
happy that Mary got married, if Tim did not share in the knowledge that Mary had ever
been engaged, or that she had ever been in a relationship (or even worse, if Tim did not
know who Mary was) (pp. 9-10).
The fact that they usually relay old information directly relates to their default use of the
subjunctive. One of the uses of the subjunctive is to communicate information that is of
low value (e.g., old or known information) (e.g., Faulkner 2021a, 2021b; Gregory and Lunn
2012; Lunn 1989; Mejías-Bikandi 1998; Quer 2001; Sessarego 2016). However, although
this is normatively the case, there are some situations in which the conversational
participant(s) will be unfamiliar with the topic of the emotive-factive complement. In these
cases, the speaker may consider the embedded proposition to be informative since it
involves information that is new to the addressee(s) at which it is directed. In such
instances, the speaker may opt for the indicative in order to assert and, consequently, add
the content of the factive complement to the common ground. In this way, the hearer is
able to incorporate the new proposition into their mental model. Thus, whereas the
indicative may be acceptable in a context like that of (47), where the information being
discussed is new to the addressee(s) in question, it is unlikely to be acceptable in (48), a
proposition that is described as already being known to the hearer or reader (i.e., old
information).
(47) ‘Informative’ Emotive-factive Complements
[Comment extracted from an internet forum titled: Daily menu for babies from 6 to 9
months. This mother is speaking about the eating habits of her child. The information
she is sharing is new to the readers.]
Hola, mi bebe tiene 7 meses, está bien en
Hello, my baby has.PRES.INDIC.3SG 7 months, be.PRES.INDIC.3SG well in
TRIS FAULKNER
89
el peso y el tamaño para su edad, pero me preocupa
the weight and the size for his age, but me worry.PRES.INDIC.3SG
que no le agrada mucho la comida.
that not him please.PRES.INDIC.3SG much the food.
Todavía toma leche materna.
Still take.PRES.INDIC.3SG milk maternal.
‘Hello, my baby is 7 months, he is good in weight and size for his age, but it worries
me that food doesn’t please him much. He is still on breast milk.’
(48) ‘Uninformative’ Emotive-factive Complements
[Comment extracted from a blog about religion. The man in question is asking his
pastor a question after having spoken to him many times about his relationship with his
girlfriend.]
??Tengo ahora más de 3 años saliendo con mi novia
Have.PRES.INDIC.1SG now more of 3 years date.GER with my girlfriend
y como ya sabes, es atea
and as already know.PRES.INDIC.2SG be.PRES.INDIC.3SG atheist.
¿Es malo que salgo con ella?
Be.PRES.INDIC.3SG bad that go-out.PRES.INDIC.1SG with her?
‘I’ve been with my girlfriend for 3 years now and as you know, she’s atheist. Is it bad
that I’m dating her?’
It is clear that the proposition in example (47) is informative (i.e., new to the addressee),
both through the contextual details, as well as through the subject’s use of the greeting
Hola ‘Hello’. In contrast to this, the proposition of example (48) is explicitly described as
being known to the addressee(s) at which it is directed; i.e., uninformative → old. It is in
situations (like that of 47), in which the speaker is sharing a factive proposition that is not
assumed to already be known to the hearer or reader, that assertion with the indicative may
be appropriate. In contexts like that of (48), where the speaker is discussing information
that is familiar to the conversational participants, variability with the indicative is unlikely.
In uninformative contexts (48), the use of the subjunctive is the default since the
complement is assumed to contain information that need not be added to the common
ground; i.e., information that is presumed to already be shared amongst the members of the
conversational environment.
4.2.2 Negated Indicative-normative Predicates
Negated epistemic complements are one example of a negated context in which the
subjunctive is considered the norm. They relay the speaker or matrix subject’s doubts
regarding a particular situation or happening. Thus, in example (49) below, the matrix
subject John is negating the possibility that Mary was offered the position. Based on John’s
experiences (e.g., through knowing Mary, how the job market functions, or about the
particular establishment that Mary applied to, etc.), the likelihood that Mary got the job is
low. In other words, to John, it is unlikely that Mary got the job.
(49) Juan no cree que María haya
Juan not believe.PRES.INDIC.3SG that Mary have.PRES.SUBJ.3SG
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
90
conseguido el trabajo.
got.PP the job
‘John doesn’t believe that Mary got the job.’
As was previously stated, in many cases, negation is considered a trigger for the
subjunctive (e.g., Borrego et. al 1989; Farkas 1992b; Giorgi and Pianesi 1997; Portner
2018; Quer 1998, 2009). The general pattern is, therefore, that negated epistemics take
subjunctive complement clauses. However, as seen below, they may allow for the
indicative, especially if the speaker’s goal is to assert a rejected or refuted, previously
mentioned statement. In this way, the negated proposition gets incorporated into the
common ground. For instance, in example (50) to follow, the presence of an indicative
negated epistemic complement serves to assert a contrasting opinion; i.e., the speaker
rejects the idea that the matter relates to the exploitation of gas. In example (51), the
contrastiveness of the negated proposition is also made explicitly clear via assertion with
the indicative. In rejecting the idea that all men are equal, the speaker uses the contrastive
yo ‘I’ in order to point out that their opinion differs from a previously mentioned statement.
