Review of Epistemic Modality
AUTHOR: Pietrandrea, Paola
TITLE: Epistemic Modality.
SUBTITLE: Functional Properties and the Italian System.
SERIES: Studies in Language Companion Series 74
PUBLISHER: John Benjamins
Heiko Narrog, GSICS, Tohoku University, Japan
This is a book on epistemic modality in Italian and in general. It is
organized into 9 chapters and a conclusion. Chapter 1 introduces
the ''notional category of epistemic modality''. In this chapter, epistemic
modality is distinguished from related categories and concepts, such
as deontic modality, mood, illocution, reality status and evidentiality.
Furthermore, the relationship between modality and subjectivity is
discussed in some detail. The author views modality essentially as a
pragmatic category, and concludes in defining it as ''[a] performative
category expressing the speaker's genuine opinion towards the
modalized proposition'' (p. 39) (it should be noted here that this
definition is strictly speaking circular, as it already contains the
term ''modalized''). Chapter 2 provides ''a typological classification of
epistemic systems.'' Drawing on data from Palmer (1986), Pietrandrea
(P) identifies five parameters along which epistemic modality in a
specific language can be characterized typologically:
(1) ''specific vs. parasitic'' marking of epistemic modality, that is,
whether a modal marker is specialized on marking epistemic modality
or piggybacks on a different category,
(2) degrees of certainty (e.g. necessity vs. possibility),
(3) genuine epistemicity vs. inferential evidentiality,
(4) the degree to which evidentiality is differentiated between direct
vs. indirect, and reported vs. non-reported evidence, and
(5) degrees of performativity.
It should be mentioned here that P shares an understanding of
evidentiality increasingly common in European linguistics, in which
evidentiality is seen as a notional category expressed in many
markers that traditionally were viewed as epistemic modal markers.
This notion of evidentiality contrasts with the notion advocated by
other scholars such as Aikhenvald (Aikhenvald 2004), for whom only
languages with a grammaticalized system of evidentiality, and
obligatory presence of this category in the sentence, have
evidentiality. Chapter 3 introduces ''epistemic modality in Italian'',
identifying a large number of epistemic modal expressions in this
language, among which only the modal verbs and the epistemic future
(a mood) are classified as ''grammatical'' (p. 67).
Chapter 4 deals with ''semantic oppositions'' among these grammatical
epistemic modal expressions. Deve 'must', for example, is said to
express a higher degree of certainty than può 'can', and the same
deve contrasts strongly with the epistemic future in that the former has
a distinct evidential nature while the latter is the only purely (non-
evidential) epistemic marker in Italian.
Chapter 5 offers ''a typological characterization of Italian epistemic
modality'' along the parameters presented in Chapter 2. It adds to
chapter 4 a more detailed view of the evidentiality of Italian epistemic
modals, which, in the authors view, form a ''complex'' evidential
system, distinguishing both direct vs. indirect evidence and reportive
vs. non-reportive evidence.
Chapter 6 discusses ''inflectional and distributional constraints'' on
epistemic modality. It shows the degree to which epistemic modal
markers in Italian are constrained in their interaction with tense and
person marking, and the degree to which they can occur in conditional
and interrogative contexts. The relative freedom with which especially
the modals can occur in such collocations and constructions points to
their low perfomativity. According to P, this can be ascribed to the fact
that ''in Italian evidential forms are borrowed and employed for the
expression of epistemic modality.'' In other words, the author views the
Italian epistemic modals as essentially evidentials.
Chapter 7 discusses ''aspectual constraints on the propositional
contents''. Here it is claimed that epistemically modalized propositions
have to be aspectually ''incomplete'', that is, they must either belong to
the actional class of ''states'' in the sense of Vendler, or they must be
marked as progressive, habitual or perfect. The notion of
incompleteness builds on work by Desclés.
Chapter 8 continues this argument by claiming
the ''metapropositionality'' of epistemic modality. In a layered structure
of clause, epistemic modality only takes full propositions, which must
be incomplete'' into as complements while deontic modality takes units
of a lower layer as complements, namely predications, which
are ''complete''. This fact is explained by a ''semantic projection'' of
incompleteness by epistemic modal markers on their complements.
