Although all languages are unique and present certain puzzles and challenges for theoretical analysis, it is unarguably true that Icelandic is one of the most researched languages in the Germanic family. From the standpoint of its syntax, Icelandic is a very interesting language, with aspects of its word order, clause structure, agreement patterns, and case system that have aroused much theoretical interest and debate since the onset of the generative enterprise. In this volume, Höskuldur Thráinsson takes on the task of presenting an overview of linguistic phenomena related to the syntax of Icelandic. From the outset, it goes without saying that this volume is more than just a mere “syntax guide,” but rather contains a rich description of both data and relevant theoretical issues and controversies that have sparked interest in these constructions in theoretical circles. The design of each chapter is bifurcated into two main sections; the first provides a descriptive overview of a particular grammar topic while the second involves a discussion of the theoretical and comparative issues surrounding the aforementioned data.

With regard to the contents of the volume, Thráinsson spends the first two chapters introducing the reader not only to Icelandic, but also to languages that are closely related to it; most notably, other Scandinavian languages and, especially, Faroese. The first two chapters contain a detailed description of basic structural descriptions of the clausal architecture of Icelandic in discussions of issues such as the default order of constituents, positions of finite vs. non-finite verbs, the nature of V2, object position and functional categories, and the placement of adverbs and syntactic structure. Even in these initial chapters, Thráinsson spares no effort in the introduction of theoretical controversies surrounding these basic cartographic data. Take, for instance, his discussion of the formal properties and theoretical relevance of negation phrases in Icelandic (see section 2.2.5). Here Thráinsson engages the reader into the controversies of whether or not the negative particle ekki ‘not’ should be considered a specifier of a functional projection (that is, NegP) or whether this particle should be considered the head of this projection. To justify his stance on this matter, Thráinsson relies on seminal arguments from the theoretical literature (coupled with robust empirical examples). Concerning the debate regarding the specifier vs. head argument centering around the negative particle ekki ‘not’, Thráinsson cites J—nsson (1996), demonstrating that the negation ekki ‘not’ can be modified, for example by alls ‘at all’, and objects can be shifted across this modified negation, and it can apparently as a whole undergo so-called stylistic fronting (as suggested by the data in 1 and 2 below):

1. (1)

   1. a.

   2. b.

2. (2)

   1. a.

   2. b.

Thráinsson does an excellent job of walking through any particular “controversy” that might exist, weighing both sides of the issue, and in the end providing some commentary that serves as food for thought for those who want to investigate these and related issues further. Regarding the specifier vs. head status of ekki ‘not’, Thr'ansson considers proposals by scholars such as Zanuttini (2001) and Christensen (2003) who propose that ekki could either function as a head or maximal projection (in Icelandic as well as related languages such as Norwegian and Swedish). The data sited above introduce various diagnostics to test whether negation in a given language functions as a head or maximal projection in a given language.

Continuing our discussion of the relevance of these data between 1 and 2 mentioned above, a second point—exemplified by the data fragment in 3—is that the negation ekki can be topicalized (that is, trans-posed to the left periphery of the clause).

1. (3)

   1. a.

   2. b.

As a result of the possibility of topicalizing negation—a process generally regarded to move elements in Spec, CP—it should be the case that only specifiers and not heads should be eligible candidates to participate in this process (following standard assumptions regarding movement; see also Platzack 1988 for similar arguments regarding inte ‘not’ in Swedish). Turning to the data in 4, J—nsson (1996) considers how the concept of Negative Object Movement, or Negative Object Shift in Icelandic (already previously mentioned in section 2.1.5 of this book) support the analysis that negation is not a head property in Icelandic (or related Scandinavian languages).

1. (4)

   1. a.

   2. b.

Thráinsson describes Negative Scrambling along the lines of 5.

1. (5) Negative Scrambling

   1. a. is not subject to Holmberg's Generalization but Object Shift (OS) is.Footnote ¹

   2. b. Negative Scrambling applies to objects of prepositions and to prepositional phrases but OS does not.Footnote ²

According to the definition in 5, Negative Scrambling moves a negative object across a non-finite main verb when an auxiliary is present, but OS does not apply at all when an auxiliary verb is present (that is, according to Holmberg's Generalization, see note 1). As a matter of fact, the data in 6 suggest that Negative Scrambling is obligatory based on the observation that enga b—k ‘no book’ cannot remain in situ. If the object remains in situ, we get the sentence negation ekki and the relevant negative polarity item.

1. (6)

   1. a.

   2. b.

   3. c.

   4. d.

Relying on J—nsson's (1996) observations and analysis, Thráinsson concurs that since Negative Scrambling (see 5) differs in so many ways from OS this suggests that it is movement to a different position. Although the discussion at this point transitions into a larger debate about other puzzling qualities of Negative Scrambling, here we have evidence of how Thráinsson adopts a non-controversial, largely theory-neutral discussion of the data and related conceptual issues of Icelandic (and closely related languages) data. Thráinsson maintains this tone through the entire volume; patiently introducing interesting data, pointing out theoretical and descriptive controversies, and providing a detailed, yet non-threatening commentary of the issues discussed.

The remainder of the volume focuses on topics in Icelandic syntax that are quite often motivated by theoretical interest. These topics include (but are not limited to): the order of elements within clausal architecture (for example, in the noun phrase and verb phrase); case, agreement, and grammatical relations; passives, middles, and unaccusatives; expletive constructions; focusing and (heavy) NP-shift; finite and non-finite complements and adjuncts; and pronouns, reflexives, and empty categories. Anyone familiar with topics in Germanic syntax, in particular in the generative tradition of the Principles & Parameters framework, will see the motivation for the inclusion of these topics. As demonstrated by the (brief) representation of Thráinnson's discussion of the specifier vs. head and the negative particle ekki ‘not’, Thráinnson much to his credit does not avoid any particular issues because they may be contentious. Another example of this can be found in his discussion of the -st suffix that simultaneously functions (most likely) as an inflectional suffix in the construction of “true” middle voice constructions (7) versus the combination of this suffix with nouns and adjectives acting as a derivational suffix (8).

1. (7)

   1. a.

   2. b.

2. (8)

   1. a.

   2. b.

   3. c.

The inability of the predicate to be modified by a volitional adverbial element viljandi ‘on purpose’ (which is a common test used to distinguish “true” middle voice constructions cross-linguistically), demonstrates that the -st suffix in 7a is a middle voice construction. In contrast, the -st suffix behaves like a derivational suffix in the data in 8. Regardless of what formalism one employs, such data (along with the other examples cited in section 5.1.5) offer a significant challenge to an easy analysis. The data cited in this volume serves as the ideal starting point for any detailed, thorough analysis of this and related issues.

Although Thráinsson does an excellent job of avoiding the pitfall of getting overly involved with theoretical discussion, the reader who comes to this volume with little theoretical background will naturally encounter some difficulties with the text. This comment is not intended to be a mark against this volume; on the contrary, it seems to be a natural by-product of compiling such a comprehensive volume of syntactic data that addresses both descriptive and theoretical concerns. Those who are more familiar with other theoretical backgrounds (for example, HPSG and LFG) should have little trouble navigating through the text and making use of the rich descriptive and theoretical descriptions of Icelandic syntax.

This volume is remarkable in its scope and coverage and is a wel-come addition to the literature on the topic of Icelandic syntax and related topics. Perhaps the most valuable asset this volume has to offer is its detailed bibliography, which equips its reader with the resources to continue serious inquiry into these issues. The rich sources of empirical data, the enlightening theoretical discussions, and the comprehensive bibliography make this volume an extremely valuable source on Icelandic syntax to be consulted for years to come.