It is commonplace in linguistics to argue that human language is unique in being able to represent itself, an insight that has underpinned much of the early work undertaken by sociolinguists in the area of language attitudes and folk linguistics. However, the increasingly interdisciplinary nature of sociolinguistics, informed not least by recent work in (critical) discourse analysis and language ideology studies, has highlighted the extent to which this central metalinguistic function represents more than a mere cognitive ability to reflect on language “as object.” This is because it is through metalanguage (ML) that we are able to convey our ideas not only about what language is, but what we think it ought to be. As such, ML shows itself to be more than simply self-serving: It is inherently ideological. Seen in conjunction with an increasing awareness of the nature of language as not only socially contextualized but also contextualizing, the notion of metalinguistic competence is therefore closely bound up with hegemonic struggle at a particular point in history (late modernity) when many traditional social boundaries might well be being dismantled but where new ones are constantly emerging in their stead – a process to which language and discourse are indisputably central.

Metalanguage: Social and ideological perspectives is divided into four sections. In Part One, “Approaches to metalanguage,” the various contributors flesh out the main theoretical issues in relation to the study of ML. Nikolas Coupland & Adam Jaworski begin with what is itself both a well contextualized and highly contextualizing contribution that aims to synthesize the various approaches to ML as deployed in different domains of linguistic research, such as language attitudes and folk linguistics, language representation (i.e., Bakhtinian notions of “double voicing” but also work in stylistics on the representation of speech, thought, and writing), the poetics of style and stylization, and the language-ideological approach emanating from linguistic anthropology. Central to their reflections is the extent to which it is actually theoretically desirable, let alone possible, to uphold the distinction between language, on the one hand, and the so-called meta-zone, on the other, not least when it is clear that the relationship between the two is consistently one of symbiosis as opposed to parallelism. It is in this context that Coupland & Jaworski identify the language-ideological notion of “iconization” as a key theoretical tool in the transition from a descriptivist to a critical stance on social meaning in language, with its attendant shift from notions of naturalness, orthodoxy, and common sense to questions of intentionality and impact. This, in turn, suggests the concept of metalinguistic competence to be similarly in need of recalibration, when one considers that judgments of competence are always formulated in the context of extant “orders/regimes of discourse.” A good example of this is provided in a later chapter by Adam Jaworski & Itesh Sachdev, which examines the metalinguistic evaluations provided by a number of UK schoolteachers writing references for their pupils' applications for a place at university. Against the backdrop of a discursive regime that typically values talk over silence, the authors uncover a disturbing gender difference in the assessment of the communication skills of male and female applicants, with the latter emerging as less likely to be characterized by their teachers in terms of the positive qualities of loquaciousness and geniality.

Sadly, there is not the space here to comment on the individual chapters of the remaining three sections of the book, which gradually move the theme of ML forward from “ideological construction” to “social evaluation” to “stylization,” by which time it becomes increasingly vexing to try to distinguish between ML in its pure sense of “language about language” and notions of metapragmatics, metadiscourse, metacommunication, and metasemiotics, where language is employed in the representation and construction of a plethora of ideological values. Suffice it to say, however, that the quality of all contributions is high, and the diversity of areas on which the theoretical debate over ML has been brought to bear is impressive, ranging from the more traditional themes of language evaluation in the sociolinguistic sense of language attitudes to the functions of metalanguage/metadiscourse in the context of advertising and shopping.

The collection is brought to an apposite conclusion with a reflective commentary by Deborah Cameron. Here Cameron returns to the key question of the desirability of maintaining the language/metalanguage divide, which most contributors have problematized but by and large chosen to uphold (at least for the purposes of this book). Cameron also identifies two recurrent sociolinguistic dimensions of ML, its moral and ludic functions. She then revisits the all-important question of “folk” versus “expert” conceptualizations of ML introduced earlier in the volume by Coupland & Jaworski, rounding the discussion off nicely with a personal account of the way in which professional self-reflexivity on the part of linguists can (indeed must) inform our discussions of the real-world applicability of our own academic linguistic knowledge. In this so-called real world, Cameron notes how folk notions of metalinguistic competence are typically rooted in the belief that language should be used to “tell it as it really is,” whereby any perceived deviation (for example, on the part of the media) is likely to be denigrated as manipulative and hence ideological (in the folk sense of the term). Yet while many professional linguists nowadays might want to critically explore, and most likely reject, the notion of language as a mere conduit for the communication of a fixed and preexisting reality, Cameron shows how, outside the confines of academe, such an insight may prove at best irrelevant and at worst counterproductive. Thus, describing how she was once approached by a group of psychiatrists frustrated with what they saw as vacuous and misguided changes in the language of the National Health Service (NHS) bureaucracy (for example, “patients” were referred to as “customers”), Cameron found herself confronted with the typical folk discourse whereby language was failing adequately to reflect the real world: The psychiatrists correctly contended that the word “customer” could only apply where patients had a choice of locations to direct their custom, which was clearly not the case. Though certainly able to achieve much in terms of raising general metalinguistic awareness among the psychiatrists in question, Cameron soon discovered that her own expert imperative to argue the case of language as not merely constituted by reality but also constitutive of it was largely beside the point. What the psychiatrists required to be truly empowered was not theoretical-linguistic enlightenment that would invert their arguably naïve view of the language–reality relationship. What they needed from their expert witness – as committed academic – was a heightened degree of critical metalinguistic awareness that would afford them the ability to tackle NHS managers within that self-same folk discourse. It is a fitting story with which to conclude this book, and one that echoes Cameron's longstanding view of the need to engage with those outside of linguistics on their own metalinguistic terms if we are, as linguists, to see our insights usefully applied in the real world.

In conclusion, Metalanguage offers valuable insights into how we might, as linguists, achieve such real-world relevance, for it is only by understanding the theoretical form and function of ML that we can begin to reflect on our own ability to act in the midst of conflicting folk and expert discourses of language. The book first began life in 1998 as a Round Table in Sociolinguistics organized by the Cardiff University Centre for Language and Communication. However, it is considerably more than what publishers typically fear as an ad hoc collection of diverse papers arising from a conference. Metalanguage is a theoretically formidable and thoroughly edited collection by leading academics in the fields of sociolinguistics, discourse analysis, and media/communication studies. It is also a fine contribution to the already impressive “Language, Power and Social Process” series edited for de Gruyter by Monica Heller and Richard Watts.