The assertion of the negated complement with the indicative, thus, ensures that the new
and contrastive idea gets added to the current common ground.
(50) ‘Informative’ Negated Epistemic Complements
[Comment taken from a blog about the exploitation of gas. The readers are seeing for
the first time the president’s response regarding the issue.]
Hay gente que trata de politizar este asunto.
Have.PRES.INDIC.3SG people that try.PRES.INDIC.3SG politicize.INF this issue.
No es verdad que tiene que ver con
Not be.PRES.INDIC.3SG truth that have.PRES.INDIC.3SG-to-do with
explotación petrolera, sostuvo el presidente.
exploitation gas, maintain.PAST.INDIC.3SG the president.
‘There are people who try to politicize this issue. It is not true that it has to do with gas
exploitation, maintained the president.’
(51) ‘Informative’ Negated Epistemic Complements
[Comment extracted from a blog titled: I like being a woman. The blogger in question
is informing readers about something she read regarding how men treat women
nowadays. The readers don’t know anything about the issue.]
Yo no pienso que los hombres son todos
I not think.PRES.INDIC.1SG that the men be.PRES.INDIC.3PL all
iguales. Hay hombres que quieren
equal. Have.PRES.INDIC.3PL men that want.PRES.INDIC.3PL
casarse y que quieren abrirnos puertas.
get-married.INF and that want.PRES.INDIC.3PL open.INF-us doors.
‘I don’t think that all men are equal. There are men who want to get married and who
want to open doors for us.’
This is why a statement such as (52) to follow would be dispreferred. Since (52) involves
a negated proposition that already formed part of the common ground (i.e., the fact that it
TRIS FAULKNER
91
may be unnecessary to carry out the particular spending ‘today’), assertion with the
indicative is likely to be deemed redundant or unnecessary. Indicative, negated epistemic
complements are most likely to be acceptable when the speaker intends to add a new,
contrastive perspective to the discourse environment.
(52) ‘Uninformative’ Negated Epistemic Complements
[Comment extracted from a blog about religion. The author of the comment is
reiterating the fact that all of the blog’s participants are friends, even if they have
different points of view. The blog’s readers are already well-acquainted with this
perspective.]
??No creo que los que estamos en protesta
Not believe.PRES.INDIC.1SG that them that be.PRES.INDIC.1PL in protest
somos enemigos de los que están de acuerdo.
be.PRES.INDIC.1PL enemies of them that be.PRES.INDIC.3PL of agreement.
‘I don’t believe that those that are in protest are enemies of those who are in agreement.’
4.2.3 Dubitative Clauses
Like negated epistemic clauses, dubitative complements present a speaker or matrix
subject’s doubts regarding a particular situation or happening. Put differently, they reveal
the belief that the subject holds concerning the likelihood of the proposition being
discussed. For instance, in example (53) below, the matrix subject John is negating the
possibility of Mary having been offered the position. Thus, according to John’s experiences
(e.g., having seen Mary’s application or applied to the same job himself, etc.), Mary likely
did not get the job.
(53) Juan duda que María haya conseguido el
John doubt.PRES.INDIC.3SG that Mary have.PRES.SUBJ.3SG gotten the
trabajo.
job.
‘John doubts that Mary got the job.’
Although Spanish dubitatives have the subjunctive as their default, speakers may opt
for indicative in order to call the hearer’s attention to the negated proposition (i.e., by
adding it to the common ground). For instance, in example (54) below, the indicative is
considered acceptable if the speaker intends to make known their disagreement with a
previously mentioned statement (e.g., Smead, 1994). Thus, the use of the indicative signals
the contrastive effect of the doubted proposition.
(54) Dudo que el profeta vuelve.
Doubt.PRES.INDIC.1SG that the prophet return.PRES.INDIC.3SG
(implica que alguien ha
(implies.PRES.INDIC.3SG that someone have.PRES.INDIC.3SG
afirmado previamente: “el profeta vuelve.”)
affirm.PP previously: “the prophet return.PRES.INDIC.3SG.”)
I doubt that the prophet is returning: implies that someone had previously
mentioned that the prophet was returning.’
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
92
(Bell 1990: 92 as cited in Smead 1994)
In using an indicative, dubitative complement, the speaker does not take for granted that
their disagreement is background information that the hearer already knows. Alternatively,
their intent is to assert the negated proposition so that it gets added to the hearer’s mental
model. If the subjunctive were instead to be used (55), it would not necessarily be inferred
that it had previously been affirmed that the prophet was returning. Contrastiveness would,
therefore, not inevitably be manifested.
(55) Dudo que el profeta vuelva.
Doubt.PRES.INDIC.1SG that the prophet return.PRES.SUBJ.3SG
‘I doubt that the prophet is returning.’
