A common way to draw the line between semantics and pragmatics is to argue that semantics deals with meaning abstracting away from context, whereas pragmatics deals with meaning in context. But what actually is ‘context’? Stalnaker’s book provides an impressive answer to this question. First and foremost, the book is a profound exploration of the notion of ‘common ground’. Although this notion can be traced back to Grice, who explains that certain propositions have ‘common-ground status’ (Grice Reference Grice1989: 65), it is due to Stalnaker to devise a comprehensive account of common ground (see Stalnaker Reference Stalnaker1973, Reference Stalnaker, Munitz and Unger1974, Reference Stalnaker and Cole1978, Reference Stalnaker2002). This account, as Stalnaker points out, is essentially Gricean in that it is motivated by the Gricean notion of speaker meaning (2). Since speaker meaning is defined in terms of intentions and their recognition, Stalnaker argues that communication is only possible if there is a relatively rich background shared by the participants of the conversation. This background is the aforementioned common ground.

Altogether, the book consists of eight chapters and an appendix. In what follows, I first summarize the main content of each chapter and then provide a critical evaluation of the book.

In the first chapter, three different notions of context are presented, one of them being the notion of common ground. According to Stalnaker, this notion of context is to be understood as a body of available information that can be accessed by all participants of a conversation (24).

However, there is more to say about what constitutes common ground, which is why the second chapter is completely devoted to elaborating this notion. Stalnaker argues that the notion of common ground should be explained without making use of any conventional rules of language (35). The main reason for this is that the notion of common ground is not confined to a range of linguistic phenomena but plays a central role in any kind of communicative practice whether it involves language or not. In this regard, the so-called process of ‘accommodation’ is of great importance, particularly in cases where a speaker is presupposing something not known to the hearer. For example, a speaker may say that she has to pick up her cat at the veterinarian, even though the hearer did not know that she had a cat. In this case, there is good reason for the hearer to believe that the speaker has a cat, since the speaker is presupposing that. Accordingly, the hearer accommodates her beliefs. This process of accommodation is not based on a conventional rule of language, but takes place for rational reasons that pertain to the conversation at hand.

The third chapter deals with the diversity of presuppositional phenomena. Among these phenomena are, for example, change of state verbs (John stopped beating his wife), factives (He is surprised that Ramona voted for Obama), definite descriptions (The supermarket is closed) and other referring expressions (I want to introduce you to my wife). Given that these phenomena, which are grouped together in a single category of presuppositional phenomena, are very different, Stalnaker argues that the attempt to treat all of them as cases of truth-value gaps runs the risk of inferring too quickly from surface phenomena to theoretical conclusions about what explains these phenomena. In other words, rather than looking for a uniform explanation of all presuppositional phenomena, Stalnaker argues that it is better to explain them case by case; while some of these phenomena can be explained semantically, others can be explained pragmatically (70).

The fourth chapter is concerned with the role common ground plays in the process of utterance interpretation. Stalnaker claims that there is a permanent interaction between the process of semantic composition and the process of pragmatic reasoning. More precisely, the common ground, which is permanently changing throughout the process of utterance interpretation, contains pragmatic information that interacts with the given semantic composition rules (79). In view of this, one problem addressed in the chapter is the so-called ‘projection problem’ for presupposition. Assuming that presupposition is a semantic relation, this problem has been understood as the problem of specifying the compositional rules for presupposition. However, Stalnaker argues that this is not the right way to pose the projection problem. Based on the view that presupposition is a pragmatic relation, he regards the projection problem as the problem of revealing some generalization about the relation between what speakers presuppose when they use a complex sentence and what they presuppose when they use a part of the same sentence to make a statement (95).

In the fifth chapter, Stalnaker explains that speakers and hearers use the information in the common ground to form what they say and to interpret what is said. However, the fact that the information is available to them must itself be part of this information. Stalnaker demonstrates how such ‘self-locating’ information is represented in the common ground (108).

The sixth chapter is also concerned with self-locating information, but deals with the way this information is expressed in conversation. As Stalnaker points out, there are linguistic devices that help the participants of a conversation to stay on the same page in relation to self-locating information. Among these linguistic devices are modal expressions such as may, might and must. Stalnaker argues that his notion of common ground helps to understand how these expressions actually work and what their role in communication is (126).

In the seventh chapter, Stalnaker sheds light on the contrast between indicative and subjunctive conditionals. In particular, he is concerned with the grammatical devices that mark this distinction and serve to signal that certain presuppositions are suspended (176).

The final chapter is devoted to the account of semantic relativism, since the notion of common ground devised in the book is friendly to this account. The main idea behind semantic relativism is that what appears to be an absolute claim turns out to be a relational claim. However, there are more radical accounts of semantic relativism. For example, new semantic relativism states that propositions themselves have truth-values only relatively (192).

Turning to a critical evaluation of the book, it should be understood from the start that the notion of common ground is far from uncontroversial. Over the past years, several objections against this notion have been raised. In my view, two in particular are worth discussing here in that they demonstrate some disadvantages of Stalnaker’s position.

The first objection relates to Stalnaker’s view that the speech act of assertion serves to add propositional content to the common ground. Although nobody would deny that assertions are normally used to add content to the common ground, there are clearly cases in which this does not happen, such as cases of phatic communion (see Abbott Reference Abbott2008: 532). In the book under review, Stalnaker makes no serious attempt to adjust his view on assertive speech acts, which thus remains rather idealized and simplified.

The second objection worth discussing here is based on the observation that there are cases in which speakers and hearers pay no attention to the common ground. Rather, they form and interpret what is said from their own perspectives (see Keysar Reference Keysar2007: 71). Given that such egocentric cases of communication seem to undermine the notion of common ground, it is surprising that Stalnaker does not address the growing literature on this topic.

Despite these weaknesses, the volume addresses several intriguing questions. For example, there is the question of whether the output of a semantic theory should be a truth-value or a proposition. Assuming that semantics is a part of a general theory of communication, Stalnaker favours the latter. This is a convincing position, in that a propositional output allows one to speak about information as well as about the exchange of information.

All in all, Stalnaker offers a worthwhile exploration of the notion of common ground which is thoroughly Gricean in spirit. In particular, the attempt to develop an account of common ground without relying on conventional rules of language is highly impressive; I would argue that Stalnaker is well justified in claiming the autonomy of pragmatics, since context matters in any kind of cooperative activity, whether or not it involves the use of language.

In sum, the book represents an imposing contribution to linguistic research in semantics and pragmatics, and it is essential reading for anyone interested in these domains.