Review of Clause Linking and Clause Hierarchy
EDITOR: Isabelle Bril
TITLE: Clause Linking and Clause Hierarchy
SUBTITLE: Syntax and Pragmatics
SERIES: Studies in Language Companion Series 121
PUBLISHER: John Benjamins
Paul Isambert, University of Paris 3, France
This volume is a collection of studies of clause-linkage in a wide set of (often
little-studied) languages, focusing especially on subordination. The authors
investigate strategies as diverse as converbs, special conjugations, or markers
of information structure, in many different language families. The focus is,
quite obviously, on diversity.
The book's introduction by Isabelle Bril is a brief review of the properties of
coordination and subordination, followed by a summary of the chapters.
PART I: Syntactic terminology and typological methods
In ''Clause linkage and Nexus in Papuan languages,'' William A. Foley
reinvestigates what he previously called ''cosubordination'' (Foley and Van Valin,
1984), a conjunction of clauses with features from both subordination (one
clause depends on the other for inflection) and coordination (there is no
embedding). Foley now considers such structures to be instances of coordination,
with the particularity that S constituents are coordinated under a common IP
node, whereas usual coordination links IP constituents. However, not all
inflection features behave similarly; negation in Tauya (Papuan), for instance,
attaches to some of the verbs in a chain, while illocutionary force has scope
over the entire chain. Polarity is thus construed as belonging to a lower level
in the IP projection.
Balthasar Bickel's approach in ''Capturing particulars and universals in clause
linkage: A multivariate analysis'' is expressed simply as ''measuring instead of
reducing diversity'' (p. 55). Bickel remarks that trying to account for similar
(but not equivalent) clause-linking constructions in e.g. Swahili (Niger-Congo)
and Belhare (Sino-Tibetan) in the same way is arbitrary, while not acknowledging
their similarity (positing two language-dependent structures) lacks heuristic
value. Instead, the author breaks up those structures into sets of variables
used to describe 69 constructions from 24 languages and thus measures how much
they resemble each other; correlations between features can also be measured.
The resulting picture casts doubts on such well-defined concepts as
''subordination'' or ''coordination''; while a cluster of structures one could tag
as ''subordination'' emerges, chaining constructions do not gather together but
spread over a continuum of structures with varying scope properties.
PART II: Clause-chaining, converbs, masdars, absolutive constructions, etc.
In ''Specialized converbs and adverbial subordination in Axaxdərə Akhvakh,'' Denis
Creissels investigates the use of a rich set converbs in a Nakh-Daghestanian
language to express adverbial subordination (manner, temporal relations, etc.).
Many of those forms are etymologically transparent; the author hypothesizes that
grammaticalization is constantly at work, rapidly renewing the inventory of
converbs, also witnessed by the many synonymous forms.
In ''Finite and non-finite: Prosodic distinctions on Budugh verb stems,'' Gilles
Authier investigates the use of verb-stems (uninflected for tense and mood) in
Budugh (Daghestanian) syntax. Those forms can both depend on nouns or other
predicates or be used independently in injunctive sentences. Interestingly,
although Budugh displays a rich verbal morphology, the distinction between the
two uses is made by prosody, the forms adopting either the prosody of nouns
(when dependent) or that of finite verb forms (when independent).
Christiane Pilot-Raichoor's ''Converbs and adverbial clauses in Badaga, a
South-Dravidian language'' studies adverbial subordination in a language of
Southern India, expressed with converbs, adjectival participles and a quotative
marker. As in Akhvakh, Badaga uses both a contextual converb (without a fixed
meaning) and specialized converbs. Nonetheless, no semantic nuance is
particularly tied to one or another construction.
In ''Coordination, converbs and clause chaining in Coptic Egyptian: Typology and
structural analysis,'' Chris H. Reintges investigates coordination in an
Afroasiatic language spoken in the third to thirteenth centuries. While a
symmetric conjunction is often used, Coptic Egyptian also employs special verb
conjugations, often ambiguous between coordination and subordination. Also,
those forms pertain to information structure, displaying focus-sensitivity
unlike neutral declarative clauses.
