Review of The Syntax, Semantics and Pragmatics of Spanish Mood
Date: Tue, 21 Jan 2003 14:56:58 -0700
From: Jorge Porras <jorge.porras@SONOMA.EDU>
Subject: Review of Haverkate (2002), The The Synatx, Semantics, and ..
Haverkate, Henk. (2002). The Synatx, Semantics, and Pragmatics of Spanish Mood.
John Benjamins Pubishers Co., paperback, ISBN 90 272 5347 1 (Eur) / 1 58811 252
6 (US) 235pp.
Jorge E. Porras, Sonoma State University, Rohnert Park, CA
DESCRIPTION OF THE BOOK
In his introduction (p. 1), Haverkate states that this book is a revised,
Extended and updated version of his 1989 study, written in Dutch. It provides
A tripartite description and explanation of the syntax, semantics, and pragmatics
of the modal system of contemporary Peninsular Spanish (basically, indicative
and subjunctive). Specifically, the syntactic analysis covers modal variation in subordinate and non-subordinate clauses. The semantic analysis covers the role
of the truthfunctional categories of realis, potentialis, and irrealis. The pragmatic
analysis covers speech act theory in terms of Grecian maxims, presupposition,
relevance, and politeness. Haverkate states that the present book has been
especially written for researchers and advanced-level students of Spanish
The book consists of a short introduction and six chapters, the last one being
The conclusion. Chapters vary considerably in length, from three to four pages (1,
2 and 4) up to 141 pages (chapter 5). It also contains a list of references and
several indexes (corpus, subject, proper names, and lexicion: pp. 225-235). In
his conclusions (pp. 197-98), Haverkate states that his book is a critical
discussion, revision, and elaboration of previous approaches to the study of
Spanish mood, where a new semantic classification of clause-embedding
predicates is proposed.
Chapter 1, "Modal categories of the Spanish verb: Levels of analysis",
Discusses five levels of analysis for the Spanish "modos verbales": Phonetic,
morphological, syntactic, semantic, and pragmatic. About the phonetic level,
Haverkate claims that supra-segmental features play a major role, due to
differences in pitch and sentence stress, which he illustrates with an example:
"Cierra la puerta Juan" (Juan closes the door) and "Cierra la puerta, Juan¡"
(Juan, please close the door). In the morphological level, Haverkate equates
the modal paradigms of Spanish to those of tense, aspect, gender, number,
and case, but he distinguishes the imperative from the subjunctive by their
functional role (p. 4). Syntactically, he makes an analysis of subordinate and
non-subordinate clauses, with more attention put on the latter. He notes that
the occurrence of the imperative is restricted to non-subordinate clauses. While
the indicative is not subject to this constraint, the main working area of the
subjunctive is the subordinate clause. Haverkate makes a tripartite classification
of Spanish mood in terms of noun, adjective, and adverbial clauses, and he claims
that, semantically, "the truthfunctional categories of realis, potentialis, and irrealis
play a major part in interpreting the variation of the indicative and the subjunctive".
He also claims that mood selection in noun clauses is dependent to a large extent on
the lexical class-membership of the embedding predicate" (p. 4).
Chapter 2, "Functions of the modal categories of the Spanish verb" outlines the
features of the pragmatic component of the modos verbales. Haverkate considers
these categories as expressions of illocutionary functions, according to the
theory of speech acts (Searle 1976). Assertive speech acts are defined in terms
of the speaker's intention, and thus "serve the purpose of convincing the hearer
that the speaker commits himself/herself to the truth of the proposition
expressed", while directive speech acts " influence the intentional behavior
of the hearer in such a way that the latter carries out the action specified by the
proposition" (p. 7). In non-subordinate clauses, the indicative mood characterizes assertive sentences; the imperative typifies directive sentences. In turn, the
subjunctive is mainly restricted to subordinate clauses. This mood, however, characterizes a type of speech act not explicitly mentioned by Searle: oraciones
optativas, that is, sentences expressing wishes or desires that cannot be fulfilled by human agents, such as ¨"Muera el general¡" (Away with the general), "Viva el presidente" (Long live the president).