The proposition in (54) can, thus, be considered informative in that its contrastiveness
foregrounds its newness (i.e., it brings a new perspective to the discourse context the
negation or refutation of a previously mentioned belief) and assumed importance to the
speaker. This is also why the indicative may appear in a sentence like (56) to follow. In
this particular example, the speaker indicates in two ways that the dubitative statement is
contrastive. The first means by which they do this is through the use of La verdad es que
‘The truth is that’, which naturally implies some contradiction. The second, is through the
addition of the follow-up statement: Sin embargo, yo no estoy de acuerdo con este punto
‘However, I am not in agreement with this point’. In the follow-up statement, the speaker
expresses their disagreement with the previous answer they provided by means of the
contrastive yo ‘I’, the use of sin embargo ‘however’, as well as through the explicit
mentioning of their disagreement with the point – yo no estoy de acuerdo… ‘I am not in
agreement’. What is inferred from this statement is, thus, that a previous point was made
regarding the presumed longevity of email marketing. Said statement was then followed
by the contrastive dubitative proposition which refuted the claim that said modality would
be long-lasting. Finally, in order to show their agreement with the implied previous
statement (that email marketing would have a lengthy life), the speaker follows up by
affirming their disagreement with the negated proposition. Thus, like example (54), the
dubitative proposition of (56) is informative, in that it is contrastive and, thus, brings
something new to the addressee’s mental model.
(56) ‘Informative’ Dubitative Complements
[Opinion taken from an internet forum that has to do with marketing. The computer
specialist that wrote this comment is informing the readers of the weak aspects of email
marketing.]
La verdad es que muchos dudan que el
The truth be.PRES.INDIC.3SG that many doubt.PRES.INDIC.3PL that the
email marketing va a seguir vivo. Sin embargo, yo no
email marketing go.PRES.INDIC.3SG to continue alive. However, I not
estoy de acuerdo con este punto.
be.PRES.INDIC.1SG of agreement with this point.
‘The truth is that many doubt that email marketing is going to stay alive. However, I
do not agree with this point.’
(Adapted from Bell 1990: 92 as cited in Smead 1994)
TRIS FAULKNER
93
Both (54) and (56) demonstrate that the speaker’s choice to use an indicative, dubitative
complement may result from the intent to bring the contradicted proposition into the
common ground. Therefore, in a context like that of (57) to follow, we can assume that the
indicative would be redundant since assertion is unnecessary. It is only when the negated
complement contrasts an implied or explicit previous assertion that the speaker is likely to
opt for the indicative.
(57) ‘Uninformative’ Dubitative Complements
[Extract from an email about the changes that should be implemented in order to
increase the number of readers of an online magazine. The author is directing this
message to the creators of the magazine, so everyone (the addressees) already knows
that there had been little interest in a certain recently published interview.]
??Dudo que muchos lectores tenían interés
Doubt.PRES.INDIC.1SG that many readers have.PAST.INDIC.PL interest
en la entrevista con ese artista.
in the interview with that artista.
‘I doubt that many readers had interest in the interview with that artist.’
4.4 Modal Clauses
Modal meanings differ based on their STRENGTH (Portner, 2018). For instance,
whereas must and it is necessary that are strong modals, may and it is possible that would
be characterized as weak (Portner, 2018). Strong modals tend to be called NECESSITY
modals, while weak ones are deemed POSSIBILITY modals (p. 12). Possibility modals
such as ser posible que ‘to be possible that’ reflect the speaker or subject’s uncertainty
towards the subordinate proposition. Because they express uncertainty, they tend to be
described as requiring subjunctive complement clauses (e.g., Romero, 2012).
(58) Es posible que María haya
Be.PRES.INDIC.3SG possible that Mary have.PRES.SUBJ.3SG
conseguido el trabajo.
get.PP the work.
‘It is possible that Mary got the job.’
However, Spanish speakers may opt for indicative, modal clauses if the aim is to assert
their commitment to the embedded complement. Unlike emotive-factives and the other
verbs of uncertainty (negated indicative-normative predicates and dubitatives), the use of
the indicative in a possibility clause serves primarily to add a proposition that the speaker
or subject is confident about (more likely or realistic) to the common ground. For instance,
in the three uses of es posible que ‘it is possible that’ to follow, all taken from the same
source, only one clause appears in the indicative.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
94
(59) Cualquier persona que está presionando demasiado
Whichever person that be.PRES.INDIC.3SG pressure.GER too-much
a. es posible que tiene otras intenciones
be.PRES.INDIC.3SG possible that have.PRES.INDIC.3SG other intentions
diferentes a los sentimientos que tú tienes o
different to the feelings that you have.PRES.INDIC.2SG or
b. es posible que esté pensando
be.PRES.INDIC.3SG possible that be.PRES.SUBJ.3SG think.GER
en ejecutar una estafa que acabará con usted
on carry-out.INF a fraud that end.FUT with you
perder su propio dinero.
lose.INF your own money
‘Whoever is too pressuring, it is possible that s/he has different intentions
from the feelings that you have or it’s possible that s/he is thinking about
carrying out a scam that will end with you losing your money.’
(Davies’ Corpus del Español, 2016)
c. Es posible que la persona trabaje
Be.PRES.INDIC.3SG possible that the person work.PRES.SUBJ.3SG
por teléfono en un país extranjero, tratando de encontrar formas de
by telephone in a country foreign, try.GER of find.INF forms of
obtener una tarjeta verde o acceder a su cuenta de banco o
obtain.INF a card green or access.INF to your account of bank or
un vuelo a su país, o el dinero para un vuelo que nunca
a flight to your country, or the money for a flight that never
va a usar.
go.PRES.INDIC.3SG to use.INF
‘It is possible that the person works by phone in a foreign country, trying to find
ways to obtain a green card or access your bank account or a flight to your country,
or the money for a flight that s/he is never going to use.’