The last chapter, chapter 9, presents a diachronic hypothesis about
the development of epistemic meanings in deontic modal markers. P
contrasts four different assumptions about the semantic relation
between deontic and epistemic modality in the same modal marker,
namely homonymy (Palmer 1986), metaphoric change (Sweetser 1990
and others), conventionalization of conversational implicatures
(Traugott 1989 and others), and context-driven interpretation (Heine
1995 and others). The last approach (by Heine) fits P's own research
approach and her data best, and she essentially claims that ''the
reinterpretation of modal operators is the result of a reinterpretation of
their semantic scope from predicational to propositional'' (p. 187). It is
remarkable, however, that the epistemic meanings are already
present as far as P goes back in Italian/Latin language history (cf. p.
198). Therefore, it is not possible for P to make an argument based on
the actual analysis of historical data, and she has to rely on internal
Contents-wise, this book can be divided into three parts, namely
chapter 1, which provides the definition of the topic, chapters 2
through 5, which delineate Italian epistemic modality, and the last
three chapters, which deal with the interaction of modality with other
categories, most notably aspectuality. In terms of quality, that is, as an
evaluation, I would not hesitate to divide the book into a weak first part
from chapter 1 to 3, which is fairly informative and interesting to read,
but scores low in terms of rigor and analyticity, a strong part close to
the end, from chapter 6 to 7, which offers fresh and intriguing data
and hypotheses, and, finally, chapters 4, 5 and 8, which stand
somewhat in between quality-wise. In the following few paragraphs, I
will substantiate my evaluation, especially with respect to the weak
and the strong chapters.
Chapters 1 to 5 are the general chapters of the book. They are worth
reading because of the good command of the literature on modality
from a functional perspective that the author displays, and her ability
to pick out topics which are currently relevant, and present them
concisely. Her conclusions are largely either common-sensical or
intuitively acceptable. On the other hand, the manner these
conclusions are reached and the arguments are built up is neither
analytic nor rigorous. I'll give a few examples. On the first page of the
first chapter (p. 6), the category of modality is introduced through
quotes from Benveniste, Givón, Bally, and Palmer. This is done so in a
manner as if these linguists were referring to the same concept.
However, even if they used the same label, each of them had (or has)
a different concept in mind. In particular, the concept of Bally is only
partly compatible with the common concepts of modality in current
linguistics. The confusion culminates on the top of the following page
where the author contrasts ''modality'' with ''deontic modality''
and ''dynamic modality''. Is this also in the sense of Benveniste, Givón,
Bally and Palmer? Later in this chapter the reader will learn that for
the author only epistemic modality is true modality, and deontic
modality etc. don't deserve the same label. This should have been
clarified first, and then epistemic modality could have been relabeled
as ''modality'' and new labels could have been used for deontic and
dynamic modality. Also, it would have been interesting to know where
Benveniste, Givón, Bally and Palmer converge in their concepts, and
where they differ, and how the author's own concept contrasts to the
others. Instead, uncertainty about the author's use of terms prevails
throughout the introductory chapter 1, while from chapter 2 on,
surprisingly, the traditional labels are used fairly consistently. Another
remark is in place here about the author's use of the
term ''irrealis.'' ''Irrealis'' is dismissed as a concept for the definition of
modality (2.4.), but in doing so, the author seems to conflate irrealis as
a form category and as a semantic concept. It is only in the latter
sense that ''irrealis'' becomes a candidate concept for defining
modality (cf. van der Auwera and Schalley 2004).
