Louise Cummings's Clinical pragmatics is a valuable resource. The book has two main aims: to survey pragmatic disorders and existing treatments of them, and to evaluate the field. It meets the first aim very effectively. With regard to the second aim, the book has some noteworthy flaws, but is nonetheless a useful contribution to an important emerging discipline.

Cummings considers a host of pragmatic deficits, describing both developmental disorders and deficits acquired in adulthood. Chapter 2, ‘A survey of developmental pragmatic disorders’, is devoted to the former. She explains in detail the symptomatology of Specific Language Impairment (SLI), Autism Spectrum Disorders (ASD), emotional and behavioral disorders, including especially Attention Deficit Hyperactivity Disorder (ADHD), and numerous varieties of mental retardation (e.g., Down's, Fetal Alcohol Syndrome, Williams' Syndrome). As for acquired deficits, in Chapter 3, ‘A survey of acquired pragmatic disorders’, Cummings provides a detailed overview of the pragmatic effects of, among other things, left and right hemisphere damage (whether due to stroke, other lesions or trauma), and the pragmatic deficits characteristic of schizophrenia and Alzheimer's.

Beyond presenting the symptoms, Cummings provides an overview of the various suggested causes. In particular, in the fourth and fifth chapters (‘The contribution of pragmatics to cognitive theories of autism’ and ‘The cognitive substrates of acquired pragmatic disorders’) she explains at length the posited roles of Theory of Mind, Weak Central Coherence, and Executive Function. Building on this, she catalogues in Chapter 6, ‘The assessment and treatment of pragmatic disorders’, a host of existing assessment tools and treatment regimes.

In terms of the second aim of the book, evaluation, Cummings writes:

In essence, Cummings alleges two such failures. She points out, with some justice, that many studies lack a clear theoretical rationale: collectively, she insists, we need to work towards explanation of the deficits, with less energy spent on unending description of the symptoms. Related to this, many clinical studies include too much under ‘pragmatic behaviors’ – indeed, some treat ‘clinical pragmatics’ as if it studied every imaginable impediment to effective communication. This, suggests Cummings, has hampered assessment and treatment. On both fronts, the proposed remedy is a more global, multidisciplinary perspective, informed by pragmaticians (29).

So much for my survey. I turn now to my own evaluation.

Clinical pragmatics demonstrates an excellent command of the literature, covering an extraordinary range of fields. The explications of the symptoms and their posited causes, and of the existing interventions, are clear and comprehensive. What is more, Cummings is scrupulous about the complexities, and about what researchers do not know. Beyond serving as a concise reference work, the book is also an excellent departure point for deeper investigations into various pragmatic deficits, affording both the necessary background and appropriate pointers to further reading.

I would highlight three other very positive features of the ‘survey’ part of the book. Cummings presents novel and intriguing explanations of pragmatic deficits. For example, she reviews studies which tie pragmatic impairments to temporal processing constraints and/or phonological short-term memory. In particular, SLI correlates strongly with slow auditory processing in infants, and with verbal short-term memory deficits (50f., 79). A similar correlation between troubled phonological memory and pragmatics was found in Down's Syndrome (74). Second, she identifies highly specific neurological sites for particular pragmatic deficits:

(See also 100, 104 and 154 for more on the neural correlates of sarcasm, metaphor and illocutionary force recognition respectively.) Third, and finally, Cummings reviews numerous studies which place pragmatic tasks on a continuum of difficulty: direct speech acts; indirect speech acts; metaphor; irony; deceits; and, finally, failure recovery (see, for example, 28 and 148).

This exhaustive survey does have its downside: sometimes too much detail is provided (as, for example, illustrated in the quotation immediately above!); not all of the literature reviewed seems necessary, given the overarching goals of the book; and there is some repetition. These, however, are minor quibbles.

The more serious negatives pertain to the evaluative portions of the book. I want to focus on a bipartite complaint of hers, namely that researchers in clinical pragmatics have not properly delimited pragmatics, and that this has impeded proper assessment and treatment. I was not convinced on either count. In support, Cummings offers numerous case studies. To be clear, there are lessons to be learned in these sections. She does find genuine mistakes in certain works; she uncovers numerous cases where researchers failed to give proper due to the abilities of their subjects; and she is right to point out that, given how inherently flexible and holistic pragmatic interpretation is, the explicit teaching of pragmatic rules cannot eliminate pragmatic deficits (though, pace Cummings's pessimism, such training can improve performance). However, the errors she highlights do not, I think, actually trace to confusions about the nature and scope of pragmatics. They are orthogonal to that issue.

Moving beyond the (non-probative) case studies, Cummings's plaint rests upon her idiosyncratic understanding of the ‘truly pragmatic’ (about which more below). Her criticisms in this regard strike me as mostly terminological. She does indeed establish that various studies are not targeting the truly pragmatic in her sense. But, first, she does not show that clinicians are making a genuine error thereby; and, second, she does not establish that choosing an alternative conception of the pragmatic makes for poorer interventions.

Before summing up, I would like to dwell for a moment on the nature and scope of pragmatics. Cummings's own characterization strikes me as unmotivated. Essentially, she provides a list of pragmatic speech: speech acts, including indirect ones; particular and generalized conversational implicatures; pragmatic presupposition; deixis; and non-literal usage as in proverbs, metaphor and irony. Cummings then proposes that underlying the items on her list is a ‘pragmatic competence’. It has four key cognitive hallmarks, namely being rational, inferential, and global, and involving charitably putting oneself in the other person's shoes. Moreover, this ‘pragmatic competence’ is specific to the production and interpretation of language. (Clinical pragmatics then pertains, by definition, to problems internal to this competence.) Myself, I doubt her inventory forms a natural class of behaviors; still less is there reason for thinking that there exists exactly one mechanism, a free-standing ‘pragmatic competence’, that underlies all and only the list's members.

A more plausible view (first proposed, I believe, by Asa Kasher) is discussed by Cummings herself. The idea, in a nutshell, is that there are two kinds of pragmatics. ‘Decoding pragmatics’ pertains to linguistic devices which are, as a matter of their form or content, context-oriented. This includes, for example, use-theoretic meanings for mood and politeness markers; lexically specified presupposition; indexicals and other context-sensitive formatives; language-specific rules for turn-taking; etc. Here there really is a competence: knowledge of pragmatic encoding. ‘Inferential pragmatics’, in contrast, pertains to interaction effects of linguistic competences with cognitive capacities applicable to communication generally. Here would fall, for instance, conversational implicature, along with ‘live’ metaphor and irony.

Crucially, neither the decoding nor the inferential variety is characterized by a pre-specified list. Instead, to find what is ‘truly pragmatic’, one begins with pretheoretically plausible paradigm cases, then looks for what holds them together. It is only upon finding the underlying essence – which may not in fact be shared by some of the initially plausible candidates – that the class of the genuinely pragmatic may be identified.

Approaching things this way, clinical pragmatics becomes still more fascinating: it may provide evidence about what that underlying essence is. Qua philosopher of language, this constitutes my ‘holy grail’: not merely uncovering the cognitive causes of deficits, but drawing an inference about underlying cognitive mechanisms in neurotypical pragmatics – using this, in turn, as a new tool for finding, rather than stipulating, where the semantics–pragmatics boundary lies.

In sum, I learned a great deal from Clinical pragmatics. The book is exceptionally strong on describing the state of the art, both in terms of the nature and causes of various pragmatic impairments, and the currently available treatments. There are weak points, including in particular Cummings's own overarching twofold critique of the field. Nonetheless, she is right that a more global perspective is needed, and this important book lays the groundwork for it.