Norma Schifano’s Verb Movement in Romance provides an extremely thorough and systematic treatment of the variation in the height of movement of the verb across a wide-ranging survey of Romance varieties and contexts, as well as a detailed explanation of how such movement is triggered and parametrized.

After the introduction in Chapter 1, Schifano devotes two chapters meticulously testing the height of verb movement across Romance. Schifano diagnoses the position of the verb through its placement with respect to a dozen different adverbs, which she assumes are arranged as in Cinque’s (Reference Cinque1999) cartographic hierarchy. Chapter 2 is dedicated to a sample of varieties across the Italian peninsula, while Chapter 3 covers French, Romanian, Spanish, Valencian Catalan, and both European and Brazilian Portuguese. Each chapter is divided into two sections: the first presents evidence for the ‘default’ placement of the verb in each variety, while the second discusses ‘microvariation’ in the location of different inflected forms of the verb, including auxiliaries, non-finite forms, and tense, aspect, and mood inflectional forms. Schifano argues for four targets of verb movement in Romance, as follows: high (French, Romanian); clause-medial (Sardinian and several other Italian varieties); low (European Portuguese and several Italian varieties); and very low (Spanish and Catalan).

The empirical coverage of these two chapters is stunning, but the downfall of a thorough presentation of data like this is that it is quite repetitive and overwhelming. More aid to the reader could have been provided in this respect, such as a brief survey of the descriptive generalizations and theoretical consequences at the beginning and repeated throughout, to help the reader contextualize the data. A dialect map of Italy would have also been a welcome aid. Nonetheless, the empirical contribution of this survey of data across Romance is quite valuable as a reference text.

The diagnostics used by Schifano assume cartography, an assumption which may not be well-founded. For example, Van Craenenbroeck (Reference Van Craenenbroeck2009) provides a whole volume discussing alternatives to cartography. On the other hand, though, Schifano’s diagnostics depend only on the empirical generalizations made within cartographic analyses about the ordering of adverbs, not on any of the theoretical claims. Schifano herself calls it an ‘empirical tool’ (4), and adds more caveats to her use of cartography throughout the book. Thus the use of the cartographic ordering of adverbs is likely valid in this instance, to the extent that these adverbs are universally ordered cross-linguistically (or at least within Romance).

In Chapter 4, Schifano implements a theory accounting for the different heights of verb movement attested across Romance. This theory has three main points, which I will summarize and evaluate in turn: (i) a comparison between the height of movement and the morphological richness of the inflectional paradigms of the verbs; (ii) a theory about how the movement of verbs can be triggered by formal features in the absence of morphological richness; and (iii) an analysis of how the attested variation in verb movement can be incorporated into a theory of parameters.

First, Schifano argues that the height of verb movement is correlated with the paradigmatic instantiation of tense, aspect, and mood of a given Romance language, where paradigmatic instantiation is defined as follows:

Schifano proposes that the four positions identified as the targets of movement in Chapters 2 and 3 correspond to the Mood, Tense, Aspect, and Voice fields. The verb in a language must raise to one of these fields if that value is not paradigmatically instantiated. Thus, movement is a syntactic strategy to license the tense, aspect, or mood interpretation of the verb if it is not formalized within the morphology. For example, Northern Regional Italian has separate synthetic indicative, subjunctive, and conditional paradigms; therefore, mood is paradigmatically instantiated, and the verb does not need to raise to the mood field. However, mood is marked through an analytic form in Romanian, and through forms syncretic with the indicative in French; in these two languages, mood is not paradigmatically instantiated, and the verb must raise.

Although the author begins with seemingly clear cases of correlation between paradigmatic instantiation and verb movement, as the chapter progresses, the analysis becomes tenuous at times. Although Schifano is always able to justify her classification of whether a given paradigm is paradigmatically instantiated or not, at times the justifications seem post hoc, rather than predictive. For example, although the paradigmatic instantiation of mood and tense are determined by comparing contrasts across the tense and mood paradigms, paradigmatic instantiation of aspect is determined by the distribution and use of a single paradigm, the perfect aspect. Furthermore, Schifano’s categorization process takes gradient empirical facts, such as if some but not all of the items in a paradigm meet the criteria for PI or if some paradigm is in the process of being lost, and uses these gradient properties to determine a binary value for paradigmatic instantiation, without criteria as to how much gradience is permitted. Moreover, although Schifano provides evidence that there is a correlation between paradigmatic instantiation and the height of verb movement, this evidence does not necessarily indicate that paradigmatic instantiation (or lack thereof) is what causes the verb to raise. Other possible analyses include that verb raising tends to result in the morphological expression of synthetic forms, or that some third process causes both verb raising and paradigmatic instantiation.

Second, Schifano implements her analysis through a theory of formal features. She argues that tense, aspect, and mood are all formal features in Romance. If one of these categories is paradigmatically instantiated in a given language, then it is an interpretable feature, and no movement is necessary. But if it is not paradigmatically instantiated, then it is uninterpretable, and it probes and attracts the interpretable feature of the relevant category (tense, aspect, or mood) on the verb.