Chapter 3, "The modal structure of non-subordinate clauses" (pp. 11-40),
Resumes topics outlined in the previous two chapters, that is, the different
Illocutionary functions fulfilled by the imperative and the (optative) subjunctive,
on the one hand, and the indicative, on the other, in non-subordinate clauses.
Such functions include, for the imperative and optative subjunctive, speech acts
Defined by the world-to-words direction of fit, which "serve the purpose of bringing
About a state of affairs not existing at coding time" (p. 11).
This chapter consists of an analysis of the three categories, in corresponding
subdivided sections, following a tripartite sentence-level model: syntactic,
semantic, and pragmatic. Here, and throughout the book, Haverkate illustrates
with examples taken from a corpus, and critical revisions on previous research
on each topic. It is shown that selection is determined by the illocutionary function
of imperatives. After Dick 1980, he distinguishes four types of predicates:action,
process, position, and state, with two distinctive features underlying this
typology, control and dynamic; he concludes with a survey that only action and
position verbs may be inflected for imperative mood. For the pragmatic
function, Haverkate argues that, unlike English, the Spanish imperative is
commonly used to make a request, not an order.
The section dedicated to indicative sentences, (p. 21), begins by making a
comparison between the paradigmatic systems of indicative and imperative.
While the imperative consists of two paradigms only, affirmative and negative,
The indicative is made up of nine paradigms which, in contrast to imperative, are
distinguished by a set of semantic parameters bearing on the categories of tense,
aspect and truth value (present, future, pluperfect, etc. Here, four parameters are
specified for the points of the realis-irrealis scale: real, possible, probable, and
unreal. The author also makes comments on a set of adverbs and adverbials that
can modify the assertive force of indicative sentences: "indudablemente"
(undoubtedly), "de seguro" (surely), and "con certeza" (with certainty), which
are compatible only with indicative moo. As for the pragmatic function of
indicative sentences, Haverkate discusses some typological matters such as the
non-existence of an interrogative mood and the fact that assertives area crucial
part of verbal interaction, hence their wide verb membership, and that commissives
are the mirror image of directives
The section on subjunctive sentences is devoted mostly to show the considerable
differences existing between the paradigmatic systems of subjunctive and
indicative (nine for the indicative and only for the indicative); for example,
there are no subjunctive counterparts of the preterito indefinido and the
preterito imperfecto. A syntactic contrast of the two modal systems is made
with comments on clitic order and subject behavior, among others. The semantics
of subjunctive sentences is tackled from two different perspectives: the lexical
meaning of the predicate and the realis-irrealis scale, to which Haverkate adds
another scale: realizable and non-realizable wishes. In analyzing optative
subjunctive from a speech act point of view, the author identifies for Spanish a
world-changing function and an expressive function.
Haverkate next examines assertive subjunctive sentences, which are formally
determined by the occurrence of dubitative adverbs or adverbials. To cite just
one example, "Tal vez lo haya visto" (May be he - SUB - has seen it), is a
statement referring to a state of affairs which, according to the realis-irrealis
scale, should be qualified as hypothetical or possible, an interpretation stemmed
from the lexical meaning of tal vez. The author assumes, more generally, that the
use of the subjunctive in non-subordinate clauses is relatively infrequent, something
that is paralleled by limited illocutionary potential.
Finally, as for the pragmatic function of subjunctive sentences, Haverkate
Defines optative sentences, within a speech-act frame of reference, at the
Linguistic output of a particular class of directives. That is, optative speakers
utter a wish to become true in a future world For example, he shows that
pluperfect optatives may function as indirect speech acts and that the
non-subordinate subjunctive serves the purpose of either softening or
strengthening the force of the speech act. This past subjunctive also has the perlocutionary effect of expressing more than conditionals. The conclusion from
this section is that research on this topic calls for a distinction between three
levels of analysis.
Chapter 4, "Que-sentences", deals with a brief examination of some syntactic,
semantic, and pragmatic properties of this category. Syntactically, que-sentences
could be viewed as a hybrid category sharing properties of both subordinate and
non-subordinate sentences. On one hand, que may function as a complementizer
and, on the other, these sentences are not embedded in a matrix clause.
Pragmatically, they perform a variety of illocutionary functions as they may be
used to express assertions, orders, promises, and wishes. An example of an order (emphatic directive) is: "¡Que se siente Ud¡" (Sit down, I say¡); "¡Que se lo pagare
todo¡" (I shall pay you everything¡), expresses a promise (emphatic assertive).