Chapter 2 promises a typological classification, but the
term ''typological'' is used here in a very loose sense. Examples from a
number of genealogically unrelated languages are offered, but largely
they are second hand quotes taken from Palmer (1986) or other
textbooks, and there is no indication that the choice of languages is
systematic. Also, when it is stated that some phenomena are rare (e.g.
languages with specific markers for epistemic modality), and others
are frequent (languages with parasitic markers for epistemic modality),
no numbers from specific language samples are provided. The reader
is forced to believe the author instead of being invited to verify her
evidence. In this chapter, specific marking of epistemic modality is
contrasted with parasitic marking. However, a confusion of form and
meaning categories seems to take place, as parasitic markers are
characterized as ''auxiliary or semi-auxiliary modal verbs, generally
used also for the expression of deontic necessity and deontic
possibility; verbal moods; verbal tenses; clitics or non-affixed particles;
complementizers'' (p. 41). Logically, there is no reason, why a clitic or
non-affixed particle should not be a specific marker for epistemic
Chapter 3 introduces a large number of epistemic markers,
constructions and lexical items in Italian. Based on Lehmann (1985),
the author presents a catalogue of criteria to decide which of them are
grammaticalized. However, with the exception of a brief demonstration
of tense inflection on capace che 'may' and si vede che 'I see' (p. 64-
66), it is not shown how these criteria are applied, and only the
conclusion on which of the forms are grammaticalized is provided.
Showing the actual application of the criteria and its results would
have immeasurably increased the value of this chapter, and the
credibility of the conclusions provided.
I wish to close the critical part of my evaluation with a remark on P's
definition of modality. It is entirely legitimate to define modality as a
pragmatic notion, and in terms of ''speakers' opinions''. However, if it is
done so, this definition should be followed through consequentially. If
the defining criterion is the expression of speakers' opinions, then
lexical and grammatical classes such as mental attitude verbs,
epistemic adverbs and evaluative adjectives and constructions should
be presented as the core of the category, and the question how far
these expressions are grammaticalized or only lexical should not play
a decisive role. This direction in modality studies has been pursued
more consequentially by scholars like Maynard, e.g. Maynard (1993).
Instead, the mental attitude verbs and the epistemic adverbs are
mentioned but dismissed rather easily as non-grammatical (a point,
which, at least in the case of epistemic adverbs would deserve closer
investigation), and evaluative expressions are not mentioned at all. P
very conventionally picks out the modal verbs and future mood
marking as the typical epistemic markers. In order to do so, it would
have been more appropriate to stick to a more conventional definition
of modality, from a semantic, rather than a pragmatic point of view,
based on concepts like factuality, validity or relativization of the
proposition, e. g. Kiefer (1987)), Narrog (2005). The reviewer himself
has raised the question before whether it is possible at all to define a
grammatical category in pragmatic terms like ''speakers' attitudes''
or ''speakers' opinions'' (Narrog 2005). P's introductory chapters do
not present a strong case in favor of such a definition.
The best part of the book comes in chapters 6 and 7. In chapter 6,
finally some systematic language data are presented, showing, for
example, the constraints on the use of Italian modals with respect to
grammatical person. The hypothesis of the connection between
epistemic modality and ''incompleteness'' presented in chapter 7, in
the very terms in which the author defines it should be original to her,
and has not been discussed in this explicitness in any other book in
the functional literature on modality. However, the connection between
different types of modality and aspectuality has already been explored
from a different theoretical perspective, particularly by Abraham (e.g.
Abraham 1995), but also by other authors -- see Leiss 2002: 75-76;
P's book antedates Leiss' article (see below) -- and it would have
been interesting to read how P's approach and conclusions differ from
previous ones. Nevertheless, a detailed analysis of this topic from a
functional perspective is still rare, and with this book, P has secured
her place in a discussion which is certain to draw even more attention
in the future. In this chapter, P presents a thorough theoretical
discussion and a good number of tests and examples. In one central
point, however, her hypothesis appears to be problematic, if not
flawed. In analogy to Kiparsky and Kiparsky's factive predicates, which
render their complements factive, she suggests that epistemic
operators ''project'' a property of incompleteness on their
complements. The comparison appears to be misguided. Factive
predicates indeed seem to make the sentences that they take as their
complements factive, but epistemic operators apparently cannot make
their complements incomplete. On the contrary, they have to ''select''
complements which are already incomplete, that is, either stative, or
marked as progressive, habitual etc..
