AUTHOR: Ariel, Mira
TITLE: Pragmatics and Grammar
SERIES: Cambridge Textbooks in Linguistics
PUBLISHER: Cambridge University Press
Dinha T. Gorgis, Jadara University, Irbid, Jordan
This book attempts to resolve a number of intriguing problems related to the
complex relationship between grammar and pragmatics. While maintaining the
mainstream conviction about the grammar/pragmatics division of labor, the author
seems to be more willing to announce a happy marriage between the two, though
admitted (explicitly and/or implicitly) to be at times uneasy bedfellows, than
keep them single. If they are kept absolutely distinct from each other, as
suggested by standard analyses, Ariel argues that accounts for grammaticization
and semanticization will not be possible (cf. pp. xiii; 257). The book opens
with a preface followed by an introductory chapter and six more chapters
distributed over three parts, each of which addresses questions intended to be
answered on the basis of her mostly natural linguistic/discourse examples
collected from various sources, mainly representative of Hebrew and English.
Original examples and their glosses can be found at:
www.cambridge.org/0780521559942. A cursory look at the rich list of references,
followed by two indexes, is indicative of the tremendous efforts she has exerted
on preparing the book.
Chapter 1: Grammar, pragmatics, and what's between them (pp. 1-24) draws our
attention to the fact that although we need to draw a distinction between
grammar, as correlated with a set of codes, and pragmatics, as correlated with
different types of inferences, to which part I is devoted, we equally need to
account for how inferences cross over and become codes. In addition to being
content with ''the now well accepted assumption that we always communicate by
combining codes (grammar) with inferences (pragmatics)'' (p.3), which means that
inferences constitute an integral part of grammar by definition, the author is
committed to the issue of cross-over which is believed to have serious
implications for grammaticization and/or semanticization and hence the on-going
development of current grammars. Ariel appeals to Grice's pragmatic theory; for
since it is assumed ''that every act of communication is actually inferential''
(p. 4), i.e. involving additional or complementary interpretation, it follows
that all pragmatic theories are essentially Gricean. As such, the author's
position in this book ''is contra the assumption made by other linguists, that
there must be a purely grammatical literal meaning which corresponds to the
complete utterance (usually assumed to be a single complete proposition) ...
[for it] may well be a combined grammatical/pragmatic representation in most
cases'' (p. 24).
Chapter 2: Distinguishing the grammatical and the extragrammatical: referential
expressions (pp. 27-67) focuses on some referential expressions, mostly definite
expressions and pronouns. In order to provide arguments for drawing the division
of labor between interpretations and language use conditions whose
grammatical/pragmatic identity is opaque, presuppositions are given a special
treatment because the available literature about their assignment is divided.
The heated argument, based on case studies and ''contentious'' examples, shows
that some aspects of the use of presuppositions are semantic, others, pragmatic.
Certain other referring expressions seem to require explanations by making
appeal to extralinguistic principles but, as evidence shows, a grammatical
convention is also involved. Preferred Argument Structure constraints as well as
register-specific referring expressions are claimed to fall in between, i.e.
neither encoded nor inferred. In Ariel's words, the conclusion is that ''one
aspect of the use of definite descriptions is grammatically encoded, the other,
pragmatically inferred'' (p. 44).
In part I (pp. 25-109), we are told that ''some interpretations are implicated,
some are explicated, and yet others are only potential truth compatible inferred
interpretations'' (p. 25), which chapter 3 (pp.68-109) handles in conjunction
with codes. For the purpose of assignment, Ariel selects a number of issues,
e.g. conjoined clauses with 'and' and some scalar expressions such as 'most' and
'all'. While Gricean pragmatists variously rely ''on the criterion of truth
conditionality for distinguishing grammatical and extragrammatical
interpretations, taking truth-conditional meanings as semantic and
nontruth-conditional meanings as pragmatic'' (p. 69), Ariel's position is
consonant with Relevance theoreticians ''who apply the code/inference distinction
strictly'' (p. 69). For them, Grice's conventional implicatures, which are the
domain of pragmatics, constitute coded meanings and hence semantic in essence.
