This remarkable book consists of nine chapters: an introductory exposition dealing with three key, but disparate, notions, three chapters on name theory, and five on an analysis of what the author calls the grammar of names formed in Old English (OE), by which she means exclusively given (non-hereditary personal) names. Hereafter, following the author's lead in her title, these are referred to in this review simply as names. Some of the reasons for calling the book remarkable are positive but others not so. The book is by no means one to be picked up and dipped into; it is a difficult read requiring sustained attention and a range of expertise.

The introduction consists of three sections, which do not unify smoothly. The first is on the nature of an onomasticon or name-lexicon. Colman emphasizes on p. 1 that such a concept is necessarily a matter for individual human beings, but nevertheless seeks to persuade us that there is a way of approaching it theoretically as an abstract concept that may be associated with a language, just as a lexicon is. The second section is a brief note about sources of data. The chapter concludes with a section on the expression of gender in name data culled from Anglo-Saxon sources. This last section (1.3) is perhaps the oddest, in that it is an extended lucubration on the expression of gender in names on coins of the Anglo-Saxon period, which concludes with the author's firmly expressed belief that any apparent gender ambiguity in such names is illusory. The section does not relate straightforwardly to the main thrust of the book, although the tension between grammatico-lexical and biological gender emerges as a sub-theme of chapter 8.

The three chapters of part I (2–4) deal with the relationship between names and words, specifically nouns. Chapter 2 considers the possible equation of the concepts name and noun, in order to reject that equation firmly in chapter 3. Did anyone still believe it was applicable? Undoubtedly many names in many languages do have the appearance of nouns (e.g. in their inflectional properties), but many are syntactically phrasal, and in their primary referential function all of them operate as noun phrases, or the equivalent notion in a range of theoretical frameworks. Colman restricts her analysis to one-word names, because that is all that Anglo-Saxon culture had, if we ignore the issue of by-names and nicknames. In chapter 4, the notion of the referentiality of names is explored, followed by a discussion of their lexicology and of suitable principles for their lexicography. Names are abstract objects (i.e. generated by a grammar, like lexical items and sentences), and placed in the real (mental) lexicons of users by acts of bestowal (unlike sentences).

OE names may consist of either two elements (dithematics), one element with lexical suffixation or suffixation which may indicate paradigm class (suffixed monothematics), or one element alone (monothematics), in each case with the potential addition of inflectional suffixes. The five chapters of part II (5–9) present an analysis of name elements and their grammar, etymologies and “semantics”. Chapter 5 introduces the attested name elements and their combinatory possibilities, which are not semantically driven, but veer in the direction of randomness, a direction mitigated only by certain gender-based considerations and some other culturally driven restrictions. In chapter 6 the formation of lexical items in OE is analysed using a Dependency Grammar (DG) approach, in preparation for the comparison and contrast of the formation methods employed for names in chapter 7. Chapter 8 deals with paradigmatic variation in name structure. The concluding chapter 9 presents a brief summary of the analytic framework, with a ‘sample onomasticon’ showing how the system devised might be operationalized, using selected name elements classified by a matrix of seven grammatical features. The book ends with the usual apparatus of a reference list and indices of authors, the personal names discussed, and subjects.

Even on the basis of this sparse description, the reader will get an idea of the breadth of scholarship and the originality of what has been attempted. Not many scholars have a simultaneous thorough command of name theory (here drawn from linguistics, e.g. the work of Willy Van Langendonck and Bill Nicolaisen, rather than philosophical logic, except that the shadow of J. S. Mill hovers over chapter 2 like a broadly benevolent but sometimes uncomfortable ghost), and of grammatical theory and Germanic philology. Colman is to be commended on the ambitiousness of the way she has set out to synthesize findings over such a broad range.