PART III: Subordination, informational hierarchy and referential hierarchy
Isabelle Bril's ''Informational and referential hierarchy: Clause-linking
strategies in Austronesian-Oceanic languages'' is a survey of subordination and
coordination in a large set of related languages, with an emphasis on
information structure. The author shows how topic and focus markers (themselves
often derived from connectors) are used as subordinators, as are
reference-tracking markers like demonstratives. Sentence-level syntax and
discourse devices are thus shown to be intrinsically tied to one another.
In ''Comment clause: Crossing the boundaries between simple and complex
sentences,'' Zygmunt Frajzyngier studies the marker ''wá'' in Wandala (Central
Chadic), which introduces a comment clause (in the Prague School sense of topic
and comment, not a parenthetical clause). Such a clause may follow a variety of
contexts: from topicalized noun phrases to verbs of saying to conditional
clauses. Hence the occurrences of ''wá'' cannot be predicted from what precedes
it, and simply indicates that something is to follow about what has been said
In ''Deixis, information structure and clause linkage in Yafi' Arabic (Yemen),''
Martine Vanhove shows how the two markers ''raʕ'' and ''ta'' used for deixis and
topic and focus particles are also employed in clause-linking. Although Yafi'
Arabic has other subordinating devices, information structure provides new
strategies which then grammaticalizes. Interestingly, the two markers have very
different origins, ''raʕ'' being the imperative of ''see'' while ''ta'' is a
demonstrative, and both go through the intermediate stages of presentative and
In ''The role of the Berber deictic and TAM markers in dependent clauses in
Zenaga,'' Catherine Taine-Cheikh studies the divergent evolution of the particle
''ad'' in Zenaga and the rest of Mauritanian Berber. The marker has many identical
functions, such as demonstrative, copula, connector, etc., but some of its uses
(conditional marker and future) exist only in Zenaga and are difficult to link
to the original demonstrative. The problem is solved thanks to interactions with
the TAM system and with the Aorist in particular.
Evangelia Adamou's ''Deixis and temporal subordinators in Pomak (Slavic, Greece)''
investigates the use of a deictic suffix to mark temporal subordination. The
choice of the suffix indicates the relation between the process and the
discourse situation, whereas the absence of that suffix marks the absence of
In ''Correlative markers as phoric 'Grammaticalised Category Markers' of
subordination in German,'' Colette Cortès studies markers (of determinative
origin) binding clauses (the markers holding the place of the subordinate clause
in the main clause), specifying semantic and pragmatic relations in complex
sentences. She argues against treating them as superficial, optional markers,
and considers them instead as fundamental clause-binding devices.
PART IV: Informational hierarchy and TAM markers' functions in clause-linkage
In ''Focus, mood and clause linkage in Umpithamu (Cape York Peninsula,
Australia),'' Jean-Christophe Verstraete shows how Umpithamu (Paman) makes up for
its lack of clause-linking markers by using resources from other domains: for
instance, potential mood marks a conditional relation, while prominent arguments
are used to relate a sentence to the previous one; in some cases the marker
contributes only semantic information (subordination being marked by shared
arguments), in other cases subordination is the result of an inference based on
Stéphane Robert's ''Clause chaining and conjugations in Wolof: A typology of
parataxis and its semantics'' investigates paratactic linkage based on
conjugations (related to tense, enunciation, or focus) in Wolof (Niger-Congo).
Various combinations produce various interclausal semantic effects: succession,
causality, opposition, etc. Such chaining yields more or less integrated
structures, from simple juxtaposition to full syntactic dependency.
In ''Pragmatic demotion and clause dependency: On two atypical subordinating
strategies in Lo-Toga and Hiw (Torres, Vanuatu),'' Alexandre François studies two
Oceanic languages where, despite no lack of subordinators, two TAM markers are
used to encode clause dependency: the Subjunctive lacks illocutionary force, and
the Background Perfect denotes presupposition. Clauses thus marked can't be
stand-alone utterances, and must be interpreted as dependent on other clauses.
Jacqueline Leroy's ''Tense-mood concordance and clause chaining in Mankon (a
Grassfields Bantu language)'' explores the use of verbal constructions to mark
syntactic and semantic links between clauses. Explicit coordinating or
subordinating markers can generally be done without.