Towards the end of the chapter, however, Ariel admits that it is ''a delicate
matter to distinguish the truth-compatible from the encoded'' (p. 108).
In part II: Crossing the extralinguistic/linguistic divide (pp. 111-256), the
author takes the reader to an arena where the grammar/pragmatics interface is
likely to take place. Central to this part is the argument that ''pragmatics,
together with other extragrammatical triggers, provides the raw materials and
impetus for grammar'' (p. 111). Following such an argument, mainly based on the
complex relationship between diachronic and synchronic facts, the reader may at
this junction draw two assumptions: (1) the linguistic/extralinguistic divide is
a prerequisite for the grammar/pragmatic divide; and (2) our current and future
grammars can equally contribute to the building up of foreseeable grammars and
hence witnessing an ever running process of grammaticization (or
grammaticalization, if you wish).
Chapter 4: Grammar, pragmatics, and arbitrariness (pp. 117-148) begins by
addressing the most crucial question, viz. whether grammar is
extralinguistically motivated or arbitrary. In introductory linguistic courses
we often tell our students that grammar is a self-contained system. If
autonomous, then it must be arbitrary. We do this for the sake of purely
describing form-function correlations without seriously embarking on possible
motivations. Yet, if we, as researchers, are not prepared to accept the simple
fact that ''grammar is a natural historical product'' (p. 148), then it follows
that human history is ''chaotic'' and ''arbitrary''. In fact, one of the main
reasons why certain linguistic phenomena seem to us arbitrary is the lack of
evidence. Don't we often frown at innovations in language use, which are likely
to end as conventional codes by which others abide gradually?
Chapter 5: All paths lead to the salient discourse pattern (pp. 149-211)
addresses a number of questions intended to show how ''grammar is constantly in
the making'' (p. 149). A number of factors, viz. (embodied) cognition,
sociocultural norms and inferential practices (pragmatic enrichments), not to
mention grammar itself (cf. p. 150), may all conspire in the creation of new
codes. It is undeniable that the world in which we live, ''the world filtered by
the human cognitive make-up'' (p. 151), can have a considerable impact on the
molding and re-molding of our grammars. In Ariel's words, all of these
''constitute an integral part of communication, because the linguistic code is
forever under-determinate'' (p. 166). It is to be noted that the influence of
extralinguistic factors on grammar is not direct. Rather, it is the mediating
salient discourse patterns, of which only a small set (as deemed necessary)
undergoes conventionalization (cf. p. 211).
Chapter 6: The rise (and potential fall) of reflexive pronouns demonstrates why
and which salient discourse patterns actually do turn grammatical. Current
English reflexive pronouns are used as an example to show that ''earlier
grammaticizations do not preclude newer ones, and newer ones do not always
cancel out old ones'' (p. 213). The English reflexive pronoun, ''which we are used
to thinking of as a type of referring expression, is historically tied to two
quite distinct grammatical categories. It evolved out of an originally emphatic
adjunct, which acquired argument status. As such, it became a marked form, used
for marked interpretations'' (p. 253). But once it is used intransitively, as is
sometimes the case today, ''the reflexive pronoun will lose not only its
interpretative markedness, but also its argument status (once again)'' (p. 253)
and hence the widespread use of the nonreflexive form, i.e. regular pronouns, as
perhaps a more marked use in the future.
The few introductory pages in part III: Bringing grammar and pragmatics back
together (257-308) round off the whole argument run in parts I and II. While
they remind the reader of what has been achieved so far, on the one hand, they
are meant to enable us to link the previously discussed meanings, e.g. the
so-called 'conveyed' and 'bare' meanings, with an intervening (or intermediate)
'basic level' meaning, captured by the fashionable, but highly controversial,
'what is said' concept, on the other hand. This third level of meaning
representation is the topic of chapter 7: Grammar/pragmatics interfaces (pp.