(Davies’ Corpus del Español 2016)
What is observed is that, whereas uses b and c describe more inconsistent, less probable
situations, use a relays a proposition that is highly likely. Use b discusses the possibility
that an online individual who messages frequently may be trying to swindle the addressee
out of their money. Use c talks about the likelihood that the person works by phone in a
foreign country, with the intention of scamming the addressee out of a green card, their
bank account information, or money for a flight. The modal expression in use a, however,
states that it is possible that the party doing the pressuring has intentions that differ from
or do not coincide with those of the addressee; i.e., the addressee’s intention to seek out a
romantic relationship. Therefore, in comparison to uses b and c, the likelihood that the
proposition of use a will materialize is higher, which, consequently, makes it more realistic.
In other words, the complement of the first modal clause depicts an event that has a higher
likelihood of being true or transpiring. This makes the proposition a more suitable
TRIS FAULKNER
95
candidate for the speaker to commit to and assert. Another interesting example is that of
(60) below. In this case, the use of the indicative seems to relate to the announcing of a
new possibility.
(60) Es posible que va a decir
Be.PRES.INDIC.3SG possible that go.PRES.INDIC.3SG to say.INF
que usted es una persona mala persona…
that you be.PRES.INDIC.3SG a person bad person
‘It is possible that s/he is going to say that you’re a bad person.’
(Davies’ Corpus del Español 2016)
Thus, with respect to modal clauses, it appears that the use of the indicative primarily
has to do with the commitment or confidence that the speaker has regarding the truth of the
complement. The more likely the proposition is to happen (the closer it is to the actual
world), the more possible assertion with the indicative becomes. Accordingly, in these
cases, the informativeness of the proposition pertains to its higher likelihood.
5 Revisiting Lozano’s (1972) Two Subjunctives
In the previous sections, we proposed that Spanish has both a required and a default,
overridable subjunctive. However, as mentioned in section 1, the idea that Spanish has a
division between subjunctives is not a novel concept (e.g., Bello 1847; Gili Gaya 1960;
Lozano 1972; Quer 1998, 2010). The theory most related to the present approach is that of
Lozano (1972). Lozano explains that there are two main features of the Spanish
subjunctive, [+optative] and [±dubitative], each of which can later be split into several
other sub-categorizations (p. 76).
According to Lozano (1972), the category [+optative] “obligates subjunctives and is
related to the semantic characteristics of volition, intent, persuasion, or obligation” (p. 76).
The sub-categories that fall under [+optative] include optative-imperative (e.g., pedir que
‘to request that’; ordenar que ‘to order that’), optative-impersonal (e.g., es importante que
‘it is important that’; es mejor que ‘it is better that’), and optative-emotional (e.g., me alegra
que ‘it pleases me that’; querer que ‘to want that’). When it comes to the feature
[±dubitative], Lozano states that it “may or may not obligate subjunctives and is related to
[the] semantic characteristics of doubt, unreality, and probability” (p. 76). Under the
[±dubitative] main umbrella are dubitative-personal (e.g., dudar que ‘to doubt; negar que
‘to deny that’) and dubitative-impersonal predicates (e.g., es posible que ‘it is possible
that’; es dudoso que ‘it is doubtful that’). In terms of the differences between the two
categories, Lozano affirms that they behave dissimilarly under negation (p. 77). While the
feature [+optative] ‘always’ requires subjunctives, “the feature [±dubitative] may or may
not obligate subjunctives if the matrix verb in the main clause is preceded by the preverb
no
7
” (p. 77). He provides the following sentences as examples of this distinction.
7
It is important to point out that negation in Spanish may take forms other than that of the use of the
preverb no; for example, nadie ‘no one’, nunca ‘never’, etc.
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
96
Table 2: The Optative-Dubitative Distinction According to Lozano (1972)
8
THE FEATURE [+OPTATIVE]
THE FEATURE [±DUBITATIVE]
(61) Quiero que lo hagas.
Want.1SG that it do.SUBJ.2SG
‘I want you to do it.’
Dudo que lo hagas.
Doubt.1SG that it do.SUBJ.2SG
‘I doubt that you’ll do it.’
(62) No quiero que lo
Not want.1SG that it
hagas.
do.SUBJ.2SG
‘I don’t want you to do it.’
No dudo que lo hagas.
Not doubt.1SG that it do.SUBJ.2SG
‘I don’t doubt that you’ll do it.’
(63) *No quiero que lo
Not want.1SG that it
haces.
do.INDIC.2SG
*‘I don’t want you to do it’
No dudo que lo haces.
Not doubt.1SG that it do.INDIC.2SG
‘I don’t doubt that you’ll do it.’
(64) *Quiero que lo haces.
Want.1SG that it do.INDIC.2SG
*‘I want you to do it.’
*Dudo que lo haces.
Doubt.1SG that it do.INDIC.2SG
*‘I doubt that you’ll do it.’
Similar to the present study, Lozano’s (1972) two subjunctives include one that is
volitional and required (‘optative’ desideratives and directives), and another that is
somewhat variable (‘dubitative’ dubitatives, negated epistemics, verbs of possibility).