Chapter 8 on diachrony should have made mention of Traugott and
Dasher (2001). Traugott's position is identified by P with her 1989
paper, but chapter 3 of the 2001 book gives a more developed and
explicit scenario of change in modal meaning. From the list of
references it becomes apparent, however, that the actual writing of
the book must have stopped at some point in 2001. It was accepted as
a PhD thesis written in Italian in 2003, and the English translation has
come out in 2005, apparently without any update. I wish I would not
have to mention this, but typos, grammatical and stylistic errors are
pervasive to an extent that I cannot remember to have seen in any
book published by Benjamins. I'll give a few examples:
''expressd'' (p. 11),
''you my [may?] have been tired'', ''is specular [peculiar?] to the
behaviour'' (p. 172),
''It has been submitted [suggested?] in this chapter'' (p. 205).
Throughout the book spaces are missing after punctuation marks, and
in some places there are two spaces instead of one. Some typos may
seriously hamper understanding, e.g.: '''more direct' is 'less reliable'''
(p. 101) should read '''more direct' is 'more reliable''', or, ''This entails
the Italian system being strongly oriented towards the epistemic pole
of the epistemic-evidential axis'' (p. 107), where the author in fact
means ''the evidential pole of the epistemic-evidential axis''. It appears
that the book was not proofread prior to publication. In fact, the
introduction does not mention any proofreading.
It is the task of any review to be critical and point out potential
problems. In this review I have mentioned several such problems.
They should not distract from the fact that overall this book is a very
valuable contribution to the discussion on modality from a functional
perspective, which in some parts presents data that should be of
interest to anyone researching modality. In the opening chapters, the
author reveals a good instinct for topics which are currently relevant in
the discussion on modality. In chapters 6 and 7 she comes up with an
in-depth analysis of the relationship between modality and other
grammatical categories, particularly aspectuality, resulting in the
presentation of intriguing data and hypotheses.
Abraham, Werner (1995) Deutsche Syntax im Sprachenvergleich.
Grundlegung einer typologischen Syntax des Deutschen Tübingen.
Tübingen: G. Narr.
Aikhenvald, Alexandra Y. (2004) Evidentiality. Oxford: Oxford
Heine, Bernd (1995) Agent-oriented vs. Epistemic Modality: Some
Observations on German modals. In Bybee, Joan and Suzanne
Fleischman (eds.) Modality in Grammar and Discourse. Amsterdam:
John Benjamins, 17–53.
Kiefer, Ferenc (1987) On defining modality. Folia Linguistica XXI/1, 67–
Lehmann, Christian (1985) Grammaticalization. Synchronic Variationa
and Diachronic Change. Lingua e Stile 20/3, 303-319.
Leiss, Elisabeth (2002) Explizite und implizite Kodierung von
Deontizität und Epistemizität: Über die grammatische Musterbildung
vor der Entstehung von Modalverben. Jezikoslovije 3/1-2, 69-98.
Maynard, Senko K. (1993) Discourse Modality. Subjectivity, Emotion
and Voice in the Japanese Language. Amsterdam: Benjamins.
Narrog, Heiko (2005) On defining modality again. Language Sciences
27/2. pp. 165-192.
Palmer, F. R. (1986) Mood and Modality. Cambridge: Cambridge
Traugott, Elizabeth Closs (1989) On the Rise of Epistemic Meanings
in English: An Example of Subjectification in Semantic Change.
Language 65/1, 31–55.
Traugott, Elizabeth & Dasher, Richard (2001) Regularity in Semantic
Change. Cambridge: Cambridge University Press.
Van der Auwera, Johan & Ewa Schalley (2004) From optative and
subjunctive to irrealis. In: Brisard, Frank et al. (eds.) Seduction,
community, speech: A Festschrift for Hermann Parret. Amsterdam:
John Benjamins, 87-96.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Heiko Narrog is an associate professor at Tohoku University, Japan.
His research interests include historical linguistics, syntax and
semantics, modality, linguistic typology, and the Japanese language.