In ''Clause dependency relations in East Greenlandic Inuit,'' Nicole Tersis
studies the use of verbal morphology to mark subordination in Tunumiisut
(Eskimo-Aleut). In narratives, such dependency can run on entire clause chains,
and actually only a minority of clauses are independent, thus turning dependency
into a discursive phenomenon. The author also addresses the similarity of the
verbal forms and possessive noun phrases, but dismisses (at least in synchrony)
the idea of reducing the verbal clause to a nominal form.
In ''Coordination and subordination: 'Áma' in Bulgarian dialectal Greek,'' Eleni
Valma studies a polysemous marker used for both coordination and subordination.
As a subordinator, it signals either temporality or hypothesis (depending on the
clauses it links). As a coordinator, the marker denotes opposition, and has
probably evolved out of contact with Turkish.
The evaluation of this volume will be twofold: on the one hand, a criticism of
the editorial work; on the other hand, a praise of its diversity.
The editorial work can be criticized in both form and contents. Concerning form,
the book would have benefited from better proofreading. For instance, at the end
of his introduction (section 1), Gilles Authier outlines the rest of his pape ,
enumerating the first, second, third … sections, which actually are sections 2,
3, 4 … and mentioning sections that do not exist. In Catherine Taine-Cheikh's
paper, section 3.4 follows section 4.3.3. Frequently, the list of abbreviations
at the end of each chapter is incomplete or some entries are misplaced. The
morpheme-by-morpheme glosses are sometimes vague: in Christiane Pilot-Raichoor's
paper, example 25 has ''bandale'' translated as ''come.HYP'' (i.e. conditional),
while in example 27 it is translated simply as ''come'' (although it's still the
conditional form). As a final example, the topic index is unusable: the entry at
''discourse'', for instance, refers to a hundred of undifferentiated occurrences:
one sends to ''the discourse function of the converbal clause'' (p.121), the next
to ''Budugh discourse'' (p.162), then to ''discourse settings'' (p.167), and so forth.
As for contents, what the volume lacks is a thorough introduction. One would
have appreciated an in-depth analysis of the concepts at hand: coordination,
subordination, converbs, masdars etc., along with a summary of theoretical
stances; failing that, it is sometimes hard to compare the contributions,
because one doesn't really know if the authors investigate the same phenomena,
all the more as they sometimes rely on different (and incompatible) theories.
I'm not saying the authors should have been forced to adopt one theoretical
framework; but the editor should have made clear, from the very beginning, how
different approaches concur or conflict.
One could say that the first part of the book is the theoretical introduction I
was expecting. But Foley's paper is much too parochial, using concepts that make
little sense outside generative grammar, and it can't lay the ground for the
chapters that follow (only a few allude to generative grammar, but none really
enters the theoretical field). On the other hand, Bickel's refusal to reduce
diversity makes sense in a cross-linguistic perspective only, and the other
contributions generally focus on one language.
Yet there is another, good reason why no general picture emerges from the book
(which incidentally seems to confirm Bickel's position): the diversity of
languages. Indeed, the terse summaries above don't do justice to the book's main
quality: presenting a wealth of data on numerous and diverse (and often
underdocumented) languages. Each chapter introduces the reader to a new language
with its own constructions. Many papers take time to present the language's
grammar, even when it's not directly relevant to the issue at hand.
From Tunumiisut to Wolof, from the Daghestanian family to the Austronesian
languages, and from morphological details to discourse strategies, the book is a
fascinating ride through (and a wonderful tribute to) the many ways human beings
express themselves, focusing here on one issue that could be summed up simply
as: how do you express two things conjointly? The chapters are often not much
more than raw catalogues of forms and uses, but that's precisely their strength:
they provide valuable data instead of uncertain concepts. That is also the
reason why, I believe, this volume is an invaluable addition to any linguist's
bookshelf, a reference you'll come back to again and again.
Foley, W. A. and Van Valin, Jr., R.D. (1984), _Functional Syntax and Universal
Grammar_, Cambridge University Press, Cambridge.
Nedjalkov, V. (1995), Some typological parameters of converbs, in ''Converbs in
Cross-linguistic Perspective,'' M. Haspelmath and E. König (eds.), 97-136, Mouton
de Gruyter, Berlin.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Paul Isambert has just completed his PhD at the University of Paris 3,
France. He's currently working on grammaticalization and discourse
structure, especially concerning topic shifts and anaphora.