Haverkate places sentences such as "Que en paz descanse", (God rest her soul),
"Que descanses" (Sleep well), and "Que lo pases bien" (Good luck to you), in a
separate class, optatives. These sentences typically function as politeness formulas.
He explains that "(d)ue to their conventional use, they lack the emphatic or
reiterative strength typifying assertive and directive que-sentences" (p. 43).
Chapter 5, "The modal structure of subordinate clauses", contains a detailed,
thorough account of the indicative and the subjunctive in the subordinate
clause within the framework of the tripartite distinction between noun,
adverbial and adjective clauses Four syntactic functions of noun clauses are
considered: Subject, direct object, prepositional object, and nominal predicate,
in each of which both the indicative and the subjunctive may appear. The
prototypical complementizer of noun clauses is the conjunction que. In
introducing adverbial clauses, Haverkate finds that they differ from noun
clauses in important ways. First, they do not fill argument slots but maintain a
peripheral relation with the main predicate of the sentence; secondly, they use
many more conjunctions than just two in the noun clause (que and si). The last
type of subordinate clause, the adjective (or relative) clause differs, at the same
time, from the preceding two in that the two former operate at the level of the
sentence, whereas the latter operates at the hierarchically lower level of the noun
Haverkate introduces this section with a review of relevant literature on the
subject. He claims, for example, that Terrell and Hooper 1974 analysis is not
optimal because it lacks appropriately defined taxonomic criteria, and is not
maximal because it does not specify the class of clause embedding predicates.
These two terms come from his own proposal for the semantic classification of
clause-embedding predicates: "Optimal means that the classification reflects in
both a consistent and a coherent way that part of the world that is described
by the predicates in question". Maximal means that, "it includes all
clause-embedding predicates of the language" (p. 132). Basically, Haverkate
shows that clause-embedding predicates share the property of providing
information on the set of mental processes that characterize intentional human
behavior. In chronological order, the first class dealt with is the acquisition of
knowledge predicates, which describe the processing of perceptual and
conceptual information. The second class consists of predicates describing the
storing and assessing of the input information; there are two subclasses here:
cognition and evaluation predicates. The third class describes those categories
involved in the output of Intentional behavior, with three subclasses denoting
causative acts, mental acts, and speech acts For the cognition predicates,
Haverkate cites examples with verbs such as saber (to know): epistemic; creer
(to believe): doxastic), and dudar (to doubt): dubitative. He makes an important distinction between subject meaning and speaker meaning, which accounts for the
use of the indicative in the sentence, "Mi novia duda que soy millonario" (My
girlfriend doubts that I am - IND - a millionaire), which leads to the interpretation
that the speaker (not the subject) of the sentence, as a polyphonic source, presupposes the factuality of being a millionaire.
Another important distinction made by Haverkate is between foreground (or
focalized) and background (or de-focalized) information, to refer to
information which is either new or already known to the hearer (cf. Tomlin 1985).
This concept is used, for instance, in his analysis of evaluation predicates.
(both rational and emotional). About the latter, for example the subclass describing positive attitude, Haverkate states that the use of emotional predicates, gustar (to
like), encantar (to enjoy thoroughly) and alegrar (make happy "does not focalize the
propositional content of the subordinate clause but the evaluative judgment on
that proposition denoted by the main clause" (p. 95). This implies that the use of
the subjunctive corresponds with a low degree of information value. Some
conclusions about the modal distribution in the subordinate clauses of Spanish
cognition and evaluation predicates are: In sentences containing an evaluation
predicate, backgrounding of the content of the embedded proposition requires the
use of the subjunctive, whereas foregrounding of the content requires the indicative.
Epistemic and doxastic predicates select the indicative; dubitatives select the
subjunctive if a judgment of true or false can be made; if the proposition is
deemed true, the indicative is used.
Next, action predicates are analyzed. They describe the different output
Categories of intentional human behavior. Syntactically, there are those which do not take complement clauses: viajar (to travel), bailar (to dance), salir (to leave);
And those which do: impedir (to prevent), pensar (to think), informar (to inform).
