I am delighted to review this thought-provoking book because it provides a solid view of semantics, creating a reliable path for students who find it hard to orient themselves in the forest of controversies concerning the semantics/pragmatics debate. In fact, while a number of scholars have defended the view that semantics is context-independent, others have claimed that pragmatics intrudes into truth-conditional meaning and that, without such intrusions, it is not possible to arrive at the propositional form of a statement.

The book by Emma Borg is an intriguing discussion of the semantics/pragmatics debate in which various (diametrically opposed) views are considered and the reasonable point is defended that semantics is both truth-conditional and largely independent of contextual effects. According to the author, words and their syntactic combinations determine the truth-conditions of sentential meaning and semantics should be considered formally tractable. In Chapter 1, Borg discusses the idea of minimal semantics as well as the ideas at the heart of various forms of contextualism: contextualism of the radical or of the moderate type. In this chapter, she discusses the role played by context-shifting arguments in contextualist positions (see discussion below). In Chapter 2, Borg discusses the relationship between Minimalism and psychological evidence. She considers various objections to Minimalism on the basis of the idea that Minimalist positions conflict with our normal intuitions about the meanings of utterances. In Chapter 3, Borg defends a view of Minimalism in which meaning is seen as propositional and she defends this perspective against opposing views, such as moderate contextualism and relativism. Chapter 4 focuses on reference and demonstratives. Borg attempts to defend a view of reference that expunges speaker’s intentions. To do so, she explores new theories such as the theory of mirror neurons, the idea that there is a class of neurons which are activated when a person performs an action. In Chapter 5, Borg considers semantic Minimalism in the light of Chomsky’s semantic internalism, the view that semantics only provides conceptual representations. She argues that there is a connection between sentential and word meanings and the world and that, at least in simple cases, sentences are about the world. She rejects semantic internalism as she views the proper task of semantics to beto allow us to talk about the world. Chapter 6 considers various methodological arguments against minimal semantics. Borg here argues that there is no inconsistency in combining semantic Minimalism with an approach that allows the possibility of materials which are not voiced at the surface level – that is to say unarticulated constituents (a case in point would be ‘I am not ready (for that)’).

This book can be seen as an antidote against radical versions of contextualism. Contextualists have often threatened the autonomy or independence of semantics by using examples such as the following: ‘Mary would like to drink some beer’. John opens the fridge and says ‘There is some beer in the fridge’ noting some drops of beer on the fridge’s floor. Intuitively, according to scholars like Travis (Reference Travis1989), such a statement should be considered false because the speaker’s intention is to refer to some bottles of beer and not to the stuff ‘beer’. The speaker’s intentions shape the proposition expressed. Instead, if Mary is cleaning the fridge and notes some drops of beer on the fridge’s floor, the utterance ‘There is some beer on the fridge’ would count as true.

Although there may be more problematic and controversial examples that seem to threaten the autonomy of semantics, I believe the example above is a good platform to start the discussion, because there is a tension between the pragmatically enriched statement ‘There are some bottles of beer’ and the minimal interpretation ‘There is some stuff in the fridge which is “beer”’. Even so, the minimal interpretation remains basically correct (true) and not false. I suggest that it is misguided to try to push the virtues of contextualismon the basis of these debatable examples, and I agree with Borg that many of the examples used by contextualists in order to motivate their theory of language use can be explained in a different way, without jeopardising the idea that there is a semantics and that it is truth-conditional.

As I said, there are more controversial and difficult examples, to be discussed, when one deals with the semantics/pragmatics debate. One of these is the following:

It appears that this sentence, uttered on different occasions/situations, might mean different things, such as ‘John is ready for the exam’ or ‘John is ready to leave’. Examples like this one are taken by contextualists to show that semantics is under-determined. Indeed, we are not surprised that some expressions are in need of contextual saturation or completion. Even Minimalists like Borg or Herman Cappelen and Ernie Lepore (e.g. Cappelen & Lepore Reference Cappelen and Lepore2005) accept that in language there are a number of contextual expressions (this, that, he, the local bar), but they reject the idea that the set of these contextual expressions should proliferate. Thus, they are reluctant to accept that ‘ready’ should be a genuinely contextual expression. The argument by Cappelen & Lepore, surprising though it may appear, is that ready means ‘ready’ full stop and that, only as part of speech act meaning augmentations, it means e.g. ‘ready for the exam’. Thus, Cappelen & Lepore are careful to say that minimal semantics does not coincide with speech act (final) interpretation. I have always thought that the position that ready just means ‘ready’ is too dogmatic and will not manage to persuade contextualists that this is the correct view. Aware that dogmatism is no good in philosophical arguments, Borg tries to strengthen the Minimalist’s position by injecting some syntax into the proposal and by claiming that there is a syntactic constituent which is ‘unheard’ or implicit. Thus, Borg accepts the position that ‘ready’, after all, should be construed as ‘ready for something’ (presumably she is resorting to the theory of implicit arguments, well known in syntax). This position is reminiscent of, though not coincident, with Capone’s (Reference Capone2008b) position. It might, prima facie, seem to be a good step in the right direction, but is not itself devoid of problems. Borg is aware of an argument by Gauker (Reference Gauker2012). Gauker notes that while ‘Tipper is ready’ implies ‘Tipper is ready for something’, ‘Tipper is ready for something’ does not imply ‘Tipper is ready’. Borg does not agree with Gauker’s intuitions. However, her intuitions, this time, do not seem to me completely reliable. These problems could be amended if one accepted the position by Capone (Reference Capone2008b), according to which ‘I am ready’ contains an implicit PP (prepositional phrase) like ‘for that’. Positing such a PP need not amount to enlarging the set of contextual expressions since the demonstrative that is already part of it and the preposition for has an invariant meaning. Thus, with maximal parsimony one has achieved the result of explaining how ‘ready’ can be contextually saturated, without being itself a contextual expression. The facts about implicit arguments are independently needed for syntax to work (and many other sentences should be explained by positing them, (e.g. The ship was sunk to collect the insurance)).