Whereas optative predicates reject any variability with indicative, dubitative expressions
may allow for the indicative when negated. Thus, while Lozano’s optative subjunctive is
obligatory, his dubitative subjunctive is default. However, different from the present paper,
Lozano’s variably subjunctive predicates (‘dubitatives’) do not incorporate emotive-
factives. Instead, he positions evaluative and impersonal expressions under the optative
group of predicates, with which alternations with indicative do not occur. Additionally,
although Lozano mentions that mood variation occurs with negated dubitatives (e.g., no
dudar que ‘to not doubt that’, no pensar que ‘to not think that’), he claims that it is
disallowed with affirmative verbs of doubting (e.g., dudar ‘to doubt’). Thus, the
complements of both affirmative dubitatives and emotive-factives are described as
prohibiting the use of the indicative.
Contradicting Lozano’s (1972) claims are, however, the findings discussed in section
4.2; both emotive-factive (4.2.1) and (affirmative) dubitative clauses (4.2.3) may accept
the indicative. Whereas with emotive-factives, this is most likely to occur if the speaker
intends to assert a proposition that is addressee-new, with dubitatives, it may occur if the
speaker intends to assert a negated complement that is new and/or contrastive. Thus, in
both cases, the informativeness of the proposition increases the likelihood that the speaker
will opt for assertion. Accordingly, unlike Lozano, who describes variation between moods
8
First published in Faulkner (2021b).
TRIS FAULKNER
97
as stemming solely from the relationship between predicate type (optative vs. dubitative)
and negation, the current theory proposes that it results from the interaction between
predicate type (group 1, subjunctive-1-selecting predicates vs. group 2, subjunctive-2-
taking predicates), assertion, and (un)informativeness. In this way, all subjunctive
predicates with which variation may occur are accounted for and grouped together, whether
affirmative or not; i.e., group 1 non-realistic, preference-based predicates (e.g.,
desideratives, directives, and purpose clauses), and group 2 emotive-factives and verbs
of uncertainty (i.e., negated indicative-normative predicates, dubitatives, and expressions
of possibility).
6 Conclusions
In this paper, I argued that Spanish has two subjunctives; subjunctive 1 and subjunctive
2. We stated that subjunctive 1 is selected by the core, subjunctive, volitional predicates
(e.g., desideratives, directives, and purpose clauses), while subjunctive 2 is found in the
complements of emotive-factives and verbs of uncertainty (negated indicative-normative
predicates, dubitatives, and modals). Subjunctive 1 is licensed by predicates which cannot
assert p or ¬p since the propositions contained in their complements are non-realistic. That
is, propositions which cannot be assumed to materialize in the actual world. Subjunctive 2,
on the other hand, co-occurs with predicates whose complements can depict propositions
that are at least somewhat realistic. Propositions that are somewhat realistic are those which
are not far from the common ground. I argued that these include situations that are tied to
truth or experience (as related to the speaker or matrix subject). Whereas emotive-factives
are tied to truth, since their complements are presupposed (e.g., ‘She is happy that Mary
got the job’ presupposes the truth of Mary getting the job), verbs of uncertainty are used to
depict the likelihood of the proposition, as assessed by the speaker or matrix subject’s
experiences (e.g., ‘I doubt that/I don’t believe that/It is (im)possible that Mary got the job’,
given what I know about Mary or the person with whom she interviewed). With these types
of complements, assertion with the indicative is possible. We, thus, described subjunctive
2 as the default of predicates which can assert p or ¬p. As we have already pointed out,
emotive-factives are a hybrid case, in that they can both assert p or ¬p and compare p or
¬p. However, in spite of their hybridity, we placed them in group 2 (subjunctive-2-taking
predicates), since, like verbs of uncertainty, their propositions may be realistic (thus,
making it possible for subjunctive to be replaced by indicative). Finally, we saw that in
many cases, assertion with the indicative is influenced by the informativeness of the
complement (how new, important, contrastive, and/or how likely the proposition is). The
end result of the affirmative or negated complement’s assertion is that it gets added to the
common ground and, consequently, to the hearer’s mental model.
Tris Faulkner, Ph.D.
Assistant Professor
Spanish Linguistics
Department of Spanish Language and Literature
Dewing Hall: DEW203B
Kalamazoo College
1200 Academy St
Kalamazoo, MI 49006
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
98
tris.faulkner@kzoo.edu;
trisfaulkner.phd@gmail.com
ORCID ID 0000-0002-9099-0057
References
Alonso-Cortés, A. M. (1981). Gramática del subjuntivo. Madrid, Ed. Cátedra.
Bello, A. (1847). Gramática de la lengua castellana. Buenos Aires, Losada.
Becker, M. G. (2010). Principles of mood change in evaluative contexts: The case of
French, in Modality and Mood in Romance : Modal interpretation, mood selection,
and mood alternation, edited by Martin G. Becker and Eva-Maria Remberger,
Berlin, New York: De Gruyter, 2010, pp. 209-234.
https://doi.org/10.1515/9783110234343.2.209
Blanch, Lope J. M. (1958). Algunos usos de indicativo por subjuntivo en oraciones
subordinadas. Nueva Revista De Filología Hispánica (NRFH), 12(3/4), pp. 383-
385. https://doi.org/10.24201/nrfh.v12i3/4.3061
Blake, R. (1981). Some Empirically Based Observations on Adult Usage of the
Subjunctive Mood in Mexico City, in James Lantolf & Gregory Stone (Eds.),
Current Research in Romance Languages, pp. 13-22. Bloomington, IN: Indiana
University Linguistics Club.