In chapter 1, Colman establishes the position on which the rest of the book depends: namely that an onomasticon is not a list of names like an old-fashioned phone book but a set of elements and principles for constructing names. This immediately raises a difficulty. It is clearly apposite when considering OE with its host of lexically and structurally transparent names, but less immediately so when considering languages whose associated names are typically (though not exclusively) opaque in these respects. Modern English is one such language, in which Lily is untypically lexically transparent and Basil irrelevantly so (few men – or women – are named after the pot-herb, I suspect). Even OE had a slew of names which are still unexplained a century after they were first consigned to lists of ‘unintelligibles’ by Mats Redin in his Uppsala dissertation of 1919; Colman returns to such names in chapter 5, section 4. Colman's theoretical characterization of the onomasticon as a highly structured device for generating name-forms is therefore not wrong-headed, but applies differentially across languages. It applies strongly in the case of OE; and it may seem to in the cases of Yoruba and Arabic, but in the former practically any sentence might in principle serve as a name, and in the latter the formulaic names are governed by the finite list of the Qur'anic attributes of God. It applies weakly in the case of Modern English, pre-nineteenth-century Finnish, Hindi (for users of the language who are not Sanskrit scholars), and in general in the case of languages whose onomasticon is etymologically that of another language or an earlier phase of speakers’ own. This inevitably undermines the project on which Colman has embarked, because one has a sense of the theory being set up in order to handle languages at the OE-like end of the scale, whose rarity has not been assessed, and it does not and cannot pass muster as a general theory of onomasticology. Assuredly, any such theory will include the type of constructs and devices that Colman presents, but they are not the be all and end all.

Taking Colman's position simply as a prelude to writing a grammar of OE names, as the title promises, we are on surer ground – at least, up to a point. The author has made a lifetime's study of OE names, particularly as exemplified on Anglo-Saxon coinage, notably her contribution to the Sylloge of coins of the British Isles (British Academy, 2007) and her monograph Money talks (Mouton de Gruyter, 1992). Part II, the description of her dataset of OE moneyer's names, involves a competent and sophisticated application of the theoretical apparatus described in part I. Chapter 5 introduces representative attested name elements, and chapter 6 outlines the constructional principles of OE lexical items for the purpose of applying those principles to name structures, as noted above. Accordingly, chapter 7 points out the similarities between word- and name-structures in OE, and between the processes to which they are subject. Appropriate attention is given to variation in the attested forms of names, with a solid discussion of how such variation relates to stress-based compression and reduction. The author observes that names and lexical words differ in that, in lexical words, reduction is a ‘concomitatnt [sic] of semantic obscuration’, but that similar processes applying to names occur as a result of the process of becoming a name (p. 219). Reduction gives rise to morphemes of suffixal status in words, but not in names. These statements are backed by the idea that names have no denotation, and that illustrates a fundamental difference between the author and the reviewer. Names have no sense (I presume for both of us, since Colman is at pains to note at this point that deuterothemes – second elements of compound names – lack sense), but Colman believes they also lack denotation, which I cannot understand. The denotation of a name is the set of its potential referents, surely, just as with lexical items. Here Colman is at her most detached. I can only interpret her as meaning that she wants to view names as purely linguistic objects, things that are generated by the rules of the onomasticon, whose applicability as names – that is, as expressions for referring to things in the world at large – is a secondary matter. Looking at OE names, one might see her point. Looking at the fecundity of circumstance names given by people of other cultures, such as that of Yoruba speakers mentioned above, or of the Ambo people of Namibia, one wouldn't. Many cultures allow an enormous range of linguistic structures as names, tempting us towards the conclusion that, at one end of the spectrum, a name may have any structure at all, or none.

Chapter 8 deals with paradigmatic variation in name structure. Colman attacks, inter alia, the difficult question of whether the OE weak masculine suffix -a is appropriately classified as derivational or inflectional insofar as it appears in names. She concludes after over 40 pages of detailed argumentation, which I cannot hope to summarize in the space available, that it is not derivational, but that it participates in a system of declensional class flexibility which is a special characteristic of names.

The concluding chapter 9 presents a brief summary of the analytic framework, with a ‘sample onomasticon’ showing how the system devised might be operationalized, using 17 name elements classified by a matrix of seven features. Colman lemmatizes not attested name-forms but sample individual entries which are the onomastic equivalent of morphemes. The listed morphemes can be combined with a certain grammatical freedom, but are also equipped with item-specific semantico-pragmatic information about how they apply in relation to the sex of the bearer, and grammatical information, notably about how they are used as first or second elements in dithematics or as monothematics, which strong or weak declension-class they belong to, and their potential for forming diminutives. But herein there also lies a weakness which does not arise in the case of similar data collected ethnographically. An ethnographer can probe native speakers about the reasons for the lack of certain forms in the data and discover whether they are accidental or principled gaps. Colman is hampered by the lack of native speakers of OE in her circle of acquaintances, and cannot be sure whether the absence of a compound X + Y or of a suffixed form X–Z is a fluke or not.