261-308), which is mainly a critical review of some important accounts of 'what
is said' in the literature starting with Grice. The chapter discusses ''the
possibility that 'what is said' is not only important for grammar/pragmatics
interfaces during interaction, but also in processes in which pragmatics crosses
over to become grammar'' (p. 261). To conclude, a word of caution is in place at
this juncture. The author's position is that explicated inferences (or conveyed
meaning level), rather than implicated inferences, would serve as the immediate
impetus for, or potentially give rise to, most semanticizations and
grammaticizations (cf. pp. 306-307).
With the exception of a few invented examples, this book utilizes a wealth of
references, a comprehensive literature overview indeed, and naturally occurring
data, which enables the author to draw subtle pragmatic distinctions while
simultaneously offering a wide range of convergent/divergent perspectives, all
being undoubtedly indicative of her strong research background. The book has a
clear structure. It is equally error-free, but perhaps with the exception of the
third occurrence of the word ''implicatures'' (p. 81 fn. 13) which must be
At the outset, one might be curious to ask Mira Ariel why choosing to order
''Pragmatics'' before ''Grammar'' as the title of the book when her consistent use
of their reversal is quite evident throughout the whole book, including part and
chapter titles. One may simply attribute concept order to prominence (or focus,
if you like) which captures the main argument of the feeding relationship, i.e.
the fact that pragmatics, including whatever extralinguistic factors, is held
responsible for grammaticization/semanticization. If the initial ordering and
its reversal are never meant to invite inferences, should we say that a paradox
in ordering is optional and hence arbitrary or even meaningless? Suppose the
following conversation is valid in a possible world:
Teacher: Did you like Ariel's _Pragmatics and Grammar_?
Student: Yes, I did, but not in this order.
Teacher: How's that?
Student; Well, I'd rather place ''Grammar'' before ''Pragmatics''.
Teacher: But why?
Student: Can you ever draw inferences without an existing code, including gestures?
(cf. p. 285 for a similar example).
I do not wish to push it any further and compare it with the chicken-egg puzzle,
but I still wonder. And if an answer is provided, I'm afraid (delayed) repair
work should have been given due attention as regards (written) communication.
The second remark I'd like to make is the author's consistent choice of ''her''
for 'speaker', 'person', 'participants' (A, B, ...) and even 'dog' (cf. p. 307),
perhaps with two exceptions where in the first she refers to the addressee as
'his' (p. 49, fn. 27) and in the second both 'she' and the neutral 'they' (p.
291, fn. 29). Since she discusses 'reference' and 'anaphora', could her
extensive use of 'she' (cf. her 'frequency' throughout) be indicative of
femininization and hence susceptible to foreseeable grammaticization? One
wonders if English will ever have two competing codes, viz. a she-grammar vs. a
he-grammar, in the future.
Documentation is impressive, yet often felt extravagant and annoying at times.
The frequently squeezed references and cross-references in a work which is
intended as a textbook may have side effects, particularly on the smooth flow of
information. The book is highly theoretical and argumentative in a number of
places and hence presupposing too much background knowledge on the part of the
(average) reader. Just as certain things are not clear to the author (cf. p.
152, fn. 4), we do not expect all students to make their way through the
condensed theoretical account about, e.g. iconicity and world view (p. 152).
Equally in the absence of examples, e.g. left dislocations and their ''three
distinct discourse functions'' (p. 120), the reader would likely be unable to
follow the argument. To substantiate this particular point, I gave the following
extract to some of my M.A. students: ''one can measure the syntactic
(in)dependence of combined clauses by the (in)dependent choice of tenses,
illocutionary forces, participants, etc. Thus, causative and modal relations are
high on the cohesion hierarchy, representing one perceived event, temporal
adverbials and conditionals are intermediate, and two unrelated events
(propositions) end the cohesion hierarchy'' (p.152).
In a word, they were totally lost, simply because no example is provided. While
she is a supporter of Grice's maxims, she appears to violate those maxims,
though unintentionally, in a few places.
