LINGUIST List 12.845

Mon Mar 26 2001

Review: Rebuschi & Tuller, Grammar of Focus

Editor for this issue: Terence Langendoen <>

What follows is another discussion note contributed to our Book Discussion Forum. We expect these discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in.

If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for discussion." (This means that the publisher has sent us a review copy.) Then contact Simin Karimi at or Terry Langendoen at


  • Elena Maslova, Review of Rebuschi & Tuller, Grammar of Focus

    Message 1: Review of Rebuschi & Tuller, Grammar of Focus

    Date: Mon, 26 Mar 2001 12:34:52 -0800
    From: Elena Maslova <>
    Subject: Review of Rebuschi & Tuller, Grammar of Focus

    Georges Rebuschi & Laurice Tuller (eds.) The Grammar of Focus. Linguistik Aktuell (Linguistics Today). John Benjamins. Amsterdam/Philadelphia. 1999.

    Elena Maslova, University of Bielefeld

    This collection of papers grew out of a Paris workshop (1996) of the same name. The goal, as it is stated in the editors' introduction, is "an appraisal of the grammar of focus", after over 20 years of intense cross-linguistic investigations within the previous incarnations of the generative theory and in view of its most recent radical reformulation (Minimalist Program). The editors' introduction outlines the evolution of focus studies in the generative tradition and the current "focus issues", which are elaborated in eleven individual papers dealing with different languages and focusing on various aspects of the topic (that is, the grammar of focus).

    Manuela Ambar's ("Aspects of the Syntax of Focus in Portuguese") goal is to arrive at a unified account of various focus-marking devices in Portuguese (and, apparently, elsewhere). The proposed solution is based on the hypothesis that (any) focus requires a "topic-like" element (different from "pure topic"); accordingly, there is a functional projection called TopicFocus Phrase, and Spec, TopicFocusP is occupied by this hypothetical topic- like element. Further, it is assumed that a contrastive focus contains its own topic-like element, whereas a non- contrastive one does not; this accounts for their different positions with respect to the verb.

    Joseph Bauer ("Bound Focus or How can Association with Focus be Achieved without Going Semantically Astray?") examines the syntax of focusing particles (like 'only') in English and German. His goal is to reduce all possible constructions with such particles to a single constellation in which the particle occupies the operator position (and heads a "particle phrase"), as in 'John would *only* invite Mary'. The paper shows that the minimalist theoretical machinery does license such a reduction, and this is because overt and covert operations are treated in a strictly uniform fashion. However, the elegant theory of bound focus developed in the paper requires covert movement of entire phrases, whereas the assumption of pure feature movement leads to a host of false predictions.

    Anne Clech-Darbon, Georges Rebuschi & Anni Rialland's paper, "Are There Cleft Sentences in French?", argues for the negative answer to this question: the authors claim that there is no specific "construction" corresponding to what is traditionally called cleft sentences in French (and probably elsewhere). More precisely, they propose to analyze the post-focal relative clause as right-adjoined to an ordinary identificational clause, whereby 'ce' is assumed to contain a predicate variable. It is claimed that this analysis predicts the prosodic (duplication of "terminal intonation") and semantic (uniqueness of the focused term) properties of cleft sentences.

    Nomi Erteschik-Shir ("Focus Structure and Scope") develops a model of information structure, which she calls "f-structure theory". F-structure (an annotated structural description in which Topic and Focus constituents are marked) is intended to replace the logical form (LF) component; the scope relations in f- structures are transparent; topic quantifiers always take the wide scope. The (pragmatic) interpretation of f- structures is based on an elaboration of Reinhart's "file card" metaphor.

    Larry M. Hyman ("The Interaction between Focus and Tone in Bantu") presents a detailed description of focus- related tone effects in Bantu; examples come from a variety of Bantu languages, selected to show the range of possible focus-tone interactions. Hyman shows that there is no direct focus-tone link (that is, no examples where "semantic focus" would unambiguously condition a [+focus] tonal effect, or where the absence of "semantic focus" conditions a [-focus] effect). In reality, the focus- related tonal features are conditioned by certain grammatical configurations, which correlate (but, again, not unambiguously conditioned by) with "semantic focus".

