This work constitutes an important contribution to the historiography of English grammar and linguistics. The title gives potential readers a clear hint as to the content, which the author presents from a historical perspective and in a thought-provoking and stimulating manner, overtly tackling long and familiar controversies on linguistic categorisation.

The primary aim of the book is to examine the treatment of the preposition as a part of speech within the English grammatical tradition between the sixteenth and the eighteenth centuries, more specifically from the time of William Bullokar's Bref Grammar for English (1586) until John Dalton's Elements of English Grammar (1801). The book marks out that the English grammars to be included in this tradition are those written by native speakers of English whose texts have the purpose of describing the language. In that sense it advocates that Bullokar's work may in fact be the first English grammar.

The author does not aim to comprehensively detail the period under review, but rather to select a representative group of works, whether they are proper grammars or other linguistic works that discuss approaches to prepositions. The intended result is to offer a sample of the main trends in terms of the grammatical description of the parts of speech, particularly prepositions. In addition to recourse to the original works, both in facsimile and modern editions, the author takes into account linguistics historiography, with special attention to the exhaustive and detailed work of Michael (Reference Michael1970).

The introduction presents an analysis of the categorisation issues in certain influential grammars, namely Jespersen (Reference Jespersen1924), Quirk et al. (Reference Quirk, Greenbaum, Leech and Svartvik1972, Reference Quirk, Greenbaum, Leech and Svartvik1985) and Huddleston & Pullum et al. (Reference Huddleston and K. Pullum2002), particularly regarding the characterisation of ‘preposition’ as category. The relevant point here consists in making the reader aware of the diversity of approaches. Thus, whereas Jespersen describes adverbs, prepositions and conjunctions as a single category by virtue of a syntactic parallelism with the syntactic behaviour of verbs, Huddleston & Pullum et al. (Reference Huddleston and K. Pullum2002) propose a diffuse category of lexemes, including some which may not have a complement, like downstairs, here, now, afterwards and thereby. As a conclusion, Lundskær-Nielsen remarks on the fact that defining and characterizing categories, particularly prepositions, is by no means a settled issue, even currently.

The first chapter in the book historically reviews perspectives on the parts of speech or word classes, starting with the distinction between noun and verb pointed out by Plato. Aristotle added a third category, which he called sundesmoi, including conjunctions, prepositions and other particles. The Stoic tradition increased the list of word types to eight: noun, verb, pronoun, participle, adverb, preposition, article and conjunction. The first definition of the preposition as a category, proposed by Dionysius Thrax, points out that it is used before other elements of the sentence in compound forms and grammatical constructions. Since then, that has been the definition adopted by most grammarians and linguists. Lundskær-Nielsen introduces the treatment given to parts of speech by other Greek and Latin authors, thereby offering a good display of the first historical definitions of the category preposition. In some cases, formal aspects may be relevant in these definitions, for example the fact that prepositions are indeclinable (Varro). In other cases, semantic aspects are crucial, such as a preposition's capacity to supplement, alter or reduce the significance of its complement (Donatus). Otherwise, syntactic aspects are noted, such as the preposition's occurrence before other elements, both in composition and construction (Apollonius Dyscolus). Priscian compiles these criteria, giving a more complete treatment to the preposition as a category that conjoins all of these features. For Priscian, a preposition is:

Priscian also notes that the main property of a preposition is to be used as a separate word before case-inflected words, and in composition before both case-inflected and non-case-inflected words. Finally, Priscian observes the indeclinable character of prepositions.

Next, an interesting chapter on the evolution of grammar in the Middle Ages illustrates the introduction, well into that period, of the modus significandi by Peter Helias. Roger Bacon institutes the principle that grammar is one and the same in its substance across all languages, and only shows accidental variations in particular ones. He also establishes the principle of awarding linguistic studies a scientific status, taking up the Aristotelian viewpoint of an object of study being scientific only with universal applicability.

The treatment of speculative grammars by Boethius of Dacia and Thomas Erfurt receives remarkable care in this chapter. Speculative grammar is presented as an attempt to integrate traditional Latin grammar into a larger framework of scholastic logic and philosophy. In this context, the author spotlights Boethius of Dacia's view of prepositions as connective parts of speech with a distinct function, namely linking declined parts with the verb; according to Boethius, prepositions are indeclinable and participate in the composition of the sentence. As prefixes, they change the word they enter in composition with. Thomas Erfurt observes that prepositions get their meaning through adjacency to a form with case, linking a preposition with the event expressed by the verb. For Erfurt, prepositions precede nouns whose case they rule, and link them to the verb. On the other hand, prefixes are not considered true prepositions by Erfurt, because they form part of the word that they appear in.