Semantically, complement-taking action predicates are classified as causative,
mental, and speech acts. Syntactically, causative predicates run parallel with
desideratives in that they select either the subjunctive or the infinitive. In
either case, the indicative is excluded. Semantically, they are related, too,
because their use creates a temporal relation between the main and the
subordinate clause, defined in terms of prospectivity. Also, there is a large class of
two-place predicates, and a small class of three-place predicates. An example
of the latter is: "La policia les obligo a desalojar el lugar" (The police
obliged them to vacate the building).
Mental acts describe thinking processes. They are generated by acts of
thinking, acts of making a prediction, and acts of creating a world of belief. Some
predicates such as demostrar (to prove) are process-oriented, while others such
as inferir (to deduce) are agent-oriented. Predicting acts, in turn, imply making
an assertion on a future; an example is adivinar (to guess). Although it means an
unreal state of affairs, these predicats take the indicative because the
speaker is confident enough about that the state of affairs will become true. Last,
acts creating a world of belief, such as fingir (pretend), also perform in
indicative because the speaker conceives the states of affairs in the imaginary
world as a reality in that world. Direct discourse serves as an evidentiality marker.
In indirect discourse, however, evidentiality applies to so-called de re
representations. Assertives use predicates such as decir (to say). In both
direct and indirect speech, the mood of the original assertion is reproduced in the
complement clause. Haverkate makes it clear that agreement of mood is not
matched by agreement of tense, and he analyses examples.
The class of directive speech acts makes a distinction between interrogative
And non-interrogative acts; only the former elicits a verbal response from the
hearer. Another distinction holds between direct and indirect performances As
for the class of commissives characterized by predicates such as prometer
(to promise), garantizar (to guarantee), and jurar (to swear), the indicative is
used with direct reports, whereas in indirect discourse this mood alternates
with the infinitive. Finally, expressives enable the speaker to express a psychological state. They are realized by expressions lacking propositional content, for example: gracias (thanks) and perdon (sorry). The event described is backgrounded, so the
information is given by the performative use of the expressive main predicate.
Thus, the subjunctive is used.
The second longest section of this book is the one on adverbial clauses (pp.
133-182. The following categories are considered: Time, manner, purpose, cause,
consequence, concession, and condition. To begin with, temporal clauses are
divided into three classes, according to their simultaneous, successive, or temporal
specification: Mientras (while), despues (de) que (after), antes (de) que
(before,), cuando (when), are respective examples. The use of despues and antes
creates retrospective and prospective relation, respectively. This criterion
usually triggers subjunctive mood.
As happens with these and other conjunctions treated here, their use may
Express more than one relational meaning; mientras, for instance, not only
Expresses simultaneity, but its use also implies that the speaker looks at the states
Of affairs described from a contrastive point of view, like in the following
sentence: "Yo me alojo en una pension, mientras que el se hospeda en el
Castellana Hilton: (I am staying at a guesthouse, whereas he is ? IND ? staying
at Castellana Hilton. Some conclusions for these clauses are: with cuando, the
indicative is used in both factual and non-factual clauses. Just like mientras and
cuando, en cuanto (as soon as) and tan pronto como (as soon as), may introduce
iterative clauses; iterative aspect is emphasized by cada vez (whenever), siempre
que (whenever), and hasta que (until). Subordinate clauses marking the beginning
of the event expressed by the main clause are generally introduced by the
conjunction desde que (since).