Finally, Schifano makes the following two generalizations about paradigmatic instantiation (PI) in Romance:

Drawing on these two generalizations, Schifano suggests that the paradigmatic instantiation of categories can proceed bottom–up. If a category is found to be [ $+$ PI], the checking procedure can be stopped, as all remaining (higher) categories are also predicted to be [ $+$ PI]. Note that, in this discussion on the order of assessment of PI in Section 4.2.2, it is not clear if Schifano is discussing methodology in linguistic research or how children assess PI during the process of acquisition; if the former, I think her arguments do not apply because we should always look for negative evidence as per the scientific method, that is, evidence that a generalization is false or limited. After this discussion, Schifano also proposes a paramater subhierarchy accounting for the different heights of movement attested in Romance.

One limitation of the generalizations Schifano makes about the height of verb movement is that they are specific to Romance. Schifano herself notes this, and even argues that Brazilian Portuguese shifted outside of the Romance typology when it lost its rich agreement and null subject status. For me, this raises many questions. For example, do the generalizations in (2) ever apply outside of Romance? If so, what does that mean for languages where the verb raises directly from V to C, without stopping in T, as argued for by Roberts (Reference Roberts, Hornstein and Lightfoot1994), or in a language where aspect is rich but tense and mood are not? Or, how do the subhierarchies Schifano presents on pages 177 and 186 fit into a larger cross-family typology? This seems especially problematic when the very low movement languages are considered, which do not have paradigmatic instantiation for any TAM paradigms, but still belong to this subhierarchy. If these patterns are restricted to Romance, to what extent are they historical (since the languages are all related), rather than based on some underlying property of Human Language? Finally, how does a language learner identify that this subhierarchy applies to a given language during the acquisition process? Though this book does not address these questions in depth, the analysis presented therein is still a notable contribution to verb raising typologies. That this book led me to formulate such questions is itself evidence that it provides a foundation that may trigger further typological work outside of Romance. Furthermore, developing such a detailed typology that applies only to Romance also highlights an important methodological issue that we as linguists tend to overlook when convenient: even closely related languages with similar surface characteristics have underlying differences.

Schifano’s empirical diagnostics, as well, can be difficult to apply outside of Indo-European languages. I attempted to apply those diagnostics to K’iche’ (Mayan, Guatemala), and was unsuccessful, due in large part to the difficulty in eliciting adverbs in K’iche’. Although the order of adverbs established by cartography is widely accepted, there are languages where the order is reversed (e.g. Malagasy, Tzotzil, Quiavini Zapotec, Palauan; Pearson Reference Pearson and Svenonius1998, Travis Reference Travis, Zanuttini, Campos, Herburger and Portner2006), the result of snowballing phrasal movement of verbal constituents. The height of verb movement in such languages can likely be identified by the point at which the order reversal ceases; however, the existence of such languages adds some complexity. Is it possible that the two types of movement could be mixed in some language, and how could this be identified? Are there other operations that can scramble the order of adverbs and cause them to be an unreliable diagnostic?

Finally, in Chapter 5, Schifano revisits her empirical description by returning to some more difficult cases. In the first part of this chapter, she analyzes marked but grammatical word orders in Romanian and Spanish as caused by optional movement to the left and the low periphery, respectively, triggered by focus. She also discusses how participles, infinitives, and verbs inflected for mood might also target higher heads in the syntactic spine, in order to be licensed by aspect or finiteness. She argues that apparent adjacency requirements between the auxiliary and participle in some languages are actually caused because both target the same field (although she does not explain, or even note, that they seem to always appear in auxiliary-participle order rather than the reverse).

Schifano’s discussion in Chapter 5, especially when juxtaposed with her analysis in Chapter 4, raises some interesting issues with regards to the notion of the paradigm. In particular, the identification of paradigmatic instantiation for a category in a given language was done by comparing the paradigm system of the language as a whole. However, in Chapter 5, Schifano argues that particular verb forms with particular values may raise to different heights than the ‘default’ case in some languages. Why, then, is the default case about the whole paradigm, but the height of forms like the infinitive and subjunctive about the particular inflectional category? An even bigger question though – one that preceded and will continue on after Schifano’s book – is about the notion of paradigm itself. Schifano’s approach implicitly assumes that a paradigm is an entity in the grammar. After all, how can paradigms be compared if they do not exist? But an equally plausible approach, and perhaps even more appealing under popular morphological theories like Distributed Morphology, is that paradigms are epiphenomena, rather than grammatical entities.

In short, Verb Movement in Romance makes an enormous empirical contribution with regard to the height of verb movement across Romance languages. It surveys not only an impressive breadth of Romance varieties, but also a large number of forms and contexts within each of those varieties. Schifano also documents a correlation between inflectional richness and height of verb movement and hypothesizes a parameter hierarchy that corresponds to the attested patterns. However, both Schifano’s diagnostics and empirical generalizations are limited to Romance, and rely on the reality of the paradigm and, to a limited extent, on the cartographic entreprise.