In Sensorimotor cognition and natural language syntax, Alistair Knott provides a theory that argues that the syntactic structure of a sentence is grounded in the sensorimotor mechanisms via which we perceive and interact with the world.

After an introductory chapter outlining the volume's contents, the second chapter, “Sensorimotor processing during the execution and perception of reach-to-grasp action”, begins with a basic overview of brain anatomy and a brief explanation of the neural pathways of the visual system. This system includes the object classification pathway; the visual pathway in the visual cortex, specifically the inferotemporal cortex, which encodes the shape of the object being reached for (as well as identifying and categorizing it) and/or the agent doing the reaching. It also includes the visual pathway for attentional selection; the pathway that encodes the location of objects in the agent's environment, computing location of visual stimuli (rather than its form). The chapter then goes through detailed neuroscientific explanations of the sensorimotor pathways involved in the execution and perception of a reach-to-grasp action. Focusing on this single kind of motor action, Knott reviews the sensorimotor processes that occur when an agent observes or participates in a simple transitive action. First, the author explains the ‘reach’ neural pathway, which is concerned with converting visual representations of object location into arm movements. The reach pathway receives input from neural regions that classify the shape of observed objects and from somatosensory areas that encode the current shape of the hand as well as its tactile sensations. Second, Knott deals with the ‘grasp’ pathway, which is concerned with converting a visual representation of an object's shape into a hand movement and generating an articulation of the hand and fingers that cause the fingertips to travel to the appropriate position for grasping.

To provide a theory of syntactic structure that is grounded in sensorimotor processes, Knott argues in the third chapter, “Models of learning and memory for sensorimotor sequences”, that not only are linguistic utterances immediate reflexes of motor actions, but they may also encode actions or events that have occurred in the past. He therefore hypothesizes that the semantic representations underlying utterances come from memory. Hence, working memories and episodic memories are briefly explained. In particular, Knott explains the working-memory representation of the sensorimotor sequence. In addition, through introducing the basic neuronal model of episodic memory, he discusses the role of working memory in the encoding and retrieval of episodic memory sequences. It is argued that planning to grab a cup encodes a particular sequence of sensorimotor operations in the prefrontal cortex (PFC). At the same time, the working memory component of the PFC maintains this sequence as a ‘plan’ to grab a cup. In other words, memory preserves the sensorimotor sequence that characterizes the cup-grabbing action through working memory and/or via retrieval from episodic memory

In the fourth chapter, “A syntactic framework”, Knott introduces a syntactic account of transitive sentences grounded in sensorimotor cognition. He argues that the speakers of the world's languages all have the same sensorimotor and memory apparatus, therefore the syntactic structure of a sentence describing a particular concrete episode is the same in any language. However it is worth noting that the model presented in this book does not directly address the structure of sentences that do not relate to sensorimotor action. In the fifth chapter, “The relationship between syntax and sensorimotor structure”, Knott details, on the basis of the arguments provided in the previous chapters, the sequence of sensorimotor processes involved in experiencing the cup-grabbing episode replayed in working memory. In the other words, he supports a sensorimotor interpretation of logical form as a description of sequencing structures in the prefrontal cortex.

In chapter 6, “Linguistic representations in the brain”, Knott deals with the neural mechanisms underpinning linguistic behavior by reviewing the neural mechanisms involved in language processing. Here he aims to situate the sensorimotor framework argued for in the previous chapters in relation to language acquisition and developmental processes. Therefore, in this chapter he reviews the language development of infants. First he discusses the neural representation of phonemes; that is, how they are represented and generated in the brain. Next he discusses the mechanisms by which the brain represents the meaning of concrete nouns and action verbs. Lastly, he discusses the neural locus of syntactic processing by reviewing studies on language disorders and the neuroscientific evidence regarding phonological loops – the ability of working memory to retain short sequences of syllables and repeat them later – in people with impairments.

Based on Knott's discussions of sensorimotor processing, working memory and the neural substrates of language and its development in infants, the seventh chapter, “A new computational model of language development and language processing”, introduces a computational model of language processing and language development in infants. The aim of this chapter is to present “a computational model of how children learn the meanings of the words and a computational model of how children learn syntactic constructions” (p. 271). Here the author presents a model that explains how children learn the meaning of concrete words and also how they learn to associate working memory representations with sequences of words.

This book is of interest to linguists, more specifically to psycholinguists and neurolinguists, from graduate students to experts in these fields. It is noteworthy that the sensorimotor model (the main model that the author presents in this volume), can be considered one model among many for explaining sensorimotor representations of syntactic structures as “there are many other models that have been synthesized from the huge literature on the performance, perception and memory of reach-to-grasp actions, and no doubt many more that could coherently be synthesized” (p.153). However, the sensorimotor model in particular, Knott claims, might reintegrate Chomskyan grammar, and in particular Minimalism, into mainstream cognitive science. Additionally, Knott argues that the model that is provided here is more detailed than the competing models and is stated in more general terms, as well as incorporating many aspects of previous models of grasping. The author has provided a model in which the logical form of a concrete sentence is the description of a rehearsed sensorimotor process. In other words, whenever a speaker is producing a sentence, they are mentally rehearsing this process.