Confusion may also arise with term-distinction. For example, it's only after
reading 94 pages that Ariel asks us to see Sperber and Wilson (1987) ''for
arguments against positing a distinction between generalized and particularized
conversational implicatures'' (p.94, fn. 22). Much earlier (p. 22), she already
seems to be for the idea that ''many cases analyzed as generalized conversational
implicatures by neo-Griceans are analyzed as explicated inferences by Relevance
theoreticians ... [in which case] the inferences are non-prototypical
implicatures.'' What makes scalar implicatures involving 'most', 'not all', 'more
than half', for example, generalized conversational implicatures? Is it because
they easily lend themselves to guesses? Are particularized conversational
implicatures ''generated only in specific contexts'' (p. 98)? If 'most' were
compatible with 'all', as concluded by Horn (2006) and reported by Ariel (p.
92), 'most' would have been synonymous with 'all', which is not the case. The
quantifier 'all' was most likely not intended by the speaker, Dana, whose 'most'
turned out to be 'all'. If you asked Dana after the event: ''Did you expect
'all', without exception, would prefer square plates?'', she would most likely
answer: ''Well, not really'', or ''I guess so'', etc. If this were not the case, we
could also say that ''most likely'' is equally compatible with 'absolutely',
'certainly', 'positively', etc., which is not intended. If it is a matter of
focusing on ''informativeness'' only, then this would be fine. Still,
possibilities and probabilities of meaning dominate her overview of arguments
for/against implicature/explicature accounts of the upper bound of 'most'.
In fact, one could come to the conclusion that the best part of the book is
really much less about grammar/pragmatics interface per se than
semantic/pragmatic overlap/interface. One would have liked to see, e.g. Quirk et
al. (1985), a classic code, to be used as a point of reference for grammatical
points but, unfortunately, it is only footnoted in one place (p. 252, fn. 61).
Unlike Van Valin (2005), which looks at how syntax, semantics and pragmatics
interact in different ways across human languages, Ariel's book, though focusing
on the ''grammaticization of pragmatics'' (p. 159), is a kind of grammar that
sounds very much lexically biased. This is evident in, e.g. her support of a no
''one-to-one relationship between cognitive concepts and grammatical categories''
(pp. 167-169) as in the case of 'washing' and 'hiding', while maintaining that
''the grammar goes the discourse, rather than the cognition, way'' (p. 169) and
hence much in agreement with Hopper and Thompson (1980; 1984; 1993) who are of
the view that semantic concepts are derivative of discourse functions, rather
than the other way round (p. 172). Except for the development of S-modifiers (p.
297) and 'since' (p. 307), the rest of the argument presented in Chapter 7 is
devoted to solving the controversial question of 'what is said' (see Levinson
2000 and Ariel's summary, p. 292), which is obviously a pragmatic/semantic issue.
This is not to deny that she has tackled grammatical, including a number of
phonological/morphological, issues. In one place (p. 192), she says that ''some
strings are more useful to speakers than others''. As such, we would expect users
of Standard Arabic to feel content with the construction /sa?aktubu/ 'I'll
write' and thus ridding of the categorical boundary in the equivalent expression
/sawfa ?aktubu/. Such a preference may be ''privileged'' on the basis of
''frequency'' (cf. p. 191), but since negation is blocked in the construction,
users normally negate the expression: /sawfa lan ?aktuk/ 'I'll not write'., at
which classical grammarians would frown, though. In another place (p. 186; cf.
also p. 191), she seems to agree with Hooper's (1976) argument when she writes:
''That a potential reductive sound change applied to the original schwas in
_every_, _artillery_, and _memory_ according to their respective frequencies.
The fact that _every_ has no schwa preceding the /r/ now, but _artillery_ does,
is due to the fact that the former is a highly frequent word, whereas the latter
is quite rare. In between in terms of frequency is _memory_, which is also
phonetically intermediate, with a syllabic /r/.''