Blake, R. (1982). El uso del subjuntivo con cláusulas nominales: regla obligatoria o
variable. Actas Del Primer Congreso Internacional Sobre El Español De América.
Río Piedras. Puerto Rico: Universidad De Puerto Rico, pp. 351–360.
Bolinger, D. (1991). Essays on Spanish Words & Grammar, in J.H. Silverman (Ed.),
Essays in Spanish: Words and Grammar. Newark, DE: Juan de la Cuesta Hispanic
Monographs, pp. 207-210.
Borgonovo, C. & Prevost, P. (2003). Knowledge of polarity subjunctive in L2 Spanish, in
B. Beachley, A. Brown & F. Conlin (Eds.), BUCLD 27: Proceedings of the 27th
Boston University Conference on Language Development,1, pp. 150–161.
Somerville: Cascadilla Press.
Borrego, J., Asensi, J. G. & Prieto, E. (1989). El subjuntivo: valores y usos. Madrid:
Sociedad General Española de Librería.
Borrego, et al. (2013). Gramática de Referencia Para La enseñanza de Español: La
combinación de Oraciones. Salamanca: Ediciones Universidad de Salamanca.
Crespo del Río, C. (2014). Tense and Mood Variation in Spanish Nominal Subordinates:
The Case of Peruvian Varieties. Unpublished doctoral dissertation. Urbana-
Champaign, IL: University of Illinois at Urbana-Champaign.
Davies, M. (2016). El Corpus del Español. Retroceded from
https://www.corpusdelespanol.org/
Deshors, S. C., & Waltermire, M. (2019). The indicative vs. subjunctive alternation with
expressions of possibility in Spanish. International Journal of Corpus Linguistics,
24(1), pp. 67-97. https://doi.org/10.1075/ijcl.17038.des
Farkas, D. F. (1992b). On the Semantics of Subjunctive Complements, in P. Hirschbühler,
(Ed.), Romance Languages and Modern Linguistic Theory: Selected papers from
the XX Linguistic Symposium on Romance Languages, pp. 69-105.
Amsterdam/Philadelphia: John Benjamins. https://doi.org/10.1075/cilt.91.07far
TRIS FAULKNER
99
Faulkner, T. (2021a). Prescriptively or Descriptively Speaking? How ‘Information-quality’
Influences Mood Variation in Spanish Emotive-factive Clauses. Pragmatics, 31(3),
pp. 357-381. Amsterdam/Philadelphia: John Benjamins.
https://doi.org/10.1075/prag.19044.fau
Faulkner, T. (2021b). A Systematic Investigation of the Spanish Subjunctive: Mood
Variation in Subjunctive Clauses. Unpublished doctoral dissertation. Washington,
D.C.: Georgetown University.
García, M. E., & Terrell, T. (1977). Is the Use of Mood in Spanish Subject to Variable
Constraints?, in M.P. Hagiwara (Ed.), Studies in Romance Linguistics: Proceedings
of the Fifth Linguistic Symposium on Romance Languages, pp. 214-226. Rowley,
MA: Newbury House.
Giannakidou, A. (2017) Evaluative Subjunctive and Nonveridicality, in Joanna Blaszczak,
Anastasia Giannakidou, Dorota Klimek-Jankowska and Krzysztof Migdalski
(Eds.), Mood, Aspect, Modality Revisited: New Answers to Old Questions, pp. 177-
217. Chicago: University of Chicago Press.
https://doi.org/10.7208/chicago/9780226363660.003.0005
Gili Gaya, S. (1960). Curso superior de sintaxis española. Barcelona: Publicaciones y
Ediciones Spes.
Gili Gaya, S. (1980). Curso superior de sintaxis española. Barcelona: Bibliograf.
Giorgi, A., & Pianesi, F. (1997). Tense and Aspect: From Semantics to Morphosyntax.
Oxford: Oxford University Press.
Gregory, A. E., & Lunn, P. V. (2012). A Concept-based Approach to the Subjunctive.
Hispania, 95(2), pp. 333-343.
Hooper, J. (1975). On Assertive Predicates. Syntax and Semantics 4, pp. 91-124.
https://doi.org/10.1163/9789004368828_005
Kowal, J. (2007). La elección del modo subjuntivo en las subordinadas nominales.
Lingüística Española Actual, 29(1), pp. 45-72.
Laca, B. (2010). Mood in Spanish, in B. R. Rothstein & R. Thieroff (Eds.), Studies in
Language Companion Series Mood in the Languages of Europe, pp. 198-220.
Amsterdam/Philadelphia: John Benjamins. https://doi.org/10.1075/slcs.120.11lac
Lozano, A. G. (1972). Subjunctives, Transformations and Features in Spanish. Hispania,
55(1), 76-90. https://doi.org/10.2307/338249
Lunn, P. V. (1989). Spanish mood and the prototype of assertability. Linguistics, 27(4),
687-702. https://doi.org/10.1515/ling.1989.27.4.687
Mejías-Bikandi, E. (1998). Pragmatic Presupposition and Old Information in the Use of
the Subjunctive Mood in Spanish. Hispania, 81(4), pp. 9-41.
https://doi.org/10.2307/345807
Portner, P. (2018). Mood. Oxford: Oxford University Press.
https://doi.org/10.1093/oso/9780199547524.001.0001
Portner, P. & Rubinstein, A. (2020). Desire, belief, and semantic composition: variation in
mood selection with desire predicates. Natural Language Semantics, 28(4), pp.
