1. Introduction.

Since the 1980s (see among many others Abney 1987 and Szabolsci 1983, 1987), a substantial part of the research into the syntax of the nominal domain has been devoted to parallelisms with the verbal domain. Dorian Roehrs situates his book as being part of this tradition by claiming that demonstrative determiners and definite articles are the nominal counterparts of verbal auxiliaries. According to him, demonstrative determiners and definite articles are not only similar to auxiliaries in the respect that they are functional elements and display agreement, but they also undergo movement comparable to that of auxiliaries in the clause. Demonstrative determiners and definite articles move over adjectives, as in 1, just like auxiliaries move over negation as in 2b (see, for example, Emonds 1978, Pollock 1989, and Chomsky 1995).

1. (1)

   1. a. this_i nice t_i girl

   2. b. the_i nice t_i girl

2. (2)

   1. a. Peter would never have done this.

   2. b. Peter has_i never t_i done this.

This movement of definite articles and demonstrative determiners constitutes the main claim of Roehrs' book. He argues that it accounts for the following phenomena in the nominal domain in Germanic: the distribution of definite articles in the Scandinavian languages, the diachronic development of the definite article out of a demonstrative, the link between syntax and semantics of both restrictive and non-restrictive modifiers, and the distribution of strong and weak adjectival inflection.

2. Contents.

In the first chapter, Roehrs formulates the outlines of his proposal and introduces his main assumptions. In line with the cartographic tradition, he adopts a highly articulated structure of the nominal domain, one that has been greatly inspired by Julien (2002, 2005). In this structure, Roehrs' proposes that the determiner moves from its base-position, a low projection called ArtP, to D crossing adjectives, which are hosted in the specifier of projection AgrP, as in 3.

1. (3)

In the first half of the second chapter, Roehrs shows how the proposal in 3 accounts for the panchronic alternation with respect to definiteness marking in Icelandic, that is, the observation that the “diachronically related [definiteness markers] occur in different synchronic positions” (p. 2). The definite article in common Modern Icelandic originates from the same source as its counterpart in literary Modern Icelandic, namely from the Proto-Scandinavian demonstrative determiner hinn. Yet, the definite article in common Modern Icelandic is a suffix on the noun, as in 4a, while it is a freestanding definite article in literary Modern Icelandic, as in 4b (Roehrs 2009:47).

1. (4)

   1. a.

   2. b.

Simplifying somewhat, Roehrs attributes the difference between 4a and 4b to the availability of movement of the article from Art to D. Common Modern Icelandic has lost this movement. This results in a low position of the article, from which the article has become a suffix on the noun. In Literary Modern Icelandic, the definite article moves over the adjective to D, a position from which it cannot become a suffix on the noun.

In the other half of his second chapter, Roehrs tackles the problem of the doubling that the definite article displays in Norwegian, Faroese, and Swedish in the context of an adjective. In these languages, the definite article is a suffix on a noun in unmodified DPs, like in 5a. In case an adjective is added, it is however doubled by a freestanding definite article that precedes the adjective, 5b.

1. (5)

   1. a.

Roehrs assumes that D needs to be valued for definiteness. In unmodified definite DPs, D enters into an Agree relation with the article in Art, which is valued for definiteness, 6a. This values D for definiteness. The article itself remains in Art and is spelled out as a suffix on the noun. In the presence of an adjective, Roehrs claims that this Agree relation is blocked. According to him, adjectives, which he takes to be in the specifier of AgrP a projection between ArtP and DP, indirectly induce a second agreement domain (in the sense of Bobaljik & Wurmbrand 2005). As a result, Art and D are in different agreement domains and Agree cannot be established. In order to rescue the derivation, the definiteness feature on D is valued by moving the article successive cyclically from Art to D, as in 6b.

1. (6)

   1. a.

   2. b.

In order to account for differences in interpretation between the freestanding and suffixal definite article, Roehrs proposes that only a subpart of the definite article is moved to D. The part that moves is spelled out as the freestanding determiner. The part left behind is realized as the definite suffix.

In his third chapter, Roehrs discusses the interaction between definiteness marking and the restrictive and non-restrictive readings of modifiers like adjectives and relative clauses. Following ideas from Heim & Kratzer (1998), he claims that this interaction readily follows from his proposal in case the definite article is taken to be a scope-marking element. In order to derive the restrictive readings, the copy of the definite article in D is interpreted. Roehrs assumes that modifiers like adjectives and relative clauses are attached to some level between ArtP and DP. The copy in D therefore has these modifiers within its scope. As a result, interpreting this copy gives rise to a restrictive interpretation of these modifiers. For the non-restrictive readings, Roehrs claims that the copy of the definite article in Art is interpreted. This puts adjectives and relative clauses outside of the scope of the definite article, explaining their non-restrictive interpretation.

In his final chapter, Roehrs discusses weak adjectival inflection, that is the phenomenon that the adjectival inflection expresses in some contexts (like definite DPs) less gender, number, and case distinctions. Roehrs claims that this is again due to the movement of the definite article. According to Roehrs, elements like the definite article are marked for triggering impoverishment. As such, they delete the gender, number, and case features on the adjective when they are moved from Art to D.

3. Evaluation.

The main strength of Roehrs' book is that it tries to account for a number of seemingly unrelated phenomena by a single operation in the syntax, namely movement of the definite article. In doing so, Roehrs offers the reader some interesting new insights in the syntax of the Germanic nominal domain. These insights, like for instance the connection between determiner movement on the one hand and weak adjectival inflection, and the interpretation of modifiers on the other hand, are original and very promising for further research.

The greatest merit of Roehrs' book is at the same time also its main weakness. If one tries to account for many phenomena by one simple general claim, there is always the risk of leaving unexplained some of the finer details of the account. Unfortunately, Roehrs did not manage to avoid this risk in all cases, leaving some aspects of his account unspecified. For instance, Roehrs introduces the very interesting idea that movement of the determiner from a low to a high position brings about weak adjectival inflection in Germanic (see above). Roehrs' implementation of this idea is that the Germanic determiners that trigger weak adjectival inflection carry an impoverishment feature. Roehrs, however, does not specify any further details of this impoverishment feature. Why would determiners in the Germanic languages for instance have this feature, while those in the Romance languages do not? It also seems uneconomical, and hence undesirable, to first license gender, number, and case features on the adjective and then delete these features later on in the derivation by some mysterious impoverishment feature.

Another aspect of Roehrs' account that might be in need of some further refinement concerns the trigger for moving the definite article from Art to D. On Roehrs' account, it is not clear why the article would actually need to move as high as D. Roehrs proposes that ArtP in the presence of an adjective delineates an agreement domain. Given that D is then in a separate agreement domain, Agree between D and Art is blocked. The definiteness feature of D is instead valued through means of movement of the definite article to D. As shown in 6b above, this movement makes an intermediate landing in Agr, the head of the projection hosting the adjective. Under Roehrs' analysis, Agr and D are, however, within the same agreement domain. Movement to Agr would therefore in principle render valuation of the definiteness feature on D possible through Agree between D and the definite article in Agr. It therefore remains unclear why the definite article has to move further on to D.

Despite these points that may need some further refinement, Roehrs' book is a good scholarly work that will be of interest to anyone with an interest in the syntax of the nominal domain.