    Sarah D. Kennelly ("The Syntax of the P-Focus Position in Turkish") explores a sub-class of focused elements (P- Focus, P for "presentational"), which includes (minimally) non-specific objects (as indicated by the absence of the Accusative marker) and interrogative elements. It is proposed that non-specifics and wh- elements constitute a "natural class"; they are described as existential quantifiers of a free-choice partial function. These elements (when focused) are taken to occupy an "F" position, which is adjoined to VP. The analysis is motivated primarily by the fact that P-Focus only has a cumulative reading with respect to the subject (i.e. no scope distinctions).

    Ayesha Kidwai ("Word Order and Focus Positions in Universal Grammar") examines four languages: Hindi-Urdu, Malayalam, Western Bade and Tangale. The paper argues against positing a "Focus Phrase" and develops a minimalist theory of [+FOCUS], which is argued to be [P(honetic) F(orm) [+Interpretable]] and may therefore survive to the interface unchecked, which explains the cross-linguistic heterogeneity of focus-marking mechanisms (word order, morphology, intonation). It is proposed that structures derived by PF-movement are interpreted at a level distinct from the LF, Domain D(iscourse), which is located at the edge of the PF- component.

    Alain Kihm ("Focus in Wolof. A Study of What Morphology May Do to Syntax") claims that (contrastive) focus sentences in Wolof are bi-clausal clefts (just like their English counterparts). This similarity (or even identity?) is blurred by the morphological complexity of the copula in Wolof. On the other hand, Wolof draws a formal distinction between two structures ("focus" vs. "presentational") which happened to coincide in one formal structure in English (and other languages). Interestingly, although the analysis of "real" (i.e. focus) clefts is essentially the same as suggested by Clech-Darbon et al. for French, yet the interpretation is quite different: not only the existence of clefts is not rejected, but this structure is taken as a kind of universal "focusing" structure (the cross-linguistic differences being due to morphology alone).

    Jacqueline Lecarme ("Focus in Somali") examines the distribution of "focus markers" (particles 'baa/waa') in Somali. She shows that this distribution cannot be (fully) accounted for in functional terms, which is taken as evidence that these particles are not discourse markers at all. It is argued that the "focus markers" are overt 'root' complementizers; the focus position is a structural Case position (providing cofigurational positions in a "non-configurational" language).

    Jon Ortiz de Urbina ("Focus in Basque") reviews various approaches to focalization in Basque, where focus phrases must occur immediately to the left of the verb. It is argued that there is a functional projection to which both foci and wh-words move. The problem is to identify, in a thoroughly head-final language like Basque, some left-headed functional category which may serve as an appropriate landing site for both operators and heads. Ortiz de Urbina compares different proposals for such phrase.

    Jamal Ouhalla ("Focus and Arabic Clefts") examines focus constructions in Moroccan Arabic, which consist of a focused constituent and a free relative clause linked by a pronominal copula. Since free relative clauses can be analyzed as noun phrases, this structure is identical to that of equative sentences. On the other hand, free relatives can only have a set reading; i.e. they cannot denote an individual. Therefore, focus constructions are best analyzed as existential statements over choice functions: the focus phrase is the individual obtained by applying a choice function to the free relative denoting a set.

    An important question (which the book does not answer) is what is actually being studied. In the introduction, the editors carefully and deliberately avoid formulating any definition of focus; they claim, however, that the book shows that "[t]here [...] appears to be consensus on what the focus issues are, which surely amounts to progress in our understanding of this grammatical phenomenon" (p. 17). Let alone the (partly incompatible) theoretical solutions (after all, as the editors correctly observe, these solutions are strongly theory-dependent), there appears to be no consensus on a more basic descriptive question: how focus constructions are identified in each specific language? Since the formal heterogeneity of "focus constructions" analyzed in the volume is beyond any doubt, the grounds for such identification can only be functional; what is, then, the function (call it "FOCUS") that identifies a language-specific device as something to be subsumed under the title "The Grammar of Focus"?