343–393. https://doi.org/10.1007/s11050-020-09167-7
Quer, J. (1998). Mood at the Interface. The Hague: Holland Academic Graphics.
Quer, J. (2001). Interpreting mood. Probus, 13(1), pp. 81-111.
https://doi.org/10.1515/prbs.13.1.81
THE TWO SPANISH SUBJUNCTIVES: THE REQUIRED AND DEFAULT SUBJUNCTIVES
100
Quer, J. (2009). Twists of mood: The distribution and interpretation of indicative and
subjunctive. Lingua, 119(12), pp. 1779–1787.
https://doi.org/10.1016/j.lingua.2008.12.003
Ramsey, M. M., & Spaulding, R. K. (1956). A Textbook of Modern Spanish: As Now
Written and Spoken in Castile and the Spanish American Republics. New York:
Holt, Rinehart and Winston.
Real Academia de la Lengua Española (RAE). (2011). Nueva gramática de la lengua
española: Manual. Madrid: Espasa Libros.
Romero, M. (2012). Mood Selection in Romance Complement Clauses. SinFonIJA, 5, pp.
1-6.
Ruiz Campillo, J. P. (2001). El subjuntivo es lógico: una actividad de concienciación.
Cervantes, 2, pp. 33-36.
Ruiz Campillo, J. P. (2008). El Valor Central del Subjuntivo, ¿Informatividad o
Declaratividad? marcoELE, Revista de Didáctica ELE, 7, pp. 1-44.
Sessarego, C. (2016). A Discourse-Pragmatic Approach to Teaching
Indicative/Subjunctive Mood Selection in the Intermediate Spanish Language
Class: New Information versus Reformulation. Hispania, 99(3), pp. 392-406.
https://doi.org/10.1353/hpn.2016.0067
Silva-Corvalán, C. (1994). The gradual loss of mood distinctions in Los Angeles Spanish.
Language, Variation, and Change, 6, pp. 255-272.
https://doi.org/10.1017/S095439450000168X
Smead, R. (1994). Sobre La Elección De Modo Para Las Dubitativas Factivas y Las
Cognoscitivas/Aseverativas Negadas: Tendencias Entre Monolingües Y Bilingües.
Hispanic Journal, 15(2), pp. 271-289.
Stalnaker, R. (2002). Common Ground. Linguistics and Philosophy, 25(5/6), pp. 701-721.
https://doi.org/10.1023/A:1020867916902
Studerus, L. (1995). Mood Variability in Border Spanish. Bilingual Review, 20(2), pp. 99-
113.
Terrell, T., & Hooper, J. (1974). A Semantically Based Analysis of Mood in Spanish.
Hispania, 57(3), pp. 484-494. https://doi.org/10.2307/339187
Villalta, E. (2008). Mood and gradability: An investigation of the subjunctive mood in
Spanish. Linguistics and Philosophy, 31(4), pp. 467–522.
https://doi.org/10.1007/s10988-008-9046-x
Waltermire, M. (2017). Mood variation after expressions of uncertainty in Spanish. Studies
in Hispanic and Lusophone Linguistics, 10(2), pp. 349-372.
https://doi.org/10.1515/shll-2017-0011
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
It is generally put forth that Spanish has the subjunctive as the required mood in the complements of emotive-factives ( alegrarse de que ‘to be happy that’), desire verbs ( querer ‘to want’), verbs of uncertainty ( dudar ‘to doubt’), modals ( ser posible que ‘to be possible that’), causatives ( hacer que ‘to make that’), and directives ( recomendar que ‘to recommend that’) (e.g., Real Academia Española 2011 ). However, in spite of these traditional rules, it has been observed that some of these environments allow for the indicative ( Blake 1981 ; Crespo del Río 2014 ; Deshors and Waltermire 2019 ; Gallego and Alonso-Marks 2014 ; García and Terrell 1977 ; Gregory and Lunn 2012 ; Kowal 2007 ; Lipski 1978 ; Silva-Corvalán 1994 ; Waltermire 2019). The current study explored one such environment; emotive-factive clauses. Results showed that the presuppositions that speakers hold regarding the knowledge that their addressees possess influence the mood that they select. This, thus, demonstrates the important role that pragmatics plays in the occurrence of mood variation.
Article
Full-text available
Mood selection properties of desire verbs provide a rich source of evidence regarding the semantics of propositional attitudes. This paper approaches the topic by providing an analysis of crosslinguistic variation in the selection patterns of the desire verbs ‘want’ and ‘hope’, focusing on Spanish and French. There is no evidence that the meanings of ‘hope’ and ‘want’ differ between these languages, and yet in Spanish esperar ‘hope’ and querer ‘want’ both take subjunctive, while in French only vouloir ‘want’ selects subjunctive and espérer ‘hope’ strongly prefers the indicative. The inclination of ‘hope’ toward the indicative is manifest also in other Romance languages. Previous theories tie mood selection tightly to the verb’s modal backgrounds and do not anticipate such variation. We explain the consistency in mood selection with ‘want’ versus the variation with ‘hope’ in terms of two key ideas: (i) moods are modal operators that encode different degrees of modal necessity, and (ii) modal backgrounds can be manipulated by the grammar. In terms of (i), we argue, building on the comparison-based theory of mood, that the indicative is a strong necessity operator, while the subjunctive encodes a weaker necessity. Regarding (ii), we propose that two backgrounds may function as one under certain well-defined circumstances. Our proposal supports the decompositional approach to attitude verbs, where mood is responsible for the quantificational force traditionally attributed to the verb.