After digesting this table in section 9.5, and having been instructed for over 250 pages in how an onomasticon works, the reader might be forgiven some frustration on being told that ‘regularity, in the form of sheer familiarity (to the point of tedium perhaps) can be imagined as obscuring, for certain names for a given speaker, the discreteness of these elements … a late Old-English speaker may have had the whole Godwine sequence onomasticized’. That is, whilst the name Godwine may well have been compounded into existence by rule at some point in history, that fact may have no practical consequences in the actual act of naming or in the referential or vocative uses of the name of an actual person. It seems to me that the author has got things upside down in pursuit of a rather austere formal tidiness. Or that perhaps in an onomasticon like that of OE some things are one way up and others the other way up. She opens the possibility that a name bestower might simply say ‘I name this child Godwine’. Three interpretative options are open: (1) simple bestowal, as hinted here; (2) partially etymologized bestowal: the child is given a name because in some family naming with the deuterotheme -wine has become traditional (though not obligatory, i.e. not rule-governed); (3) bestowal of a ‘possible name’ after consultation of a Colmanian onomastic grammar. Given those alternatives, it would appear that there is no more reason to think that naming is driven by a grammar than by a list of previously given names and some analogies which are partly linguistic and partly genealogical; or in other words, heterogeneous heuristics for namegiving.

The author and I have disagreed in print regarding the question of the categorizability of names into semantic or pragmatic subclasses, and therefore also about the nature of an onomasticon. Whilst I acknowledge the justice of the criticism directed at my own emerging analytic framework for the meaning of names on pp. 68-9 of this book, I venture to point out that I quickly moved on from the position which Colman sketches there. I have published a full critique of challenges to my developing thinking in Onoma 52 (2017), and this is not the place to do battle about it, except to note that Colman subscribes to the idea that names fall into culturally defined categories (notably in this book human and non-human, male and female) whilst she knows that in fact the allocation of names to classes is probabilistic and prototypical, though she wishes to deny it (e.g. p. 46).

The final judgement must be that the book, for the many virtues of its parts, does not succeed as a unity. It is theoretically sophisticated, the knowledge of the data displayed by the author is impressive, the application of the chosen framework to the data is at once thorough and meticulous. For all that, it starts from a flawed premise that what is being illustrated is a general theory of onomastic grammar. If it is not that, it must be a theory intended only to apply to OE. This ‘theory’ can be drafted in as an essential part of an overarching theory which is more sensitive to the variety seen in systems of onomastic practice world-wide, but that is something which we still await. It seems odd indeed to start from refining a theory which is an offshoot of DG into a general scheme and then to apply it to data culled from a single defunct language. I remain keen to see a sophisticated theory emerge from the empirical consideration of data from a range of cultures.

The reviewer feels the lack of the human dimension quite sorely. Naming conventions are not just a matter of grammar. Nothing in the book explains, for example, the attested popularity of certain compound names, like Wulfsige or Eadweard, versus the thin representation of others, which is a non-trivial matter, as noted above in the discussion of Godwine. The author does not explore – and does not set out to explore – the key onomastic dimension of naming-after. Such things make it an exercise in a rather detached linguistics, and not in onomastics – a shame given the presence of the word culture in the book's subtitle. It is not as though Colman is completely insensitive to the human dimension; as early as section 2.3.3, she discusses a range of issues from circumstance names (e.g. babies being named from local events accompanying their birth, such as a volcanic eruption – not in Anglo-Saxon England!) and the socially indexical nature of some modern English names. But in the discussion of OE names, little of such awareness is deployed. The discussion more or less stops short after the paragraphs in section 5.1 dealing with the characteristic lexical sources of OE name elements, and what such elements reveal about the Anglo-Saxon culture of the book's subtitle.

As might be expected from Oxford University Press, the book is beautifully presented and there are very few typos.