Manner clauses, such as como (as) and a medida que (as) allow both moods to
Appear as fillers of the modal slot of the clause, since it must be attributed a
realis or an irrealis interpretation. Purpose clauses, Haverkate explains, are
intrinsically related with cause, consequence, concession, and condition-
indicating clauses, and they refer to the output of intentional behavior. The
typical conjunction of this class is para que (in order to). Cause and consequence
clauses are both dominated by the indicative pattern, whose primary function is
to enhance the relevance of the statement made in the main clause. Typical
conjunctions are porque (because) and por (lo) tanto (therefore). After the
tripartite analysis of these clauses, with the customary critical revision of the
literature, Concessive clauses are introduced by a diversity of conjunctions and
connectors, some of which are: si bien (although), por mas que (no matter how),
por mucho que (no matter how much), and (aun) a riesgo de que (even at the risk
that). The most frequently used and extensively studied is aunque (although). From
a truthfunctional point of view, this conjunction shows a maximum distributional
potential, since it is employed to introduce realis, potentialis, and irrealis
clauses. However, por mas que and por mucho que, for example, only take the
Finally, the analysis of conditional clauses. Haverkate points out that these
clauses are fundamentally distinct from other types of adverbial sentences in
that the main and the subordinate clause are strictly interdependent. This can
be seen by comparing the following two sentences, where their coordinative
constructions cannot be paraphrased: " Si tu padre viene, me marcho yo¨(If your
father comes - IND -, I shall leave) vs. "Tu padre viene; me marcho yo" (Your father
is coming; I shall leave). Haverkate shows that, consistently, indeed, the si (if)
conjunction takes the indicative, and embarks himself in a truth functional analysis of
conditional sentences. It yields a threefold classification based on the realis,
potentialis or irrealis interpretation of the causal link between the protasis
(the subordinate clause), and the apodosis (the main clause). Haverkate.
concludes this section with three isolated phenomenon involving morphological, syntactic, and semantic facts.
The next two sections close the book: Adjective clauses (pp. 182-193) and de
que-clauses (194-196). As for adjective clauses, a fundamental point of
departure is the traditional distinction between restrictive and non-restrictive
clauses, illustrated as follows: "Los alumnos que vivian lejos llegaron tarde a la
escuela" (The pupils who lived far away arrived at school late). This sentence is
restrictive or especificativa. If the clause "que vivian lejos" is between
commas (or pauses), the sentence becomes non-restrictive or explicativa .
Restrictive clauses perform a referential function; non-restrictive ones do not.
The realis character of the non-restrictive proposition brings about the use of
the indicative mood (with two exceptions, not included here).
The analysis of restrictive clauses is more complex. Haverkate examines their
syntactic properties in terms of mood distribution as well as their semantic
characteristics in terms of the realis, potentialis and irrealis parameter. He
discusses the following four situations: (I) The existence or non-existence of
the entity referred to is focussed in. (II) The entity referred to may exist in a
virtual world (III) The entity referred to cannot be identified. (IV) It is
not obvious whether the entity referred to fits the description given. Haverkate
concludes that, despite statistical research showing indicative predominance,
modal distinction should be described in terms of contrastive distribution.
The last section of Chapter 5, de que-clauses, is short in length, but not
easy. It deals with a "minor category" of noun clauses which, despite their resem
blance to adjective clauses, have a quite different grammatical status. Haverkate
compares two sentences, one with an adjective clause and one with a noun clause:
"La idea de que me hablaste ayer me parece de dificil plasmacion" (The idea you
talked - IND - to me about yesterday seems to me dificult to implant) vs. "La idea de
que el presidente dimita inmediatamente no es compartida por todos" (The idea that
the president should - SUBJ - immediately is not shared by everyone). While the que
element in the relative clause is a relative pronoun that functions, among other
things, as an anaphoric link to the antecedent, and also as a prepositional
complement of the predicate hablaste, in the noun clause it is a complementizer
with no referential meaning.
Haverkate cites other examples and, following Solano-Araya 1987, he points out
that the lexical meaning of the complement-taking noun, rather than the realis or
the irrealis interpretations of the complement proposition, may trigger the use of
a specific mood. However, H. does not buy Fente´s 1997 proposal that there
exists a tendency in formal speech registers to use the subjunctive . Furthermore,
the complement-taking noun is subject to the constraint that it expresses abstract
meaning. Instead, Haverkate supports Guitarte´s 1984 proposal that in
sentence-initial position, the el hecho (de) que clause typically provides
thamatic (i.e, presupposed) information. Haverkate provides the example: "El heco
de que venga a vernos significa que nos tiene afecto" (The fact that he is -SUB -
coming to see us means that he has affection for us), and observes that since the
front clause is de-emphasized the subjunctive is used.
This book clearly satisfies the goals set up by its author. Furthermore, it
presents a rigorous and detailed linguistic analysis of Spanish mood. It is
well written, well thought out, and consistent in approach and theoretical
conception. Certainly, it proves to be a rich source of information and serious
analysis about the Spanish indicative and subjunctive. No doubt, it constitutes
an important contribution to the field.