The book by Borg shares an unappealing quality with the work by Cappelen & Lepore (Reference Cappelen and Lepore2005). In that work, reference was considered part of minimal content. While other eminent semanticists (e.g. Higginbotham Reference Higginbotham1985) have done their best to expunge reference from the domain of semantics, by considering abstract semantic rules (e.g. the meaning of the sentence Mary is happy is the following: if Mary refers to x, then the statement consisting of the sentence Mary is happy is true in case x is happy), it seems clear that both Cappelen & Lepore and Borg are doing their best to retain all the troubles deriving from including reference in the domain of semantics. It is clear that, at one point, when Borg discusses the issue of demonstratives, she has problems (which in my opinion cannot be easily dismissed) in trying to expunge speaker’s intentions from the interpretation of demonstratives. (We may at this point remind readers that this topic was already tackled conclusively, it appears, by a contextualist like Carston in her important monograph (Carston Reference Carston2002), although there is no hint at that discussion in Borg’s book.) First of all, Borg tries to defend an anti-intentionalist story of demonstratives by mentioning a semantic rule according to which demonstratives should be accompanied by a demonstration (assuming that the demonstration is enough to establish the intended referent), but then voices some problems, relating to the possibility of not using the demonstration (e.g. a soldier faints in a line of soldiers and I say ‘that soldier …’, without using any accompanying demonstration; even so, in this case, the utterance would be interpretable because the referent is salient enough). Even if the most promising route for treating demonstratives semantically fails, Borg resorts to other routes, such as salience and also simulation theories, including the theory of mirror-neurons. Salience is, like everyone knows, a pragmatic notion and it is not easy to isolate it from the notion of speaker’s intentions. Simulation theories may also contain pragmatic components, as they may rely on abductive reasoning. The theory of mirror-neurons seems to Borg the most attractive one, because it does not require a level of abductive reasoning. However, she has not conclusively justified the view that mirror-neurons should be involved in securing reference through demonstratives. Even if they may offer ways to simulate other people’s minds in connection with simple actions (e.g. eating an apple), it is not clear that this should work in the case of demonstratives, especially if one considers the intractable cases, those where the demonstrative is not accompanied by a demonstration (this is the case that generates contextualist worries). Since there is no action to be interpreted, mirror-neurons presumably would not fire.

The best way to avoid these problems is to rely on Higginbotham’s idea that semantics should be sufficiently abstract and should be obtained through generalization. Thus, it would be best to exclude reference from the picture. This is a criticism which applies to Cappelen & Lepore’s version of Minimalism, and one which applies to Borg’s extreme defence of that picture too. It is true that Borg demonstrates that speaker’s intentions need to be regimented somehow (e.g. I cannot refer, while looking with a group of friends at a red Porsche, to the white Porsche parked round the corner). It is possible that there are discourse rules for restricting reference, but this is not the same thing as saying that there are discourse rules for establishing reference when a demonstrative is uttered. It is reasonable to assume that pragmatic reasoning is involved in the attempt of fixing the referent of a demonstrative, when there is no demonstration or when the demonstration is not enough for establishing reference. In any case, demonstratives are only a sub-set of the set of contextual expressions, and even if their use is, admittedly, regimented to a greater extent than bare pronominals, the problem of contextual expressions like pronominals is not resolved by a discussion of demonstratives. Speaker’s intentions, indeed, intrude into semantic interpretations when we consider reference fixing.

The strongest part of the book is the one dealing with modularity of mind and the attempt to segregate semantics/syntax from pragmatics (the idea is discussed in Chapter 1). I agree with Borg that semantics (and syntax) require formal mechanisms for interpretations and that such mechanisms may be part of a module that interacts with the mind-reading module but is separate and autonomous. Perhaps something needs to be added on how the representations of semantics and those of pragmatics are combined and perhaps the author could have discussed the theory of merger representations by Jaszczolt (Reference Jaszczolt2016).

The book as it is, even if it discusses important aspects of the semantics/pragmatics debate, has concentrated on aspects which can more easily be reconciled with a rigid view of semantics. After reading this book, though, we know almost nothing of all the serious work done on the semantics/pragmatics or syntax/pragmatics debate by Salmon (Reference Salmon1986), Jaszczolt (Reference Jaszczolt1999), Huang (Reference Huang1994), Levinson (Reference Levinson2000), and Capone (Reference Capone2008a, Reference Capone2016), among others, and that certainly requires taking a more critical stance towards the theory proposed by Borg. It is possible to claim that contextualism can be reconciled with a view like the one by Higginbotham on semantics and even with a view like Borg’s according to which semantics should aim at providing propositional representations at least in simple cases; however, something should have been said on this attempt to reconcile Minimalism with contextualism. This silence or reticence somehow reveals a lack of interest in the other part of the picture: abductive reasoning, how rationality can constrain pragmatic interpretation, and how the semantic potential can be boosted by a complementary pragmatics.

Even so, Borg has provided a salutary reminder to pragmaticists that semantics is indeed about representing the world, that semantics is a formal theory of meaning, that semantics needs a modular treatment, that semantics can be independent from pragmatics to some degree, and that semantics should mainly be treated without reference to speaker’s intentions. Taken with a grain of salt, this book will be really beneficial to all readers, including students and scholars. Furthermore, the author of this book has proven to be an excellent communicator and should be a model to us all.