Lundskær-Nielsen confers special relevance to an alternative view (Ramus, Sanctius), emerging in the Renaissance, in which prepositions are identified with adverbs within a single category. That approach influenced subsequent English grammarians, mainly due to their re-examination of criteria for the determination of primary grammatical categories. The author goes on to remind us of the influence of some Italian scholars who both developed the idea of the preposition as a sign of case and launched the debate on the distinction between case markers and true prepositions (Bembo, Giambullari, Dolce). The relevance of these remarks becomes obvious to the reader in subsequent chapters, since both the identification of prepositions with adverbs in a single category and their treatment as case markers appear very frequently in later English grammars. It is noticed that the largely influential grammar by Lily (A Short Introduction to Grammar, 1549), even though specifically on Latin, also introduces the idea that prepositions turn into adverbs when they occur alone.

In his analysis of English grammars, the author proceeds from the division established by Michael (Reference Michael1970), mainly based on two criteria: the first distinguishes between grammars that follow the Latin system and grammars that base their description on the vernacular; the second is chronological, distinguishing grammars that show some modification of the Latin system before and after 1700. Lundskær-Nielsen offers some data on the influence of Latin on early English grammarians. According to Michael's classification, nearly 75 per cent of all English grammars up to 1800 (though some with certain modifications) follow the Latin model in their treatment of word classes in English. In this context, the author points out a gradual process during the eighteenth century whereby English grammarians began to challenge the traditional parts-of-speech divisions and consequently suggest alternative systems more relevant to English than Latin.

With regard to prepositions, Lundskær-Nielsen notices that, despite the category appearing as primary or secondary in most grammars based on the Latin model, hardly any of the vernacular systems, and few of the modified ones, give prepositions such a prominent status. Instead, most of these include prepositions in the category of particles and a few of them in that of adverbs or connectives. The author observes a clearly emergent pattern through the period, by virtue of which prepositions are interpreted as particles, together with adverbs and conjunctions, and this account becomes increasingly the norm in accordance with the linguistic data found in English (p. 100). At this point, the author vindicates the English grammatical tradition of the period, which provides the basis for Jespersen's account of particles, given that present-day linguists only trace this account to Jespersen, and no further back. An obvious effect of this evolution is the reduction of the number of primary parts of speech in English grammars compared to the Latin canon. The vernacular systems tend to show a fourfold division, in most cases substantives, adjectives, verbs and particles. The trend began with Greave's De Grammatica Anglicana (1594), which shows a strong influence from Ramus, and is the first English grammar to reduce the number of parts of speech.

Lundskær-Nielsen's survey scrutinises a selection of fifty texts, noting in each case what they say about the parts of speech and particularly about prepositions. Excluded from his survey are those he considers to be ‘too repetitive or unimportant’ (p.101).

The first grammarian surveyed, William Bullokar (1586), already introduces some peculiarities in contrast with Latin grammar, though still strongly influenced by it. Lundskær-Nielsen expounds Bullokar's definition of the preposition in positional and relational terms as a part of speech used before another element, and governing an accusative case set immediately after it. In addition, Lundskær-Nielsen observes the grammarian's acceptance of the stranded use in relative clauses. Another novelty noticed by Bullokar is the use of prepositions in composition, not only as prefixes, but also as suffixes added to adverbs, such as in thereon and hereafter. Altogether, Lundskær-Nielsen remarks on Bullokar's innovative view of English grammar whereby a vernacular can be described on its own terms and thus differently from Latin.

The second work analysed by Lundskær-Nielsen is Greaves's De Grammatica Anglicana (1594). Though written in Latin, this is the first English grammar where the preposition is included as a subgroup in the same category as adverbs. He observes that Greaves displays a mixture of adherence to, and independence from, the Latin model.

These analyses of early English grammars are of paramount importance in ensuring a proper understanding of the subsequent evolution of certain aspects of English grammar, such as the frequent categorisation of particles as a single word class, or the idiosyncratic positioning of prepositions in English relative clauses.

In his review of seventeenth- and eighteenth-century authors, Lundskær-Nielsen clearly highlights those aspects of each grammar that are to influence later work, as well as those adopted from tradition. Thus, his reviews are not limited to a summary of contents; rather they show the reader how the different perspectives on particular issues evolve. For each author, Lundskær-Nielsen brings in brief biographical details that reveal academic and historical circumstances about the personage.

With regard to content, the author follows a thread in the entire book, such that he summarises and demonstrates the stance of each grammarian in relation to each parts-of-speech issue, their number, their function with respect to each other, and their classification. Secondly, he comments on the place of prepositions in each author's classification, as well as their definition as a category.