Some critical comments follow, which should not affect the overall high quality
Of the book. Two topics are treated somewhat hastily, considering their
importance, On one hand, chapter 4, Que-sentences, appears to fall short in
exploring some pragmatic approaches such as ellipsis resolution in conversational discourse, particularly in relation to mood selection. It needs some expansion, for
example, on the semantic analysis of its logical form. On the other hand, section
5.4, on de que-clauses, needs more clarification on the typological status of the
category involved. Also, it is not quite clear that the mere position of the el hecho
que-clause determines mood selection (p. 194). Both orders sound grammatical:
"El hecho de que venga a vernos significa que nos tiene afecto" (The fact that he is
- SUB - coming to see us means that he has affection for us), and "Significa que nos
tiene afecto el hecho de que venga a vernos". What seems clear, however, is that the indicative is excluded at sentence-initial position, but not at final position, which is consistent with the (de)-focalization theory. This fact should put in perspective the convenience of a pragmati consideration in this respect.
Also mentioned should be the fact that the analysis of conditional clauses
Calls for a more comprehensive account. One point that comes to mind is a
treatment of conditionals in terms of modal variation in the apodosis: "Si tuviera
dinero, compria, compraba, comprara un coche" (If I had money, I would - COND,
IND, SUB - buy a car), (See e.g. Silva-Corvalan 1982); Sohrman 1991); another
point may emerge in connection with tense (see e.g, Tynan 1997). Similarly, the
analysis of the imperative, although relevant in regard with the speech act theory,
needs to be tackled more aggressively from a morphological point of view. A question
to pose would be whether there are imperative parameters the way there are imperative
rules. (See e.g, Harris 1997). By the way, morphological accounts are frequent
throughout the book, which suggest that a consideration about a fourth level of
analysis could be useful.
Although this reviewer does not concur with the grammaticality of some examples
In the book, it is proper to acknowledge that Peninsula Spanish differs from Latin
American Spanish in potentially important respects, and also that dialectal
variation has been taken into consideration throughout the book. In a different
vein, a semantic disparity is observed in the concept deonti c, as used by
Haverkate, on one hand, and by F. R. Palmer 1986, on the other. Haverkate uses
This term to refer to predicates that, "express the necessity that a certain change
In the world take place" (p. 91), and he cites the expressions ser preciso, ser
imprescindible, and ser conveniente, Palmer, in turn, uses it (in opposition to
epistemic) to refer to volition predicates such as querer, desear, and the
like. The latter are termed desideratives by Haverkate. (See Porras 1990 for an
analysis of Spanish subjunctive in Palmer´s terms).
A final comment on the organization of the book is in order. This reviewer
found it a little difficult to read through the chapters. This is probably due to the
scarcity of summaries and diagrams. A more reader-friendly layout would be to
divide the book in three parts, one for each level of analysis, and to make a
more balanced distribution of chapters and sections.
Harris, James (1997). "There is no imperative paradigm in Spanish", in
Martinez-Gil, Fernadoñ Morales-Front, Alfonso, eds .Issues in the phonology
And morphology of the major Iberian languages. Washington, DC: Georgetown UP
Porras, Jorge E. (1990). "Analisis semantico del uso del subjuntivo espanol",
in Discurso hispanico 7.2, pp. 387-394
Serrano, Maria J. (1996) "El subjuntivo ?ra y ?se en oraciones condicionales",
in Estudios filologicos 31, pp. 129-140
Silva-Corvalan, Carmen (1982) "Conditional for subjunctive in Old Castile", in
Maccauley, Monica et ale, eds. Proceedings of the Eighth Annual Meeting of the
Berkeley Linguistics Society. Berkeley: UCB UP
Sohrman, Ingmar (1991). Las construcciones condicionales en castellano
contemporaneo. Upsala: Upsala UP
Tynan, John; Delgado Lavin, Eva (1997) "Mood, tense, and the interpretation of
conditionals", in Dirven, Rene, ed. On conditionals again. Amsterdam:
| ABOUT THE REVIEWER:
Jorge Porras is Associate Professor of Hispanic linguistics at Sonoma State University (Rohnert Park, California), and serves as coordinator of the Spanish Program. His research interests include phonology, morphology, Spanish for native speakers in the US, and sociolinguistics (especially, Spanish-based Creoles and language contact).