My position is that in _every_ primary stress is retained, whereas stress would
shift to the pre-antepenultimate syllable if the schwa were reduced in
_artillery_. Since the majority of multisyllabic words in English receive their
primary stress on the antepenultimate syllable (my dictionary-based conclusion),
due to its prominence, the schwa must be maintained in the penultimate syllable
as constrained by English phonology elsewhere.
The ''surprising facts'' (p. 254) for which she posits seven stimulating questions
are not very much different from the ones called ''ugly facts'' (p. 195) in
essence; for there's no question about their pragmatic motivations, including
whatever extralinguistic forces. At this point, one may wonder if there is any
well-defined borderline between pragmatic and extralinguistic factors. Why not
call any phenomenon that does not count as an integral part of the code, i.e.
grammar, just extra or non-linguistic? I agree with her on the rather minor role
played by language academies (p. 182, fn. 28), including prescriptive
grammarians, on the process of grammaticization/semanticization, particularly in
the Arab world, but what about the globalization of English (cf. Englishes),
borrowing, translation, immigrants' use of English as a second language, the
media, the co-existence of bilingual and multilingual communities, for example?
Which of these can be categorized as linguistic and/or extralinguistic or pragmatic?
Ariel, after all, is fully aware that the ''road to grammar is still not
obstacle-free'' (p. 209). That said, Ariel's use of ''potential'' so frequently
that factors affecting linguistic change will forever remain competitive and
hence justifying the ''paths'' she has chosen for the bride and groom, named
grammar and pragmatics, will undoubtedly pave the road for other researchers to
follow up the numerous challenging issues which await verification in light of
fresh data secured from other languages. Her reference to Haspelmath (2004),
being written at the same time as her chapter 6 ''without either knowing about
the other'' (p. 213, fn. 4), is a sign of honesty and credibility, which we are
compelled to admire. Though found (sometimes) a bit tiring at certain
crossroads, I must say I enjoyed reading the book.
Haspelmath, Martin. 2004. A frequentist explanation of some universals of
reflexive marking. Paper presented at the ''Reciprocity and reflexivity –
description, typology and theory workshop''. Free University of Berlin, October 2004.
Hooper, Joan B. 1976. Word frequency in lexical diffusion and the source of
morphophonological change. In W. Christie, ed. _Current progress in historical
linguistics_. Amsterdam: North-Holland, 96-105.
Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in grammar and
discourse. _Language_ 56: 251-299.
Hopper, Paul J. and Sandra Thompson. 1984. the discourse basis for lexical
categories in universal grammar. _Language_ 60: 703-752.
Hopper, Paul J. and Sandra Thompson. 1993. Language universals, discourse
pragmatics, and semantics. _Language Sciences_ 15: 357-376.
Horn, Laurence R. 2006. The Border Wars: a neo-Gricean perspective. In Ken
Turner and Klaus von Heusinger, eds., _Where semantics meets pragmatics_.
London: Elssevier, 21-48.
Levinson, Stephen C. 2000. _Presumptive meanings: The theory of generalized
conversational implicature_. Cambridge, MA: MIT Press.
Quirk et al. 1985. _A comprehensive grammar of the English language_. London:
Sperber, Dan and Deirde Wilson. 1987. Précis of Relevance: Communication and
cognition. _Behavioral & Brain Sciences_ 10: 697-754.
Van Valin, Robert D. 2005. _Exploring the syntax-semantics interface_.
Cambridge: Cambridge University Press.
ABOUT THE REVIEWER
Dinha T. Gorgis has been teaching several linguistic modules, including English
grammar, discourse, pragmatics and translation, at a number of Arab universities
since 1975, and is currently professor of linguistics at Jadara University in
Jordan. He is chief editor of STJ, member on the editorial (advisory) boards of
Linguistik and TLJ online, IPrA and WATA member. His latest publications are: ''
English and Arabic conceptual metaphors of anger: Implications for translation''
(STJ, 2008) and ''The translation of Arabic collocations into English:
Dictionary-based vs. dictionary-free measured knowledge'' (Linguistik, 2009,