For every author in his collection, Lundskær-Nielsen pinpoints the relevant aspects that make their work noteworthy at their particular historical moment, or in relation to the bulk of the English grammar tradition. The reader enjoys, for instance, a fair exposition of: Wallis's view of prepositions as indicators of logical connections in discourse; Dalgarno's distinctions with regard to the polysemy of the preposition in in a basic local or spatial sense and other derivative senses like the temporal one; Locke's view on prepositional polysemy; Wilkins's distinction between universal grammar and the particular grammars of languages; and Lowth's, Elphinston's and Ward's fine-grained semantic analyses.

His accolades notwithstanding, Lundskær-Nielsen does not evade sharp criticism of authors deemed important or influential by historiography. In the case of Wilkins, although some aspects of his work may be relevant for linguistics, like his proposed notion of a universal grammar, his description of prepositions is harshly criticised. The author points out the opaqueness of criteria and the vagueness of Wilkins's notional distinctions, the centrality of the Latin model at listing the prepositions, and his lack of exemplification. On balance, Lundskær-Nielsen also calls attention to positive achievements in Wilkins's work, namely the distinction between causal and local prepositions, as well as between those only indicating movement (directionality) versus those indicating only rest (place) and those which may indicate both. Wilkins specifies that prepositions denoting movement cannot indicate rest (to, from, into, out of), while those indicating rest may imply movement in the complement (at, off, within, without, above, below). At this point, Lundskær-Nielsen validates Wilkins, in that his approach foreshadows certain modern accounts where the locative (L) is embedded in goal (G) (Bennett Reference Bennett1975).

It is interesting to see how the treatment of English grammar evolves in contrast to Latin. Although some of the seventeenth-century authors already mention that the English language rather than Latin is their object of study, their systems, concepts and terminology nonetheless are still based on Latin grammar. By and large, the new more autonomous perspective gains ground during the eightenth century. Thus the idea progressively spreads that a preposition behaves as, or turns into, an adverb when it is not followed by a noun. Furthermore, Maittaire is noted as discerning the phrasal verb construction for the first time, a theme that appears more frequently later on during the eighteenth century.

In his concluding summary, Lundskær-Nielsen compares his study with the grammatical systems established by Michael (Reference Michael1970), but questions the idea of a vernacular system defined as one which postulates only the categories of noun, adjective, verb and particle. As he asserts, the article also differentiates Latin and English. Likewise, he raises the issue that many of the works analysed have a pedagogical vocation, having been written by scholars or teachers with the purpose of helping students. The gradual shift towards prescriptive grammar in the eighteenth century is also emphasised.

To conclude his work, the author reminds the reader of five predominant aspects in the treatment of prepositions during the period studied:

• • Prepositions as relational elements: the appreciation is common that prepositions and particles have a connective function that gives cohesion to discourse and vividness to style. Most authors both in the seventeenth and eighteenth centuries stress that prepositions relate two elements, thus they characterise the category as a connective.

• • The interaction between prepositions and cases (government): another common pattern is the characterisation of some prepositions as case markers. Of designates the genitive, for example, and to the dative. This trend began in the Italian grammars of the sixteenth century and it is assumed by most of the grammarians analysed in this book.

• • The distinction between separable and inseparable prepositions, as well as the ongoing questioning of the latter as prepositions, although the term ‘prefix’ does not appear in the period studied: the inclusion of Latin prefixes in the category of prepositions, due to the influence of Latin grammars, is common for some authors, given that these forms are also prepositions in Latin (ex-, pre-, sub-, trans-). As a consequence, other English prefixes (a-, mis-) were also included in the category. Lundskær-Nielsen adds that some English prepositions (under, over, after) can function as prefixes too (understand, overlook, afternoon).

• • The position of the preposition in relative clauses: the stranded use of English prepositions in relative clauses already features with the first grammarians of the seventeenth century.

• • The distinction of different senses of the same preposition, starting with the motion-versus-rest distinction, and then adding causal usage, plus figurative, metaphorical and abstract usage.

In sum, the author is exceptionally successful in his intention to discuss a usefully representative selection of the relevant issues surrounding prepositions as a category, their syntactic use, and their function and meaning, rather than offering an exhaustive study of all the available material from the period. Especially praiseworthy is the extensive and practical bibliography of secondary sources, far exceeding the references used in the text, along with the care and rigour over footnotes throughout the book. All in all, Lundskær-Nielsen contributes a very helpful tool for linguists, grammarians and historians alike.