Article
Full-text available
This study explores the indicative vs. subjunctive alternation in Spanish subordinate clauses following epistemic adverbials and expressions of possibility. Anchored in semantic-pragmatic and variationist theoretical frameworks, traditional research on mood alternation in Spanish remains largely experimental in nature. In contrast, we adopt a corpus-based multifactorial methodology to investigate 4,199 occurrences of fourteen expressions of possibility extracted from the Corpus del Español (e.g. caso de que, poder ser que, por si acaso, posiblemente, etc.) annotated contextually for structural, semantic and stylistic variables. Methodologically, we conduct an exploratory multiple correspondence analysis followed by a confirmatory binary logistic regression to examine whether/how the linguistic contexts affect mood variation. Overall, the results indicate that previously unexplored semantic factors (such as the inherent lexical aspect of verbs in subordinate clauses) significantly influence mood variation in Spanish. Ultimately, our results suggest that subjunctive uses are less uniform and more prone to internal variation than indicative uses.
Article
Full-text available
According to prescriptive grammars of Spanish, subordinate clauses after expressions of uncertainty require the subjunctive when introduced by the complementizer que ‘that’ (e. g., No creo que venga (S). ‘I don’t believe that he’ll come.’) but require the indicative when introduced by the complementizer si ‘if/whether’ (e. g., No sé si viene (I). ‘I don’t know if he’ll come.’). The use of these moods in actual speech, however, is highly variable. Using a database of conversational Spanish from 10 cities in Latin America and Spain, 428 tokens of subjunctive and indicative mood use were examined according to the following linguistic and social factors – subject agreement, intervening material, frequency of the subordinate verb, age, sex, and education level. Of these variables, only intervening material is statistically significant in the conditioning of mood variation. The use of the subjunctive is preferred after the complementizer que when there is no intervening material separating it from the subordinate verb. In order to further explore possible interactions between individual factors, a Joint Multiple Correspondence Analysis using R was conducted. Results from this exploratory analysis suggest that the use of the subjunctive correlates with young, male, and highly educated speakers whereas the indicative correlates with older, female, and less educated speakers. The variable use of mood after expressions of uncertainty in Spanish may be socially motivated, as these results suggest. As such, the impact of social variables on mood variation should be explored in future studies.
Article
Full-text available
En este número no se incluyeron resúmenes ni palabras clave.
Thesis
The link to the full text of this Ph.D. thesis is available at Georgetown University's Institutional Repository: http://hdl.handle.net/10822/1062622 Standard Spanish grammar states that desideratives (querer que), directives (aconsejar que), purpose clauses (para que), causatives (hacer que), emotive-factives (alegrarse de que), negated epistemics (no creer que), dubitatives (dudar que), and modals (ser posible que) embed subjunctive complement clauses. However, in spite of these classifications, some predicates exhibit a certain degree of mood variation. For instance, emotive-factives can take indicative complements (Crespo del Río, 2014; Faulkner, 2021). Similar variability between the moods may also come about in negated epistemic (Bolinger, 1991), dubitative (Blake, 1981), and modal clauses (Deshors and Waltermire, 2019). When both moods are acceptable, the choice between the two conveys a difference in meaning (Kanwit and Geeslin, 2014, 2017). Corroborating this hypothesis are findings from Faulkner (2021), which demonstrated that the mood variation occurring in emotive-factive clauses relates to the INFORMATIVENESS of the embedded proposition; i.e., propositions that are addressee-new versus addressee-old. The objective of the current dissertation is to investigate the standardly subjunctive environments in which the use of the indicative may become acceptable. Two hundred and twenty-three native Spanish-speakers completed a 128-item Acceptability Judgment Task (AJT), pertaining to the use of each mood in eight traditionally subjunctive complements (desiderative, purpose, directive, causative, modal, dubitative, negated epistemic, and emotive-factive clauses). Participants rated separate instances of subjunctive and indicative based on the (un)informativeness of the complement proposition. Statistical analyses of the ratings revealed that subjunctive is always preferred over indicative in traditionally subjunctive environments. However, certain predicates were more receptive than others to taking indicative clauses. Whereas preference-based expressions (desideratives, directives, purpose clauses, causatives) require subjunctive complements, emotive-factives and verbs of uncertainty (negated epistemics, dubitatives, modals) may accept indicative if the speaker intends to add the affirmative or negated proposition to the common ground. The likelihood that this occurs increases when the proposition (affirmative or negated) is informative. Since this was the case regardless of the participants’ countries of origin, said findings suggest that the Spanish mood system involves a split between two types of subjunctives, one that is required in preference-based contexts, and another that is default and can be replaced by the indicative. INDEX WORDS: Spanish, Verbal Mood, Mood, Subjunctive, Indicative, Mood Variation