    To begin with, Lecarme claims explicitly that the phenomenon she is studying (in Somali) does not encode "FOCUS" (and, for that matter, no "fixed discourse function"). This is shown by examples where "focus particles" are attached to something that is clearly not "FOCUS". Notably, the focus particles pass the "question- answer" test employed as the crucial diagnostic by some other authors. Moreover, Hyman makes virtually the same claim about Bantu "focus-marking" (there is no perfect match between "form" and "function"). Yet the interpretations are entirely different: In Bantu, "FOCUS" (Hyman calls it "semantic focus") is apparently assumed to play a significant role in the choice of an appropriate grammatical construction, although the correlation is always imperfect due to the intervention (mediation) of grammar. In Somali, the choice of a construction is a "matter of optional selection"; it is claimed that focus markers "in most cases do not trigger any special pragmatic effect" (p. 284). This claim is followed by a really intriguing statement: both A and B (where A and B are sentences differing in "focus- marking") are "normal, unmarked clauses, simply because *it* is the only structure that is available" (ibid., emphasis mine). The question is, what is the antecedent of "it", i.e. what is the *only* structure available (A, B, or both)?

    Interestingly, other case studies presented in the book do not seem to be concerned about potential mismatches between "FOCUS" and the syntactic phenomena under discussion at all. Notably, Lecarme's paper is virtually the only one that invokes some authentic discourse data - to show that a function-based account would not explain the distribution of "focus markers". As the vast literature on information-packaging phenomena in traditions other than the generative one shows, even a cursory look at the actual discourse contexts of *any* construction always reveals the absence of a "perfect" form-function correlation (assuming, of course, that the function is defined independently of the specific construction). The fact that the contributors to this book do not look at actual data and base their analyses on sets of constructed examples would be, in itself, unproblematic: after all, different traditions are interested in different aspects of linguistic phenomena. The problem is that at least some of them implicitly make unjustified assumptions on the discourse function(s) of these phenomena: they are believed to encode the (undefined) function of "FOCUS" (once more, otherwise it is unclear how formally different constructions are identified). This may well be true - in some sense, but certainly not in the strict sense imposed by Lecarme. In any event, this belief must be justified by something beyond the authors' intuitions (cf. "we *believe* that the cleft structure supplies this sort of semantic information too" (Clech-Darbon et al., p. 103; emphasis mine)). Alternatively, focus constructions may be defined (and identified) without reference to their discourse function(s); but then again, it would be nice to have an explicit definition, if only to be sure that it applies to all constructions under discussion.

    As it seems, the skepticism about the possibility of scientific (linguistic) description of "meanings" leads, at least in this case, to an undesirable result: the overall impression is that whether or not a language- specific construction will be identified as a "focus construction" depends more on the theoretical stance taken by the scholar (*within* the generative tradition) than on the properties of the construction itself. If Lecarme were to apply her strict criteria to other languages, she would find no focus constructions whatsoever. If Ambar happened to study Somali data, she would posit a TopicFocus phrase (and probably something else), etc. And Hyman would attest an imperfect form- function correlation in each case (and this is probably the only conclusion that really satisfies the criterion of descriptive adequacy).

    Biographical statement As is probably clear from the critical remarks above, I do not belong to the generative tradition; I have studied information-packaging phenomena in a couple of languages known for their "focus-prominence" (most importantly, the Yukaghir languages of Northern Siberia), as well as cross-linguistically. I am interested both in the actual pragmatic function(s) of such phenomena and the cross- linguistic structural correlates of these functions, as well as in the methodological basis of cross-linguistic comparability of the relevant constructions.